Semidesert Grassland

The semidesert grassland complex of climax range communities comprises the most xeric grasslands in North America. There are two general regions of semidesert grassland in North America both of which are associated with the Basin and Range physiographic province and major climatic deserts therein. The larger and better known of of the two regional semidesert grasslands (grassland complex) is that affiliated with the "warm" Chihuhuan and Sonoran Deserts. The Chihuhuan-Sonoran regional unit of pre-Columbian climax range vegetation dominated by grasses occurred as both 1) a transition zone (= ecotone) between semiarid prairies and plains grasslands and natural desert scrubland and 2) a mosaic of smaller grasslands and areas of Chihuhuan and Sonoran Deserts. Pre-Columbian semidesert grasslands of the Chihuhuan and Sonoran Regions generally extended from portions of Trans-Pecos Texas and Chihuhua across Arizona and Sonora into southern California. Expansion (desertification by invasion of woody plants) of Chihuhuan and Sonoran Desert vegetation onto virgin range on which natural climax vegetation was xeric grassland has resulted in expansion of disturbance climax of desert scrub and corresponding reduction of climax semidesert grasslands. The main outline of original pre-white man and disturbance-retracted margins of the xeric semidesert grasslands were and remain more-or-less conterminous with the semiarid mixed prairie and shortgrass plains grasslands of the Great Plains on the north and east. To the west and south the drier semidesert grasslands are "constant contact" with the various subdivisions of the Chihuhuan or Sonoran Deserts.

The second (and considerably smaller) of the regional semidesert grassland formations is that associated with the "cool" Great Basin Desert and, marginally, the "transition" Mojave Desert. Great Basin semidesert grasslands were considerably smaller and much less frequent in ocurrence as potential natural (pre-Columbian) range vegetation. The widely scattered and smaller semidesert grasslands of the Great Basin occurred between the semidesert grasslands and desert shrublands of the Chihuhuan-Sonoran Regions to the south and the shrub-bunchgrass steppe of the Columbia Plateau to the north. Current generalized distribution coincides roughly with pre-Columbian boundaries as they have been interpreted in studies and maps of natural vegetation. Relatively limited occurrence, along with small and restricted patial scale-pattern, of these dry grasslands is even more so on existing Great Basin range where desertification and expansion of human settlement has been pronounced. In fact--as was explained below--semidesert grasslands of the Great Basin were often completely ignored in detailed treatments of semidesert grasslands. Semidesert grasslands of Intermountain North America were recognized and described briefly by some of the better-known and more experienced students of Great Basin vegetation, but overall these xeric grassland ranges received but a pittance of the coverage given their counterparts to the south.

Technical note on syntax- Deserts are deserts, grasslands are grasslands. One of these biomes cannot be the other anymore than a forest can be a desert. One would never speak of a desert forest. So why use the oxymoron, "desert grassland"? The difference in life form of dominant plants is greater between grassland and desert (herbaceous species vs. woody plants) than between forest and desert (woody plants dominate both). The range vegetation described below is semidesert (a usage deemed to be self-explanatory and one of long-standing precise usage) grassland. Traditional titles in the historic literature such as "desert grassland" and, especially, "desert plains grassland" were shown when such titles were essential in discussion of particular papers or perspectives of specific authors.

Chihuhuan and Sonoran Deserts and semidesert grasslands are closely allied--in both space and in ecological, especially successional, relations-- such that the two major units of range vegetation cannot be satisfactorily described or understood without considering both simultaneously (at least by way of introduction). Beginning with Shrevve (1917, p. 123) through Clements (1920, ps. 145, 162, 163, 169), Weaver and Clements (1938, p. 525), Shreve (1942), and Dick-Peddie (1993), this latter following Shreve (1942), vegetation scientists recognized a regional or zonal vegetation entitled variously the Desert-Grassland Transition (Shreve, 1917), Desert Grassland Transition Region (Shreve, 1942). desert savanna (Shantz and Zon, 1924), and desert shrub grassland (Darrow, 1944). Dick-Peddie (1993, ps. 106-107, 131-132) discussed this large ecotone under both Desert Grassland and Chihuhuan Desert as if this transition zone encompassed both of these divisions of different biomes.

Shreve (1917, map plate, 123) described Desert-Grassland Transition as: "A regioin intermediate in character between the Grassland the Succulent Deserts of Texas and Arizona". From descriptions by the above cited authors the boundary seemed to be more in doubt between the Chihuhuan (than Sonoran) Desert and semidesert grassland. Community appeared to be "in the eye of the beholder". [Hint to Gleasonians: this would seem to be an outstanding example of vegetation as evidence for the individual theory of plant associations (Gleason,1917, 1926).] That both Chihuhuan and Sonoran Deserts are affilitated and related to semidesert grassland is incontrovertible, but the nature of the relationship (evolutionary, successional, floristic, etc.) remains unknown and, as with some much about this range vegetation, open to interpretation.

Over the last 100 to 300 years the successional-community development relationship has been "one way" with the multi-faceted forces of desertification causein conversion of climax semidesert grasslands into desertscrub. In this on-going range retrogression desert shrubs continue to invade the grassland so that the desertland increases while grassland decreases. This desert encroachment and commensurate loss of semidesert grassland has been especially dramatic along the contiguous edge of the Chihuhuan Desert and the semidesert (Chihuhuan) grassland, a transition zone or desertscrub-semidesert grassland ecotone, where vast acreages of disclimax (= disturbance climax) Chihuhuan Desert have developed over the last one to two centuries.

Most of this degradation of semidesert grassland into desert (= arid shrubland or arid scrub) was induced by actions of European man so that the woody invasions constitute an anthropogenic disclimax desertscrub (ie. a man-made desert). The major adverse human impact responsible for this range deterioation has traditionally and consistently implicated overgrazing (and some overbrowsing of palatable woody range plants) by livestock (Clements, 1920, ps. 145-146, 148; Weaver and Clements, 1938, ps. 525-526, 537; Dick-Peddie, 1993, ps. 107-109; 131-132; Holechek et al. 2004, p. 101; Gibbens et al. 2005, ps.665-666). Other factors that have been implicated range from relatively recent severe or extraordinary drought, suppressation of fires or changed fire regimes, and increasing atmospheric carbon dioxide. It was even postulated that certain dominant grasses like black grama (Bouteloua eriopoda) could only retain their dominance during cooler climates such as minor Ice Ages that ended a century ago ((Gibbens et al., 2005, 666). Consensus (by no means unanimity) of scientific judgment clearly implicates past overgrazing by livestock (especially during early Spanish-Mexican and frontier Anglo-Saxon open range ranching eras) and periodic, prolongued droughts as major causes for range degradation including brush invasions. It is almost a given that these two factor resulted in compunded if not synergistic impacts that favored increases in woody species and decreases in palatable climax grassses.

Among affects of livestock grazing the two most obvious were 1) dissemination and perhaps even scarification of mesquite (Prosopis glandulosa) seed when the flavorable and nutritious legume fruit ("bean pods") were eaten and 2) shifting of competitive and survival advantage from climax grasses to less palatable species (especially shrubs and suffrutescent plants) when grasses were defoliated --often at repeated continued high degrees of use so as to constitute overused-- and woody plants were ungrazed or only lightly browsed. The latter phenomenon was explained well by Burgess in McClaran and Van Devender (1995, ps. 46-49).

Other workers (Humphrey, 1958, ps. 51-62; McClaran and Van Devender (1995, ps. 131-141, 242, 244, 249,251-252) have made much of the interaction of livestock grazing and fire. With heavy utilization of herbaceous species by livestock in early ranching eras there was inadequate fuel for range fires. Fire typically induces comparatively more stress on woody plants having their perennating parts within "easy reach" of fire (phanerophytes and chamaphytes) than on herbaceous species like the grasses with perennating plant parts at or below soil surface (cryptophytes or geophytes and hemicryptophytes). Humphrey (1958, ps. 51-62; 1962, ps. 155-165) considered in depth the role of fire in semidesert grasslands. Humphrey was a "born-again" range-burner and concluded that the semidesert grassland was not, contrary to the Clementsian view, a climatic climax but "a subclimax maintained by fire", and he was unequivocal in his conclusion (Humphrey, 1962, ps. 163-165).

Upon subsequent and more investigations it became obvious that the Humphrey hypothesis was too general. Burgess in McClaran and Van Devender (1995, p. 45) concluded that it was incorrect "... to assume that fire is a universal feature of desert grasslands". The various species of plants in this range type vary greatly in their response to fire. Most importantly, grass species in semidesert grasslands vary in their response to fire (Burgess in McClaran and Van Devender, 1995, p. 45; McPherson in McClaran and Van Devender, 1995, ps. 134-141 passim). For example, tobosagrass (Hilaria mutica) tolerates fire and recovers quickly from it whereas black grama recovers slowly following burning (Burgess[citing several studies] in McClaran and Van Devender, 1995, ps. 45-46). These are the two dominant plants on many semidesert grasslands in climax condition; tobosa dominating clay swales and black grama the dominant on sandy upland ranges. A fire burning across one large range on one ranch could thus have drastically different effects on depending on range site (say, swale vs. sandy loam).

The grazing X fire interaction on semidesert grassland has an interesting "twist" when certain exotic grasses that were introduced as complementary pasture or as a species for revegetation can invade semidesert grasslands. Lehmann lovegrass (Eragrostis lehmannii), which was discussed herein under the Introduced chapter of Grasslands, has proved invaluable for revegetation of abandoned cropland and drastically altered lands such as mine spoils, soil and water conservtion, and generally as a naturalized range grass. It also has the ability to become a serious weed when under certain contitions it invades native semidesert grasslands, sometimes those in high successional status. Lehmann lovegrass has the ability to dominate semidesert grassland ranges following disturbances like drought and fire because it reseeds readily whereas sexual reproduction is more limited and recovery by asexual reproduction is slower in the native climax grasses. Range fires at high frequencies selectively favor (shift the competitive advantage to) Lehmann lovegrass over natives, especially those with a fairly low tolerance to fire (eg. black grama, the regional dominant grass). Under such situations proper livestock grazing by reducing fuel loads can help maintain native grassland ranges. Elimination or even reduction of domestic grazing could result in deteriorated semidesert grassland ranges by aiding establishment (invasion by) undesired non-native grass species like Lehmann lovegrass (Roundy and Biedenbender in McClaran and Van Devender, 1995, ps. 293-294Roundy and ). This is an example of the necessity of management in range management.

While all contributing causes and interactions for range retrogression will likely never known, and therefore will continue to be debated, one ecological phenomenon that no one knowledgeable of semidesert grassland ranges challenges is past and present encroachment of woody vegetation into grassland. This on-going woody invastion has been to the extent that the Chihuhuan Desert has expanded northward and eastward while the Sonoran Desertscrub has made similar advances including expansion into foothill slopes that were formerly desert grassland ranges. So rapid has been this vegetational change-- again, much, probably most, was human-induced-- that much of the desertification has occurred during the course of a human lifetime.

Worse, it seems probable that this range degradation will continue. The combination of deep taproot and wide-reaching laterals make root systems of many of indigenous desert shrubs (see Bender, 1982, figs. 9-2 and 9-5, ps. 423 and 430; McClaran and Van Devender, 1995, figs. 2.1, p. 34) much more competitive and better able to survive drought than the native grasses.After discussing superiority of shrub root systems, Gibbens et al. (2005, p. 666) applied the state and transition model and disequilibrium concept (Friedel, 1991 and concluded that "... the change from grassland to shrubland has passed a threshold ... and manipulation of grazing pressure will not reverse the process".

To some extent this was textbook knowledge decades ago. In the confusing terminology of the elaborate, geologic age-based Clementsian system Weaver and Clements (1938, ps. 525-526, 537) said much the same thing when they wrote that numerous shrub species had "...so increased by overgrazing as to simulate a scrub climax" and "... a disclimax of annual grasses ... and forbs" with a desert scrub that had "...greatly increased during the historical period ... " and having "...all the appearance of a true climax". Writing 67 years before Gibbens et al. (2005), and using the monoclimax theory, Weaver and Clements (1938) provided a less conclusive discussion, but the very designation of a "disclimax" or the appearance of a climax said the same thing (only in the code of another system to explain vegetation dynamics). With disturbance climax (= disclimax) described as one of several other kinds of climax and with many examples provided (Clements, 1936, ps. 265-266; Weaver and Clements, 1938, p. 86) is was clear that disturbance-induced shrubland which replaced semidesert grassland was 1) the now potential vegetation and 2) it was probably permanent in human-time scale. Whitfield and Anderson (1938) wrote in the monoclimax-based system of their day and designated a "creosotebush disclimax" and a "mesquite-acacia postclimax" both of which had increased "as a result of disturbance". These early range conservationists were more optimistic regarding recovery of the natural desert grassland vegetation "if properly handled" (Whitfield and Anderson, 1938, p. 176), but they documented conversion of grassland to shrubland having understories of annual weeds. The Clementsian disclimax concept was retained and used in the biotic community classification of Brown et al. (1998 ps. 12, 40) with "desertified" semidesert grassland designated Shrub-Scrub Disclimax Series.

Given the de facto type conversion of semidesert grassland to anthropogenic (disclimax) Chihuhuan Desert it was seen as axiomatic that such NeoChihuhuan Desert range should be treated under the Chihuhuna Desert chapter (Shrublands). That treatment included discussion of the Chihuhuan Desert-semidesert (Chihuhuan) grassland ecotone or transition zone in the context of desertification.

Treatment of semidesert grassland range cover types in this Grasslands chapter was limited to climax grasslands, including grass-shrub savannas. Coverage was consistent with rangeland cover types recognized by the Society for Range Management (Shiflet, 1994). Some of the climax (potential natural vegetation) forms of semideset grassland that have been recognized and described by other accepted authorities (eg. National Park Service) were not included in Shiflet (1994). Some of these omitted semidesert grasslands (cover types) were covered below.

There are several forms (range cover or dominance types) of semidesert grassland and semidesert grass-shrub savannahs. These are zonal (= regional) and/or "patchy" (local in occurrence) with distribution of the range types often determined by physiographic (eg. such topographic features as slope and aspect), elevational (which may coincide and interact with physiography), edaphic (including soil texture, structure, fertility, and frequently above all else, water-holding capacity), and precipitational or thermal conditions.

Choices as to which grassland range types to include under a chapter entitled Semidesert Grassland was based on vegetation historically described as "desert grassland" or "desert plains grassland" (Clements, 1920, ps. 144-149; Weaver and Clements, 1938, ps. 525-526; Oosting, 1956, ps. 333-334; Humphrey, 1958, Shelford, 1963, ps.359-364; Daubenmire, 1978, ps. 202-204; Vankat, 1979, ps. 169-170; Dodd in Gould and Shaw, 1983, ps. 355-356; Sims in Barbour and Billings, 1988, ps. 280-281; and Dick-Peddie, 1993, ps. 106-110, 116-120). This corresponds approximately to the broader vegetation map unit, Grama-Tobosa Shrubsteppe of Kuchler (1964, p.58). Unfortunately (in this author's opinion) The Desert Grassland (McClaran and Van Devender (1995), which should by all appearances be the magnum opus on this grassland, so "mixed-up" semidesert grassland with the geographically and floristically aligned mixed prairie, especially shortgrass plains, and adjoining Chihuhuan and Sonoran Deserts that even those experienced with semidesert grassland ranges were left wondering whether the vegetation described was indeed "desert grassland, mixed shrub savanna, shrub steppe, or semidesert scrub?" (Burgess in McClaran an dVan Devender, 1995, p. 31).

It seemed unavoidable that grassland vegetation included below would be somewhat arbitraty and therefore controversial. For the record, however, range types chosen for inclusion followed the classic or traditional delineations of the previous ecologists listed above. The main defining criterion for these semidesert grassland range types (ie which types to include hereunder) was dominant species. This was seen as the least arbitrary basis (and certainly a difficult-to-debate one) given that cover types are by definition dominance types (Whittaker, 1975, p. 128-129; Whittaker, 1980, ps. 59, 67-79). and that Society for Range Management rangeland cover types (Shiflet, 1994), which "covered" semidesert grassland vegetation, were titled and described as to major or defining (ie. dominant, associate, and indicator) plant species.

Dintinction of semidesert grassland from mixed (mid-grass) and shortgrass prairie vegetation will always be the most difficult separation among closely affilitated range cover types due to commonality of almost all grass species in most range sites of these three broad, general forms of semiarid grasslands. Use of dominance as the key criterion, along with associate and indicator species, permitted separation of semidesert grassland range types from the mixed and shortgrass range types of plains-mesa grassland and, in eastern portions, from those of Chihuhuan Desert (Dick-Peddie, 1993, ps.104-110, 113-120, 131-132, 140-141).Thus, climax grasslands dominated or co-dominated by blue grama (Bouteloua gracilis), sideoats grama (B. curtipendula), curly mesquite (Hilaria belangeri), buffalograss (Buchloe dactyloides), New Mexico feathergrass (Stipa neomexicana), little bluestem (Androgpgon scoparius= Schizachyrium scoparium), or sand dropseed (Sporobolus cryptandrus) --regardless of climatic, edaphic, topographic, or any other habitat features-- were, with but few notable exceptions, automatically included with and treated as "plains-mesa grassland", Great Plains and foothills of various mountain systems of Dick-Peddie (1993, ps. 104-106), and covered under the chapters, Mixed Prairie and/or Shortgrass Plains.

Such separation was consistent with traditional interpretation and treatment of "desert grassland" or "desert plains grassland" by the many workers cited earlier. It differed from that by Whitfield and Beutner (1938) and Whitfield and Anderson (1938) who divided the "desert plains grassland" into the Bouteloua-Hilaria faciation (black grama-tobosa community) and Hilaria-Bouteeloua faciation (curly mesquite-blue grama community). This was consistent with the "pure-milk" of Clementsian terminology, but inconsistent with interpretation of grassland vegetation by Clements (1920) and Weaver and Clements (1938) and their "descedents", all of whom included Great Plains and foothill-mesa range communities dominated by curly mesquite and/or blue grama in the mixed and shortgrass prairies.

Precise placement-- based on dominants (= dominant plant species) and, to lesser extent, indicator species-- means that semidesert grassland could develop "side-by-side" (contiguous) with mixed prairie, Chihuhuan or Sonoran Desert or, even rarely, pinyon pine-juniper woodland. This is consistent with desert range cover types exiting immediately adjacent to riparian forest types, plains-mesa grassland types in contact with pinyon-juniper woodland types or, even, montane forests, and alpine range types conterminous to subalpine forests. Any given range or pasture (grazing unit) could, if sufficiently large or occupying mountainous terrain, include any number of such range cover types and, more yet, range sites.

Proximate location of one range plant community type to relative to other range community (even if these communities are of different biomes like grassland vs. desert vs. forest) means absolutely nothing with regard to designation and description of range cover types, except as this is related to ecotones between or among them. In the case of such transition zones, the range vegetation may be a distinct-- a "hybrid"-- range community requiring designation as its own range cover type. Most of these ecotones as between semidesert grassland and desert or semidesert grassland and woodland are savannas. Herein such composite range plant communities were identified, named, and described as savannas.

All range types and range sites of semidesert grassland include woody (many of them succulent) species so that a shrub and/or tree layer can be interpreted as being present so that the vegetation could be regarded as a savanna. It was explained earlier that some students (eg. Shreve, 1942) viewed this entire zone of range vegetation (= the regional climax) as an ecotone or ecotonal (transition) area or region. In the view of most students, including this one, the arboreous or arborescent species like the soaptrees and Spanish daggers (Yucca spp.) and joint-fir or Mormon tea (Ephedra spp.) are so widely spaced that semidesert grassland, except in true transition zones between adjacent grassland and shrubland, is part of the general grassland biome or formation. Dominant grass species are beyond doubt affiliated and aligned taxonomically and floristically with the Great Plains grassland. Shrubs and, infrequently, small trees are no more prevalent on semidesert grassland than, for comparison, are large trees and shrub thickets on tallgrass prairie.

Major types of semidesert grassland include:

1. Black grama grassland- Black grama has typically been interpreted as the regional dominant of semidesert grassland. The black grama-defined range cover type occurs primarily on sandier soils on upland range sites. This is a predominant landscape or terrain throughout much of the Basin and Range physiographic province so upland black grama range is the regionally dominant form or range type of semidesert grassland.. In this type black grama is the obvious dominant where major associates include tobosagrass, bush muhly (Muhlenbergia porteri), mesa dropseed (Sporobolus flexuosus), and red threeawn (Aristida longiseta) or some member of the purple threeawn (A. purpurea) complex. Miscellaneous species vary with range site and can include more mesic "mid-grass" species such as sideoats grama (Bouteloua curtipendula) and cane bluestem (Andropogon barbinodis= Bothriochloa. barbinodis). Exact species composition and structure of a given range plant community depends on range site and range condition (successional statue). On soils overlying gypsum black grama is frequently co-dominant with alkali scaton (Sporobolus airoides) to form what was described by the Society for Range Management (Shiflet, 1994) as the Black Grama-Alkali Sacaton rangeland cover type (SRM 702).

On upland semidesert grassland ranges in excellent condition black grama develops into nearly "pure stands" and forms a consociation with shrubs of liliaceous or gramineious form. Dick Peddie (1993, ps.116-118, 131-132, 140) concluded that presence of soaptree (Yucca elata) and longleaf jointfir or longleaf Mormon tea (Ephedra trifurca) were diagnostic (= climax) or indicator species whose presence distinguished semidesert grassland. The forb known as desert holly (Perzia nana) and the short fluffgrass (Tridens pulchellus) were also interpreted as possible indicator species of semidesert grassland. Dick-Peddie (1993, ps. 131-132) felt that these species, even as rare remnants, were relict (climax) members that served to distinguish semidesert grassland from anthropogenic Chihuhuan Desert.

Black grama is also a major componment and often dominant species of Great Plains grasslands. The Socity for Range Management (Shiflet, 1994) described Black Grama-Sideoats Grama and Blue Grama-Sideaots Grama-Black Grama rangeland cover types (SRM 703 and SRM 707), but these are not range types of the semidesert grassland. Mere presence of black grama does not denote semidesert grassland.

2. Tobosagrass flats- Tobosa makes up the second most widespread and common rangeland cover type of semidesert grassland. It frequently develops as a consociation on swales having clay soils. Such range sites are usually called by such generic names as "tobosa flats", "tobosa swales", or "tobosa clay flats". A characteristic distinctive shrub on tobosa-dominated lowlands is littleleaf sumac (Rhus microphyllum). Tobosagrass is an associate to black grama on sandy uplant range sites. The two species are sometimes co-dominant in ecotone communities where upland and lowland range sites meet.

Black grama and tobosa comprise the two major range types of the of what could be described as the semidesert grassland division of the North American grassland biome or the semidesert grassland subbiome. These two grassland range types were frequently placed together as a larger climax unit such as the Bouteloua-Hilaria faciation (Whitfield and Brutner, 1938; Whitfield and Anderson, 1938) in language of the Clementsian model (Clements, 1936). Warnock in Shiflet (1994, p. 92) remarked that previous workers had recognized a Bouteloua-Hilaria mutica association.

The Society for Range Management (Shiflet, 1994) lumped the distinct, and usually spatially separated, black grama and tobosa range types into the Grama-Tobosa- Shrub rangeland cover type (SRM 505), but recognized three variants thereof: pure black grama grasslands, tobosa grasslands, and blue grama-black grama grasslands. Apparently the latter develop at upper elevations adjacent to pinyon pine-juniper woodlands or as a transition form to mountain scrub (Gibbens in Shiflet, 1994, p. 65). In the current publication blue grama-black grama (or any other widely distributed blue grama-dominated grassland) was interpreted as part of the plains-mesa grassland (Dick-Peddie, 1993) and treated thereunder.

Historical "quickie": In review of cover types (Range Types under Literature Review herein) it was noted that range cover types are dominance types that basically originated with the vegetation unit, association, especially as it was interpreted, modified, and applied by the Clementsian school. Numerous of the recognized rangeland and forest cover types are synonymous with Clementsian associations when faciation was ignored. Such is the case with Bouteloua-Hilaria mutuca association and Grama-Tobosa-Shrub rangeland cover type (SRM 505). As originally envisioned and strictly arranged, the Bouteloua-Hilaria faciation (Whitfield and Brutner, 1938; Whitfield and Anderson, 1938) was a subunit of the Aristida-Bouteloua association (Clements, 1920, 144-149) which described the entire Desert Plains Grassland as an association (Weaver and Clements, 1938, p. 525). In everyday practice Clementsians did not always adhere strictly to such formal organization. The faciation tended to fall out except under formal writing and the association--always the real working unit of vegetation, became closely aligned with dominance types.

3 Sacaton and vine mesquite (Panicum obtusum) flats- Flat, mesic (and often saline, sodic, and/or gypyseous) bottomlands, including playas, in the Basin and Range sometimes create unique semidesert grasslands. These constitue range sites (eg.Salty Bottomland, Salty, Gyp in the Trans-Pecos Region), but the collective of such range sites has been recognized as a rangeland cover type. In fact, it appears that there are three appropriate Society for Range Management (Shiflet, 1994) rangeland cover types: Alkali Sacaton-Tobosagrass (SRM 701), Black Grama-Alkali Sacaton (SRM 702) noted above, and Vine Mesquite-Alkali Sacaton (SRM 725). Both of these were listed under Southern Great Plains Cover Types, but that should not be taken too literally as to geography or seen as a precluding technicality. These three plus the Grama-Muhly-Threeawn type (SRM 713) and a semidesert variant of Grama-Bluestem (SRM 714) should have had been listing under Southwestern Cover Types as well as Southern Great Plains Types (Shiflet, 1994 p. v). All three of the "flats" types plus SRM 713 and SRM 714 occur outside the Great Plains and into the Basin and Range physiographic province and its semidesert grassland ranges. The three alkali sacaton types sometimes exists as consociations of vine mesquite or alkali sacaton and as alkali sacaton- Wright's sacaton (Sporobolus wrightii) grasslands. Wright's saqcaton may be locally dominant. On semidesert grassland ranges in Basin and Range physiography the Alkali Sacaton- Tobosagrass cover type is typically a transition between the more saline sacaton flats and the heavier clay tobosa flats. Distinction between SRM 701 and the more "pure" form of tobosagrass swales, the tobosa variant of SRM 505, is often difficult and an arbitrary matter of interpretation.

4. Mixed grass semidesert grassland prairie- This range cover type is probably the third most widespread form of the semidesert grassland following black grama uplands and tobosa swales. It was described by the Society for Range Management (Shiflet, 1994) as the Grama-Muhly-Threeawn (SRM 713). In describing this rangeland cover type Warnock in Shiflet (1994, p. 92) emphasized that while this grassland vegetation had been included as part of the Bouteloua-Hilaria mutica association it had been recognized separately as the Bouteloua-Muhlenbergia-Aristida association even before 1900. Dominant and associate species (and genera) vary widely because this range type developes on habitats varying from bottomlands to ridges with soils ranging from clay loams to gravel. One of the diagnostic characteristics of this type is the absence of tobosagrass except where this vegetation joins tobosa flats. This semidesert grassland range type is unique in that blue grama can dominate on deeper soils in valleys. Panicoid grasses are sometimes represented by cane bluestem and silver bluestem (Andropogon saccharoides= Bothriochloa saccharum).

Species composition, structure, and physiogonomy of this type is such that it "forms a prairie" (Warnock in Shiflet, 1994, p. 92) which is quite distinctive (hence inclusion of "prairie" in the above title by the current author). Likewise, number of grass species is greater in this vegetation than in any other major form or expression of the semidesert grassland (thus the descriptive "mixed grass" in the above title).

5. Sand dunes savanna- This is one of the more restricted forms of semidesert grassland. It includes the Sand Hills range site, but even when stabilized the topography is hummocky sand (duneland) and not of the height or diameter of sand hills per se. This range type is very similar to the "shinnery sands" and has a floristic affinity with it through such species as sand sagebrush (Artemisia filifolia) that contributes to the shrub layer and savanna physiogonomy and grasses like various dropseeds (Sporobolus spp.), little bluestem and, very rarely, giant sandreed (Calamovilfa gigantea). Other shrubs that contribute to the savanna structure and physiogonomy include the ubiquitous honey mesquite (Prosopis glandulosa), longleaf Mormon tea, and Yucca spp.

This sand dune grass-shrub form of semidesert grassland obviously has affinities with the Shinnery Sands ()YOn the example of this climax sand dune range vegetation found and photographed by the author broom was an associate shrub to sand sagebrush. This is a mixed shrub savanna (a grassland with several shrub species), but sand sagebrush was dominant.

This almost striking range cover type was overlooked in the Society for Range Management Rangeland Cover Types (Shiflet, 1994). A Giant Dropseed-Sand Sagebrush or, maybe, just Dropseed-Shrub should be added and described in any future edition of Rangeland Cover Types under Southwestern Cover Types (Shiflet, 1994). There are undoubtedly several variants of this range type, probably with more consistency in dominant grass (Sporobolus) species than in shrubs.

Introduction to 6. 7. and 8: Arrayed in small-scale zonal range communities from basins and plains up hills and mountain foothills there are often three distinctive semidesert grassland types. The lowermost of these is the Chihuhuan Desert-semidesert (Chihuhuan) grassland ecotone, a transition zone between the creosotebush-dominated desert plains and the chino grama-dominated foothill grassland immediately above the ecotone. The chino grama grassland is a counterpart range cover type to the regionally dominant black grama grassland, the major form or type of semidesert grassland, that forms a "patchwork" of vegetation with the Chihuhuan Desertrscrub. The third and highest elevation semidesert grassland type in this zonal "threesome" is the sotol-lechuguilla-grama grassland. Chino grama is generally the dominant (or, at least, most abundant) grass, but sideoats, blue, black, and hairy grama are also common as are the threeawns and cane bluestem. The sotol-lechuguilla-grama grassland is commonly adjacent to and elevationally just below the oak-pinyon-juniper woodland. Chihuhuan Desert-semidesert (Chihuhuan) desert grassland ecotone and sotol-lechuguilla-grama grassland can both be interpreted as savannas (shrub steppes is an apt description for both), but these are nonetheless grasslands.

6. Soto (Dasylirion spp, mostlty D. leiophyllum)-lechuguilla (Agave lechuguilla)-grama grassland- This is the chino (grama)-lechuguilla association of Warnock (1974, p. 47) and the sotol-grassland formation or, simply, sotol grasslands of Wauer (1980, ps. 22-23, 32). Warnock (1974, p. 47) explained that in certain areas this is a black grama-lechuguilla association. Wauer (1980, ps. 22-23, 32) omitted lechuguilla from his vegetation name, but he limited his description and title to the one (admittely a large one) locality of Big Bend National Park. Wauer (1980, p. 28) discussed lechuguilla and described it as being "found almost everywhere on the desert and ranges far up the mountins ..." and "... into the lowlands". It seemed as if Wauer (1980) omitted lechuguilla from his sotol-grasslands because the locally cosmopolitan presence of lechuguilla negated its ecological relevance in designation of a plant community. But since when has that stopped such inclusion when a dominant? If it did, there would be no plant formations, associations, or cover types with "creosotebush" or "mesquite" in their titles!

MacMahon (in Barbour and Billings, 1988, ps. 251-152) considered the D. leiphyllum and Yucca spp. "community type" to be "a transition to grasslands". MacMahon (in Barbour and Billings, 1988, ps. 249-250) interpreted A. lechuguilla to be associated more with the creosotebush-ocotillo-grass savanna of a lower elevation zonal vegetation.

This author observed numerous range plant communities of the lechuguilla-grama association (Warnock, 1974) and the sotol-grassland formation (Wauer, 1980) and found both succulent shrub species present--and typically as co-dominant shrubs-- of the vast majority of these communities. In other of these grass-shrub savanna-like grasslands only one or the other of Dasylirion or Agave species was dominant. Therefore, it appeared logical to use the descriptive designation of sotol-lechuguilla (or vice versa)-grama grassland. Once again the problem of dealing with description and interpretation of grass-shrub vegetation was encountered. Was this a savannah or a grassland range community? Are savanna and grassland mutually exclusive (like desert and grassland) in case of this particular vegetation? How about viewing and designating this range community as a succulent shrub steppe? Or maybe a liliaceous shrub steppe? Even the adjective "liliaceous" is controversial because some plant taxonomists "lump" Agave, Dasylirion, and Yucca in the agave subfamily of the lily family (Liliaceae) while other authors (Powell, 1988) "split" them out into the agave family (Agavaceae).

Vegetation classification can be as controversial as plant taxonomy withthe dilema of "lumping vs. splitting" range or forest communities, but this hillside to foothill grassland community is strikingly different from the black grama and tobosagrass communities of basins and upland plains. The distinctive presence of (shared dominance by) Dasylirion and Agave species at higher elevations and on sloping (often steeply) topography versus widely spaced dispersion of the shrubs Yucca elata and Ephedra trifurca on grama-tobosa grasslands on lower plains and valleys clearly separated the sotol-lechuguilla-grama grasslands from other semidesert grasslands, including the next three semidesert grassland cover types that are also found on hillsides and mountain slopes.

Sotol-lechuguilla-grama grasslands are the highest elevtion semidesert grassland types in the zonal gradations extending from Chihuhuan Desertscrub-semidesert grassland ecotone through chino grama foothill grasslands and ending at lower edges of the oak-pinyon pine-juniper woodland. Mixture and local dominance of grass species, usually one or more of the gramas, changes with elevation as well as with slope and aspect, soil, and overstorey plant species. Sacchuista or bear-grass (Nolina spp.; primarily N. texana and N. erumpens) is also frequently common in this range type, but its presnece in greater amounts (at larger coverage) has been interpreted as indicating some degree of range retrogression due to disturbances like overgrazing and drought.

Somehow the obvious (and very conspicuous) sotol-lechuguilla-grama range cover type that had already been described and named by a trained botanist (Warnock, 1974, p. 47) sin a National Park Service publication (Wauer,1980) was left out of the Society for Range Management(Shiflet, 1994) Rangeland Cover Types . This was an unfortunate and rather glaring omission. Sotol-Lechuguilla-Grama needs to be given formal recognition if there is a second edition.

7. Chino grama foothill grassland- One of the most intriguing of the semidesert grassland range types is one that for whatever reason(s) has been so infrequently described (a local "best-kept secret") and yet is so important in specific rangeland localities. This is apparently due to the fact that the "star of the cast" is an "unsung hero" except at an area scale. Black grama and tobosa are the widely known and highly celebrated featured species of the semidesert grassland. Understandably. They are regional or widespread dominant species. They even have their own entitled rangeland cover types and, as explained above, Clementsian faciation. These two have name recognition and get all the "big billing" along with (though to somewhat lessser degree) the most widely distributed associate species such as sideoats and blue grama and, to the west, Rothrock grama (Bouteloua rothrockii). Even the little runt fluffgrass comes in for its share of the limelight, even if taxonomists change its genus as frequently as their underwear.

But chino grama or, even less revealing, chinograss? Who but the locals ever heard of this one? Yes, the unsung Bouteloua ramosa got due justice in an outstanding regional grass manual (Powell, 2000, ps. 15, 217-217), but this species was often confused with the less valuable, more restricted gyp grama (B. brevisita) including in both of the agrostology encyclopedias of Texas (Silveus, 1933, p. 419; Gould, 1975, p. 351)! For the record, the ever-accurate classic Botany of Western Texas (Coulter, 1891-1894, p. 530) as well as the invincable Warnock (1974, ps 26-27) and meticulous Powell (2000, ps. 216-219) correctly distinguished between these two species based on morphological features and locations. Gyp grama is limited to-- surprise-- gypyseous soils. Those who would know plants must also know the land.

Chino grama frequently forms consociations on rocky foothill ranges similar to those of black grama on upland plains and tobosagrass on swales, but more commonly chino grama is a dominant of mixtures of grasses such as cane bluestem, other gramas, and threeawn species. Chino grama is often co-dominant with black grama to which it seems similar in soil water requirements.Rangemen identifying individual grasss plants on such ranges have to pay careful attention because these two species superficially resenble each other.

Chino grama-dominated grasslands can be almost completely herbaceous or include scattered individuals of woody species including creosotebush, cacti (primarily Opuntia species), Acacia species, ocotillo (Flouquireia splendens), mariola (Parthenium incanum), ceniza (Leucophyllum frutescens), and Big Bend silverleaf (L. minus). In the chino grama grasslands (in contrast to the desertscrub-semidesert grassland transition zone described below) these shrubs are of such exceedingly low density and cover that there is no savannah appearance whatsoever.

Maxwell (1968, ps. 108-109) explained the value of "chino grass", but did not offer a scientific name. Neither did the National Park Service (1983, p. 81) which described it as "the dominant grass of the Big Bend foothills". The restricted range of this type and its defining species were assuredly one reason for lack of appreciation of both.

This slight in recognition included the Society for Range Management publication, Rangeland Cover Types (Shiflet, 1994). If there is a second edition of this reference chino grama-dominated grasslands whould be included. The published cover type could encompass the Chihuhuan Desertscrub-semidesert grassland ecotone described immediately below with both types designated as variants was was done in Shiflet (1994) for a generic black grama-tobosa cover type.

8. Chihuhuan Desert-semidesert (Chihuhuan) grassland ecotone- It was explained above in this introduction that Shreve (1917, 1942), Clements (1920), and Weaver and Clements (1938) regarded much, if not all, of what is today entitled the Chihuhuan Desert (following naming by Shelford, 1942) and much of the semidesert grassland as a transition zone between the Chihuhuan Desert and this grassland.on the eastern edge and between the Sonoran Desert and the grassland on the western boundary of this immense ecotone. The actual nature (climax composition, structure, function, etc.) of this range vegetation remains unclear. The decades-old conundrum is complicated by the mosaic of vegetation characteristic of Basin and Range terrain and by affinity of semidesert grassland to Great Plains grasslands and the occurrence of semidesert grassland to the generally more mesic north and east of the Chihuuhuan Desert.

At local scale there are numerous ecotones (vegetational transition areas) formed when communities of creosotebush (Larrea tridentata)-dominated Chihuhuan Desert on the plains of basins unite with grama-dominated grasslands on bajadas and foothills. Ecotones so formed are relatively narrow corridor-like in shape. These are unique range plant communities that are a composite of lower, generally xeric shrubland and higher, less xeric grassland. Grasses are primarily of cespitose (bunchgrass) form and cresotebush, ocotillo (Flouquieria splendens), and other shrubs grow as individual plants so the general appearance of the plant community is of a savanna. With nearly universal dominance by creosotebush, the regional dominant of Chihuhuan and Sonoran Deserts generally, the savannah could be called a creosotebush shrub steppe.

This ecotonal vegetation is clearly more of a grassland in appearance and, almost surely, in productivity as well given density, cover, and growth of grasses that make up the herbaceous layer.

As was the case for chino grama foothill grassland (the grassland portion of this savanna) and sotol-lechuguilla-grassland this range cover type was not included in Rangeland Cover Types produced by the Society for Range Management (Shifllet, 1994). This oversight should be corrected when and if there is a second edition by addition of the this range type. At least it could be "lumped" in with chino grama foothill grassland (as was done with black grama and tobosagrass) and fall under Southwestern Cover Types.

9. Mixed grass (grama-bluestem-cottontop) hillside grassland- The author used this designation for an array of varied grassland communities, all of which have a "common core" of species, on hillsides and lower mountain slopes. Grassland vegetation appeared to be more diverse in species composition and structure on drier west and south slopes and hillsides.This range cover type encompassed several Soil Conservation Service range sites described within the SCS Desert Grassland vegetation zone including Igneous Hill and Mountain, Limestone Hill and MountainSandstone Hill and Mountain, Foothill Slope, Draw, and Gravelly Hill.

Overall this is the most mesic expression of semidesert grassland. It is one of the most diverse in number of grass species, but extremely consistent in which of these species are present. Sideoats grama and cane bluestem are dominant species on all of these range sites. Arizona cottontop (Trichachne californica), plains bristlegrass (Setaria leucopila), and tanglehead (Heteropogon contortus) appear in almost all of these published range site descriptions. Green sprangletop (Leptochloa dubia) is absent as an associate species only on the drier, more gravelly soils. Black grama is, of course, common to all. Chino grama is essentially absent or appears in only trace amounts on all of these range sites.

This combination of Gramineae species and the absence of chino grama is a unifying basis to a unique grassland community on hill- and low mountain-side habitats.

This is the Bouteloua-Andropogon-Trichachne post-climax community (sideoats grama-feather grass-Arizona cotton grass type) of (Whitfield and Brutner, 1938, ps. 32, 35-36).

Woody species are also diagnostic, but somewhat less consistent in occurrence than grass species. Shrubs are always present even in the climax state. These woody species are also diagnostic, but are more varied among range sites and as to slope and aspect. Mariloa (Parthenium incanum), range ratany (Krameria parvifolia), and Acacia species are key indicator shrubs. The more common cacti are the cholla and tasillo forms of Opuntia (Cylindropuntia subg.) Skeletonleaf goldeneye (Viguiera stenoloba) is also a frequent dominant shrub.

While the sideoats grama-cane bluestem-Arizona cottontop range type develops on hill- and mountain sides it is not typically restricted to elevational zones as are the preceding three semidesert grassland types. Rather the grama-bluestem-cottontop dominated grassland usually develops along entire slopes (or nearly so). It may be that local physiographic features (eg. directional aspect or configurational arrangement of terrain) are determinative in development of this most mesic expression of semidesert grassland.

There is no Society for Range Management (Shiflet, 1994) rangeland cover type for this range type. If there is a second edition of Rangeland Cover Types this vegetation should be described as Grama-Bluestem-Cottontop Shrub under Southwestern Cover Types along with those existing in Shiflet (1994). This range community-- complete with emphasis on grazing use-- was named and described in the literature decades ago (Whitfield and Brutner, 1938, ps. 32, 35-36).

All standard treatments of semidesert grasslands including The Desert Grassland (McClaran and Van Devender (1995), the what-should-have-been-but-was-not comprehensive reference on the subject, limited treatment of the driest grasslands in North America to those of the Chihuhuan and Sonoran Regions of the Basin and Range physiographic province. Semidesert grasslands of the Great Basin Region of the Basin and Range province have either been ignored (eg. McClaran and Van Devender, 1995) or gotten short-shrift treatment. Great Basin semidesert grasslands are much more restricted (less extensive in geographic distribution) and are of relatively less economic importance when compared to the immense extent of Great Basin scrublands, pinyon pine-juniper woodlands, and mountain forests. Regardless, semidesert grasslands of the Great Basin do exist, and these natural grasslands are distinctive from those of the general Chihuhuan-Sonoran zone as well as being economically important (greater grazzing capacity on a unit of land basis relative to most Greatr Basin Desert range). Comprehensive treatment of North American semidesert grasslands deems inclusion of those varied dry grassland cover types.

Coverage of semidesert grassland range herein was divided as to those of 1) Chihuhuan and Sonoran and 2) Great Basin vegetation.

Semidesert grasslands of the more southern portions of the Basin and Range physiographic province developed both as 1) zonal (likely ecotonal) range vegetation between Great Plains grasslands (mixed and shortgrass prairie range types) and the Chihuhuan Desert as well as mountain grasslands and the Sonoran Desert and 2) vegetational mosaics of grassland and among Chihuhuan and Sonoran shrubland (desert). Thus spatial distribution and pattern of Chihuhuan and Sonoran semidesert grassland range varies from broad bands or belts (zonal vegetation occupying relatively large terrestrial space) to "patches" or a "patch-work" dispersed within the overall surrounding desert scrub. The latter spatial arrangement in context of Landscape Ecology consist of semidesert grasland patchs in a desert matrix. It was explained previously that the spatial extent to which Chihuhuan and Desert Desert vegetation has invaded (encroached upon or replaced) pre-Columbian climax semidesert grassland is unknown. Scientific or empirical evidence (research), which has been consistent with anecdotal evidence (eg. land surveyer notes, rancher accounts, explorer and traveler diaries), has conclusively shown that there has been (in some instances continues to be) extensive desertfication and loss of grasslands to scrubland (desert).

There is considerable variation within the Chihuhuan-Sonoran semidesert grassland complex. Unfortunately the extent and kind of this variation, especially as to various grassland communities as reflected in range cover (= dominance) types, was not presented in recent ecological works on the subject as for example in hoped-for encyclopedic coverage in The Desert Grassland (McClaran and Van Devender, 1995).

Technical note: the next five sets of paired photographs compared JPEGs made from 35mm Kodachrome slides by an Epson Perfection V700 Photo color scanner (photographs on left) and a Hewlett-Packard Scan Jet 7400C-Scan Jet XPA color scanning unit (photographs on right). Color was most accurately reproduced by the Epson Perfection V700, but this Epson unit produced (added) an unnatural "sparkle" or "frosted" appearance to the foliage which the Hewlet-Packard Scan Jet 7400C did not add. Conversely, the Hewlet-Packard unit added a bluish-tone to foliage that the Epson did not.

Conclusion: neither the Epson Perfection (perfection it most certainly is not) with its additional "shotgun-shot sparkle" nor the Hewlet-Packard Scan Jet with its added scanned bluish tinge were completely satisfactory for the task of scanning with accurte color. Likewise, both of these cheaply made, imprecise pieces of equipment automatically cropped edges of slides. The only reason that the author used these pieces of "gyp-aggo" apparatus was that they are about all that is available for scanning decades-worth of slides taken prior to introduction of digital imaging (which incidentally has myriad problems of its own as well).

Science (education, in general) alsways involves trades-offs and compromises.

8. Tobosagrass clay flat on an ecotone- A small tobosa swale situated among larger areas of ridges in a transition zone (an "overlap" area) among western Rio Grande Plains, eastern edge of Trans-Pecos Basin and Range, and marginal contact with semiarid Edwards Plateau. There has been inconsistent mapping of vegetational (= land resource) areas of Texas (Gould, 1962, ps. 1, 11; Correll and Johnston, 1979, map 1, ps. 9-10, 12), potential natural vegetation (Kuchler, 1964, map in Garrison et al., 1977), and Texas level III ecoregions (Griffith et al., 2004) with respect to western boundary of Edwards Plateau and eastern boundary of Chihuhuan Desert. This tobosagrass flat and surrounding stand of honey mesquite, huisache (Acacia farnesiana), and Texas persimmon (Diospyros texana) were on range that was included in the western Edwards Plateau by (Gould, 1962, ps. 1, 11; Correll and Johnston, 1979, map 1, ps. 9-10, 12), but in the eastern contacts of Chihuhuan Desert according to (Kuchler, 1964, map in Garrison et al., 1977) and (Griffith et al., 2004). There was not even one single plant of creosotebush (Larrea tridentata) in the range vegetation shown here so designation of this range plant community as being in the Chihuhuan Desert was not overly plausable.

The swale or clay flat supporting a consociation of tobosagrass did provide some textbook examples of this range plant community (even if it was a small parcel of grassland) that was used for instructive purposes (with qualifications duly noted). For practical illustration of the tobosagrass-dominated semidesert grassland the designations of vegetational units were consistent with those cited above. FRES No. 40 (Desert Grasslands Ecosystem). K-48 (Gramagrass-Tobosagrass Prairie). Not specifically described by SRM cover type; variant of SRM 505 (Grama-Tobosa Shrub). Tobosa Grass-Scrub Series 143.12 of Chihuhuan (Semidesert) Grassland biotic community 143.1 of Brown et al. (1998). Level III ecoregion of Texas was transition between Chihuhuan Deserts and Playas Ecoregion 24a of Chihuhuan Deserts and Semiarid Edwards Bajada Ecoregion 31b of South Texas Plains (Griffith et al., 2004).

Val Verde County, Texas. October, phenological stage of tobosa varied from anthesis through grain-ripe to grain-shatter.

9. "And the green grass grew all around, all around; and the green grass grew all around" (American folk song)- Characteristic sward of tobosagrass on a swale showing the semi-cespitose habit of this rhizomatous perennial eragrostoid grass. Individual plants of tobosagrass typically grow in close proximity to each other so as to form an open sod or turf, especially on more mesic habitats (eg. swales and larger bottomland sites), so that stands of tobosa have a bunchgrass-like appearance. This is in contrast to the "kissin' cousin", curly mesquite (H. belangeri), which is stoloniferous and forms a more nearly closed turf or complete sod.

This small stand or consociation had been lightly grazed by cattle such that individual plants and the open sod or turf were more obvious. These photographs were "picture perfect" examples of proper degree of use, which in the semiaridity or aridity of this range environment was on the light side of moderate. This range professor reminded his student readers that moderate grazing (= defoliation) is not automatically synonymous with proper grazing any more than is light or heavy degree of grazing defoliation proper or improper. Proper use factors are not only species-specific, but even site-specific for species-specific standards or guidelines. A key diagnostic guide to proper degree of use (the first and usually the most important of the Four Cardinal Principles of Range Management) is presence of individual shoots that were ungrazed and, usually, matured to flowering and grain production. Reproduction of perennial grasses like tobosa is usually asexual (by tillers and rhizomes in H. mutica) so seed production is not necessary nor usually important for regeneration of this valuable climax species (decreaser on swale sites). Rather, presence of sexually reproductive shoots is a very good sign to graziers that the tobosagrass range was properly managed as to degree of use. Plus, this degree of grazing defoliation with subsequent seed-set best provided for the option of establishment of seedlings that can fill in interspaces among existing tufts of older plants. Seedling establishment permitted presence of new genotypes (the opportunity for on-going genetic adaptation) of this remarkable and valuable semidesert species.

Val Verde County, Texas. October, phenological stage of tobosa varied from anthesis through grain-ripe to grain-shatter. FRES No. 40 (Desert Grasslands Ecosystem). K-48 (Gramagrass-Tobosagrass Prairie). Not specifically described by SRM cover type; variant of SRM 505 (Grama-Tobosa Shrub) variant. Tobosa Grass-Scrub Series 143.12 of Chihuhuan (Semidesert) Grassland biotic community 143.1 of Brown et al. (1998). Level III ecoregion of Texas was transition between Chihuhuan Deserts and Playas Ecoregion 24a of Chihuhuan Deserts and Semiarid Edwards Bajada Ecoregion 31b of South Texas Plains (Griffith et al., 2004).

10. Beautiful, and outstanding example to boot- Individual plant of tobosagrass that was grazed during current growing season to proper degree of use as determined by general qualitative appearance, but not on specific quantitative guidelines (assuming that such even exist). Key features of proper grazing included: 1) adequate photosynthetic tissue (both leaves and culms) remaining following grazing, 2) some shoots that were permitted to mature to sexually reproductive (= flowering) stage, 3) irregular, domed shape of grazed plant rather than a heavily cropped one having a flat, highly hedged appearance, and 4) overall degree of utilization (composite of defoliation for whole plant) that was somewhat on light (conservative) side of the total or maximum utilization that would enable the plant to survive. In regards feature #4, management of this tobosagrass range was such that some "surplus" feed was left for an emergency, risk, etc. that might befall plant, animal, or ranching firm. The conservative or cautious grazing use would come closer to permitting higher levels of herbage production under environment stress, especially drought. The rangeman responsible for husbandry of this range had hedged his bets and left "a little extra" grass just in case "things get tight". Such wise stewartship is good ranch as well as good range management.

The rest of the explanation for the bountiful flowering of tobosa on this range as shown in these photographs was an extremely wet late summer and early autumn. Numerous, well-spaced showers and cool temperatures resulted in near ideal soil moisture conditions. Students take note: even abundant rainfall and other features of a fabulous plant-growing environment would not have permitted the profuse flowering and lush foliage of this tobosagrass had it been grubbed to the ground by overuse. Vice versa, even if growing conditions had been adverse for tobosa on this swale there would still have been some sexual reproduction in the stand given the wise use management--especially proper degree of utilization--it received.

Val Verde County, Texas. October, phenology ranged from anthesis to grain-shatter stages.

11. Nothing much purtier to a desert grassman- A large speciment of tobosagrass at peak bloom on a swale in Rio Grande Plains savanna-Chihuhuan Desert ecotone. This plant had been properly grazed (defoliation of outer edge of grass still evident in foreground) and, with blessings of late-growing season rain showers, a high number of shoots had developed to sexual maturity.

Val Verde County, Texas. October. On this one plant phenological stage varied from anthesis through grain-ripe to grain-shatter (with more soil water from each new shower ever more shoots advanced to flowering stage).

14. Ring muhly or ringgrass (Muhlenbergia torreyi)- This cespitose grass grows on plains and mesa grasslands of the Great Plains through the semidesert plains grasslands of the Trans Pecos Basin and Range province. As indicated by it's common name ring muhly exhibits a unique growth habit. This bunchgrass produces secondary shoots only in the form of tillers such that asexual reproduction takes place around edges of the existing mother plant. The result is an ever-expanding circular pattern of growth. When the oldest tillers (mother and first daughter shoots) that were in the center of the plant were shed (ollowing their senescence and death) the result was a "ring" of tillers. Ring muhly shoots often have characteristically curled leaves.

Ring muhly is a common but not a major forage-producing species. It is, however, often readily taken by livestock. Ringgrass is undoubtedly valuable for erosion control and watershed protection. It's growth pattern was an apt illustration of the concept of clonal organism.

18A. Black grama semidesert grassland in "mint condition"- With proper management-- in this case, limited to proper grazing management--even after a decade or more of severe drouth this vegetation was still as virgin as it gets. Species composition, community structure, and ecosystem function was the same as if human eyes had never rested on it. This was almost a single grass species-community with sparsely scattered evergreen shrubs. There were also scattered individual plants of sideoats grama (like the larger cexpitose grass plant in center foreground) and cane bluestem. Tobosagrass grew on the isolated microsites of small depressions in the land surface.

Students were again reminded that black grama semidesert grassland is by far the most common kind or form of what has traditionally been called "desert grassland. Black grama is the regional or zonal dominant grass of the greater Chihuhuan-Sonoran Deserts Basin and Range Region.

Presidio County, Texas. June, early estival aspect prior to onset of summer rains. FRES No. 40 (Desert Grasslands Ecosystem). K-48 (Grama-Tobosa Prairie). SRM 505 (Grama-Tobosa Shrub). Grama Grass-Scrub Series 143.11 of Chihuhuan (Semidesert) Grassland biotic community of Brown et al. (1998, p. 40). Chihuhuan Deserts- Chihuhuan Desert Grasslands Ecoregion, 24 b (Griffith et al., 2004).

18B. Species diversity on "perfect" black grama semidesert grassland- Species diversity on the black grama form or expression of semidesert grassland is typically very low. It is a classic Clementsian consociation across vast areas. Other grass (and a few forb) species are present, but in small to almost trace amounts. Two of the more frequent of associate grass species include sideoats grama--State Grass of Texas--, which was conspicouosly dominant locally in the foreground, and cane bluestem, the conspicuous, somewhat decumbent grass with larger shoots in center midground. Both sideoats and cane bluestem are mid-grasss species (in contrast to shortgrass species like black, hairy, and blue grama). Given presence of both mid- and shortgrasses this range vegetation could be described precisely as "mixed prairie", but that designation as a proper title has usually been reserved for Great Plains grassland communities that also often include tallgrass species. Kuchler (1964, in Garrison et al., 1977) entitled this range type "prairie". It fits.

"Savannah" would be another arbitrary distinction that could be used to describe this range vegetation, specifically to characterize for instructive purposes presence of woody plants (the larger of which have a tree-like form). In this author's judgment, semidesert grassland is-- as historically described by the word "grassland"-- a true grassland (ie. vegetation and a range cover type in the grassland biome). Dick-Peddie (1993, p. 107) cited other workers who described this vegetation as a "shrub-steppe" . It is important for those new to this range vegetation to understand that presence of the soaptree yucca so prominent in this photograph and longleaf joint-fir or longleaf Mormon tea (Ephedra trifurca) are indicator and defining species to climax black grama semidesert grassland. Dick- Peddie (1993, ps.116 ) carefully explained this key floristic feature. That key botanical phenomenon was underscored here and now by the current author as well.

Presidio, County, Texas, early estival aspect prior to convectional summer rains. FRES 40 (Desert Grasslands Ecosystem). K-48 (Grama-Tobosa Prairie). SRM 505 (Grama-Tobosa Shrub). Grama Grass-Scrub Series 143.11 of Chihuhuan (Semidesert) Grassland biotic community of Brown et al. (1998, p. 40). Chihuhuan Deserts- Chihuhuan Desert Grasslands Ecoregion, 24b (Griffith et al., 2004).

18C. Black grama-dominated semidesert prairie- Soaptree yucca, State Flower of New Mexico, is an indicator of climax semidesert grassland of the black grama form (Dick-Peddie, 1993, ps. 131-132) and presents a savanna-like physiogonomy to this wonderfully adapted range vegetation. The semi-open sward seen in this and the next photograph is also a obvious physionomic characteristic of semidesert grassland, especially of the black grama form. As shown in photographs below, black grama, like tobosagrass, spreads (often vigerously and extensively) by stolons. The semiaridity of climate and xeric nature of soils is such that the soil surface is not covered to the extent that it is in more mesic areas and with other stoloniferous species like buffalograss (Buchloe dactyloides) or even the larger rhizomatous bunchgrasses. Buffalograss, blue grama, bush muhly, mesa dropseed, and perennial threeawns are typically present even on virgin state black grama range like this one, but their overall density and cover is miniscule other than at microsite scale. A high percentage of these grass species (eg. bush muhly, mesa dropseed, and threeawns) are strictly cespitose and produce no stolons.

Presidio County, Texas. June, early estival aspect (prior to summer rains necessary for meaningful "green-up"). FRES No. 40 (Desert Grasslands Ecosytem). K-48 (Grama-Tobosa Prairie). SRM 505 (Grama-Tobosa Shrub). Grama Grass-Scrub Series 143.11 of Chihuhuan (Semidesert) Grassland biotic community of Brown et al. (1998, p. 40). Chihuhuan Deserts- Chihuhuan Desert Grasslands Ecoregion, 24b (Griffith et al., 2004).

18D. Overall appearance of black grama-dominated semidesert grassland range- This climax vegetation illustrated physiogonomy and botanical compostion produced by presence of both herbacesou and woody species. In addition to soaptree yucca, longleaf Mormon tea or joint-fir was prsent in this slide. The herbaceous layer was "pure" black grama and the sward had the patchy or spotty appearance produced by this species. (Even though black grama reproduces to large degree via stolons this species still creates a semi-open herbaceous canopy.)

Dick Peddie (1993, ps. 116, 131-132) explained that presence of certain diagonostic (= indicator) species on what was Chihuhuan Desertscrub provided evidence that the vegetation had been semidesert grassland which had most likely been converted to desert by human-induced shifts in vegetation. Indicator species included longleaf Mormon tea (in contrast Torrey Mormon tea [Ephedra torreyana]), desert holly, and fluffgrass (Triens pulchellus). Presence of both soaptree yucca and Mormon tea added another layer to this range plant community and resulted in rang vegetation that had a savanna-like appearance and a shrub-steppe physiogonomy. These evergreen, narrow-leaved shrubs also produced a unique visual aspect to the vegetation.

Presidio County, Texas. June, early estival aspect (prior to onset of summer rainy season). FRES No.40 (Desert Grasslands Ecosystem). K-48 (Grama-Tobosa Prairie). SRM 505 (Grama-Tobosa Shrub). Grama Grass-Scrub Series 143.11 of Chihuhuan (Semidesert) Grassland biotic community of Brown et al. (1998, p. 40). Chihuhuan Deserts- Chihuhuan Desert Grasslands Ecoregion, 24b (Griffith et al., 2004).

19. Sward of black grama (Bouteloua eriopoda) - Outward physical appearance of a sample of a stand (= a colony might be more appropriate) of black grama. On a lightly grazed (by cattle) range that was in Excellent range condition class at the end of first growing season following the Long Drought of the 1990s through early 2000s. Many of the eragrostoid (= chloroid) grass species (those in Gramineae subfamily, Eragrostoideae) have shoots that live for more than one growing season (ie. have aboveground living tissue that persists for two or more years). This is not to be confused with the biennial life cycle (they are perennial species) or with semi-woody (either suffrutescent or becoming woody) plants. Rather, some shoots of some of these grasses under certain growing conditions simply do not die completely and instead maintain some meristematic tissue in culms or buds. Plants of certain grass species enter (or persist) into a condition that might be described as "semi-dormancy" from which they are able to emerge and (or in which they) "green-up" (ie. produce new tissue that is soon capable of photosynthetsis). This is undoubtedly an evolved opportunistic adaptation for survival under sporadic growing conditions such as aridity, periodic heavy precipitation, widely fluctuating temperatures, etc.

Plants in such a state of "suspended perenniality" (to coin a purely descriptively instructive term) not only respond quickly to episodes that are favorable for growth and/or reproduction, they also retain biomass of higher nutritive value (in contrast to the nutritionally leached herbage of plants and species that do not maintan such tissue and instead "die back to the ground" at the end of each growing period. This is a major reason why available herbage of most panicoid grasses like Andropogon, Bothriochloa, Panicum, and Paspalum species is less nutritions and palatable than that of of some eragrostoid grasses like certain species in Bouteloua, Chloris, and Spartina. Biomass of buffalograss also falls in this latter group.

The phenomenon just described can be visually observed by the "looks sorta green" (pale or "dim" green or gray-green) color of shoots of these species (again, at least under certain conditions). This condition and coloration was readily seen in the black grama shoots shown in this slide. Even though summer rains had not commenced at time of photographing these plant there was enough residual soil moisture (from winter and spring precipitation) that lower parts of last year's shoots were still partly green. This plus new growth of this year's young shoots provided the coloration shown here. (This is another reason for using non- or low-color enhancing film or neutral film like Kodachrome.)

Students should also take note of the open appearance of black grama sward. Black grama is not a cespitose (bunchgrass) species. Instead it is stoloniferous and a sod-forming grass. Black grama reproduces asexuallly by tillers as well as stolons but the resultant growth pattern and appearance is of an open or patchy sod with substantial bare soil among dense "clumps" of a perennial grass that spreads most effectively by runners (stolons).

20. Plant of black grama or wooly grama- This "plant" of black grama consist of a larger "clump" (group of tillers) behind three smaller "clumps". The three smaller "bunches" are clones or modules ("daughter plants") produced asexually from the larger "parent" (or "mother") plant, the latter of which was almost undoubtedly a "daughter plant" from its "parent plant". Assuming that the "parent" of the "daughter plants" was itself a clone or module of a preceding parent, each of these vegetatively or asexually produced "plants" (each module) was a ramet of one genet, the original genetic (sexual) individual. This illustrated a modular organism. Only the genet, the plant produced by sexual reproduction (exchange of gametes) was a unique or "one-of-a-kind" organism. How many years, decades, or, even, centuries ago this sexually generated ancestor came into existence is known but to God.

Black grama reproduces by seed (from a caryopsis) very rarely. Prof. Ken A. Valentine spent over forty years involved in research on the Jornada Experimental Range and New Mexico State University College Ranch. Valentine (personal communication) claimed that he had never seen a single black grama seedling.

Presidio County, Texas June.

21. Shoots of black grama- Another common name for Bouteloua eriopoda is wooly grama. This basis for this descriptive name was the abundant pubescence on the internodes (hence, wooly or fuzzy stems). This pubescence and a unique shade of green color on the culms is the distinctive idenfiying feature of this regional dominant and extremely valuable range plant.

Big Bend Ranch State Park, Presidio County, Texas. June.

22. Detail of black or wooly grama shoot- Some sheaths on black grama shoots cover (enclose) almost all of the internode of the culm. While rangemen refer to the wooly "stem" of wooly grama it is actually the wooly leaf. Leaves are frequently absent or missing from portions of black grama shoots thus exposing the culm that is often a light yellow color. This alternating green and yellow color pattern was visible in the preceding slide. Interspersed with the green and yellow pattern of living shoots is the dull gray color of dead-but-still-attached shoots like the one featured in this shot. This shoot was a tiller, a vertical or upright intravaginated shoot.

Presidio County, Texas. June.

23. Black grama runners- In contrast to the tiller in the preceding photograph, a runner (stolon is the scientific term) is a horizonal extravaginated shoot. In this photograph two stolons were laying atop and across a third (though less visible) stolon. All three stolons were living, but as is often the case for shoots of this species they looked "more dead than alive". Three nodes with emerging new clones or modules (appearing as buds) were visible. Plants were just initiating new growth at onset of summer rainy period.

Presidio County, Texas. June.

24. "Throw out the life line"- True to the words of the old hymn by E.S. Ufford black grama has grown the life line for another generation of this mostly clonally reproducing mainstay of the semidesert grassland ranges. Several stolons were visible on the surface of this sandy semidesert soil. A large module or ramet was developing from the node of the foremost runner. This "daughter plant" was shown in more detail in the second slide.

Presidio County, Texas. June.

25. "Putting down roots" or "a hen and two chicks"- Two stolons from an established black grama plant have each produced a "new plant" (a module= clone= ramet). Each "daughter plant" became established (one or maybe two growing seasons prior) and was waiting for rain at beginning of the summer rainfall period. Summer is when most precipitation occurs in the climate of the Trans-Pecos Basin and Range.

New modules were developing on another stolon (to left of established ramets) to be ready for next summer's rain.To a range plant and a rangeman, "There's always next year". The asexual reproduction shown here took place during last phases or just after one of the most severe droughts in weather records. Grazing management must allow for such reproduction: "Take care of the grass and it will take care of you".

Presidio County, Texas. June.

26. Inflorescences of black or wooly grama- The infloresecence type of eragrostoid grasses (those in subfamily, Eragrostoideae) is a raceme. Spikelets are borne on pedicels on the rachis (adjective, pedicellate). That black grama produces grain was shown here; that such grain is less productive of new black grama plants was shown in preceding photographs of asexual reproduction.

Presidio County, Texas. June. (Spikelets had persisted on the rachis throughout course of winter and spring suggesting to any range-wise rangeman that there had been but scant precipitation.)

27. Mesa dropseed (Sporobolus flexuosus)- Local stand of mesa dropseed. This is one of the most common and important species on recovering black grama range (deteriorated range undergoing plant succession). Mesa dropseed appears to respond to disturbance primarily as an increaser (Pieper and Beck,1990), but on many, if not most, range sites in the semidesert grasslandmesa dropseed is an important component of the climax range vegetation (Soil Conservation Service, undated; range site descriptions). Mesa dropseed is especially important on depleted black grama range that was converted into Chihuhuan Desertscrub disclimax. On such "man-made" deserts mesa dropseed may be a major grass species along with the less palatable perennial threeawns and the palatable bush muhly which frequently grows near creosotebush.

Elephant Mountain Wildlife Management Area, Brewster County, Texas. June, early new growth ("green-up") at beginning of summer rainy period.

28. Individual plant of mesa dropseed- Specimen of mesa dropseed with the "sprawling" habit frequently developed in this species. The twisted or rolled leaves are a common characteristic of Sporobolus species. The grass's "next door neighbor" was the invasive, but short-lived suffrutescent composite, broom snakeweed.

New Mexico State University College Ranch, Dona Ana County, New Mexico. June, early new growth ("green-up") at onset of summer rainy period (dead plant material was biomass produced the previous summer).

29. Mesa dropseed (Sporobolu;s flexuosus)- Whole plant--shoots; aboveground portion-- of mesa dropseed (first slide) and details of shoots (second slide) of mesa dropseed, the principal increaser on many degrated former black grama-dominated semidesert grassland ranges. All conspicuous plant parts (all but small green basal and the few green adult leaves) were produced the preceding summer-fall growing season. The bent or flexuous leaves are a feature common to most Sporobolus species, but apparently they were so prominent in this specific species as to serve as basis for the specifid epithet.

The deep sandy soil of this mesa dropseed specimen probably was the edaphic basis of a black grama semidesert grassland ecosystem that was largely replaced by the adjacent but distinct Chihuhuan Desert due to human action. This location was directly on the El Camino Real, the first European road in North America (ran from Santa Fe to Chihuhua City). Grazing by draft animals (horses, mules, asses, oxen) used for trade and travel along this route for nearly 300 years was the most probable cause of human-induced desertification.

On such drastically altered landscapes and ecosystems mesa dropseed is a keystone species, and the key to whatever recovery of vegetation, soil etc. is possible.

New Mexico State University College Ranch, Dona Ana County, New Mexico. June, plant organs were of both the previous growing season and the first part of current growing season.

30. Local stand of sixweeks grama (Bouteloua barbata)- On this spot (microhabitat) several plants of the ephemeral sixweeks grama had germinated and just completed their short life cycle. Sixweeks grama is one of the few annual grass species that is native to semidesert grassland and desertscrub in the Chihuhuan Region. Perenniality is a major plant adaptation to harsh environments like desert, alpine, saline, and rocky ecosystems. Under severe disturbance (eg. anthropogenic causes that highly modify natural habitats) plant species with the annual life cycle pattern become more competitive with those having biennial and perennial life cycles. This is the concept of therophytes. Therophyte refers to one of Raunkiaer's life form groups that has the annual or ephemeral life cycle so that growth and sexual reproduction are completed quickly during short periods when conditions are favorable and that then survive unfavorable periods as seed or fruit.Therophytes are typical of deserts, especially disturbed desert habitats, tilled land, and overgrazed pastures. The seed or fruit is the plant organ most capable of surviving severe stresses like extreme temperatures, dryness, and long periods of dormancy. Therophytes are usually pioneer species on severely disturbed land. This initial repopulation of plants is essential to plant succession to facilitate invasion and establishment of plant species of higher successional order. Facilitation is the contemporary term for the preceding synonymous term of reaction proposed by F.E. Clements.

On almost all overgrazed ranges of about every range type annual plants become more plentiful with continuing retrogression, the retrograde movement of a biotic community toward communities of lower succeessional order (= lower seral stages). With continuing retrogression annual or ephemeral species often become community dominants. The sixweeks grama plants shown here were growing on land that had been a main trail along El Camino Real, the oldest European road in North America, that had been subjected to about every human activity imaginable, not least of which was livestock grazing. Climax dominants like black grama and tobosagrass were displaced by plant species of lower successional order as range retrogression moved climax black grama semidesert grassland to anthropogenic Chihuhuan Desert. The rangeland at this location is at static or maybe upward range trend as human disturbance has been reduced or eliminated. Sixweeks grama was doing "yeoman's service" to improve the range.

Under such successional conditions annual and ephemeral species-- at least native species-- are not weeds. They are essential for healing of the range. Ephemerals and annuals are as ecological invaders because they do invade and colonize disturbed rangeland. The diagnostic successional designation of invader does not always imply noxious plant. Yes, under many conditions invader certainly may mean weed or brush, but when the impacts of invaders are those of beneficial reaction (= facilitation) benefits of invaders exceed their adverse impacts. If such ephemerals are weeds they are virtuous ones.

Ephemeral in this usage refers to annual plant species that have especially short life cycles.Ephemeral grasses are often called "sixweeks grasses". Another "sixweeks grass" that is native to the semidesert grassland range type is sixweeks threeawn (Aristida adscensionis). It is usually less common than sixweeks grama on sandy soils.

New Mexico State University College Ranch, Dona Ana County, New Mexico. June, seed-shattter stage of phenology.

31. Sixweeks grama (Bouteloua barbata)- Example of one of the most common ephemeral grasses on degraded semidesert grassland and Chihuhuan Desert ranges. This specimen had just had its moment in the sun: it recently completed its short life cycle culminating with shedding of grain. It had achieved the ultimate biloogical success: it reproduced and propagated its kind. The amber color of these plants (captured by ever-true color Kodachrome) is typical of most annual grasses.

On an old trail of El Camino Real, New Mexico State University College Ranch. Dona Ana County, New Mexico.June, seed-shatter phenological stage.

32. Bush muhly or hoe grass (Muhlenbergia porteri)- Along with black grama, bush muhly is one of the most palatable--hence most apt to be grazed out--climax grass species of the semidesert grasslands. In western, more xeric areas of the Chihuhuan Region bush muhly is an associate to black grama. Bush muhly is frequently observed to grow in association with and even outcompete creosotebush on the Chihuhuan Desert-semidesert (Chihuhuan) grassland ecotone. The phenomenon (complete with photograph) was described in the Chihuhuan Desert chapter herein. Powell (2000, p. 177) stated that the common name of bush muhly was derived from the association of this palatable species with protection-providing shrubs. That is logical, but the bush-resembling general appearance of this grass would also seem a likely origin of the adjective, "bush". Whitfield and Anderson (1938, p. 174) stated that bush muhly "... forms bush-like masses". Whitfield and Anderson (1938, p. 174), Gould (1951, p. 201-202), and Humphrey (1960. ps. 67-69) remarked that in the origina range vegetation (prior to advent of ranching) bush muhly was extremely abundant, but that this extremely palatable native grass was at time of their writing found almost entirely under protection (presumedly from excessive livestock) of shrubs.Wooton and Standley (1915, p.69) listed the common name of M. porteri as "mesquite grass" (obviously a confusing name) because in southern New Mexico where it was so common "... it is nearly always found growing in the shade of mesquite bushes ...", but even then cattle "...will force their way into the mesquite to reach the grass". The relationship between bush muhly and woody plants was unclear from the descriptive literature.

The common name of "hoe grass" (Whitfield and Anderson, 1938, p. 174; Humphrey, 1960, p. 67) was even more intriguing than origin of the adjective, "bush". Numerous of the semidesert grassland grass species were harvested and sold for hay. Such hay was mostly black grama, chino grama, and tobosa, but it is likely that any grasses associated with these were also used for hay. "Hoe" could have been in reference to hay harvest.

El Paso County, Texas. June.

33. Another example of bush muhly- This specimen was growing in the Trans Pecos Basin and Range province on an ecotone between semidesert grassland and Chihuhuan Desert. The fiercely competitive feature of bush muhly was evident from this plant that was growing around a plant of tarbush (Florensia cernua) which had died as a result of grass-shrub competition. Observant students will note that there were four (or five) live plants of tarbush growing in close proximity around the dead tarbush. These live tarbush plants did not have to compete with bush muhly (= other tarbush plants did not have tarbush growing with them).

Railway Ranch, Upton County, Texas. Mid-October; muhly was entering winter-dormancy. .

34. Shoots of bush muhly- Detail of shoots on the bush muhly plant shown in the preceding slide. The decumbent-and-geniculate combination shape of bush muhly shoots allows ready idenftification of this species. Unfortunately, after several seasons or even years of weathering certain plants of black grama and even tobosagrass closely resemble bush muhly. Younger and active-growing plants of bush muhly have a conical habit or even a pyramidal-shape. Both dead (or apparently so) and live (green) shoots on this plant were readily visible in this photograph.

El Paso County, Texas. June, at beginning of summer rain season.

36. Burrograss (Scleropogon brevifolius)- One of the least palatable (and best disturbance-adapted) perennial grasses on semidesert grassland and Chihuhuan Desert ranges is this monotypic genus. Burrograss typically a dioecious (less commonly a monoecious) species, but with several perfect florets within pistillate spikelets and the whole inflorescence being viewed as a contracted panicle or spicate raceme (Gould, 1975, ps. 240). Staminate spikelets are below pistillate spikelets, an arrangement termed gynaecandrous (Shaw, 2008, p. 181).

Burrograss has a large and discontinuous species range that extends from the Sonoran and Mojave Deserts in California and Nevada, semiarid grasslands of Colorado (Shaw, 2008, p. 181) south into Chile and Argentina in South America (Gould, 1975, p. 242)

37. Fluffgrass (Tridens pulcehllus= Erineuron pulchellum= Dasyochloa pulchella; taxonomists better make up their minds on this one's name because they are running out ways to modify the root pulchel.)- This is one of the smallest, lowest stature perennial grass species on semidesert grassland. Fluffgrass has traditionally been interpreted as an ecological invader being lumped in the same company as annual grasses and forbs, most threeawns, and small perennials like red grama (Bouteloua trifida).

There is a twist to this oft-told story however. Dick-Peddie (1993, ps. 132) felt that fulffgrass served as an indicator species much like desert holly (explained above). Either of these herbaceous species might be relict species of semidesert grassland such their presence on Chihuhuan Desertscrub would indicate a desertscrub disclimax. While general occurrence of fluffgrass (and desert holly) was a very small (perhaps trace) proportion of climax semidesert grassland types (eg black grama grassland) this density, cover, etc. was extremely important as a clue to potential natural (= virgin= pre-Columbian) vegetation. Such is the value of an indicator species. It need not be a dominant, associate, or otherwise major species, just a diagnostic one.

El Paso County, Texas. June.

38. Fluffgrass at peak standing crop (such as it was)- This happy little semidesert dweller was "all swole-up" with the pride of any new parent. The "fluff" is the cottony pubescence on mature spikelets. Spikelets are arranged in fascicles, a term denoting a cluster or bunch as applied to culms, leaves, or, as in this instance, inflorescence branches (Gould and Shaw, 1983, p. 374). .Here again this enigmatic little range plant had another twist to its tale. The clumps of the inflorescence (fascicled spikelets) were themselves on clumps or bunches of shoots (and developing roots) that arose from stolons branching out from the parent plant. The little bunches of shoots topped with fascicles were ramets from the main (parent) plant which was an apparent genetic individual or genet. Each flowering module or clone was reproducing sexually (ie. there was prounced asexual and sexual reproduction simultaneously in the same generation).

In the theory of natural selection (= survival of the fittest) fitness is defined as the capacity to reproduce, specifically to reproduce deoxyribonucleic acid (DNA). The fittest organism (genotype, species, etc).is the one that reproduces the most offspring to perpetuate that genotype, species, etc., and that DNA. Increased fitness (more offspring; more DNA) can only come through genetic change. This requires recombination of genes which can only come from exchange and union (fertilization) of gametes (ie. sexual reproduction). The truest reproduction of existing superior genotypes (DNA having superior fitness) comes from cloning, which in plants is vegetative or asexual reproduction.

The fluffgrass plant shown here was hedging its bets and reproducing by both means so as to: 1) optimize the chances of producing improved genetic individuals (genets) and 2) perpetuation of proven (environmentally tested) genotypes through clones (ramets). And that was as good as the brushed twitch of any treasured tail.

The details of this phenomenon were readily visible in the second slide. Cute, hugh?

Presidio County, Texas. June, full-bloom phenological stage.

41. Longleaf joint-fir or longleaf Mormon tea (Ephedra trifurca)- Presence of this plant on a sandy loam plain range that had remnant patches of black grama and scattered plants of tobosagrass, bush muhly, and fluffgrass was successional evidence that this rangeland had once been black grama-dominated semidesert grassland. The range vegetation was a Chihuhuan Descrub disclimax (Brown et al., 1998). For over two centuries before it was surveyed by the US General Land Office this land was on El Camino Real, the first European road in North America (linked Santa Fe to Chihuhua City). After occupation and domination by Anglo culture this rangeland was part of the Public Domain and subjected to open range ranching.

The pre-European (climax) vegetation in this area was a mosaic of black grama-tobosa semidesert (Chihuhuan) grassland and Chihuhuan Desertscrub with large patches of "overlap" (ie. ecotone; ecotonal vegetation). At larger (say, landscape or even regional scale) the virgin vegetation of this entire region was interpreted and mapped as Desert-Grassland Transitionis (Shreve, 1917). Is it any wonder that after 300 years of what assuredly was overgrazing (first by beasts of burden along a trade-travel corridor that was followed by a Public Domain grazing commons) the range ecosystem was converted byr semidesert grassland to Chihuhuan Desert?

Based on evaluations of range analyses this generalized "reconstruction of the crime" was published by numerous workers, including some range survey data collected from the land shown in this photograph. The case against overgrazing and animal disseminatioin of mesquite seed was not clear-cut (and there were certainly other factors including drought), but as was cited in the introduction of this chapter the role of livestock grazing in desertification was implicated by almost every range scientist who published assessments of this range vegetation over the last 80 years. Students were referred to the up-to-now-conclusive study and synopsis by Gibbens et al. (2005).

Dick-Peddie (1993, p. 116, 140) showed Ephedra trifurca as a "diagnostic (climax) member of Desert Grassland". By contrast E. trifurca was not a climax (diagnostic) member of Chihuhuan Desertscrub and E. torreyana was not a diagnostic member of Desert Grassland. Presence of E. trifurca in the shrub-dominated range plant community shown here (taken in conjuction with presence of larger areas of relict black grama) strongly suggested that this range vegetation was once black grama semidesert grassland. E. trifurca served as an indicator species, a useful concept and application that goes back in the literature at least as far as the monographic Plant Indicators (Clements, 1920)..

42. Branches, leaves, and seeds of longleaf joint-fir or Mormon tea- Size-and-scale along with details of bark and leaf arrangement of E. trifurca were presented in these photographs. E. trifurca can usually be readily distinguished from E. torreyana, with which it often grows, by the branching pattern. Branches of E. trifurca are either solitary or arranged in whorles at shoot nodes (readily seen in the first of these two photographs) whereas branches of E. torreyana are usually arranged in groups of three (Vines, 1960, ps. 39-40) Ephedra leaves are typically reduced to scales. "Longleaf" refers to branches and not the actual scalelike leaves.

Shoots of Ephedra species are browsed only infrequently, mostly under severe overuse/overgrazing when used by cattle and under more varied conditions but with hardly more frequency by the smaller ruminant browsers. The Forest Service summarized that longleaf jointfir was "valuelaes for forage, except, perhaps as an emergency ration" (Forest Service, 1940, BB73). Dayton (1931, p. 13) reported that cattle browsed E. trifurca in winter, but he may have confused this with another Ephedra species as he described E. trifurca as "a rather small bush 2 feet high". The specimen in the preceding photograph and those from which these branches grew were roughly four feet in height.

Ephedra species are dioecious. The two plants that produced the leaders (forester and rangeman term for branches of browse plants) in the two slides shown here were females. Naked seeds surrounded by bracts were present on both plants, but were readily visible only in the second slide.

New Mexico State University College Ranch. Dona Ana County, New Mexico. June, seed-ripe stage.

43. Close-ups of the seed and surrounding structures of longleaf Mormon tea seed- Details of the ovulate (female) cones of Ephedra trifurca. The brownish, parchment-like structures are bracts which surround the naked seed. Ephedra species are gymnosperms (hence proper use of the term cone) and do not bear flowers. Students should not confuse the cone bracts with petals of an angiosperm corolla.

New Mexico State University College Ranch. Dona Ana County, New Mexico. June.

44. Staminate cones on Torrey joint -fir or Mormon tea (Ephedra torreyana)- Pollen (polliniferous) cones on branches of shoots from a male plant of E. torreyana. The pollen-bearing male structures are referred to as "stamens" (hence, staminate cones) or as staminate catkins even though gymnosperms are not flowering plants and therefore do not, strictly speaking, have stamens. (Benson, 1979, p. 608) referred to these male organs as microsporophylls, each of which has five or six sporangia with the whole structure being subtended by paired scales.

Midland County, Texas. April, full-pollen stage..

45. The following series of three photographs represented trhe mixed grass semidesert grassland prairie. This is generally one of the most xeric forms of true or "straight" semidesert grassland (ie. grassland in contrast to savanna--the various forms of ecotone or transition zone between Chihuhuan Desertscrub and semidesert grassland-- or to shrub-steppe). This grassland actually has the physiogonomy of a prairie. Warnock (in Shiflet, 1994) was quoted in the introduction to this chapter as stating that this range vegetation "forms a prairie". The range plant community shown in the next three slides was an example of the Grama-Muhly-Threeawn (Bouteloua-Muhlenbergia-Aristida) association described by Warnock (in Shiflet, 1994). This range vegetation was a bunchgrass prairie or steppe. Even though some of the gramas, especially the regional dominant black grama, are stoloniferous and reproduce primarily by asexual means the sward has an open appearance with substantial bare ground among grass clumps. (This was shown and explained for black grama-dominated grassland range earlier.) Muhlenbergia and Aristida species are almost always strictly cespitose species (bunch grasses that have only tillers, vertical shoots). Thus the clumped physiogonomy of this beautiful prairie.

El Paso County, Texas. June.

45A. Physiogonomy of mixed grass semidesert grassland prairie- Black grama and the cespitose threeawn (wiregrass) and muhly species formed this bunchgrass prairie "smack-dab" in the middle of the Chihuhuan Region. This xeric prairie was "engulfed" in the prevailing Chihuhuan Desertscrub. How much, if any, of this desert vegetation was man-caused disclimax was unknown. One thing was certain: this was virgin (= climax) range vegetation. Even more remarkable was location of this relict vegetation. This pristine range plant community was on the old El Camino Real, the Spanish-Mexican trade route from Santa Fe to Chihuhua City, less than two miles from the Rio Grande.

Dominance was shared fairly equally among Wrights' threeawn (Aristida wrightii), black grama, and bush muhly. Wright's threeawn was the most common (apparent density and cover), but the author did not feel that this was so obvious as to declare Wright's threeawn as the dominant. Besides, there were also substantial numbers of plants and plant cover from red threeawn (Aristida longistea). Plus, so many of the threeawn plants had shed their spikelets that positive identification was impossible. In addition to black grama, hairy grama, annual sixweeks grama, and even a few sideoats grama plants were also present. Bush muhly was the most common Muhlenbergia species, but ring muhly and sand muhly (M. arenicola) were also "present and accounted for" though in small proportions. Presence of the latter two species and sideoats grama was of more ecological importance as indicator species than as major constituent species.Mesa dropseed was a major species and fluffgrass was common enough that these were locally important ingredients in the graminaceous stew.

Very infrequent shrub species included soaptree yucca, creosotebush, and mariola (Parthenium incanum). The most common forb at this season was annual scorpionweed (Phacelia integrifolia). Dried "remains" of tumble mustard (Sisymbrium altissimum) attested to general abundance of this cool-season alien (and weedy) annual forb during winter and spring.

El Paso County, Texas. June; estival aspect for shrubs but dormancy stage for dominant perennial grasses prior to initiation of summer rainy period. FRES No. 40 (Desert Grasslands Ecosystem). No appropriate unit in Kuchler (1964) or Kuchler in Garrison et al. (1977). SRM 713 (Grama-Muhly-Threeawn). No appropritate biotic community in Brown et al. (1998). Chihuhuan Deserts- Chihuhuan Desert Grasslands Ecoregion, 24 b (Griffith et al., 2004).

45B. Prairie in the desert?- Yes, sort of. To be precise it is prairie in the semidesert or grassland-desert transition of Shreve (1917). Prairie is grassland which by definition cannot be desert, but the prevailing vegetation of this region is Chihuhuan Desert and the vegetation surrounding this grassland was desert. This was another example of the patchwork of range plant communities or the vegetational mosaic of this extensive range region.

This was a "picture-perfect" example of the grama-muhly-threeawn bunchgrass prairie form of semidesert grassland. The predominant gramagrass was black grama. Hairy grama and sideoats grama were present though scarce, especially when compared to black grama. Bush muhly was the major Muhlenbergia species, but sand muhly and even sand muhly were also in this grass community. The most common grass in this photograph was Wright's threeawn, but red threeawn was also common and after seed-shatter in some individuals it was not possible to distinguish among Aristida species. Mesa dropseed was distributed more-or-less uniformly throughout this range plant community. The predominant shrub in the background was soaptree yucca.

Area not occupied by plants (= "bare soil") was conspicuous and of considerable size relative to area covered by plants.This is characteristic of grasslands composed of cespitose grass species (bunchgrasses).

El Paso County, Texas. June, but still primarily the dormant season for perennial grasses before onset of summer rains. FRES No. 40 (Desert Grasslands Ecosystem). No appropriate unit in Kuchler (1964) or Kuchler in Garrison et al. (1977).for this range type.SRM 713 (Grama-Muhly-Threeawn). No appropriate Brown et al. (1998) biotic community. Chihuhuan Deserts- Chihuhuah Desert Grasslands Ecoregion, 24b (Griffith et al. 2005).

45C. Interior appearance of a grama-muhly-threeawn semidesert prairie- This inside view of an example of the Bouteloua-Muhlenbergia-Aristida association was a composite of the species make-up and botanical structure of this rangeland cover type. Species visible included Wright's threeawn, mesa dropseed, black grama, and bush muhly. The latter two species were represented by small (and presumedly younger) individuals. The shrub in the left rear corner of the photograph was mariola.

El Paso County, Texas. June; still dormancy for dominant perennial grasses before beginning of summer rains. FRES No. 40 (Desert Grasslands Ecosytem). No appropriate Kuchler (1964, in Garrison et al., 1977) unit. SRM 713 (Grama-Muhly-Threeawn). No appropriate Brown et al (1998) biotic community. Chihuhuan Deserts- Chihuhuan Desert Grasslands Ecoregion, 24 b (Griffith et al., 2004).

46. Wright's threeawn (Aristida wrightii)- This bunchgrass was one of the dominants in the Grama-Muhly-Threeawn cover type (SRM 713). Wright's threeawn was also an indicator species that helped distinguish this from other rangeland cover types in the semidesert grassland.

The second slide presented the inflorescence of Wright's threeawn. Basis of "threeawn" was conspicuous.

El Paso County, Texas. June, dormancy stage, just prior to initiation of current growing season (immediately before "green-up").

47. Sand dunes and related small land forms like hummocks appear frequently but sporatically throughout portions of the general Chihuhuan and Sonoran Desert Regions, including some of the semidesert grasslands and related range vegetation. The following three photographs were used as a series to represent the sand dunes savannah form of semidesert grassland. This range vegetation has not been described-- at least not with any regularity-- in major works devoted to semidesert grassland or related vegetation. For example, it was not noted in The Desert Grassland (McClaran and Van Devender, 1995) and it was represented by the more general form of sand dune shrubland by Warnock(1974, Fig, 23).

This is one of the more restricted and smaller spatial (acreage-wise) kinds of semidesert grassland. Floristic composition and structure is unique (and, as students will soon see, picturesque).

47A. Sand dune savanna- Exterior view showing physiogonomy, structure and botanical composition of of a deep sand, semi-stabilized dunes supporting a savanna of shrubs and dune-adapted perennial grasses. Dominant grasses were giant dropseed (Sporobolus gigantea) and mesa dropseed with spike dropseed (S. contractus) an associate herbaceous species. Shrubs included sand sagebrush (Artemisia filifolia) and broom dalea or broom pea (Dalea scoparia), both of which were local woody dominants (and appeared as gray, symmetrically rounded bushes in this photograph). Associated shrubs included longleaf jointfir and a few individuals of plains yccca (Yucca campestris). The co-dominant herbaceous species were giant dropseed (Sporobolus gigantea) and mesa dropseed. Apparent dominance of shrubs was an example of aspect dominance more than of actual dominance based on proportion of plant cover.

The most conspicuous perennial forb was leatherweed or leather-weed croton (Croton pottsii). Annual forb species consisted of Texas croton (C. texensis), trumpet-, paleflower- or whiteflower gilia (Gilia longiflora= Ipomopsis longiflora), scorpionweed, and tumble mustard.

Dominance based on role and control (domination) of the range plant community was not known, but it seemed apparent based on percentage cover (basal or foliar) and "guestimated" annual production of plant material (biomass was not measrued) that dropseed species were the plant community dominants. This was a savanna, but more of the form of a shrub steppe and not a desertscrub-grassland ecotone. The author interpreted this range vegetation as climax or, at least, a seral stage that approached climax. Giant dropseed was seen as an indictor species and keystone species the dominance by which was strong evidence for a range plant community that was close to potential natural vegetation.

Hudspeth County, Texas. June, early growth of perennial herbs (just before onset of summer rainy period). FRES No. 40 (Deert Grasslands Ecosystem). No specific units from Kuchler (1964, in Garrison et al., 1977), Brown et al., (1998), ofr Society for Range Management (Shiflet, 1994). Chihuhuan Deserts- Chihuhuan Desert Grasslands Ecoregion, 24b (Griffith et al., 2004). Sand Hills range site, Desert Grassland vegetation zone (Soil Conservation Service descriptions).

47B. Species composition of sand dune savanna semidesert grassland- "Sample" of range vegetation on the relatively restricted deep sand or sand dune savanna form of "desert grassland". In the foreground giant and mesa dropseed were clearly visible. An extra large giant dropseed was to left-rear of the largest plant (the silvery gray shrub), a specimen of broom dalea or broom pea. Several sand sagebrush plants were in right background. Plains yucca and longleaf joint-fir were also present in smaller proportions based on both foliar and plant density. Annual forbs included scorpionweed and trumpet gilia. Leather-weed croton was the obvious perennial forb.

Hudspeth County, Texas. June, still early in warm-season growing period as prior to most of summer rains.FRES No. 40 (Desert Grasslands Ecosystem). No appropriate units from Kuchler (1964, in Garrison et al., 1977), Brown et al. (1998), or Society for Range Management (Shiflet, 1994). Chihuhuan Deserts-Chihuhuan Desert Grasslands Ecoregion, 24b (Griffith et al., 2004). Sand Hills range site, Desert Grassland vegetation zone (Soil Conservation Service published description).

47C. Sand dune savanna semidesert grassland- Inside a sand dune range community dominated by giant dropseed and mesa dropseed and with sand sagebrush and broom dalea as associate species. The two large grass plants with both previous season dead herbage and current growing shoots were goiant dropseed. The smaller grasses in the foreground were mesa dropseed and spike dropseed. The largest shrub was broom pea or broom dalea. Background shrubs were mostly sand sagebrush.Also in background were some smaller plants of plains yucca Broom dalea and sand sagebrush could be interpreted as dominants such that perennial grasses and shrubs were co-dominant. More than likely, however, Sporobolus species exceeded woody plants by standards of peak standing crop (or primary productivity), cover, and density.

Students must remember that woody plants like the shrubs shown here have aboveground perennial parts so as to be larger (at least, appear larger) most of the year than herbaceous plants which die back to the ground at end of each growing season. In this photograph giant dropseed plants consisted of sprawling, sparse, dead shoots produced the previous growing season plus new growth (visible as green shoots) produced in early part of the current growing season. Yet giant dropseed--at time of photograph-- appeared smaller than plants of sand sagebrush and broom dalea which had many years (growing seasons) of accumulated growth. It current growing season biomass (live plant weight produced in any one year) of grasses and shrubs was measured and compared at peak standing crop (point or stage of maximum biomass) it would probably be found that giant dropseed produced more plant tissue and underwent greater total photosynthesis than shrubs (ie, giant dropseed would have greater gross primary productivity and/or net primary productivity). This might be true for the smaller mesa and spike dropseed as well. If, as seemed likely, grasses individually or collectively produced greater mass of plant material in a single year, it could be assumed that Sporobolus species were dominant over shrub species even though the plants of shrub species had greater cover and plant mass that had been accumulated over many growing seasons. Furthermore, most of the total (accumulated and current; the sum of) organic matter (measured or estimated as mass) of woody plants was dead. Such dead accumulated plant material is necromass and not biomass (weight of living plant tissue).

Role or impact of plant cover and plant material as well as primary productivity must be accumulated in determining or deciding upon dominance. It is possible, for instance, that shrub tissue (necromass and biomass) stored so much of available soil nutrients or cast enough shad edue to greater cover that shrubs like sand sagebrush or broom dalea had a controlling influence on the more procductive grasses. Alternatively, it is possible that the turnover of herbaceous grass organic matter--including that of roots and shoots-- controlled nutrient cycling to the extent that grasses dominated shrub growth and productivity. Or, maybe, the shallower roots of grasses absorbed soil water (precipitation) more effeciently or effectively than the shrubs. Maybe, shrubs had the advantage in regards soil water aand its use due to deeper, more expansive root systems.

Short forbs in center were leather-weed croton. The small white composite in center midground was plains blackfoot (Melampodium leucanthum). Scorpionweed was also present as at left corner of photograph.

Hudspeth County, Texas. June, still early in spring-summer growing season for perennials. FRES No. 40 (Desert Grasslands Ecosystem). No appropriate units of Kuchler (1964), Brown et al. (1998), or Society for Range Management (Shiflet, 1994). Chihuhuan Deserts- Chihuhuan Desert Grasslands Ecoregion, 24b (Griffith et al., 2004). Sand H:ills range site, Desert Grassland vegetation zone (Soil Conservation Service published description).

52. Another form of Trans-Pecos semidesert grassland- Unique grassland community at extreme eastern edge of the Trans Pecos Basin and Range Vegetation Area of Texas. One of the most xeric forms of mixed grass (mid and shortgrass species) steppe had developed along the westernmost margins of Great plains mixed prairie and beginning of the Semidesert Zone. This transition grassland had more features of the semidesert grassland than High Plains mixed prairie fbased on criteria of plant species composition, physiogonomy, and biomass production. This was an example of a gypseous (gypsum-containing) soil variant of semidesert grassland.

The dominant species was bush muhly with associate species varying from tobosagrass, gyp(sum) grama (Bouteloua breviseta), and/or cane bluestem (Andropogon barbinodis). Other locally important climax grasses included Hall's panicgrass (Panicum hallii), alkali sacaton, and sideoats grama. The most common herbaceous invaders were burrograss (Scleropogon brevifolius), fluffgrass, and sixweek's grama (Bouteloua barbata). Red threeawn (Aristida longiseta) and Wright's threeawn (A. wrightii) were well-represented. Successional status of these perennial Aristida species was not known, butthey were most likely increasers in this range type and on this range site.

.Tarbush and creosotebush were the co-dominant shrub species though tarbush was more abundant with greater density, cover, and consistency. These two shrubs were unquestionably members of the climax plant community, perhaps so much as to create a natural savanna or savanna-like grassland. .Tarbush and creosotebush had increased due to overgrazing, oil development, cessation of fire, and perhaps such natural disturbances as great climatic fluctuations two-thirds to three-fourths of a century prior to this time. A chemical brush control using tebuthiron, N-[5-(1,1-dimethylethyl)-1,3,4-thiadiazol-z-yl]-N,N-dimethylurea, had killed most of these woody co-dominants on this range, but remaining cover probably resembled that of the pre-white man vegetation. Jimmyweed (Haplopappus heterophyllus) was another shrub on this semidesert grassland range, but this compositae (of the aster tribe, Astereae) was not common. Nor was broom snakeweed which was also present only on occasional spots and then not abundant.

Part of this range seemed to be as much, if not more, of a transition between semidesert grassland and Chihuhuan Desert than of mixed prairie and semidesert grassland (see below).. This may have been due to greater degree of range retrogression and brush invasion over that portion of the range. This pasture was all the same range site.

The second of these two slides was a closer-i9n view of the left foreground of the first slide. Cane bluestem was conspicuous in both of these photographs.

Railway Ranch, Upton County, Texas. Mid-October; peak standing crop and grain-ripe stage of major grasses. FRES No.40 (Desert Grasslands Ecosystem). K-48 (Grama-Tobosa Prairie). SRM 505 (Grama-Tobosa Shrub). Mixed Grass-Scrub Series 143.124 of Chihuhuan (Semidesert) Grassland biotic community 143.1 of Brown et al. (1998, p.40). Gyp range site. High Plains- Arid Llano Estacado, 25k (Griffith et al., 2004).

53. Ecotonal semidesert grassland- Semidesert grassland dominated by bush muhly with tobosagrass, cane bluestem, and gyp grama local associates. Other grasses included Hall's panicgrass, alkali sacaton, and sideoats grama as other decreasers or increasers; red and Wright's threeawns that were probably increasers; and burrograss, fluffgrass, and sixweeks grama, all invaders. Tarbush and creosotebush were co-dominant shrubs though tarbush was first of these two. Jimmyweed and broom snakeweed were incidental shrubs or subshrubs.

This was a portion of the same range as that presented immediately above. There were more live plants of tarbush and creosotebush than on other parts of this particular pasture. This range had been treated with tebuthiron, a substituted urea herbicide, to kill woody plants and it was not known if the portion featured in the foreground of this slide had been treated or treated as effectively as portions with more dead plants of these two shrubs. Alternatively, some dead tarbush and creosotebush plants appeared to have killed due to competition from bush muhly. This is a widespread phenomenon throughout much of the Trans Pecos Region when reduced grazing pressure and other factors permit establishment of bush muhly beneath or, at least, in close proximity to these two dominant shrubs. Two examples of this shrub-killing competition were shown below.

Irrespective of explanation, greater cover and density of tarbush and creosotebush in conjuction with relatively high foliar cover and density of native grasses created a semidesert grassland savanna or a savanna-like semidesert grassland that was a transition--an ecotone--between the climaxes of semidesert grassland and Chihuhuan Desert. It was self-evident (or nearly so) that the Chihuhuan Desert had advanced eastward and expanded into rangeland that had previously been the so-caled desert grassland. This was an example of desertification, the cause(s) of which were unknown. Overgrazing, oil field activity, and elimination of fire (not in any particular order) were all obvious contributing factors, but it is possible that a unique combination of unusually cold and dry weather approximately 70 years earlier could have set into motion the processes of range retrogression. Therefore, desert advancement could have been partly or even largely natural. Alternatively, it could have been largely anthropogenic so that more desert-like vegetation, including the savanna aspect, was a disturbance climax.

Railway Ranch, Upton County, Texas. Mid-October; peak standing crop and grain-ripe stage of major grasses. FRES No.40 (Desert Grasslands Ecosystem). K-48 (Grama-Tobosa Prairie). SRM 505 (Grama-Tobosa Shrub). Mixed Grass-Scrub Series 143.124 of Chihuhuan (Semidesert) Grassland biotic community 143.1 of Brown et al. (1998, p.40). Gyp range site. High Plains- Arid Llano Estacado, 25k (Griffith et al., 2004).

54. Waging war on desert invaders- Tarbush and creosotebush, Chihuhuan Desert co-dominant invaders on a transition semidesert grassland, were attacked by man using tebuthiron and plants via intense competition. Creosotebush (center foreground) flanked on left and right by a single plant of tarbush. This range had been depleted through a number of contributing factors including cessation of fire, oil field activity, overgrazing (order of importance not known) and, perhaps, aberrant atmospheric conditions roughly 70 years earlier. Vegetational results of this range deterioration were increase in cover of native shrubs (primarily tarbush and creosotebush) and decline in abundance (density, cover, distribution or frequency) of decreaser and increaser species of grasses. Range managers resorted to improved grazing management and chemical brush control in attempts to reverse the trend in species composition and structure of the range vegetation. The result was an outstanding kill (apparent root kill at that) of brush (mostly tarbush followed closely by declines of creosotebush) and commensurate increase in grass abundance. This was most apparent for bush muhly followed by increased cover and density of tobosagrass, cane bluestem, and alkali sacaton as other major decreasers along with greater abundance of gyp grama and Hall's panicgrass.

Given that some grazing (of improved management practices) by cattle took place on this range it could not be determined the relative degree to which death of brush and increased abundance of grass was due to herbicidal treatment versus improved grazing management. It is common knowledge among rangemen working throughout this area of the Trans Pecos Basin and Range that when bush muly can establish beside, beneath, etc. adult plants of tarbush and creosotebush this grass commonly kills the shrub. Among numerous management practices that permit establishment of bush muhly is better grazing management such as reduced stocking rate, change in season of use, and deferment of grazing. It is commonly assumed that flourishing of bush muhly and death of shrubs is a result (partial result) of competition, for whatever resource is unknown.

It will be observed that shrubs in the restricted area shown in foreground of this photograph had not been treated with tebuthiron, at least not effectively as in the background where most tarbush and creosotebush had been killed. This area appeared to be outside the rangeland that had received chemical treatment. Observant observers will note that relatively small plants of bush muhly were growing beside adult tarbush and creosotebush. Would these bush muhly plants have been larger and more robust if the shrubs had been killed by tebuthiron? Or had bush muhly established due to improved grazing practices and subsequently the shrubs were slowly dying from intense competition? God, not us, knows. And we will never know without carefully designed experiments.

Railway Ranch, Upton County, Texas. Mid-October; peak standing crop and grain-ripe stage of major grasses. FRES No.40 (Desert Grasslands Ecosystem). K-48 (Grama-Tobosa Prairie). SRM 505 (Grama-Tobosa Shrub). Mixed Grass-Scrub Series 143.124 of Chihuhuan (Semidesert) Grassland biotic community 143.1 of Brown et al. (1998, p.40). Gyp range site. High Plains- Arid Llano Estacado, 25k (Griffith et al., 2004).

55. Grass victorious- A degraded semidesert grassland cattle range composed of bush muhly (the dominant), tobosagrass, cane bluestem, alkali sacaton, gyp grama, Hall's panicgrass, sideoats grama, red threeawn, and Wright's threeawn with successionally excessive abundance of tarbush and creosotebush, co-dominant shrubs, had been treated with tebuthiron a number of years prior to these photographs. A rough visual estimate was that fully three-fourths (probably more) of the two major shrubs had been killed and a commensurate increase in grass cover, biomass, density, etc. had occurred. Obviously this was a phenomenal transformation in the range vegetation, the extent of range improvement as estimated by the increase in range condition class. No wonder this range demonstration was used as a case history that became a highly successful promotional example of the effectiveness of tebuthiron. A well-earned tip of the hat.

A salute rendered, it must be noted that scientifically it cannot be concluded that such improvement is due--either in part or wholly--to the herbicide treatment. The trial was confounded because this tebuthiron-treated range was also grazed by cattle under better grazing management during the herbicide experiment. Was the obvious, undeniable improvement in range condition (an outcome of secondary plant succession) due to application of tebuthiron or to better grazing management or some combination thereof (or to neither one and instead other though unknown variables). This situation is always the case when a problem is subjected to a number of joint treatments without separate treatment applications and, hence, the capability to partition out effects and affects. It is well-known by rangemen in the Trans Pecos Basin and Range Region that establishment of bush muhly in close proximity to shrubs like tarbush and creosotebush commonly results in death of the woody plants. This takes place in absence of brush control treatments. Judicious grazing appears to permit more establishment of bush muhly. Neither can it be ruled out, however, that shrubs act as nurse plants for bush muhly (ultimately resulting in shrub death). Range scientists simply do not know.

The immense plant of bush muhly in the second of these slides (and the edge of another bush muhly in far left margin) were growing around what remained of dead tarbush plants that could (might) have been killed as results from competiton with bush muhly. These plants were outside of the tebuthiron-treated range that was shown in the first of these photographs, but it was also possible that some "stray" (undirected) tebuthiron had killed the tarbush plants thereby permitting establishment and monumental growth of the "champion" bush muhlys. Again, only God knows-- from this trial anyway.

Railway Ranch, Upton County, Texas. Mid-October; peak standing crop and grain-ripe stage of major grasses. FRES No.40 (Desert Grasslands Ecosystem). K-48 (Grama-Tobosa Prairie). SRM 505 (Grama-Tobosa Shrub). Mixed Grass-Scrub Series 143.124 of Chihuhuan (Semidesert) Grassland biotic community 143.1 of Brown et al. (1998, p.40). Gyp range site. High Plains- Arid Llano Estacado, 25k (Griffith et al., 2004).

56. Sample of semidesert assembly- Gyp grama, cane bluestem, and burrograss growing in close association in the sward of a transition semidesert grassland situated between an island of Chihuhuan Desert scrub and xeric mixed prairie at eastern margin of the Trans Pecos Basin and Range Region of west Texas. Along the border of the southern and drier Southern High Plains (Llano Estacado or Staked Plains) and the eastern outliers area of the Basin and Range physiographic province there is a mosaic (patchwork pattern) of mixed prairie grassland, semidesert grassland, and Chihuhuan Desert (a shrubland type). Any number of factors can result in conversion of one or two of these range types into one or two of the other types.

The "photoplot" seen here was on the depleted and now recovering--through range improvement practices--semidesert grassland that had been used as a cattle range and oil field for decades. Usually, it can comed back.

Railway Ranch, Upton County, Texas. Mid-October; peak standing crop and grain-ripe stage of major grasses. FRES No.40 (Desert Grasslands Ecosystem). K-48 (Grama-Tobosa Prairie). SRM 505 (Grama-Tobosa Shrub). Mixed Grass-Scrub Series 143.124 of Chihuhuan (Semidesert) Grassland biotic community 143.1 of Brown et al. (1998, p.40). Gyp range site. High Plains- Arid Llano Estacado, 25k (Griffith et al., 2004).

57. Mixed grass (grama-bluestem-cottontop) hillside grassland- Represented here was another of the less widespread kinds or forms of semidesert grassland. This range type was described in the introduction to this chapter and was not be repeated here. In the example above sideoats grama and cane bluestem were co-dominant grasses. Dominance by these two decreaser mid-grass species was the defining feature of this range cover type. Other major grasses included black and chino grama, plains bristlegrass, and perennial threeawns (mostly red threeawn). Arizona cottontop, tanglehead, and bush muhly were present in much smaller amounts. Fluffgrass was also present in smaller proportions. Forbs were extremely rare, even composites, and irrelevant. Important shrubs included mariola, skeletonleaf goldeneye, whitethorn acacia, sacahuista, tasillo(Opuntia leptocaulis), and Engelmann's pricklypear (O. engelmannii var. engelmannii).

This semidesert grassland range type is typically more varied that some of the more widely occurring types (probably including the black grama semidesert grassland type, most common of all). This is undoubtedly attributable to gret diversity of habitat factors associated with hills and mountains including slope and aspect, soils and parent materials, and numerous atmospheric variables that go along with slope, aspect, and elevation.

Chinati Mountains, Presidio County, Texas. June, early estival aspect prior to typical summer rainy period. FRES No. 40 (Desert Grasslands Ecosystem). No appropriate units in Kuchler (1964, in Garrison et al.., 1977), Brown et al. (1998), or Society for Range Management (Shiflet, 1994). The Bouteloua-Andropogon-Trichachne post-climax (sideoats grama-feather grass- Arizona cotton grass) type of Whitfield and Brutner (1938). This semidesert grassland range type should be described by the Society for Range Management as Grama-Bluestem-Arizona Cottontop Shrub rangeland cover type if and when there is a second version of Rangeland Cover Types . Chihuhuan Deserts- Low Mountains and Bajadas Ecoregion, 24c (Griffith et al., 2004). Igneous Hill and Mountain range site, Desert Grassland vegetation zone (Soil Conservation Service, Soil Survey of Jeff Davis County, Texas).

58. Cane bluestem (Andropogon barbinodis= Bothriochloa barbinodis var. barbinodis)- This widely distributed mid-grass is a climax dominant (decreaser) on more mesic range sites of semidesert grasslands and Chihuhuan Desert The discontinuous biological range of this species extends from Uruguay and Argentina in South America northward to Colorado and California (Gould, 1975, p. 600).

Railway Ranch, Upton County, Texas. Mid-October; grain-ripe stage and peak standing crop.

59. Chino grama-Chihuhuan Desert shrub savanna- The Chihuhuan Desert and semidesert (Chihuhuan) grassland communities form a vegetational mosaic with each other across much of the greater Chihuhuan Region. Countless local ecotones (localized transition zones) are thus part of the "patchwork" of range plant communities across this diverse landscape that itself was seen as a Desert-Grassland Transition by Shreve (1917). There are thus smaller ectones within a larger general ecotone, all of which could be climax vegetation. Some such transition zones are range plant communities doubtless existing in some stage of retrogression as semidesert grassland has been (or is being) degraded under human influence to disclimax Chihuhuan Desertscrub.

The range vegetation shown in here was a rather obvious (at least to this author-photographer) example of climax ecotonal Chihuhuan Desert-semidesert grassland transition. The vegetation of the particular range that included the "sample" shown here extended from creosotebush plains desert in a basin to chino grama foothill grassland higher up on the foothills of north and east slopes. The range plant community that developed on this southwest-facing hillside was a "mixture" of basin Chihuhuan Desert dominated by creosotebush and ocotillo and foothill chino grama grassland. Range vegetation on this xeric "transect" that extended from bajada (bench) to rimrock of a hilltop was a "textbook exampl"e of the savanna that develops where desert and grassland of the Chihuhuan Region meet and mingle. Based on early descriptions of natural or climax vegetation by ecologists like Shreve (1917, 1942), Clements (1920), Shantz and Zon (1924) and summaries by later ecologist like Dick-Peddie (1993) this savanna was climax vegetation characteristic of one of the major general natural plant communities of the virgin (= pre-Columbian) range. This is an example of a relict or relict vegetation that can be used as a range reference area. In fact, this sample of range vegetation could probably serve as a research natural area (Laycock, 1975 generally and especially, p. 11). Relict was defined by the Society for Range Management as: "A remnant or fragment of the climax plant community that remains from a former period when it was more widely distributed. Syn. pristine" (Jacoby, 1989).

On this savanna the dominant range plant was chino grama which made up most of the herbaceous understorey. Black grama and red threeawn were associate grass species; bush muhly and cane bluestem were "also-ran" but indicator species. Dominant shrubs were creosotebush, the regional dominant of both Chihuhuan and Sonoran Deserts, and ocotillo. Other shrubs present included the euphorbia leatherstem or sangre de drago, dragon's blood, (Jatropha dioica), whitethorn acacia and, of course, honey mesquite. The very low cover of mesquite, especially in conjuction with abundance of decreaser grass species was an indication as to the pristine condition of the range vegetation on this photographic transect.

59A. Chino grama-Chihuhuan Desert shrub savanna- Transect-like view of a hillside ecotone between basin creosotebush plains desert and chino grama foothill grassland. Chino grama was the dominant of this plant community which indicated the identification of this range cover type as grassland and not shrubland (desert). Black grama and, rarely, sideoats grama, cane bluestem, red threeawn, and bush muhly were locally common. This combination of grass species demonstrated the floristic relationship of this rangeland cover type with the mixed grass (Grama-Bluestem-Arizona Cottontop-Shrub) cover type presented immediately above.

Ocotillo was conspicuous being "fully clothed" with ephemeral, drought-deciduous leaves, but perennial grasses had just begun new growth prior to onset of summer rains (hopefully). Creosotebush was dominant shrub on this savanna. Creosotebush is the zonal dominant of the Chihuhuan and Sonoran Deserts as well as much of the Mojave Desert, Presence of creosotebush as a dominant in this climax vegetation indicated the affinity of desertscrub and grassland biomes under proximite (usually adjacent) location to each other throughout this region. Forbs were so rare as to be irrelevant.

The igneous parent material of this soil was obvious on the land surface.

Bofecillos Mountains, Big Bend Ranch State Park, Presidio County, Texas. June, early estival aspect at start of "green-up" of perennial grasses. FRES No. 40. No appropriate units for vegetation in Kuchler (1964, in Garrison, 1977), Society for Range Management (Shiflet, 1994), or Brown et al., (1998). Chihuhuan Deserts- Low Mountains and Bajadas Ecoregion, 24c (Griffith et al., 2004).

59B. Transition Chihuhuan Desert-Grassland Savanna- "Overlap" of chino grama-dominated foothill semidesert grassland and plains creosotebush Chihuhuan Desert resulted in development of this chino grama-creosotebush savanna. This range vegetation could be handily described as a Chihuhuan shrub-steppe. The range plant community is a "synopsis shot" of what ecologists from Shreve (1917) to Dick-Peddie (1993, ps.106-107) described variously as "Desert-Grassland Transition", "desert savanna", "desert shrub grassland", "shrub steppe", and "grassland transition". "The transitional nature of this vegetation is apparent from these names" (Dick-Peddie, 1993, p. 106).

Prevalence of the cespitose chino grama (most of the buckskin-colored clumps) was obvious in both of these slides. The large (probably very old) chino grama highlighted or backgrounded by the large igneous rock in left foreground was a good example of the habit of this species as it appears on the range. In addition to conspicuous individuals of the dominant shrubs, creosotebush and ocotillo, specimens of leatherstem were also obvious (in foreground and near midground of both slides and midground at far right margin of slide on the right). Taller shoots of cane bluestem, a mid-grass associate species, were visible in foreground of slide on the right. Sideoats and black grama were locally abundant, but nowhere did they approach the abundance of chino grama. Red threeawn and bush muhly were also common, especially at edges of larger bare areas and close to shrubs, respectively.

Physiography of the Basin and Range physiographic province and igneous parent material of soils was apparent in these two photographs as well as ind the preceding photograph.

Bofecillos Mountains, Big Bend Ranch State Park, Presidio County, Texas. June, early estival aspect with growth of shrubs being more advanced than that of perennial grasses which were still in early stages of "green-up" (iniital annual growth).FRES No. 40. No apt units for vegetation in Kuchler (1964, in Garrison et al., 1977), Society for Range Management (Shiflet, 1994), or Brown et al., (1998). Chihuhuan Deserts- Low Mountains and Bajadas Ecoregion, 24c (Griffith et al., 2004).

59C. Detail of range vegetation in ecotone of Chihuhuan Desert shrub-foothill chino grama grassland- This range plant community was a savanna-- transition zone-- between the Chihuhuan Desert creosotebush plains (in a lower basin) and the "straight" or "pure" chino grama-dominated grassland of bajadas (benches) and slopes that were immediatly above (adjacent to) the basin. Chino gram was dominant. Black and sideoats grama were also present. All perennial grasses were still in dormancy to early growth stages as time of photograph was before beginning of summer rain.period. There were no annual grasses. Dead stems of winter annual forbs (foreground) were mostly of tumble mustard.

Desert pavement of land surface was visible.

Bofecillos Mountains, Big Bend Ranch State Park, Presidio County, Texas. June, early estival aspect at early stages of new growth in perennial grasses. FRES No. 40 (Desert Grassland Ecosystem). No appropriate units for vegetation in Kuchler (1964, in Garrison, 1977), Society for Range Management (Shiflet, 1994), or Brown et al., (1998). Chihuhuan Deserts- Low Mountains and Bajadas Ecoregion, 24c (Griffith et al., 2004).

60. Sotol-lechiguilla-grama grassland- The following series of five photographs presented one of the larger and better known kinds or forms of semidesert grassland known by such designations as the sotol grasslands (Wauer, 1980, ps.20, 22-23) or the lechuguilla-chino grama association (Warnock, 1974, p 47.). The current author included both liliaceous shrubs in his title based on previous designations like those just cited as well as his observations that sotol and lechuguilla are usually both present on this range cover type. It was remarked in the introduction of this chapter that it was somewhat odd that the Society for Range Management (Shiflet, 1994) did not provide a title and description for this rangeland cover type given that the same had already appeared in publications, including by the National Park Service and Warnock (1974) the latter of whom did provide the title and description for one of the semidesert grassland rangeland cover types in Shiflet (1994). Regardless, the distinctive and readily discerned sotol-lechuguilla-grama (usually chino grama) grassland was treated here and now.

Readers can also discern that this climax range vegetation is a savannah or, at least, is savannah-like in appearance. It could be described as a savannah. The published accounts of this range plant community generally entitled it "grassland" and not "savanna". This author followed the precidence set by these published descriptions and identified this potential natural (= climax) vegetation as grassland noting the aspect dominance quality provided by sotol and lechuguilla. It was explained above for the sand dune savanna semidesert grassland that woody plants with aboveground perennating parts are typically larger and often occupy considerably greater proportions of total plant cover than herbaceous plants, but this increase took place over a number of growing seasons (perhaps decades). This growth pattern and resultant plant habit must be taken into account when making comparisons of woody plants with herbaceous species that have no aboveground perennating portions and thus do not accumulate aboveground plant material on par (that can be compared) with shrubs and trees.

The sotol-lechuguilla-gramagrass range type is generally found as the highest elevation grassland in areas where it develops. This form of semidesert grassland is largely determined by elevation and related habitat factors including slope and soil features. In mountains as, for example the Chisos Range, sotol-lechuguilla-gramagrass occurs as elevational zone vegetation adjacent to (elevationally just below) the oak-pinyon pine-juniper woodland and immediately above the chino grama foothill grassland zone. On some hills and lower mountains (and depending on topographic, edaphic, and other factors) the sotol-lechuguilla gramagrass cover type is the uppermost zone of range vegetation. Examples of sotol-lechuguilla-gramagrass grassland on both of these elevation-determined environments were presented below.

60A. Sotol-gramagrass grassland- Lechuguilla was absent from this savanna-like, hill-occupying grassland. The dominant herbaceous range plant was sideoats grama, followed by chino and black grama, cane bluestem, threeawn species (mostly of the Aristida purpurea complex, especially A. purpurea var. longiseta= A. longiseta), buffalograss, fluffgrass, plains bristlegrass, and hairy tridens (Tridens pilosus= Erioneuron pilosum) such that there was a full-array of decreasers, increasers, and invaders from the Gramineae. Forbs were extremely limited (mostly composites). Common shrubs on this range besides the dominant sotol were tasajillo, cane cholla (Opuntia imbricata), and pricklypear (mostly Englemann's pricklypear) cactus and broom snakeweed. Spanish dagger or Torrey yucca (Yucca torreyi= Y. treculena= Y. macrocarpa) was present as isolated speciments. Mesquite and Acacia species plus sacahuista were conspicuous by their absence (for all practical purposes).

This range vegetation had developed on high rolling hills (not mountains) and occupied entire hill slopes from above valley or basin with creosotebush plains form of Chihuhuan Desert to tops of hills like the one on which this photograph was taken.

This range had just been through the Long Drought of the 1990s-2000s. Obviously it had been stocked accordingly as grass vigor was high and range condition class was or approached Excellent. There were no obvious indications that range trend was downward.

Private property, Chinati Mountains, Presidio County, Texas. June, early estival aspect with almost no new grass growth (ie. prior to "green-up"). FRES No. 40 (Desert Grasslands Ecosystem). No appropriate vegetation units in Kuchler (1964, in Garrison et al., 1977), Society for Range Management (Shiflet, 1994), or Brown et al. (1998). Chihuhuan Deserts- Low Mountains and Bajadas Ecoregion, 24c (Griffith et al., 2004). Gravelly Hill range site, Desert Grassland vegetation zone (Soil Conservation Service site description). West slope to hill top.

60B. Sotol-gramagrass grassland- Grandeur of the semidesert grassland was displayed across the rolling hills of the Mountains as sotol and a mixture of climax mid- and shortgrass formed this savanna-like grassland. Sideoats grama was the dominant grass with black and chino grama the major associates. Red threeawn (and probably other perennial threeaws), cane bluestem, plains bristlegrass, buffalograss, hairy tridens, and fluffgrass called these hills their home. Other grasses that were likely present but not detected by the photographer included tanglehead, green sprangletop, and Arizona cottontop. Forbs were essentially absent as were leguminous shrubs including honey mesquite and Acacia species as well as lechuguilla and sacahauista. In addition to sotol other common shrubs were Englemann pricklypear, tasajillo, and cane cholla along with the suffrutescent broom snakeweed. Spanish dagger (also known as Torrey yucca) was present as widely scattered individual plants.

Good example of habit of cespitose grass species (bunchgrasses), and of the pristine expression of this rangeland cover type.

Private property, Chinati Mountains, Presidio County, Texas. June, early estival aspect overall with very little new growth ("green-up") of grasses.FRES No. 40 (Desert Grasslands Ecosystem). No rlelvant unit in Kuchler (1964, in Garrigus et al., 1977), Society for Range Management (Shiflet, 1994), or Brown et al., (1998). Chihuhuan Deserts- Low Mountains and Bajadas Ecoregion, 24c. Gravelly Hill range site, Desert Grassland vegetation zone (Soil Conservation Service range site description).

60C. Sotol-gramagrass grassland- Physiogonomy, community structure, and species composition of the virgin vegetation of this range cover type of semidesert grassland. Across a seemingly endless range of hills a diverse combination of climax mid- and shortgrass species along with shrubs formed a "textbook example" of the sotol grassland. This pristine range vegetation could be interpreted and described as a sotol-gramagrass savanna, but this author followed the precident of previously published titles and viewed this climax range as grassland with a conspicuous shrub component.

Sideoats grama was the dominant grass with chino and black grama as overall associates. There were patches where other grass species ranging from cane bluestem to fluffgrass to red threeawn dominated. Buffalograss, hairy tridens, and plains bristlegrass were present in small amounts while green sprangletop, Arizona cottontop, and tanglehead were almost assuredly present though not observed by this author.

After sotol the major shrubs were cactus species including pricklypear (mostly Englemann's pricklypear), cane cholla, and tasajillo. Broom snakeweed represented suffrutescent shrubs (and composites in general). Forbs were so sparse as to be irrelevant. Torrey yucca, better known as Spanish dagger, was found only as widely scattered individuals. Lechuguilla, sacahuista, honey mesquite, and Acacia species were noteworthy by their absence from this extraordinary range vegetation.

Private property, Chinati Mountains, Presidio County, Texas. June, early estival aspect with little new growth of grasses (prior to summer rainy period). FRES No. 40 (Desert Grasslands Ecosystem). No pertinent unit of vegetation in Kuchler (1964, in Garrigus et al.l, 1977), Society for Range Management (Shiflet, 1994), or Brown et al., (1998). Chihuhuan Deserts- Low Mountains and Bajadas Ecoregion, 24c (Griffith et al., 2004). Gravelly Hill range site, Desert Grassland vegetation zone (Soil Conservation Service range site description).

60D. Sotol-lechuguilla-chino grama grassland- On this range lechuguilla likely outranked sotol in dominance based on both plant cover and density (and likely influence on range plant community and abiotic factors). On this range, chino grama was the dominant grass with black grama and red threeawn associate species. Fluffgrass was most abundant grass species invader. Sideoats grama, cane bluestem, plains bristlegrass, and bush muhly were found but were essentially absent. Also absent were leguminous shrubs (eg. mesquite) as well as creosotebush and ocotillo which were the woody dominants on adjacent lower elevation Chihuhuan Desertscrub.

This was one of several kinds of foothill semidesert grassland. Rough Mountain was at extreme left background, Parent material was pirmarily igneous.

Big Bend National Park, foothills of Chisos Mountains, Brewster County, Texas. June, early estival aspect (before onset of summer rains so very little "green-up" of perennial grasses). No appropriate vegetational unit in Kuchler (1964, in Garrison, 1977), Society for Range Management (Shiflet, 1994), or Brown et al., (1998). Chihuhuan Deserts- Low Mountains and Bajadas Ecoregion, 24c.

60E. Species-rich sotol-lechuguilla-grama grassland- Tremendous botanical diversity marked this more xeric example of the savanna-like sotol-lechuguilla grama grassland as worthy of inclusion with the preceding pristine and interesting examples of the same range cover type. Where to start? The herbaceous understorey, like the preceding examples, could be viewed as consisting of layers corresponding to height of mature grasses (eg. mid- and shortgrass species). On this drier site red grama was dominant to chino grama and cane bluestem followed by black and sideoats grama (in that order). Shrubs included sotol, lechuguilla, creosotebush, tasajillo, Englemann pricklypear, honey mesquite, cane cholla, mariola, and tarbush. Forbs were quite limited and, like grasses, still in dormancy awaiting summer rains.

Physiography (= topography) of Trans-Pecos Basin and Range was featured prominently in this photograph. Geology of this landscape was volcanic lavas, tuffs, and sedimentary rocks (MacLeod, 2002, p. 26).

Big Bend National Park, Brewster County, Texas. June, early estival aspect (prior to beginning of summer rainy period so very little "green-up" of perennial grasses). No appropriate unit of vegetation in Kuchler (1964, in Garrison, 1977), Society for Range Management (Shiflet, 1994), or Brown et al. (1998). Chihuhuan Deserts- Low Mountains and Bajadas Ecoregion, 24c.

62. Spanish dagger or Torrey yucca (Yucca treculeana, Y. torreyi= Y. macrocarpa)- Another genus of shrub or, in case of larger individuals, tree on sotol-lechuguilla grama grassland is Yuccca. No Yucca species is a dominant plant on this range cover type to the extent that soaptree yucca is on black grama semidesert grassland. Notwithstanding this situation Yucca species do occur and are locally prominent on sotol-lechuguilla grama grassland range. The more common of these species is Spanish dagger. This species is also a member of the Chihuhuan Desertscrub at lower elevations and oak-pinyon pine-juniper woodland at the higher elevation zone.

Presidio County, Texas. June.

63. Spanish bayonet, giant dagger; or Faxon yucca (Y. carnerosana= Y. faxoniana)- This Yucca species is less common than soaptree and Torrey yucca and it is more likely to grow at higher elevations in foothill and upper mountain slopes in the Trans-Pecos Basin and Range Region.

Classification of Yucca species has remained controversial. Some authors like Smith (1977) treated this genus as the agave subfamily (Agavoideae) of the lily family (Liliaceae) while other taxonomists such as Powell (1988) placed Yucca in the agave family (Agavaceae).

Presidio County, Texas. June.

64. Texas hechtia or Texas false-agave (Hechtia texensis)- One of the most unusual succulents in the Chihuhuan Desert and semidesert grassland vetetation is this member of the pineapple family (Bromeliaceae). This bromeliad is easily confused with the Agave or even Yucca species. While Hechtia is in the Monocotyledoneae it is in a separate family which includes the Spanish moss group.

This was a small group of H. texensis in a community dominated by sotol.

Bofecillos Mountains, Presidio County, Texas. June.

66. Chino grama (= "chinograss") foothill grassland- Chino grama (Bouteloua ramosa) often forms estensive monospecific-- or nearly so-- stands (Clemenentsian consociations) on mountain foothills in the Trans-Pecos Basin and Range Area. Other grass species such as black grama (mostly), sideoats and hairy grama, cane bluestem, threeawn, and bush muhly plus scattered shrubs like creosotebush, cactuses, tarbush, mariola, Acacia species, and ceniza (Leucophyllum species) are often locally important or, more frequently, present as minor components.

This kind or form of semidesert grassland is most prevalent in the mountains of Trans-Pecos Texas. Chinograma grassland primarily occurs as zones in mountain foothills or on higher hills in proximity to Chihuhuan Desertscrub and sotol-lechuguilla-grama grassland. The chinograss zone usually is somewhat "sandwiched" between these other two range plant communities being spatially below and above them, respectively. This rough zonation of Trans-Pecos Basin and Range vegetation (including the oak-pinyon pine-juniper woodland and ponderosa pine forest cover types) is apparently at least a partial function of elevation and effects related to elevation. Obviously soils, slope, directional aspect, and other factors are likely also involved.

Chino grama grassland has affinities with the adjacent zones of vegetation so much so that chino grama is often the dominant, and almost always an associate, herbaceous species in sotol-lechuguilla-grama grassland and Chihunuan Desert-semidesert grassland transition (savanna). The chino grama-dominated foothill grassland can be viewed as the "pure" or "concentrated" expression of grassland dominated by the species in contrast to the transition vegetation (ie. savanna) below and above it elevation-wise.

Chino grama-dominated grassland is considerably more restricted in location and extent (acreage covered) than black grama- and tobosagrass, the major forms (rangeland cover types) of semidesert grassland. Chino grama foothill grassland is also much less striking in physiogonomy (at least to laymen, naturalists, and non-"grassmen type" vegetation scientists) than sotol-lechuguilla-grama grassland. It is maybe for these reasons that this minor --but locally important-- range cover type not treated by those who otherwise described semidesert grassland vegetation. The Society for Range Management (Shiflet, 1994) omitted this range vegetation as a rangeland cover type. Whitfield and Beutner (1938) and Whitfield and Anderson (1938) provided the seminal and still definitive descriptive accounts of "desert plains grassland", and they did not even list Bouteloua ramosa (or a synonym) but did give five Bouteloua species. Neither B. ramosa nor a synonym was included in the classic description of by Clements (1920) in Plant Indicators. Some authorities who described vegetation and grasses of Trans-Pecos Texas did explain importance of chino grama and describe chino grama dominated-semidesert grassland (Maxwell, 1968, ps. 108-109; Warnock, 1974, ps. 26-27; National Park Service, undated, p.84; Powell, 2000, ps. 15, 217-219). None of these publications were devoted to vegetation per se or to its description and classification.

The following series consisting of four slides presented the chino grama foothill grassland range type on par with more widely distributed and better-known cover types of semidesert grassland.

66A. Chino grama semidesert grassland- Landscape scale view of chino grama-dominated foothill grassland in the Trans-Pecos Basin and Range province. Shown here was the physiography and vast expanse of basin and range landform combined with physiogonomy of the range plant community in which the predominately cespitose chino grama is the defining keystone species. Unlike the more common and wider-distributed black grama, chino grama does not have stolons and is strictly a bunchgrass. Physiogonomy and structure of grassland communities dominated by chino grama appears steppe-like with a more open sward than semidesert grassland made up mostly of black grama, tobosa, or mixtures of several species.

Black grama was locally abundant in the range vegetation shown here, but chino grama was the sole dominant. Red threeawn, bush muhly, plains bristlegrass, and fluffgrass were also present but minor species overall. Creosotebush, Englemann pricklypear, lechuguilla, and Big Bend silverleaf (Leucophyllum minus) appeared conspicuously and infrequently.

Big Bend National Park (foothills Burrow Mesa), Brewster County, Texas. June, early estival aspect (prior to beginning of summer rainy period so that grasses were still in state formancy). FRES No. 40 (Desert Grasslands Ecosystem). No appropriate units in Kuchler (1964, in Garrison, 1977), Society for Range Managemeant (Shiflet, 1994), or Brown et al. (1998). Chihuhuan Deserts- Low Mountains and Bajadas Ecoregion, 24c.

66B. Chino grama-dominated grassland- This low hillside was spatially above a range plant community of Chihuhuan Desertscrub dominated by creosotebush and ocotillo. These two shrub species along with lechuguilla and isolated Englemann pricklypear were minor species to chino grama which was obviously in "complete control" of this foothill rangeland. Large size of individual chino grama plants suggest relatively old age and high plant vigor.

The rimrock atop the hill indicated that this land had been formed under volcanic activity.

66C. Botanical diversity on chino grama foothill semidesert grassland- An array of species, life or growth forms, and groups of range plants were present on Chihuhuan Region grassland that was dominated and defined by chino grama. Was this pristine, or was it pristine? Diverse species composition and structure were striking in this relict range vegetation dominated by chino grama.

Large, orange or yellowish-brown grass clumps in center foreground were cane bluestem. Large specimens of bush muhly were to the sides and behind cane bluestem. Black and sideoats grama and red threeawn were also major, well-represented grass species. The more common shrubs included Spanish dagger (Yucca treculeana= Y. macrocarpa), Englemann's pricklypear, cane cholla along with Acacia and Mimosa species. Sotol and lechuguilla were present but rare. Notice that this rarity contrasted with sotol-lechuguilla grasslands. Forbs were "few and far between" (essentially of no obvious importance).

This range dominance type clearly had a floristic affinity with both the mixed grass (grama-bluestem-Arizona cottontop) hillside graasland and the Chihuhuan Desert-chino grama savanna. Dominance by chino grama (versus sideoats grama) and the relatively smaller proportions of plant cover by other species distinguished this grassland cover type. This was consistent with published descriptions by local authorities and observers as cited in the aabove introduction.

Big Bend National Park(foothills, Burro Mesa), Brewster County, Texas. June, early estival aspect and prior to initiation of grass "green-up" in the current growing season. FRES No. 40 (Desert Grasslands Ecosystem). No apt units of vegetation in Kuchler (1964, in Garrison, 1977), Society for Range Management (Shiflet, 1994), or Brown et al., (1998). Chihuhuan Deserts- Low Mountains and Bajadas Ecoregion, 24c.

66D. Ultimate expression of chino grama semidesert grassland- Relict xeric foothills grassland dominated by chino grama with black and sideoats grama, red threeawn, bush muhly, and cane bluestem locally plentiful (but not dominant). Shrubs included creosotebush (most common; regional dominant of Chihuhuan and Sonoran Deserts Region), cane cholla, Englemann's pricklypear, and Spanish dagger. Occasional woody legumes (limited to Acacia and Mimosa species). No forbs "to amount ot anything".

Large specimens of chino grama suggested relatively old age (and certainly no lack of plant vigor). Chino grama is a cespitose species unlike the stoloniferous black grama that is the most common dominant of semidesert grasslands. The openness and proportion of unvegetated ground surface of this pristine chino grama grassland vegetation can be compared to the more "closed" (less bare ground) of virgin black grama grassland range shown previously.

Big Bend National Park (foothills, Burro Mesa), Brewster County, Texas. June, early estival aspect (prior to initiation of grass growth). FRES No. 40 (Desert Grasslands Ecosytem). No Kuchler (1964, in Garrison, 1977), Society for Range Management (Shiflet, 1994), or Brown et al., (1998) units applied. Chihuhuan Deserts- Low Mountains and Bajadas Ecoregion, 24c.

67. Local stand of chino grama- Appearance of cespitose chino grama or "chinograss" in a local small population. These individual plants were growing on a patch of ground that had been disturbed a few years previously. Plants were relatively small (and presumedly younger) which permitted more detailed observation of species features.

Pebbly soil surface closely resembled desert pavement. The uniform pattern of dispersion is typical of xeric environments. These plants of chino grama were dispersed uniformly as are some of the desert shrubs like ocotillo.

Chinati Mountains, Presidio County, Texas. June, and plants still dormant awaiting summer rains.

70. Racemes of chino grama- Last season's inflorescences on chino grama. Racemes of chino grama resemble those of hairy grama, but the portion of rachis that extends beyond spikelet attachment is thicker in chino grama.

Chisos Mountains (Green Gulch, Big Bend National Park), Brewster County, Texas. June.

71. Gyp(sum) grama (B. breviseta)- Sexual shoots of gyp grama growing on a gyp range site on a three-way ecotone of mixed prairie, semidesert grassland, and "island" of Chihuhuan Desert. B. breviseta and B. ramosa are very closely related. They so closely resemble each other that Silveus (1933, p. 428) and Hitchcock and Chase (1951, ps. 541-542) showed the two binomials as synonyms. Gould (1975, p. 351) followed suite and lumped them together even as he noted that plants growing on gypsum flats tended to have a different or distinct growth habit. Hitchcock and Chase (1951, p. 542) specified an edaphic environment of "gypsum sands and calcareous rocks". Since the original New Mexico Flora of Wooton and Stanley (1915, p. 87) B. brevisita and not B. ramosa have been cited for Trans Pecos New Mexico. In the equally classic, seminal, and even earlier work, Botany of Western Texas Coulter (1891-1894, p. 530) described and distinguished both B. breviseta and B. ramosa for Trans Pecos Texas. The two taxa were treated as separate species by Correll and Johnston (1979, ps. 246-247) and Powell, (2000, ps. 216-218) with the former declaring both to be present in southern New Mexico. Powell (2000, p. 216) specified that B. brevisita appeared to be an obligate glysophile (organism restricted to gypsum soils) and cited Reeder and Reeder (1980) who reported that B. breviseta was diploid whereas B. ramosa was tetraploid.

Lastly (though by no means the last word), Barkworth et al. (2003, ps. 253, 267-268) as the official--though not necessarily the correct--encyclopedic treatment of North American Gramineae, distinguished B. breviseta from B. ramosa based largely on the detailed work of Reeder and Reeder (1980). Both species were reported for Texas and northern states in Mexico, but B. ramosa was not shown for New Mexico though it was mapped in adjacent Texas counties. Barkworth et al. (2003, p. 253) specified dichotomous morphological differences as a presence vs. absence of chalky bloom on shoots and presence vs. absence of rhizomes, B. ramosa being strictly cespitose. Dichotomous habitats stated in Barkworth et al. (2003, p. 253) were limestone soils (B. ramosa) vs. gypsum soils (B. breviseta). This is perhaps one of the most distinct separations of grass species as to edaphic environment, a textbook example. These differences corresponded to differences in ploidy (tetraploid vs. diploid for limestone-derived vs. gypsum-derived soils).

Railway Ranch, Upton County, Texas. Mid-October; early stage of dormancy.

72. A needlesome grama- Needle grama (B. aristidoides) is one of three annual Bouteloua species that calls semidesert grasslands and Chihuhuan Desert its home (Gould, 1975, ps. 341, 344, 346). And yes indeed, it does closely resenble threeawn (Aristida) species. Needle grama occurs in disturbed and otherwise marginal environments from the Rio Grande Plains up through the Edwards Plateau throughout the Trans Pecos Basin and Range vegetational areas (Gould, 1975, p.341 ). The current author photographed this specimen on silver bluestem-sideoats grama mixed prairie in the High Plains vegetational area of Texas where it, the Rolling Plains, and northwestern edge of Edwards Plateau converge (Gould, 1963; Correll and Johnston, 1979, map 1).

Forage value of this like the other annual gramas varies from fair when immature to poor at maturity as seen in these photographs.

Howard County, Texas. Mid-October; grain-ripe stage and nearing end of plant life.

73. Generalized pattern of range vegetation zonation in mountains of Trans-Pecos Texas- The author's view of zones of range plant communities in the Trans-Pecos Texas Basin and Range physiography. Zonation coincided largely with elevation, but there were assuredly numerous factors at work in development of this climax vegetation that was interpreted here in terms of range cover (= dominance) types.

Ponderosa pine and oak-pinyon-juniper range types were covered in chapters of Forests and Woodland Biomes.

74. Semidesert grassland at upper elevational limit- The black grama upland form of semidesert grassland described earlier extends up the slopes of several mountain ranges in the Trans- Pecos Basin and Range province. At these higher elevations more mesic species (those more characteristic of either semiarid climates to the east and/or west of the Chihuhuan Desert) become more abundant, often becoming local dominants. This includes grasses like little and silver bluestem, sideoats and blue grama and tanglehead (Heteropogon contortus). The same pattern of plant distribution at higher elevations occurs for shrubs with Mormon tea, soaptree yucca, fourwing saltbush, and mesquite from the lower grasslands joined by mountain mahogany (Cercocarpus montanus), skunkbush sumac (Rhus trilobata), and Apache plume (Fallugia paradoxa) from the encinal or montane scrub (Dick-Peddie, 1993, ps. 123-128).

The developing thunderhead (early stage of a cumulonimbus cloud) was representative of the major source of precipitation during the summer portion of the growing season in the southern Basin and Range province. Fortunately for an area receiving only 7-10 inches of annual precipitation a high proportion of this moisture falls during the mid- to late-growing season. Most of this is delivered by local thunder showers (usually of short, but often intense, duration) in which hail may accompany rain. Some authorities interpret this summer pattern as part of the monsoonal flow characteristic of the Sonoran Desert. Thunder showers are local in nature in contrast to general or widespread precipitation from large cyclonic storms or frontal systems. Thunder showers often develop sporadically within a general high pressure system with any precipitation being formed by condensation of water that arose by evapotranspiration below. If a thunderhead builds and "lets loose" above a given locale that part of the range benefits. Otherwise it remains "high and dry" (ie. thunderstorms are a "hit and miss" affair).

75. Apache plume (Fallugia paradoxa)- Fruit and leaves of a valuable mountain browse plant. This is one of many genera and species in the Rosaceae. The rose family furnishes the largest number and the most valuable species of browse plants in North America. Apache plume is an important feed species and is tolerant of heavier browsing but it's palatability usually varies from low to moderate sometimes ranking as high as good for both cattle and sheep (Dayton, 1931, ps 50-52; Forest Service, 1940, B77). Apache plume is most important on winter range.

Lincoln County, New Mexico. June.

76. Classic example of the Upper Sonoran life zone of C. Hart Merriam in the Trans-Pecos Basin and Range— This is in the Chihuhuan Desert, but it is a transition range type from semidesert grassland to the desert shrubland to be seen below.Lechugilla (Agave lecheguilla)-Chino grama (Bouteloua ramosa)- creosotebush (Larrea tridentata= Covillea tridentata ) type. Limestone and mountains range site.

This range cover type was included here to illustrate the adjacent location of Chihuhuan Desert and semidesert (Chihuhuan) grassland and the mosaic of vegetation formed by this "patchy" occurrence of these two biomes in the greater Chihuhuan Region. This spatial arrangement is especially noticeable in Texas' Trans-Pecos Basin and Range Vegetational Area.

While this community has features of FRES Nos. 30 (Desert Shrub Ecosystem) and 40 (Desert Grasslands Ecosystem) it most closely fits FRES No. 33 (Southwestern Shrubsteppe Ecosystem) with the closest corresponding Kuchler unit being K-52 (Gramagrass-Tobosagrass Shrubsteppe[without the Tobosa]). No specific SRM cover type description; closest is variant of SRM 505 (Grama-Tobosa Shrub). Mixed Grass-Scrub Series of Brown et al. (1998). Chihuhuan Deserts- Low Mountains and Bajadas Ecoregion, 24c (Griffith et al., 2004).

77. Lechuguilla (Agave lechuguilla)- This is probably the most abundant Agave species in the Trans-Pecos Basin and Range area. In fact, lechuguilla "has one of the most extensive ranges of the agaves" (Gentry, 1982, p. 154). Lechuguilla is also one of the most easily identified (usually recognized immediately) by the high number of rosettes growing in clusters or groups. A. lechuguilla is a highly modular organism. Each of the basal rosettes is a module (a clone= ramet) of a genetic indivisual (genet). Most reproduction is asexual via suckering from rhizomes or rhizome-like structures (ie. "rootstocks"). Gentry (1982, p. 154) stated that numbers of rosettes "probably exceed those of all other native agaves". Gentry (1982, ps. 30-31) described the Agave rosette.

Incidentally this wonderful book, Agaves of Continental North America, (Gentry, 1982) is the encyclopedia and definitive reference by the man regarded as the world expert on Agave. Expensive, but well worth the price to any succulent-lover.

Lechuguilla is a poisonous range plant that has been documented to poison cattle, sheep, and goats, and--no surprise--this poisoning takes place under conditions of overgrazing. Gentry (1982, p. 157) delightfully described such mismanagement concluding that lechuguilla is "a protective agent of the range, penalizing those stockmen who, through force of circumstance or lack of foresight, decimate their resource by over-use". AMEN! The poisonous principle in lechuguilla is a saponin that causes hepatogenic or secondary photosensitization due to liver damage. References include Kingsbury (1964, ps. 56, 467-468), Sperry et al. (1964, ps. 7-8), Burrows and Tyrl 2001, 13-15), and Hart et al. (2003, 22-23).

Taxonomic treatment of Agave has been controversial. Traditionally the genus has been included in the agave family, Agavaceae, (eg. Powell, 1988), but other workers (eg. Smith, 1977) placed Agave in the lily family, Liliaceae, as an agave subfamily, Agavoideae. Incidentally, is there anything that is not controversial when it comes to plant taxonomic treatments?

Hudspeth County, Texas. June.

78. Basal rosettes of lechuguilla- Example of a cluster of rosettes, asexual modules of a genetic individual of lechuguilla. It was explained in the immediately preceding photo-caption that lechuguilla probably produces more rosettes than any other Agave species in North America. Recall from that explantion that each such rosette is a clone (= module= ramet) of the original "parent plant" which is the genet. Gentry (1982, p. 30) explained that each rosette of A. lechuguilla is a "monocarpic rosette". Each rosette flowers only once in its life (the life of that clone or module) and then dies; in fact, it begins to die as soon as the one-time flower stalk with its inflorescence begins to emerge and elongate. This condition was obvious in the flowering rosette in the cluster shown in this photograph. The inflorescence on this particular stalk was presented in the immediately succeding photograph.

With this pattern of life cycle and resource allocation, each genetic individual (the genet or actual unique plant) of lechuguilla with its ramets (rosettes) is a "multiannual" (Gentry, 1982, p. 30). Gentry's choice of terms could confuse the beginning student. Yes, the individual rosette flowers only once in its life and then promptly dies. The flowering-seed production process takes only a few weeks as flower stalk growth is extremely rapid (perhaps over a foot a day) such that this phenological development is "annual" (more like "ephemeral"). Yet it takes years (perhaps a quarter century) of reserve food storage, rosette growth, and formation of stalk primordial tissue development before flowering can be initiated. In reality each rosette is a long-lived perennial that finally flowers and then summarily dies. With asexual (vegetative or clonal) reproduction the genetically unique plant has a life span that is seemingly "endless" or "forever". The only thing annual about an Agave species is the amazing flowering and fruit production phenomenon. And that is phenomenonal.

81. Alkali sacaton-shrub savanna- Alkali sacaton was the dominant species on this saline swale, but it "kept company" with numerous shrubs including whitethorn acacia, fourwing saltbush (Atriplex canescens), lotebush (Ziziphus obtusifolia), honey mesquite, and mariola. The most common forb was the Eurasian annual weed, tumbleweed or Russian thistle (Salsola kali-tenuifolia= S. iberica= S. pesitfer= S. tragus). Other grasses included inland saltgrass (Distichlis spicata var. stricta= D. stricta), tobosagrass, cane bluestem, and naturalized Johnsongrass (Sorghum halepense).

It appeared that grass cover was considerably less and woody plant cover more than in the natural plant community. In fact this bottomland range was closely located to a mining and early settlement site so that all matter of human disturbance could have impacted this plant community. It is also probble that some, if not all, of these shrub species had been present at lower cover percentages in the virgin range vegetation. Introduced species like Russian thistle and Johnsongrass were not components of pre-Columbian vegetation, but must be assumed to be new components of the potential natural vegetation for an indefinite period of time.

Presidio County, Texas. June: warm-season perennial grass species were still in dormancy just bvefore typical onset of summer rains. FRES No 40 (Desert Grasslands Ecosystem). Natural vegetation of this small spatial dimension was not mapped by Ku;chler (1964, in Garrison et al., 1977). Variant of SRM 701 (Alkali Sacaton-Tobosagrass). Brown et al. (1998) erroneously omitted the Sacaton Series from Chihuhuan (Semidesert) Grassland under Warm Temperate Grassland, but this is what this range plant community was. Maybe there will be a second, revised, and expanded edition of Brown et al. (1998) that will include such a Sacaton series. Chihunuan Deserts- Low Mountains and Bajadas Ecoregion, 24c (Griffith et al., 2004). Degraded Salty Bottomland range site.

82. Detail of range vegetation on a local saline swale- Alkali sacaton was the dominant herbaceous species. Other grass species included inland saltgrass, tobosagrass, cane bluestem, Johnsongrass. Shrubs included lotebush (branches at extreme right margin) and (not visible in this photogrph) whitethorn acacia, mesquite, fourwing saltbush and mariola. The most common forb was Russian thistle or tumbleweed, a specimen of which was is left foreground.

Traces of encrusted salt on the soil surface were visible. It was likely that human-induced disturbances associated with local mining and settlement result in loss of grass cover and increased cover and density of shrubs, but there was no range reference area for comparison. Range vegetation shown here was still an example of one form of saline, alkaline, and/or gypsous soils on low-lying range sites generically referred to as "swales", "flats", or "sinks".

Presidio County, Texas. June: still dormant season for summer-growing perennial grasses. FRES No. 40 (Desert Grasslands Ecosystem). Smaller units of range vegetation such as this were not mapped by Kuchler (1964, in Garrison et al., 1977). Variant of SRM 701 (Alkali Sacaton-Tobosagrass). Sacaton Series under Chihuhuan (Semidesert) Grrassland 143.1 of Warm Temperate Grassland 143 if and when one is listed under a revision or expanded edition of Brown et al. (1998); an error that Sacaton Series was not so listed. Chihuhuan Deseerts- Low Mountains and Bajadas Ecoregion, 24c (Griffith et al., 2004). Salty Bottomland range site in some degree of range depletion.

87.   Alkali sacaton flat- Throughout the southwestern deserts and back into the southern mixed prairie there are some alluvial sites, especially swales and flood plains and (even more) saline or alkaline flood plains, that are dominated by alkali sacaton (Sporobolus airoides). While this species is widely distributed ranging as far as the tallgrass prairie region (eg. in Missouri) and up in the northern Great Plains of Montana, it is on overflow sites in the southern part of the arid and semiarid regions where alkali sacaton forms essentially single species stands (= consociations) which are priceless as natural sources of livestock forage, wildlife habitat, and flood control.

These “sacaton flats” or “sacaton swales” are distinct from and more widely occurring than those formed by giant and Wright’s sacaton which were shown immediately above. Also, alkali sacaton occurs in association and often as a co-dominant with other bottomland or flood plain species as for example in SRM Rangeland Cover Types 701, 702, 712, and 725. Interestingly (strangely to this author), the Society for Range Management in it’s publication on cover types (Shiflet, 1994) did not recognize the widely distributed “pure form” or “pure stand” of alkali sacaton flat. The closest would likely be a variant of SRM 701. Likewise, this range type is entirely too restricted to have been described separately as a FRES unit and too small to have been mapped by Kuchler. Sacaton flats (those of any or all such Sporobolus-dominated alluvial sites) are local range types under both FRES No. 38 (Plains Grassland Ecosystem) and No. 40 (Desert Grasslands Ecosystems), and perhaps as localized grasslands within FRES No. 30 (Desert Shrub Ecosystem).

Dick-Peddie (1993, p. 105) noted presence of swales co-dominated by tobosagrass and alkali sacaton (SRM 701) which he placed under the “Tobosa Series (swales)”, but he also showed a S. airoides-dominated flat which he presented as “Sacaton Series (swales)” (Dick-Peddie, 1993, p. 110). Alkali sacaton swales composed exclusively of this one species is a well-known range type and this cover type should have been included in the SRM rangeland cover type manual (Shiflet, 1994). 

“Proof-positive” of the existence of S. airoides consociations that form alkali sacaton flats is presented here in this photograph of a flood plain in the Texas Trans Pecos Basin and Range Region. Reeves County, Texas. October. Chihuhuan Deserts- Chihuhuan Desert Grasslands Ecoregion, 24b (Griffith et al., 2004).

89.  Border of an alkali sacaton swale- Along the perimeter of the alkali sacaton flat seen above, Sporbolus airoides was joined by two (two other) halophytes:  pickleweed (Allenfrofea occidentalis), center-most and largest plant, and glasswort or Utah samphire (Salicornia utahensis), left foreground. 

Reeves County, Texas. October.

93.   Alkali sacaton (Sporobolus airoides)- Alkali sacaton is one of the largest and most productive of the dropseed species. It often dominates alluvial or flood plain sites from the southern Great Plains into the Chihuhuan and Sonoran deserts where fertile soils and relatively high soil moisture form an edaphic environment that is postclimax to the regional vegetation. These soils are often higher in dissolved minerals thereby being saline or alkaline habitats. As indicated by its common name, alkali sacaton is well-adapted to these conditions and able to exploit the favorable soil water content. The specimen seen here is stunted by a combination of a severe drought and overutilization by cattle, but the otherwise healthy, vigorous condition of this depauperate specimen attest to the species adaptation to both conditions. (Trade secret for beginning plant photographers: diminutive specimens frequently— even typically —make the best shots of non-photogenic species like most grasses because both basal portions and flower clusters are easier to bring into crisp focus.)

Reeves County, Texas. October.

Vine mesquite (Panicum obtusum) "flat", swag, or swale- Swales and flood plains make up some of the more restrictive-- yet distinctive and productive-- groups of semidesert grasslands. This is actually a category or general grouping of several forms or kinds of "desert plains grasslands". The tobosagrass-dominated semidesert grasslands (= "tobosa flats" or "tobosa swales") comprise the most extensive of these (it was appropriately dealt with first in above coverage). "Sacaton flats" is another category of generally mesic-habitat semidesert grasslands determined and identified according to the one or two dominant Sporobolus species (and sometimes including inland or desert saltgrass and halophytes like pickleweed or fourwing saltbush). This form of swale semidesert grassland was treated immediately above.

A third range cover (= dominance) type defined by the swale, swag, flat, or flood plain topography and identified as to dominant species is that of vine mesquite.A vine mesquite consociation is, alternatively, SRM 725 (Vine Mesquite-Alkali-Sacaton), a variant of SRM 725, or a separate Vine Mesquite rangeland cover type. The latter alterntive was consistent with this author's interpretation of SRM 505 (Grama-Tobosa-Shrub) as too general and, thus, consequent separation of it into a Black Grama rangeland cover type and a Tobosagrass rangeland cover type.

97. Vine mesquite flat or swale- A consociation (Clementsian term for single-species stand) of vine mesquite developed on a mesic bottomland lying between upland sand ridges vegetated by depleted black grama semidesert grassland range. The rangeland in this and the next three photographs was on the former El Camino Real, the "Kings Road" or Royal Road" from Santa Fe to Chihuhua City. This range vegetation had somehow persisted or recovered from impact of trade and travel over almost 300 years, but upland semidesert grassland range was Chihuhuan Desertscrub disclimax. What, if any, influence upland range plant communities had on the grassland on this "flat" was not known. What was known is that vine mesquite-dominated grassland had persisted for decades under cattle grazing (background, and next slide) and protection from large animal grazing.

This vine mesquite semidesert grassland was intepreted as climax vegetation and an example of a range reference area. For all practical purposes vine mesquite was the only plant species on this range. Tobosagrass was plentiful on lower elevation and depression microsites to the right of camera range (not in photograph).

Vine mesquite in foreground was in a livestock exclosure. For decades this range vegetation had not been grazed by livestock (native ruminants could possibly have grazed inside exclosure; rodents and lagomorphs most certainly did consume some of the exclosue herbage).

New Mexico State University College Ranch, Dona Ana County, New Mexico. June, early estival aspect and pre-bloom phenological stage.FRES No. 40 (Desert Grasslands Ecosystem). No appropriate units of vegetation by Kuchler (1964, in Garrison et al., 1977) or Brown et al.(1998). Vine Mesquite variant of SRM 725 (Vine Mesquite-Alkali Sacaton). Chihuhuan Deserts- Chihuhuan Basins and Playas Ecoregion, 24b (Omernik and Griffith, 2006).

98. The difference is utilization- Fenceline difference in degree of use. Vegetation on right had been excluded from livestock entry for decades (approximately 70 years); range vegetation on left had been subjected to cattle grazing for an even longer period of time. While the difference is a stark contrast, it is a difference primarily in degree of use and not of plant species composition. Plant communities on both sides of the exclosure fence were single-species stands of vine mesquite. This should be obvious to careful viewers.

First Cardinal Principle of Range Management: Proper Degree of Use.

Even though utilization (= degree of use) seemed to be of some heavy (even suggestive of excessive) grazing use vine mesquite had persisted under this approximate grazing regime (seasonal suitability grazing at what was interpreted by on-the-graund rangemen as "moderate" rate) for many years.

Third Cardinal Principle of Range Management: Proper Season of Use.

It is likely that the greater soil moisture of this swale permitted heavy-- at least, heavier--utilization and continued survival of and comparatively high herbage yields by vine mesquite.

Brangus cattle of the New Mexico State University purebred herd. Brangus is one of the best-adapted beef breeds for semidesert grassland range due to superior rustling ability includning covering rough country and traveling relatively long distances to water. Exceptional maternal features (mothering and milking ability) is another strong point of Brangus.

Fourth Cardinal Principle of Range Management: Kind and Class of Range Animal. This includes selection of superior breeds within the proper species (Bos taurus X B. indicus cross in this instnce) and the proper sex, age, or market class (cows and calves in this case).

New Mexico State University College Ranch, Dona Ana County, New Mexico. June, early estival aspect and pre-bloom stage. FRES No. 40 (Desert Grasslands Ecosystem). Kuchler (1964, in Garrisonet al., 1977) did not map units of vegetation at spatial scale of this range cover type. Nor descriptive biotic community in Brown et al. (1998). Vine mesquite variant of SRM 725 (Vine Mesquite-Alkali Sacaton).

99. Botanical compostion and sward of vine mesquite flat- Species make-up and local physical appearance of vine mesquite consociation on a swale or mesic flat on semidesert (Chihuhuan) grassland. The only perennial plant of any abundance was vine mesquite. The most obvious perennial forb was one of the narrowleafed rosin or horsetail milkweed Asclepias subverticillata). The two species of dead plants conspicuous in the second slide were both winter annuals: tumble mustard (Sisymbrium altissimum) and an unidentifiable composite.

Noteworthy was the development of a closed canopy sward-- a turf-- of essentially 100% foliar cover in an area receiving an average of eight inches in annual precipitation. Grains on the fruiting shoot tips of vine mesquite were visible (at least at regular projection of the 35 mm slide).

New Mexico State University College Ranch, Dona Ana County, New Mexico. June, early estival aspece and pre-bloom phenological stage. FRES No. 40 (Desert Grasslands Ecosystem). No vegetation units in Kuchler (1964, in Garrison, 1977) who used larger mapping units than this. Vine mesquite variant of SRM 725 (Vine Mesquite-Alkali Sacaton). No appropriate biotic community in Brown et al. (1998): vegetation classification and mapping remains a "work in progress".

100. Stand of vine mesquite at peak standing crop- A luxurious verdent turf of vine mesquite like this one is seldom produced in the semidesert grassland, and this was no exception. This stand grew in the Western Cross Timbers of northcentral Texas, but it did illustrate appearance of vine mesquite under optimum growing conditions and the wide range of habitats, including precipitation zones, under which this panicoid species can achieve dominance.

Tarleton State Univrsity Hunewell Ranch, Erath County, Texas. July.

101. Horsetail or rosin milkweed (Asclepias subverticillata)- This narrow-leafed species of milkweed is one of the definitively proven toxic species of Asclepias. Specific references to A. subverticillata include Kingsbury (1964, p.267-268), Schmutz et al. (1968. ps. 24-25), Stephens (1980, p. 85), Burrows and Tyrl (2001, ps. 126-135 passim), and Hart et al. (2003, p. 38-39). Exact toxin in A. subverticillata remains uncertain(Burrows and Tyrl (2001, p. 132-133). From veterinary medicine standpoint the definitive authority is The Merck Veterinary Manual, but even the 50th Anniversary Edition (Kahn, 2005).

This specimen was growing in the vegetation of the vine mesquite swale inside an exclosure on New Mexico State University College Ranch. June.

It was noted in the introduction to semidesert grasslands that relatively few students of rangelands, native vegetation, etc. have realized (or recognized) that there are semidesert grasslands in the Great Basin. For example, neither Schmutz et al. (in Coupland, 1992, pgs. 337-362) nor McClaran and Van Devender (1995) in the two closest to comprehensive treatments of semidesert grasslands so much as acknowledged (let alone described) Great Basin desert grasslands. Such native semidesert grasslands do indeed exist (see below).

Intuitively this omission would appear to be a result of the rather restricted distribution-- hence limited economic and cultural importance-- of this generally arid grassland(s) with its various cover types or subtypes. As a indicator of this Welsh et a. (1993, p.7) in A Utah Flora commented: "Grasslands, those areas dominated mainly by grass species, occur widely in the state, but the total area occupied by them, when compared to that dominated by woody and herbaceous species together, is not large". These workers then described plant communities including various desert scub, chaparral, pinyon-juniper woodland, forests, alpine, and even hanging gardens but no grassland! (Obviously grasslands occupy far greater "total area" and are of more survival importance to man than hanging gardens.) Similarly (and perhaps setting a precident) in the classic Flora of Utah and Nevada Shantz (in Tidestrom, 1925. 15) did not include grasslands as plant communities other than as "alpine grassland", which strictly speaking is not grassland. In a description of the natural vegetation of Nevada Tueller (1975, p. 24) observed: "Within the Great Basin there are few natural grasslands, i.e., stands of vegetation that have no overstory of shrubs and/or trees and only a few forbs intermingled". Tueller (1975, p. 25) provided very short but poigant descriptions of some of these natural grasslands. Tueller's comments on Nevada grasslands were used to good advantage in some of the following presentations of Great Basin semidesert grasslands.

A major problem from standpoint of natural vegetation classifiction and rational organization for treatment of range cover types is whether to interpret certain native plant communities as grassland, shrubland (ie. Great Basin desertscrub), shrub steppe, or a shrub savanna. In regards some kinds of range vegetation consisting of shrubs and grass there is a continuum from what is obviously grassland with but very sparse cover or limited density (say, trace amounts) of woody plants to desert scrubland with little or no herbaceous species (including grasses). A textbook example of this phenomenon involves Palouse Prairie cespitose grassland, sagebrush (Artemisia spp.) shrub steppe, and sagebrush desert with little understorey. This array exists mostly to the north of the Great Basin. However, parallel conditions do exist widely in the Great Basin with the sagebrush- and saltbush/greasewood (Atriplex/Sarcobatus)-defined communities as well as those natural plant communities involving such shrubs as Mormon tea or jointfir (Ephedra spp.). Range plant communities that pose the real problems in this regard are not those dominated by these shrub species because they are obviously arid shrublands (= deserts). Nor is there much problem with recognized shrub steppe like the sagebrush-bunchgrass steppe that occurs in northern portions of the Great Basin. This latter range vegetation was logically placed in its singular category of savanna-like grassland. It was treated as such in this publication. Students of Great Basin vegetation have traditionally distinguished between sagebrush shrub steppe and Great Basin sagebrush desert (eg. West in Barbour and Billings, 1988, p. 212-217).

The "tough calls" are those range types that fall between what is obviously grassland and shrub steppe (a shrub-grass savanna). To a lesser degree this condition also applies to Chihuhuan and Sonoran semidesert grasslands. For example, large individual plants-- and at different densities and cover-- of Ephedra and Yucca species are part of the virgin grassland vegetation. It was noted above, that Dick-Peddie (1993, p.116) used longleaf Mormon tea or longleaf jointfir (Ephedra trifurca) as a "diagnostic (climax) member of Desert Grassland". A similar and, probably, homologous phenomenon obtains in Great Basin vegetation with Nevada Mormon tea or Nevada jointfir (E. nevadaensis) and grasses like Indian ricegrass (Oryzopsis hymenoides= Stipa hymenoides), galleta (Hilaria jamesii), and needle-and-thread (Stipa comata). The ubiquous presence of Nevada jointfir and widespread dominance of galleta when compared to the same phytoecological conditions of Ephedra and tobosagrass (Hilaria mutica) demonstrated beyond doubt to this author the homology and ecological affinity or relatedness of Great Basin semidesert grasslands to those of the Chihuhuan semidesert grassland ranges.

This perspective was apparently not shared by those in the Nevada Natural Heritage Program (2003) who recognized a Nevada jointfir/Indian ricegrass community as shrubland and not grassland. There again, the National Vegetation Classification for Nevada (Nevada Natural Heritage Program, 2003) did not ever use the designation of grassland even for what are obviously grasslands but instead applied the title of "herbaceous alliance" or "herbaceous vegetation". Likewise, Brown et al. (1998, p. 40) recognized a Chihuhuan semidesert grassland, but they did not recognize a Great Basin semidesert grassland and instead designated this range vegetation as Great Basin Shrub-Grassland (142.2) within Cold Temperate Grassland (142.). West (in West, 1983, p. 415) described one of the semidesert grassland range cover types given below and remarked that "... the type is locally known as 'semi-desert grassland'". "This rangeman author knows grassland when he sees it. The reader of this publication can decide for himself, but please view the range vegetation first.

.Great Basin grasslands existing as prairie or steppe in foothills and mountains, meadows (including those dominated by grasslike plants such as wet or flood meadows), alpine (eg. alpine turf), and riparian zones (eg. Elymus cinerus, basin wildrye, stands) are not semiarid grasslands, even though they occur within the arid zone or desert climate. With such exceptional range vegetation local habitats and marid ecological factors (eg. edaphic, topographic, fluvial features) override regional climate. Therefore, these grassland range types were not included in this chapter but rather in those that were a "better fit" (eg. Great Basin sagebrush shrub steppe was included in that chapter, wet meadows or grass/grasslike plant-dominated riparian vegetation in the meadows chapter).

Some examples of Great Basin range types that naturally fell into a "gray zone" and could arbitrarily be interpreted variously as either shrub steppe, grass-shrub savanna, or natural grassland widely scattered diagnostic (climax) shrubs were included in several chapters. For example, two rangeland types with swards (more-or-less continuous herbaceous cover) of needle-and-thread (Stipa comata) and Indian ricegrass with well-dispersed Nevada jointfir could logically be viewed as either semidesert grassland, bunchgrass shrub steppe, or less xeric forms of Great Basin Desert (arid or, maybe, semiarid shrubland). These two range types that were obviously--at least to this rangeman photographer-- climax vegetation were placed in both the semidesert grassland and the shrub steppe chapters herein with explanatory notes.

Given the arbitrary categorization of some kinds of climax range vegetation in which substantial cover of woody plants (eg. natural communities dominated by grasses but having conspicuous and greater-than-scattered cover of shrubs like Ephedra, Artemisia, or Chrysothamnus species) some examples of range types were included under both semidesert grassland and Great Basin Desert designations (in both of these chapters). This was done for convenience of readers and to acknowledge the uncertain interpretation of climax range plant communities that apparently have shared dominance by grass and shrub species (at least aspect dominance in regards to shrub cover).

103. Grassland at its grandest grandure- Probably the ultimate expression of Great Basin semidesert grassland is the Indian ricegrass rangeland cover type which is a consociation of this festucoid grass species. (This range resembled a seeded monoculture pasture as much as what was essentially a natural single-species stand.) The grassland range vegetation presented here was not only virgin it was pristine ("mint condition"), and it had been grazed by cattle as part of a Bureau of Land Management allotment! As was shown in these photographs, Indian ricegrass grassland often has Douglas or viscid rabbitbrush (Chrysothamnus viscidflorus) as the major associate shrub species with widely scattered individuals of winterfat (Eurotia lanata= Ceretoides lanata). Total cover and density of woody plants was negligible however. Noone knowledgable of natural vegetation would doubt that this was semidesert grassland in its "purest" textbook expression. Nor would any with a shread of a rangeman's soul doubt the beauty of this magnificant grassland range ecosystem and landscape.

Tueller (1975, ps. 23-24) remarked that one of the few types of natural grassland in Nevada were those consisting of "pure stands of Oryzopsis hymenoides" that developed at edges of salt desert scrub in low-lying areas. Unfortunately (and erroneously based on this author-photographer's experience and the observation of Tueller [1975, p. 25]) this climax range plant community was not listed in either the Preliminary Vegetation Classification for Utah and Nevada (Bourgeron et al., 1994) or the National Vegetation Classification for Nevada (Nevada Natural Heritage Program, 2003) but Association for Biodiversity Information (2001), which included nevada and Utah, named an Achnatherum hymenoides Herbaceous Alliance. (Apparently "herbaceous alliance" meant grassland.) The Society for Range Management (Shiflet, 1994) did not offer a name and description for an Indian Ricegrass cover type for the Great Basin.

The largest space opf bare soil in the first (upper) of these two photographs (right-center foreground) was entrance to den of Ord's kangaroo rat (Dipodomys ordii), the dominant and defining animal of this and some related Great Basin scrubland communities (see the Great Basin Salt Desert Shrubland portion of the Great Basin Desert chapter herein).

Bureau of Land Management, Fillmore Field Office, Millard County, Utah. June (early estival aspect; phenology from soft-dough to grain-ripe stage). FRES No. 40 (Desert Grasslands Grassland Ecosystem). No Kuchler designation for the climax grassland of this spatial scale. No SRM. No association-level designation by Brown et al. (1998). Central Basin and Range- Shadscale-Dominated Saline Basins Ecoregion, 13b (Woods et al., 2001).

Summary Note: Three distinct range plant communities were presented below. Based on published designations and descriptions of these (or very similar) natural plant communities the three kinds of range communities were interpreted as climax (potential natural) range vegetation. In these three kinds or types (subtypes, perhaps) of climax range vegetation the only vascular plant species that was a consistently present and usually as the dominant (at community wide scale) was Nevada Mormon tea, Nevada ephedra, or Nevada jointfir. Winterfat was the dominant shrub and Nevada ephedra the associate shrub on one of these rangeland cover types. The main two native herbaceous plant species that were consistently present as major--though not always as dominant-- species were the sod-forming grass, galleta, and the cespitose bunchgrass, Indian ricegrass. A grass common to and widespread on both Chihuhuan-Sonoran and Great Basin semidesert grasslands was the highly variable, perennial threeawn of the Aristida purpurea complex. Alkali sacaton was another grass that occurred (and inconsistently so) on semidesert grasslands in both of these diverse areas of the Basin and Range province. A major difference between these two regional semidesert grasslands was absence of a major gramagrass (Bouteloua sp.) on Great Basin types. Cheatgrass or downy brome, the ubiquitious naturalized Eurasian annual grass, was always "represented" on climax Great Basin grasslands, but it was far less common or abundant than any of the native perennial grass species. Cheatgrass was usually limited to extra-wet microhabitats such as deep cow tracks, entrances to rodent burrows, or, as shown below, at edges of harvester ant mounds.

An arborescent member of the Liliaceae that would be an approximate ecological equivalent of soaptree yucca was absent from the Great Basin semidesert grasslands. This was a conspicuous dissimilarity among the two general regional types.

The "everywhere" presence of galleta and, especially, Nevada jointfir in this Great Basin range vegetation corresponded to widespread dominance of tobosagrass on swales and omnipresence of longleaf jointfir or Mormon tea (E. trifurca) in climax vegetation of the Chihuhuan Region. The seemingly ecological equivalents (natural or climax dominant species over large expanses of natural range vegetation dominanted by grasses) of two species of Hilaria and two species of Ephedra constituted conclusive proof to this author of the close floristic relatedness of semidesert grasslands within these two vast parts of the Basin and Range physiographic province. It was emphasized previously that Dick-Peddie (1993, p. 116) interpreted longleaf jointfir as a diagnostic climax plant species of Chihuhuan semidesert grassland. Herein Rosiere extended the conclusion of Dick-Peddie (1993, p. 116) to Nevada jointfir on Great Basin range vegetation that was clearly dominated by grasses. Even though there were other climax shrub species in these kinds of natural herbaceous vegetation, Neveda Mormon tea was the only shrub that was present consistently (even if in small proportions) in these grass-dominated range plant communities.

Based on these diagnostic botanical characteristics Rosiere in this publication interpreted the Great Basin climax range vegetation presented below as semidesert grassland and not as shrub steppe. Logically some samples of these range plant communities could be seen and described as shrub savannahs (grassland with scattered shrubs), but with woody plant cover substantially less than that of such widely accepted shrub steppes as sagebrush-bunchgrass steppe. In eyes and mind of this author the physiogonomy and plant community structure were substantially different between, say, steppe co-dominated by bunchgrasses and sagebrush and the range vegetation shown below where stands of grass infrequently had an individual plant of a climax shrub species. General outward appearance, structure, and apparent species composition (relative amounts or proportions of cover or biomass) of herbaceous vegetation presented below and that of unquestioned semidesert grassland in southern parts of Basin and Range shown above undeniably portrayed the close similarity of these two regional grasslands.

Nonetheless, such conclusions unavoidably involved some value-judgments and therefore were arbitrary, at least to a point. Once more the individual reader can draw his own conclusions.

104. Bunchgrasses and sod-forming grasses in the semidesert- Mixed-grass steppe of galleta (Hilaria jamesii), alkali sacaton (Sporobolus airoides), various purple threeawns (Aristida purpurea complex), and Indian ricegrass developed across this far-flung basin with its large arroyos. This rangeland was on the upper border of and drained into the dry bed of Sevier Lake with the Beaver Mountains in the background providing a textbook example of Basin and Range topography. Range vegetation was a semidesert steppe consisting of both cespitose and sod-forming eragrostoide and festucoid grasses with widely scattered plants of Nevada Mormon tea, Nevada ephedra, or Nevada jointfir (Ephedra nevadensis).

Scarcity (almost absence) of cheatgrass (Bromus tectorum) and Eurasian crucifers (eg. Sisymbrium altissimum, Descurainia pinnata) "spoke volumes" as to climax or high seral status of this arid grassland range that was grazed by livestock as well as wildlife as part of a Federal allotment.

This semidesert grassland range was an example of "galleta-threeawn shrub steppe" described by West (in Barbour and Billings, 1988, ps.222-223, in West, 1983, ps. 413-421). The galleta-threeawn steppe range type covers larger areas east of this location such that this range was an "island example". Local dominance of alkali sacaton and presence of desert saltgrass (Distichlis stricta= D. spicata var. stricta) at lower elevation in background was consistent with a general description of Intermountain (Nevada) grassland by Tueller (1975, p. ). Presence (very sparse cover) of Nevada Mormon tea was discusssed in a Summary Note immediately above.

The combination of sod-forming clones of galleta and clumps of tillers of alkali sacaton, threeawn, and Indian ricegrass was obvious, especially in the first of these two photographs.

Bureau of Land Managaement, Fillmore, Field Office, Millard County, Utah. June, early estival aspect with phenological stages of grasses varying from dormance to "seed-ripe". FRES No. 40 (Desert Grasslands Grassland Ecosystem). K- 51 (Galleta-Threeawn Shrubsteppe). No SRM or, alternatively, Great Basin variant of SRM 712 (Galleta- Alkali Sacaton). No appropriate designation in Brown et al. (1998). Pleuraphis jamesii Herbaceous Vegetation (Nevada Natural Heritage Program, 26 September, 2003). Central Basin and Range- Shadscale-dominated Saline Basins Ecoregion, 13 b (Woods et al., 2001).

105. Mixed-grass semidesert grassland- Not exactly "a land flowing in milk and honey" but on this upland immediately above dry lakebed of Sevier Lake galleta, alkali sacaton, purple threeawn, Indian ricegrass, and Nevada jointfir formed a climax semidesert steppe. Physiogonomy and, to some extent, species composition of this range type was presented in this landscape perspective.

Desert pavement-like condition of land surface was shown in this photograph as well as the two preceding and three succeeding slides. In layman's terms this is a "hardland" or "tightland" habitat in contrast to sandy soils (soil texture classes with greater proportions of sand) of semidesert grassland range types shown below.

Millard County, Utah. Bureau of Land Management, Fillmore Field Office. June, early estival aspect with most grasses at "seed-ripe" stage and entering dormancy. FRES No 40 (Desert Grasslands Grassland Ecosystem). K-51 (Galleta-Threeawn Shrubsteppe). No SRM or, alternatively, Great Basin variant of SRM 712 (Galleta-Alkaki Sacaton). No appropriate designation in Brown et al. (1998). Pleuraphis jahesii Herbaceous Vegetation (Nevada Natural Heritage Program, 26 September, 2003). Central Basin and Range- Shadescale-dominated Saline Basins Ecoregion 13 b (Woods et al., 2001).

106. Stand of stand-alone threeawn- Individual plants (genetypes or genetically unique individuals) of purple threeawn formed a local stand of this prominently cespitose eragrostoid genus. Unlike the sod-forming dominant galleta, this and associated alkali sacaton and Indian ricegrass have asexual reproduction exclusively by tillers (vertical or intravaginated shoots).

Millard County, Utah. Bureau of Land Management, Fillmore Field Office. June, grain-ripe stage of phenology.

107. Purple threeawn- The Aristida purpurea complex in portions of the Great Basin includes threeawns that have at one time or another been interpreted as A. fendleriana, A. glauca, A. longiseta, and A. wrightii (Welch et al., 1993, p. 700). The same situation obtains for other range regions covered within the current publication as for example in the semidesert grasslands and deserts of the Chihuhuan and, to some degree, Sonoran Region. This was seen as some evidence of the affinity of semidesert grasslands in all of these regions, but it must also be remembered that the A. purpurea complex is well represented in range types of humid and subhumid climates such as tallgrass and mixed prairie grasslands.

Under more favorable environmental conditions A. purpurea is clearly a seral species (typically interpreted as an invader) whereas on semidesert grasslands of the Great Basin this taxonomic complex has been regarded as a defining dominant or associate member of the climax vegetation (eg. Kuchler [1964] map of potential natural vegetation). Even as a dominant of the climax range vegetation A. purpurea increases with heavier grazing (West in West, 1983, p.415) which is a response characteristic of an increaser or invader. Under almost all of these varied range environments A. purpurea is highly unpalatable to almost all kinds and classes of range animal. Hence, one descriptive common name is "no-eatum" (Welch et al., 1993, p. 790).

Millard County, Utah. Bureau of Land Management, Fillmore Field Offfice. June, hard-grain stage ("seed-ripe").

108. Bald in the middle- Like the narrow ring of turf around the slick dome of a bald-headed rangeman, alkali sacaton growing on semidesert grassland ranges in the Great Basin characteristically develops into large tufted plants that die out in the center (Welch et al., 1993, p. 872). Eventually the bare area of soil inside a large, ring-forming alkali sacaton plant may become a microsite or microhabitat more favorable for colonization by other plants. Thereupon, individuals of other species like galleta, Indian ricegrass, or cheatgrass invade and establish on the spot of land protected by the natural shelter of encircling sacaton tillers. The aging process of "going bald" was presented in the first photograph. Colonization by galleta and cheatgrass of the "bald knob" surrounded by sacton shoots was shown in the second photograph.

This development in the range plant community is an example of the facilitation process of plant succession (Connell and Slatyer, 1977) which, of course, is nought but an updated version (Barbour et al., 1999, p. 297) of the Clementsian model of ecesis and aggregation of invading plants due to improved habitat (growing conditions) following reaction brought about by plants previously occupying the sere (Weaver and Clements, 1938). This is an illustration of what Clements meant by "dynamic ecology".

Millard County, Utah. Bureau of Land Management, Fillmore Field Office. June, early estival aspect.

109. Bearing fruit under severe conditions- Panicle of alkali sacaton on an upland site overlooking the dry bed of Sevier Lake. Successful opportunist that it is, alkali sacaton was "hedging its bets". Rather than relying solely on asexual reproduction afforded by tillering this semidesert parent plant had also completed sexual reproduction by fruit production (which in Sporobolus species is an achene rather than a caryopsis).

Millard County, Utah. Bureau of Land Management, Fillmore Field Office. June, hard grain phenological stage.

110. Twists and turns of herbivory. Harvester ants facilitate; cheatgrass invades in response to reaction- A good headline can tell most of the story, and this incident is self-explanatory for rangemen who can "read sign". For green horns and tenderfeet (for whom much of this publication was written) a mound constructed by western harvester ants (Pogonomyrmex ) served as a "water-harvesting and collection system" with water that spilled off of the earthen cone and collected at its base serving as a favorable microenvironment for establishment and completion of the short life cycle of the invading Eurasian annual festucoid grass, cheatgrass or downy brome. This is the successional process of facilitation or reaction shown and described in the photographs and caption before last.

Millard County, Utah. Bureau of Land Management, Fillmore Field Office. June. Grain-ripe/grain-shatter (annual life cycle completed).

111. Semidesert (semi-?) grassland- An Indian ricegrass herbaceous consociation with shrub cover and/or density "enough to count". Presented here were "photo-plots" from two ranges having climax vegetation dominated by Indian ricegrass with galleta, alkali sacaton, and purple threeawn as minor grasses (present at small or even trace quantities) and having winterfat, big sagebrush, rubber rabbitbrush, and Douglas or viscid rabbitbrush as local woody associate species. Nevada jointfir or Nevada Mormon tea was always present overall, as a community associate and, locally, as a co-dominant with Indian ricegrass. The second-most widespread shrub on these two example ranges was winterfat, which was the major shrub on the next cover type of Great Basin semidesert grassland presented (shown below).

It was explained in the Summary Note above that with the species composition and structure of this climax vegetation the range plant community could be viewed alternatively (and arbitrarily) as either grassland or grass-shrub savanna. It was also explained, however, why this range dominance type was described herein as semidesert grassland and not shrub-steppe.

The Indian ricegrass-Nevada jointfir range vegetation shown here was classified as an association by the National Vegetation Classification System for Nevada (in close proximity to location of these examples) as was shown shortly. Physiogonomy and structure of this range community was strikingly similar to the black grama-Torrey jointfir semidesert grassland found in Trans Pecos and Rio Grande portions of the Chihuhuan Region, but a liliaceous shrub was lacking from this and all other observed types of Great Basin semidesert grassland.

Millard County, Utah. Bureau of Land Management, Fillmore Field Field Office. June, early estival aspect. FRES No. 40 (Desert Grassland Grassland ecosystem). No Kuchler designation. No SRM cover type. The Ricegrass Series (142.23) of Great Basin Shrub-Grassland (142.2), Cold Temperate Grassland (142.) in Brown et al (1998, p. 40).Ephedra nevadensis/ Oryzopsis hymenoides plant association (Bourgeron et al., 29 August, 1994); Ephedra nevadensis / Achnatherum hymenoides Shrubland (plant association) (Nevada Natural Heritage Program, 26 September, 2003). Central Basin and Range- Shadscale-dominated Saline Basins Ecoregion, 13b (Woods et al., 2001).

112. Nevada Mormon tea, Nevada jointfir, or Nevada ephedra (Ephedra nevadensis)- An individual plant of Nevada jointfir on each fo the two ranges shown in the two preceding slides.

Millard County, Utah. Bureau of Land Management, Fillmore Field Office. June, pre-bloom stage.

113. Limbs and branches- Branches and characteristic branching pattern of Nevada Mormon tea. On the range Nevada jointfir is distinguished from the very similar green jointfir (E. viridis) by the grayish-green color and widely spaced branches of the former. Both species are valuable browse for wildlife and, among kinds of livestock, especially sheep (Welsh et al., 1993, p. 30).

Millard County, Utah. Bureau of Land Management, Fillmore Field Office. June, pre-bloom stage.

114. Rough-lookin' but beautiful- Only the third full day of summer and this range already looked "burnt up", but only to the "newcomer". This climax Indian ricegrass-galleta-winterfat range was at peak standing crop (it had been lightly grazed) with winterfat at full-bloom. Nevada ephedra was present, but as the associate shrub. Galleta was present as the second most abundant grass while threeawn was nearly absent. Cheatgrass was present, but limited almost entirely to areas of disturbance like road ditches and berms and places of rodent and harvester ant activity. This range plant community represented the potential natural vegetation for this range site and constituted a rangeland cover type.

This range vegetation was interpreted by the author as semidesert grassland (or at least as a savanna of widely scattered shrubs) and, as such, was included in this chapter. The cespitose (tufted) habitat of Indian ricegrass (ie. a bunchgrass) combined with cover of shrublike winterfat could logically lead to an interpretation of this climax range vegetation as a bunchgrass shrub steppe. This author however felt that cover of woody and/or semi-woody plants was conspicuously less than accepted standards of a shrub steppe as, for example, in case of sagebrush-bunchgrass shrub steppe. Furthermore and, as explained below, cover and biomass of winterfat is strictly speaking as much (probably more) herbaceous than woody. Therefore, in total this climax vegetation fit a designation of grassland more than a "hybrid" of grassland and shrubland.

Indian ricegrass has a rather unique set of morphological and genetic features and, consequently, an interesting history of taxonomic treatment. Welsh et al. (1993, p. 877) presented a thorough yet succinct summary of the situation. Iinterested students were encouraged to read their explanation (it is much more objective than radical treatments by cladistically oriented taxonomists). Indian ricegrass has morphological features in common with the needlegrasses (Stipa species in traditional American nomenclature) including a sharp callus on the caryopsis and large awn on the lemma. Thus at one point in taxonomic time Indian ricegrass was included in Stipa (S. hymenoides). Later, the deciduous awn (versus permanent awn in other Stipa species and ovoid-shaped grain led agrostologists to place this species in Oryzopsis (= O. hymenoides). Subsequently Indian ricegrass was "reinstated" as S. hymenoides. This revision was soon revised so that by the chaotic and, in this rangeman's mind, radical treatment of the "new agrostology" this remarkable range grass was designated Achnatherum hymenoides. It is doubtful if this is the last name given the "puplish-or-perish" hysteria of "run-it-like-a-business" modern universities combined with the relative ease (compared to actual experiments with live plants and animals) with which plant taxonomists can get a peer-reviewed paper out of a re-interpretation of a taxon.

Unfortunately, the Society for Range Management (Shiflet, 1994) did not recognize any Indian ricegrass rangeland cover type. It was remarked periodically in this publication that recognition of range cover types by the SRM is like recognition of human award recipients: someone has to nominate and write them up. That Indian ricegrass was minus such an advocate is not evidence that there is not an Indian ricegrass dominance type. There is. Readers of this work have seen three such range cover types in this chapter.

This Indian ricegrass-winterfat range represents some of the best of winter range. In much of this region surface water is lacking and water wells are so limited that livestock grazing requires water hauling for spring or summer use or, alternatively, has to be limited to winter sheep grazing. (If sheep are supposedly so stupid why do they know to eat snow and cattle do not?). Winterfat and Indian ricegrass "cure well" (ie. retain a high proportion of nutrients during winter and other periods of dormancy). Such ranges, as well as those that are composed almost strictly of winterfat (a consociation), are also ideal for pronghorn (Antilocarpa americana). One commonly sees--usually at considerable distance-- nice herds of pronghorn feeding on such semidesert grassland ranges.

Much of the brown herbage (which seemed to "overpower" green, growing plant material) was residue from the previous growing season. Due to the "curing on the vine" feature of herbaceous material, including that of winterfat which is described as suffruticose (becoming woody or somewhat of a shrub; basal parts of the plant are woody while upper portions are largely herbaceous), even "last year's stuff" is valuable feed. To the desert-wise rangeman this is indeed beautiful range.

Millard County, Utah. Bureau of Land Managaement, Fillmore Field Office. June, early estival aspect (anthesis in winterfat, "seed-ripe" in galleta, pre-bloom in Indian ricegrass). FRES No. 40 (Desert Grassland Ecosystem). No Kuchler unit. No SRM. Ricegrass Series (142.23) of Great Basin Shrub-Grassland (142.2) of Brown et al. (1998, p. 40). Ceratoides lanata / Oryzopsis hymenoides (Bourgeron et al., 29 August, 1994); Krascheninnikovia lanata / Achnatherum hymenoides Dwarf-shrubland (plant association) (Nevada Natural Heritage Program, 26 September, 2003). ACentral Basin and Range- Shadscale-dominated Saline Basins Ecoregion, 13b (Woods et al., 2001).

115. Makes 'em "fat and sassy"- Detail of the Indian ricegrass-winterfat vegetation from the range shown in the two preceding slides. Three plants of both of these range species were clearly visible in this photograph.

Millard County, Utah. Bureau of Land Managaement, Fillmore Field Office. June: winterfat at full-bloom and Indian ricegrass at pre-bloom phenological stages.

116. Queen subshrub and rest of the royal family- Typical specimens of winterfat (in full-bloom) surrounded by shoots of Indiant ricegrass, galleta, and cheatgrass. Cheatgrass, too? Yes, of course. Royal families were infamous for their bastards and other remittance men. Present were both current and previous growing season's biomass, woody and herbaceous plants, annuals and perennials, native and exotic (but naturalized) species. Considerable biological diversity on an otherwise "dull" or "boring" range.

Winterfat is regarded as a subshrub or undershrub, a suffruticose species characterized by having a woody or semi-woody basal stem and producing new main limbs or branches during each growing season so as to appear to at least partially "die back" to or near the ground or land surface. Many more photographs, including flowering shoots, of winterfat were shown in the chapter, Great Basin Desert (winterfat cover type).

Millard County, Utah. Bureau of Land Management, Fillmolre Field Office. June.

117. Bunchgrass semidesert grassland- Example of needle-and-thread-Nevada jointfir semidesert grassland on a slope of deep sand. Other grassses present in very small proportions were galleta, Indian ricegrass, and, at even less cover, cheatgrass or downy brome. The only forb of consequence was pale evening primrose (Oenothera pallida) and it was limited to small areas of sand blowout.

Almost exclusive coverage of the herbaceous layer by the cespitose needle-and-thread resulted in a much more tufted appearance than was the case with semidesert grasslands dominated less exclusively by the cespitose Indian ricegrass. Galleta was far less common on this type than on any other kind of Great Basin grassland encountered. Cheatgrass was almost nonexistant when ompared to typical variations in cover and density of other Great Basin semidesert grasslands. Diagnostic presence of Nevada jointfir was a common feature of all types of semidesert grasslands observed in these parts of the Intermountain Region.

This obvioulsy climax range vegetation was not described in the standard works on Great Basin vegetation. The striking photogenic features of this range plant community (along with sharp callus on the Stipa caryopsis and pointed tips of jointfir branches) should have made this an obvious range cover type or subtype (ie. needle-and-thread semidesert grassland or perhaps Nevada ephedra-needle-and-thread steppe). Vegetation classification remains an unfinished task.

This semidesert grassland (or shrub steppe) range type is outstanding habitat for pronghorn a large herd of which vaminosed just as this vegetation-focused photographer stomped in on the otherwise tranquil range ecosystem.

Millard County, Utah. Bureau of Land Management, Fillmore Field Office. June, estival aspect (grain-shatter stage in needle-and-thread). No Needlegrass Series (and there should be) under Great Basin Shrub-Grassland (142.2) in Brown et al. (1998, p. 40). Closest designation in the National Vegetation Classification System would be Stipa comata Alliance (Bourgeron et al., 29 August, 1994). There should be a Heterostipa comata / Ephedra nevadaensis Shrubland (association-level natural plant community) if and when there is a published National Vegetation Classification for Utah.

118. Sticky characters- The team of needle-and thread and Nevada jointfir showed up for a range "photo-op". These two main (monopolizing is more apt) range plants of this semidesert grassland were presented as a "curtain call" of this small cast of characters. Plant species in the Great Basin are considerably less "armed" than those of other range types whose leading actors include cactus, woody legumes, and dagger-leaved yuccas. Nonetheless sharp, penetrating calluses of needle-and-thread and pointed tips of jointfir justify the high-topped boots and chinks of buckeroos fortunate enough to ride such outstanding range.

Species composition of needle-and-thread-Nevada Mormon tea Great Basin semidesert grassland.This range type was not reognized in any of the standard works on Great Basin vegetation.

MIllard County, Utah. Bureau of Land Management, Fillmore Field Office. June (post "seed shatter", pre-dormancy in needle-and-thread).

119. Not exactly an oasis, but a grassland in the desert- A desert saltgrass (Distichlis spicata= D. spicata var. stricta= D. stricta)-semidesert grassland had developed on this small basin where surface runoff water (both snowmelt and rain) from the mountain sides and inflowing water from neighboring Great Salt Lake resulted in soil moisture conditions (Clements' chresard) that were more mesic than that of the adjacent and predominant black greasewood-dominated salt desert shrub.

The lighter green or grayish-green strip behind the zone of saltgrass was a Utah samphire (Salicornia utahensis)-saltgrass zone in which samphire was the prominent species and the obvious dominant. The zone of range vegetation immediately behind the samphire-saltgrass vegetation (at base or toeslope of mountain) was a greasewood-samphire-saltgrass shrubland. (This greasewood-salt desert shrubland and adjoining samphire-grass zonal range communities were covered in the chapter, Great Basin Desert, Great Basin Greasewood.) This semidesert grassland and adjoining samphire- and greasewood-dominated range types constituted a halosere. The halosere made up of this saltgrass grassland and the other range cover types displayed here where was also described under Great Basin Desert (the section on Great Basin Greasewood). This land was traditionally refered to as an alkali flat.

All of these halophytic (adjective referring to "salt-loving" plants) plant communities were seen as climax vegetation. Shantz (in Tidestrom, 1925, ps. 19-21) used the original monoclimax model of Frederick E. Clements to describe the Salt Desert Shrub (perhaps the first usage and original literature source of this designation). Shantz interpreted the greasewood- and samphire-dominated range communities as Clementsian associations (ie. greasewood association, samphire association) while he regarded saltgrass grassland as the seral unit of associes (ie. saltgrass associes). In the original Clementsian model of plant succession any vegetation that was not determined primarily or ultimately by climate was seral. In the monoclimax model any plant communities that developed on land receiving more water than provided directly by precipitation (eg. most wetlands like marshes and riparian, floodplain, or other overflow water-derived vegetation) would be seral and regarded as a Clementsian associes. Thus designation of saltgrass associes by Shantz (in Tidestrom, 1925, ps. 19-21). Under polyclimax theory saltgrass semidesert grassland is climax range vegetation.

Range vegetation on the predominant north-slope of mountain sides was bluebunch wheatgrass-dominated bunchgrass steppe. This steppe-form of grassland has not been regarded as semidesert grassland by serious students of native vegetation. Such was also the situation here where bunchgrass range had a generally more moderate habitat (cooler temperatures, non-saline or -brackish water, non-alkali soils).

Desert saltgrass apparently was of too small a spatial scale to have a Kuchler "vegetation type" unit. Likewise the Society for Range Management (Shiflet, 1994) did include a Saltgrass rangeland cover type (probably could not find anyone to write it up). Neither was there a saltgrass biotic community in Brown et al. (1998). Apparently the saltgrass rangeland cover type (there is one--as clearly shown here and by the cited work of H.L. Shantz--even if it was not published) is a "Nobody love me" or Rodney Dangerfield "I don't git no respect" grassland. Saltgrass semidesert grassland would be a unit under FRES No. 40 (Desert Grasslands Ecosystem). Generally, Central Basin and Range- Shadscale-dominated Saline Basins Ecoregion, 13b (Woods et al., 2001).

Tooele County, Utah. June.

120. Forage and alkali- This cattle range of saltgrass-Utah samphire grassland on an alkali flat had been heavily grazed (proper use or overuse?) less than two months before this photograph was taken. Most (in some spots, almost all) saltgrass herbage had been removed by grazing thereby leaving a disproportionate cover of samphire (selective grazing).

It must be borne in mind that Utah samphire is a climax range plant. Shantz (in Tidestrom, 1925, ps. 19-21) recognized the samphire association that was made up of this perennial species and the annual western samphire (Salicornia rubra). The range considered here was a "combination" of samphire and saltgrass. However, as was shown in photographs that follow, natural vegetation that had been excluded from cattle grazing was primarily saltgrass.

Designations of this range vegetation was presented above. It is a Saltgrass rangeland cover type, but this was not recognized (or it was overlooked) by the Society for Range Mangement (Shiflet, 1994). Tooele County, Utah. June.

121. "Do you always do this or was this the first time?"- Heavy degree of use on desert saltgrass occurred on a saltgrass-Utah samphire semidesert grassland range. Cattle had been taken off this range less than two months before this photograph was taken. The ungrazed (at least hardly touched) range plants were Utah samphire. The small tufts of herbage were clonal units of desert saltgrass on stolons that remained following cattle grazing. Degree of use can be guaged by comparison with ungrazed saltgrass in the next two photographs below. It was obvious that utilization had been heavy, but whether this was overuse or proper use on this range was not determined. Precipitation amounts and frequency during preceding winter and current spring were average or slightly above. If this heavy or high degree of use is proper (as evidenced by persistence of saltgrass under this extent of utilization in most years), this grass species--at least on this range site--is remarkable for grazing tolerance. If this high defoliation (heavy use) was more intense than typical, the question is raised if this will remain a saltgrass range or if it will degrade into a samphire stand (ie. a grazing disclimax). Altenatively, current grazing management might be part of a larger grazing program in which this range will be ungrazed in subsequent growing season(s). The range examiner must always regard grazing practices and outcomes in this light. Things on the range must be taken in context of the specific range (based on range sites) and its management and management objectives.

Tooele County, Utah. June, hard-dough grain stage in saltgrass. Desert saltgrass-Utah samphire semidesert grassland. Saltgrass rangeland cover type (unpublished).

Now to the exclosure.

122. Ungrazed desert saltgrass- Sward of ungrazed saltgrass immediately adjacent to but across a five-wire fence from the range that had been heavily grazed (high degree of utilization) as shown in the preceding slide. Adjacent location ruled out differences due to soil moisture, salt, and other nongrazing variables. Mere visual comparison of this range herbage gave a rough estimate of degree of use that had to be described as heavy. Whether this was proper or imporper degree of use (hence, grazing management) was not known. Only experience and/or experimentation over a sufficient period of time can provide the answer.

There was conspicuously less cover and density of Utah samphire in range vegetation that not been grazed by cattle. It was not known how many years (grazing seasons) cattle had grazed the range shown above or had been excluded from the ungrazed part. Therefore it could not be determined if the greater proportion of samphire on grazed rangeland was due to grazing or some other factor. Clearly there was a difference, quite a difference, in species composition.

Tooele County, Utah. June, hard-dough stage in saltgrass. FRES No. 40 (Desert Grasslands Ecosystem). No Kuchler unit ("vegetation type") at this spatial scale. Saltgrass rangeland cover type, but this was not published by the Society for Range Management (Shiflet, 1994).

123. What a difference water makes in the desert- Drainage of water from both inflow of salt water out of Great Salt Lake and overland flow of fresh water from a hillside produced this pronounced zonation of range vegetation at small-scale and discrete pattern. Foreground vegetation was part of the stand of desert saltgrass (with some Utah samphire) that was shown in the immediately preceding photograph. Midground vegetation (tall, green shoots) was a narrow strip of tule marsh made up of common three square, a bulrush, (Scirpus pungens) growing in and immediately surrounding standing water. Background vegetation (barely visible) was terminus of an alluvial fan populated by black greasewood and viscid rabbitbrush.

The tule marsh and adjoining wetland vegetation was covered in the chapter, Meadows (Flood Meadows).

124. Desert saltgrass (Distichlis stricta= D. spicata var. stricta= D. spicata)- Example of desert saltgrass growing in the saltgrass-Utah samphire semidesert grassland range featured above (obviously in the exclosure) This large clump of tillers was growing as a clonal unit from a rhizome of this colony forming eragrostoid grass. Saltgrass is frequently both rhizomatous and stoloniferous. This individual was in hard-dough grain stage of phenology.

Tooele County, Utah. June.

125. Utah samphire (Salicornia utahensis)- This perennial, succulent chenopod was a second menber of the desert saltgrass-Utah samphire semidesert grassland discussed above. This species varied from being co-dominant, dominant, or subordinate (associate) to desert saltgrass. Possible role of cattle grazing was considered in above discussions.

Tooele County, Utah. June.

The following two photographs were of range vegetation along a catena on a low hillside in the eastern Great Basin. Range plant communities developed on different range sites in a zonal pattern that reflected edaphic and topographic features of the land. Two of these distinctive zonal range communities had the physiogonomy, community structure, and species composition of grassland. These were not mountain, riparian, or wetland grasslands and, instead, possessed characteristics of Great Basin semidesert grasslands including some of the dominant grass species that defined semidesert cover types presented above. These grasslands also had features in common with Great Plains semiarid mixed prairie, especially in regards to dominant grass species.

Absence of Nevada jointfir and an ecological niche or role of western wheatgrass--as either a dominant or associate species-- distinguished the more xeric (and probably, to most viewers, more obvious) semidesert grasslands from the two immediately following Great Basin vegetation that seemed to more nearly fit designation of shrub steppe.

These two examples were included in both the Semiarid Grasslands and Shrub Steppe chapters. This reflected the arbitrariness which cannot be avoided when typing some range plant communities. Also, this reflected the frequent continuum existing in natural vegetation. Inclusion of these examples of Great Basin herbaceous and/or herbaceous-woody vegetation at this point was intended to avoid confusion and to allow readers to navigate from one category of range cover types to the next.

126. When range vegetation plays by the rules- Textbook example of sagebrush shrub steppe: Indian ricegrass-mountain big sagebrush dominated rangeland with major associated grasses being needle-and-thread, western wheatgrass, galleta, and with cheatgrass a minor component species except where locally dominant on disturbed patches. Associated but widely scattered shrubs included rubber and Douglas or viscid rabitbrush and antelop bitterbrush.

Lower down on the sideslope, which and developed and functioned as a catena, western wheatgrass had displaced needle-and-thread as the main associate to Indian ricegrass and big sagebrush had replaced rabbitbrush as the dominant shrub.

All in one's perspective: this climax vegetation fits "perfectly" (at least "closely") with the ideal example of sagebrush shrub steppe. Conversely, this range plant community could be interpreted as semidesert grassland from perspectives of physiogonomy, obvious dominance by grass (cover and density), prevalence of herbaceous layers over woody plant cover, and greater biomass from grasses than shrubs. One potential basis for distinction between shrub steppe (a savanna by definition) and semidesert grassland is that of greater aridity or xericity of habitat with the latter being the driest of all major North American grasslands.

Absence of an Ephedra species from the grass-dominated Great Basin vegetation presented here in contrast to widespread occurrence of and, typically, dominance by E. nevadaensis on what are obviously semidesert grasslands (or closer thereabouts) elsewhere in the Great Basin suggested that the climax plant community above was shrub steppe not semidesert grassland. In such matters the individual reader-viewer will decide.

Juab County, Utah. June, early estival aspcet but most grasses progressing rapidly toward completion oa annual growth and dormancy. Most logical interpretation of this range vegetation in author's view was FRES No. 29 (Sagebrush Shrubland Ecosystem). K-49 (Sagebrush Steppe). SRM 402 (Mountain Big Sagebrush). Great Basin Shrub-Grassland 142.2- Ricegrass Series 142.23 (Brown et al., 1998, p. 40). Artemisia tridentata / Achnatherum hymenoides Shrubland (a plant association) (Nevada Natural Heritage Program, 26 September, 2003; there being no such classification for neighboring Utah). Central Basin and Range- Sagebrush Basins and Slopes Ecoregion 13c (Woods et al., 2001).

127. Sagebrush shrub steppe or semidesert grassland? (Or your guess as good as the next rangeman's.)- The lowestmost range plant community on the catena featured here was--depending on one's perspective and other biases--either a western wheatgrass-mountain big sagebrush steppe or a consociation of western wheatgrass having such a paucity of big sagebrush as to be a grassland. Indian ricegrass, needle-and-thread, and galleta were the other major (though only associate) grass species. Cheatgrass was "few and far between". Rabbitbrush species hardly qualified as associates to mountain big sagebrush.

On this catena-arrayed series of climax range communities western wheatgrass increased consistently until at lowest elevational zone western wheatgrass displaced Indian ricegrass as the dominant range plant. On the deeper, more mesic (or less xeric) edaphic environment the rhizomatous, colony forming habit of western wheatgrass had a decided competitive advantage over the strictly cespitose Indian ricegrass. With less bare soil available the prolific grain yields of Indian ricegrass was not the superior adaptation it is on more xeric habitats.

Juab County, Utah. June, early estival aspect but all grasses were speeding toward dormancy. The most rational perspective on this range vegetation in the author's judgment was FRES No. 29 (Sagebrush Shrubland Ecosystem). K-50 (Wheatgrass-Needlegrass Shrubsteppe). Not a close fitting rangeland cover type (Shiflet, 1994); Western Wheatgrass variant of SRM 402 (Mountain Big Sagebrush) or, descriptively but outside the Great Basin, Great Basin variant of SRM 607 (Wheatgrass-Needlegrass). Great BasinShrub-Grassland 142.2- Wheatgrass Series 142.21. Central Basin and Range- Sagebrush Basins and Slopes Ecoregion, 13c (Woods et al., 2001).

The San Luis Valley in southcentral Colorado is regarded as one of four Intermountain Parks within the Southern Rocky Mountains. Fenneman (1931, ps. 130-132) described the San Luis Valley as a structural basin situated between the Sangre de Cristo Range to the east and the San Juan Range to the west. San Luis Valley has been interpreted as part of the general Rio Grande Rift System as upper reaches of the Rio Grande flow through this basin with streams that flow into the river forming alluvial fans (Fenneman, 1931, p. 130). The surrounding mountain ranges function as barriers to precipitation and the general land form resembles that of the Basin and Range province. Aridity is the local climatic result of this land form (ie. barrier and basin) with annual precipitation ranging from approximately seven to ten inches and averaging less than eight inches. Therefore the San Luis Valley is, strictly speaking, a climatic desert within the Southern rocky Mountains (James, 1971). Trimble (2001) described the San Luis Valley as the largest, highest mountain desert in North America.

Within this mountain desert there are three major and general (large spatial-scale) range types: desert scrub dominated by black greasewood (Sarcobatus vermiculatus) and with rabbitbrush (Chrysothamnus spp) as associate shurbs; 2) desert (technically, semidesert) grassland of Indian ricegrass (usually the dominant), various muhleys (Muhlenbergia spp.), and blue grama; and 3) a savanna of black greasewood-Indian ricegrass-muhly. This savanna developed primarily on vegetated sand dunes whereas the semidesert grassland developed on non-dune land having somewhat less sandy soils and the greasewood scrub .

Location note: the greasewood scrub or greasewood-dominated desert was included in the shrubland chapter entitled Miscellaneous Scrub Types.

A beautiful grassland (or savanna in local sites) that was not given in the SRM rangeland cover types was an Indian ricegrass (Oryzopsis hymenoides)-dominated (the consistent dominant) range plant community with other major grasses being ring or blowout muhly (Muhlembergia torreyi), sandhill muhly (M. pungens), blue grama (Bouteloua gracilis), mat grama (B. simplex), and on deeper sands, needle-and-thread (Stipa comata) and with black greasewood (Sarcobatus vermiculatus) as the major woody species and plains pricklypear (Opuntia polyacantha var. rufspina) the other shrub, these two with the midgrasses sometimes forming a sparse savanna. There were occasional plants of bottlebrush squirreltail (Sitanion hystrix), but aside from species diversity this grass had little indicator value for this range type. In range plant communities where needle-and-thread was an associate to co-dominant (deeper dunes) there was very little blue grama. Forbs were scarce to non-existent.

In some areas there were fairly deep and pronounced sand dunes whereas in other areas the land surface was less hummocky, rolling , or choppy (less dune-like) and more flat in local microtopography. Needle-and-thread more-or-less replaced blue grama as an associate grass species on the deeper dunes.

This grassland and the greasewood-grass savanna were contiguous with greasewood-dominated scrubland that was the climax range vegetation on alkaline flats (alkali basins). or eru

128. General sand dune type of mixed species semidesert grassland- A semidesert grassland dominated by Indian ricegrass and with sandhill muhly (M. pungens), ring or blowout muhly (Muhlenbergia torreyi), needle-and-thread, blue grama, and mat grama as associates (varying with local microsites) that developed on sand dunes in the San Luis Valley. Black greasewood was the major woody species which ranged from absence through associate to co-dominance with Indian ricegrass. Plains pricklypear was also preent and locally common. This range vegetation varied from grassland to greasewood-grass savannah at local scale.

These two photographs presented this climax range plant community at greater spatial scale so as to show physiogonomy and the considerable variation in structure and species composition. The first photograph including a local stand of blue grama with some mat grama (foreground) surrounded by numerous plants of black greasewood and Indian ricegrass as co-dominants. This variant of the sand dune semidesert grassland prevailed on land with lower dunes and shallower soil. the The second photograph displayed the Indian ricegrass-black greasewood savana form in which needle-and-thread was the major associate species. this vegetational variant predominated on deeper sands (higher dunes).

Great Sand Dunes National Park, Alamosa County, Colorado. Mid-June (late vernal aspect). FRES No. 40 (Desert Grasslands Ecosystem). No Kuchler unit because he mapped it as K-34 (Saltbush-Greasewood) when this is semidesert grassland within K-34. No SRM unit. Closest thing in Brown et al. (1998, p. 40) was Great Basin Shrub-Grassland 142.2, Mixed Bunchgrass-Shrub Series, 142.22 and that ain't it. Arizona/New Mexico Plateau- Sand Dunes and Sand Sheets Ecoregion, 22e (Chapman et al., 2006).

129. Indian ricegrass-black greasewood savanna- A savanna of these two co-dominants with needle-and-thread as the associate species which was the range plant community most widespread on the higher dune (deeper sand) portion or form of this overall semidesert grassland vegetation that developed in the San Luis Valley. There were isolated plants of blue grama, mat grama, sandhills muhly, and ring muhly, but these were not major species (based on cover, density, general abundance) in contrast to low dune areas where black greasewood and needle-and-thread were less common.

These two photographs showed the range plant community at a smaller spatial-scale than that of the two preceding slides in order to show the interior structure of this vegetation as well as growth form and habit of Inidan ricegrass and black greasewood, co-dominants of major portions of this savanna. The next slide featured habit along with relative cover and spatial relations between the two co-dominants.

Great Sand Dunes National Park, Alamosa County, Colorado. Mid-June (late vernal aspect). FRES No. 40 (Desert Grasslands Ecosystem). No Kuchler unit because he mapped it as K-34 (Saltbush-Greasewood) when this is semidesert grassland within K-34. No SRM unit. Closest thing in Brown et al. (1998, p. 40) was Great Basin Shrub-Grassland 142.2, Mixed Bunchgrass-Shrub Series, 142.22 and that ain't it. Arizona/New Mexico Plateau- Sand Dunes and Sand Sheets Ecoregion, 22e (Chapman et al., 2006).

130. Sharing the dunes- Indian ricegrass and black greasewood were frequently co-dominants as a savanna on major portions of a general semidesert grassland type dominated overall by Inidan ricegrass with varying proportions (based on relative cover) of blue grama, needle-and-thread, sandhill muhly, ring muhly, and mat grama as associate grasses. Plains pricklypear was the second important woody species.

The cespitose (tufted or clumped) habit of Indian ricegrass, a bunchgrass, and the size and shape of black greasewood on deeper and vegetated sand dune was featured in this photograph. On adjoining alkali sinks and flats the dominant black greasewood (with minor portions of Chrysothamnus species) and an erratic herbaceous cover of galleta (Hilaria jamesii) formed local cold desert or scrubland in the San Luis Valley where this set of photographs was taken. On greasewood desert scrub individual, adult plants of black greasewood were substantially larger than on duneland savanna. Squirreltail bottlebrush, a grass of the barley or wheat tribe, joined Indian ricegrass as the other locally important grass. Although of little value as an indicator species in this range vegetation, presence of the uncommon squirreltail bottlebrush demonstrated the species diversity--especially at local scale--of sand dune semidesert grassland and savanna.

More species diversity was reviewed in the two immediately following slides.

Great Sand Dunes National Park, Alamosa County, Colorado. Mid-June (late vernal aspect). FRES No. 40 (Desert Grasslands Ecosystem). No Kuchler unit because he mapped it as K-34 (Saltbush-Greasewood) when this is semidesert grassland within K-34. No SRM unit. Closest thing in Brown et al. (1998, p. 40) was Great Basin Shrub-Grassland 142.2, Mixed Bunchgrass-Shrub Series, 142.22 and that ain't it. Arizona/New Mexico Plateau- Sand Dunes and Sand Sheets Ecoregion, 22e (Chapman et al., 2006).

131. Cast of major players on sand dune semidesert grassland and savanna- "Lineup" of botanical actors on climax vegetation of arid rangeland that developed on sand dunes in the San Luis Valley, a mountain high desert. Different local to area combinations of various cespitose grass species and two or three shrubs produced range plant communities ranging from bunchgrass prairie to a shrub savanna. The major plant species--the overall dominant--of dune vegetation was Indian ricegrass. This member of the Stipeae tribe formed extensive consociation of semidesert grassland in non-dune rangeland. The major shrub was black greasewood which was frequently co-dominant with Indian ricegrass (the most abundant, and climax, grass overall). Greasewood scrub--an area-wide, alkaline desert--was an adjoining range cover type so there were local transition zones (small ecotones) between the greasewood-dominated desert and semidesert grassland.

The "botanical actors" on this arid stage included Indian ricegrass (the most abundant, tallest, largest clumps of light green color in right foreground and left background of first slide), sandhillls muhly (most prominent as the conspicuous hemispherical-shaped groups of clumps with bent dead stalks along with darker green leaves in right foreground of second slide), ring muhly (large circular clump with sparse center in left foreground of second slide), blue grama (most covr in left fore- and midground of first slide), mat grama (not present as visible), plains pricklypear cactus (left midground of first slide; left-center midground of second slide), and black greasewood (larger shrub with dying portions in right mid- to background and small, bright-green shrub to right of pricklypear in second slide).

Great Sand Dunes National Park, Alamosa County, Colorado. Mid-June (late vernal aspect). FRES No. 40 (Desert Grasslands Ecosystem). No Kuchler unit because he mapped it as K-34 (Saltbush-Greasewood) when this is semidesert grassland within K-34. No SRM unit. Closest thing in Brown et al. (1998, p. 40) was Great Basin Shrub-Grassland 142.2, Mixed Bunchgrass-Shrub Series, 142.22 and that ain't it. Arizona/New Mexico Plateau- Sand Dunes and Sand Sheets Ecoregion, 22e (Chapman et al., 2006).

132. Two bunchgrass dominants on a semidesert grassland/savanna- Indian ricegrass (seven plants foremost in first slide, five plants at rear of second slide) and sandhills muhly (seven or eight plants in front of Indian ricegrass in second slide) on sand dune range vegetation in the San Luis valley. Indian ricegrass was the overall dominant herbaceous species (often co-dominat with black greasewood on savanna) while sandhill muhly was locally dominant to co-dominant of microsites with deeper sand. Both of these grasses have a tufted or clumped growth habit so as to have the cespitose (bunchgrass) form. Indian ricegrass is strictly cespitose having only tillers (vertical shoots) whereas sandhills muhly is semi-cespitose possessing large, coarse rhizomes from which it produces new modular units (= clones or "daughter plants"). These ramets or "sister plants" tend to grow in a circular pattern, a circumspatial arrangement of dispersion so to speak. (Ring muhly is also strictly cespitose and lacks horizontal shoots.)

It was shown in the immediately preceding two photographs and in photographs below that sandhill and ring muhlies tend to grow in a "ring" pattern. This phenomenon was explained in greater detail below. These photographs and captions remarked that in this dune vegetation grasses have the predominant bunchgrass (tufted) form due to predominance of asexual reproduction/growth via tillers rather than rhizomes or stolons which result in more of the sod-forming habit. Silveus (1933, p. 238) described both of these Muhlenbergia species as "dnesely tufted". Sandhills muhly produces new tufts as clones off of rhizomes with the whole or overall pattern being one of a rough "circle". The "ring" of ring muhly results from senescence and death of older (more interior) parts of the root crown. As the center dies out and youner tillers around the perimeter of the bunchgrass remain alive a "fairy ring" (Gould, 1975, 254; Shaw, 2008, 171). That die-off sequence explains some of the "ring" growth form of sandhills muhly, but not all of it. The tufts of tillers (ramets or clones) that arise from rhizomes of sandhill muhly eventually "close in the gaps" between themselves thereby creating a "fairy ring" even without a center die out phenomenon. Standard agrostological treatments of M. pungens (Barkworth et al., 2003, ps. 174-175; Shaw, 2008. p. 162) have overlooked or were unaware of this arrangement of vegetative growth.

Another probable factor in this circular pattern of shoot dispersion is production of new plants (new genotypes) sexually, from caryopses. An observation made on the range by this photographer was that even Indian ricegrass tends to have a circular pattern of dispersion with smaller (presumedly younger) plants around a larger (older) plant which was likely the parent of the smaller plants. This was shown in these photographs, especially the first one. Another interesting (and perhaps revealing or indicative) thing in these slides was the circular imprints in the soil surface left by windblown grass leaves. Wind "behavior" in/on duned land is a complex phenomenon resulting in, among other things, different kinds of dunes being formed. Surface wind patterns at microscale, such as on the windward or lee side or height up on a dune, could go a long way in explaining dispersal of grass grains and, thus, dispersion patterns of plants. Fruits--other than extremely light, parachute-equipped ones--tend to be dispersed in concentric zones around parent plants. Progeny plants tend to be dispropotionately concentrated around their parents greater due to this pattern of propagule dispersal (unless parent plants repel their progeny via competition or allelopathy).

There would be a tendency for many plant species to produce a "fairy ring"-like distribution of plants much like those of mushrooms or, in this case, grasses (even without die-out at the center of a tufted plant). Tolstead (1941) reported that sandhill muhly, needle-and-thread, and blue grama (among several others) produced viable seed that germinated even without influence of low-temperatue. Even if low temperatures enhanced germination, cold temperatures would not be a limiting factor as it gets "plenty cold" in the San Luis Valley which is a mountain basin desert in the Southern Rocky Mountains. Indian ricegrass produces some of the most abundant yields of grain (and of large grain size) of any North American native grass species-- as learned by the American Indians who harvested this grain and used it for a staple. Caryopses of Indian ricegrass also have limited germination due to seed dormancy, a survival characteristic that has received considerable study by range management specialists (Jones, 1990; Jones and Nielson, 1992; Jones, 2009). Various slides included in this section showed irrefutably that grains of Indian ricegrass do germinate and emerge adequately to maintain dense and, sometimes, exclusive populations of this species, including consociations-- single-species stands--over large areas of range (see below). Sexual reproduction is conspicous when there are high numbers of obviously individual plants (unique genotypes) of cespitose species.

In essence, some of the tufted plants of sandhills muhly were likely sexually produced genotypes as well as the more obvious clonal tufts arising from rhizomes. This was even more so for strictly cespitose species like Indian ricegrass, needle-and-thread, and, of course, the annual mat grama. A logical question that would make this point could be, How many of the seven or eight clumps of sandhills muhly in the forefront of the second slide are clonal or, same thing, how many are different genotypes? Contrary to what seemed the obvious, not all of clumps of Indian ricegrass are plants of a single unique genotypes. Investigations by McAdoo et al. (1983) revealed that some clumps of Indian ricegrass consisted of several genetically unique plants (ie. different genotypes) that sprouted from grain caches made by rodents. While all these plants began life as seedlings some tufts were composed of several genetic individual plants (in contrast to clumps of sandhill muhly many of which are modules or "sister plants" of the original seedling).

Great Sand Dunes National Park, Alamosa County, Colorado. Mid-June (late vernal aspect). FRES No. 40 (Desert Grasslands Ecosystem). No Kuchler unit because he mapped it as K-34 (Saltbush-Greasewood) when this is semidesert grassland within K-34. No SRM unit. Closest thing in Brown et al. (1998, p. 40) was Great Basin Shrub-Grassland 142.2, Mixed Bunchgrass-Shrub Series, 142.22 and that ain't it. Arizona/New Mexico Plateau- Sand Dunes and Sand Sheets Ecoregion, 22e (Chapman et al., 2006).

133. Reinvasion on "new land"- To an ecologist the designations of "new land" or "new ground" refers to disturbed land (often partially to largely denuded or devegetated). Disturbances vary from natural to man-made (admitting, of course, that man is part of Nature) with some of the latter including plowing, overgrazing, improper burning (or, more commonly, lack of burning), road construction, military maneuvers, recreational vehicle damage, etc. Abandonment of cropland (farm ground) results in "new land" known as "old fields" or "go-back land". Revegetation or recovery of vegetation through natural processes constitutes secondary plant succession. Denudation resulting from road-building activity frequently leaves road cuts with so little soil remaining that natural restoration has to start from the stage of parent material in which case the series of re-development of vegetation is primary plant succession.

The local denuded area shown here (and in the immediately following two photographs) was a road cut in sand dune semidesert grassland/savanna in the San Luis valley, a high desert in a mountain basin in the Southern Rocky Mountains. The bladed, devegetated area was a narrow strip of land between virgin vegetation. This arrangement or pattern in land space was shown in this photograph.

Age of this local disturbance was not known. This "scar (or wound) on the land" was relatively old. It was undoubtedly several years old. Obviously, it was being invaded by range plants most of which were sandhill muhly, ring muhly, and black greasewood with lesser cover of mat grama. Plants of the overall dominant species, Indian ricegrass, and associates needle-and-thread or, depending on local haibitat, blue grama hd not yet colonized the denuded strip of land. Invasion was the term given the Range Management profession by Frederic Edwards Clements which means the cumination, the endpoint, of successional processes (migration, ecesis, aggregation, competition, stability, etc.) resulting in movement of plants from one piece of land to another and establishment of new plants in the latter (Weaver and Clements, 1938, ps. 148, 166). Theoretically, there is no invasion in climax vegetation, only renewal or replacement of existing species (often by clonal or asexual reproduction). When land has been denuded through disturbance it becomes a "new" environment for migration of plant propagules (disseminules like fruit, seed, shoots) and their ultimate establishment (= invasion). Typically plant species that first reinvade (= pioneer or colonize) "new ground" like old fields and/or that increase under disturbances are designated invaders. With on-going development of vegetation through various plant communities over time on the sere, the piece of land on which plant succession progresses, of a given range or forest site eventually a final plant community becomes established as the climax vegetation. The climax remains relatively stability though this stability is more of a seesaw-like "dynamic equilibrium". Plant species that characteristically decline first under abnormal disturbance are given the designation of decreasers. An intermediate category of range plant species that initially increase under disturbance and then decline with continuing disturbance (and as invaders become established) are called increasers.

Invaders on severely disturbed areas like old fields are most commonly annuals surviving stressful periods as seeds or dried fruits (eg. grass caryopses) whereas decreasers are long-lived perennials with greater reliance on vegetative or clonal reproduction versus sexual reproduction (seeds, grains, spores). On the bladed road cut shown here there were very few plants of the annual and native mat grama or annual invaders like cheatgrass. Instead, the occupants were old plants of sandhill and ring muhly. Adjacent undisturbed vegetation was dominated by the climax dominants, Indian ricegrass and black greasewood along with blue grama and needle-and-thread and lesser cover of the Muhlenbergia species. This suggested that range vegetation currently occupying the road-cut disturbance was roughly mid-sere and, by extension, that perhaps the Muhlembergia species were increasers. Vallentine and Burzlaff (1964) classified sandhills muhly as an increaser and Indian ricegrass and needle-and-thread as decreasers or increasers depending on range site. Pound and Clements (1898) found that on fresh sand "blowouts" sandhills muhly came in with the second seral stage following colonization by blowout grass (Redfieldia flexuosa). Sandhills muhly persisted on blowouts where it sometimes comprised a faciation, the major subunit of a Clementsian association (Weaver and Clements, 1938, ps. 95-96). More on this in the next "installment": next slide, please.

Great Sand Dunes National Park, Alamosa County, Colorado. Mid-June (late vernal aspect). FRES No. 40 (Desert Grasslands Ecosystem). No Kuchler unit because he mapped it as K-34 (Saltbush-Greasewood) when this is semidesert grassland within K-34. No SRM unit. Closest thing in Brown et al. (1998, p. 40) was Great Basin Shrub-Grassland 142.2, Mixed Bunchgrass-Shrub Series, 142.22 and that ain't it. Arizona/New Mexico Plateau- Sand Dunes and Sand Sheets Ecoregion, 22e (Chapman et al., 2006).

134. Old rings on "new land"- Two "photoplots" provided closer-in views of sandhill and ring muhly (with a few plants of Indian ricegrass) populating an old road cut and "healing" this denudtion "wound on the range". The preceding photograph and caption provided a more distant and general view of this "site of disaster" that served as "new land" (an area of land available for revegetation). Also shown in that slide was adjoining undisturbed climax vegetation along with its caption that provided a summary of plant succession on the sere of this degraded range. That "installment lesson" ended noting that sandhill and ring muhlies as current occupants of this bladed area at mid-sere were perhaps increaser species and that plants of those two species were obviously rather old individuals.

The latter conclusion was based on knowledge of growth habit of Muhlembergia pungens and M. torreyii, sandhill and ring muhlies, respectively. Both of these perennial, eragrostoid grasses are cespitose (bunchgrass) species. Much of their growth/asexual reproduction is through addition of new tillers. An individual plant increases in basal area as new tillers are produced at the perimeter (around the outer edges) of the root crown so that older plants become larger (as measured at either basal or foliar cover). As individual plants of sandhill and ring muhly age and increase in diameter (grow wider across) through addition of peripheral tillers the older (the more interior) parts of these plants die so that eventually only a circular boundary of younger tillers remain. This growth pattern results in a uniuque and very characteristic "ring" of tillers. Hence, the common name, ring muhly, taken from the circular pattern or growth habit of clonal shoots. Some of these "rings" reach diameters of considerable size. The author stepped off some circular arrangements of muhly tillers shown in these photographs that reached from six to ten feet in diameter. Presumedly the larger plants (those with greater-diameter clonal "rings") growing at a given location are older than adjacent or closely neighboring plants of smaller size.

This circular pattern of tillers or growth habit in ring and sandhill muhly has been widely recognized and described by workers as, for example, Gould (1975, ps. 254, 255) and Shaw (2008, ps. 162, 171) This growth phenomenon has been commonly attributed to the incremental death of older more interior parts of plants (seemingly senescence of older portions of the root crown). Observations by this author found this explanation to be inadequate, at least in the instance of sandhill muhly. Although sandhill muhly is cespitose it also grows coarse, rank rhizomes (Gould, 1975, p. 254; Shaw, 2008, p. 171). The current author found many plants of sandhill muhly on the ranges presented here that had produced new clonal units (= ramets, offshoots, modules) off of their rhizomes. The rhizomatous "sister plants" or "daughter plants" had been produced in a circular arrangement so that they had grown into a "ring". This clonal "ring" was not due to senescence and death of the interior of a cespitose plant, but rather from an increase in numbers of daughter or sister offshoots which themselves grew in diameter through tiller proliferation and "when it came their turn" produced daughter modules themselves. In other words, the standard "died out in the middle" leaving a "fairy ring" description (Gould, 1975, ps. 254, 255) is an incomplete explanation for the "ring" growth habit. Again, this is the case at least for sandhill muhly. Incidentially, it appeared to this photographer that all of the extremely large rings of sandhill muhly (eg. those reaching six to eight feet across) were result of ramets growing in a circular arrangement and not due to death of interior tillers (older parts of the root crown). It seemed highly unlikely that a bunchgrass could grow to such dimensions in this high desert environment. Is was also possible that some of the clumps in the erratic " fairy circles" were other genotypes (different plants) that developed from fruits (achenes in Muhlembergia species) from sexual reproduction.

Viewers can see for themselves the individual "bunches" (ramets or "sister plants") in several of these rhizome-produced muhly "circles" or "rings" in both of these photographs, the second from the last photograph above, and in foreground of the immediately succeeding photograph.

Dominance of this bladed area (the upper slope of borrow ditch on a a road cut) by muhly species rather than Indian ricegrass, needle-and-thread, black greasewood, and (maybe) blue grama indicated that this Muhlenbergia-dominated plant community was a seral stage--perhaps mid to advanced sere--not climax range vegetation.

Great Sand Dunes National Park, Alamosa County, Colorado. Mid-June (late vernal aspect). FRES No. 40 (Desert Grasslands Ecosystem). No Kuchler unit because he mapped it as K-34 (Saltbush-Greasewood) when this is semidesert grassland within K-34. No SRM unit. Closest thing in Brown et al. (1998, p. 40) was Great Basin Shrub-Grassland 142.2, Mixed Bunchgrass-Shrub Series, 142.22 and that ain't it. Arizona/New Mexico Plateau- Sand Dunes and Sand Sheets Ecoregion, 22e (Chapman et al., 2006).

135. Miracles in the mountains- In late spring in the San Luis Valley with the Sangra de Cristo Range for a backdrop as a morning thunderstorm brought "showers of blessings" to a consociation of Indian ricegrass that formed a semidesert grassland. This Excellent condition range included amazingly few representatives of other grass species besides some widely scattered plants of squirreltail bottlebrush, needle-and-thread, and blue grama. There were also a scattered individuals of black greasewood, the dominant of desert scrub in alkali flats and local depressions in this area and a co-dominant with Indian ricegrass on savannas between alkaline desert and semidesert grassland. The main shrub on this range was apache plume (Fallugia paradoxa) a local patch of which was present in the second of these slides. There were also a few small plants of an unidentified species of rabbitbrush (Chrysothamnus sp.).

Pinyon pine (Pinus edulis)-juniper (Juniperus spp.) woodland was in the distant background up on the foot slope of the Sangra de Cristo Mountains.

Alamosa County, Colorado. Mid-June (late vernal aspect). FRES No. 40 (Desert Grasslands Ecosystem). No K unit because Kuchler did not map an Indian ricegrass unit. No SRM; no unit in Brown et al. (1998). They all missed it! Arizona/New Mexico Plateau- Sand Dunes and Sand Sheets Ecoregion, 22e (Chapman et al., 2006).

136. As grand a grassland as God ever grew, and in the desert- On the same day in late spring as shown in the preceding photographs skies cleared and the sun shone brightly on an Indian ricegrass semidesert grassland in the San Luis Valley with the Sangra de Cristo Range in the background. This was a different range from that shown in the preceding two slides though only about five or six miles down the road from that one This range was even more of a "pure" population (single-species stand) of Indian ricegrass than the one shown under a thunderhead.

There was need for but few words to describe this pristine, high-basin grassland. This consociation of Indian ricegrass shared the range with almost no other plant species except for a few widely scattered plants of rubber rabbitbrush (Chrysothamnus nauseosus) and the criciferous forb, sanddune or California wallflower (Erysimum capitatum). Indian ricegrass grasslands of this nearly single-species composition were reported by Tueller (1975, ps. 23-24) for arid conditions in the Great Basin Region.

If there was ever virgin range this was it. Simply an unbelieveable find. This spring and the preceding winter had been unusually wet so individual plants of Indian ricegrass were larger than typical, but no amount of moisture could have accounted for the immense size of these plants. They had unquestionably been managed for a number of years under light utilization. Root crowns of these plants were huge and these could only have grown to their immensity under the best of grazing management over a number of years. Size of individual plants was shown and described below.

Emphasis: this is native semidesert grassland. This was not a reseeding project.

Details of this superb specimen: Bureau of Land Management. The Foothills Allotment consisting of 7811 acres divided into three pastures. Total of 347 Animal Unit Months harvested between 1 June and 30 September by 93 head of cows (mixed breeding but black baldy sorts) with their calves. Administered out of the La Jara BLM field office. (Miss Melisa Shawcrost, persona communication).

Storm, sunshine, and sound range management showed what a semidesert grassland range can do if given the chance, but it takes all three.

Alamosa County, Colorado. Mid-June (late vernal aspect). FRES No. 40 (Desert Grasslands Ecosystem). No K unit because Kuchler did not map an Indian ricegrass unit. No SRM; no unit in Brown et al. (1998). They all missed it! Arizona/New Mexico Plateau- Sand Dunes and Sand Sheets Ecoregion, 22e (Chapman et al., 2006).

137. Details of a desert grass garden-Physiogonomy, structure, and botanical composition of the climax vegetation of a semidesert grassland range that was a consociation of Indian ricegrass. There were almost no other plant species present except for a few plants of rubber rabbitbrush and sanddune wallflower. In fact, the photographer had to search to find a local group of rabbitbrush plants. Instead a population of enormous Indian ricegrass plants formed a field-like natural grassland of almost unbelieveable purity. The remarkable size of these cespitose individuals was shown and described below.

Most of these bunchgrass clumps were made up of the same plant (one genotype) having many tillers, but perhaps some tufts consisted of shoots of several genetically distinct plants that germinated and emerged from rodent caches as documented by McAdoo et al. (1983). Either way, these large bunches had developed over several to many years and could only have reached such size under jucicious grazing management.

Bureau of Land Management, Foothills Allotment. Alamosa County, Colorado. Mid-June (late vernal aspect). FRES No. 40 (Desert Grasslands Ecosystem). No K unit because Kuchler did not map an Indian ricegrass unit. No SRM; no unit in Brown et al. (1998). They all missed it! Arizona/New Mexico Plateau- Sand Dunes and Sand Sheets Ecoregion, 22e (Chapman et al., 2006).

138. Indians' rice crop- Indian ricegrass on black greasewood-bunchgrass savanna (first photograph) and on an Indian ricegrass consociation of semidesert grassland (second photograph) in northern San Luis Valley. Cespitose habit of this bunchgrass was very noticeable. Also noticeable (if one can "read range sign") was the immense size of individual plants on semidesert grassland. Note for comparison the average size of a heap of cattle dung (typical "cow flop") in dead-center foreground of second slide. Some of these plants would have covered half of the flatbed of a one ton truck. Size of these plants growing in an arid (or, at least, nearly so) was on par with that of gigantic trees in rain forests.

Bureau of Land Management, Foothills Allotment. Alamosa County, Colorado. Mid-June.

139. Plants and a panicle- Portions of five plants of Indian ricegrass growing in a consociation of this species that comprised a semidesert grassland in northern San Luis Valley were presented in the first photograph. Second photograph was a part of a panicle of one of these plants showing the tremendously heavy grain crop produced in an extremely wet growing season. These plants and those shown in the immedately preceding photograph of the above pair of slides were giant specimens of their species growing on pristine (Excellent range conditin class) semidesert grassland that was made up almost exclusively of this one species.

Bureau of Land Management, Foothills Allotment. Alamosa County, Colorado. Mid-June, hard-dough phenological stage.

140. Ripening rice crop- Spikelets of Indian ricegrass on the widespread panicles typical of this species. Not typical was the extraordinarily heavy yield of grain that was produced in an unusually wet spring. Nonetheless, even with lower grain yield per plant or per unit of land there was such an extensive area of grassland dominated by Indian ricegrass and, as shown above, sometimes as exclusive single-species stands (consociations) that tremendous quantities of humanly edible grain could be gathered for food. This was the situation with regard to Indian ricegrass and American aborgines and, in fact, was origin of this common name.

Indian ricegrass, State Grass of Utah and Nevada, is one of the most valuable native grass species for reseeding degraded range and mined land (eg. reclamation of open surface coal mines). Unfortunately natural selection resulted in prolonged dormancy which usually has to be overcome by seed treatments such as chilling and scarification otherwise establishment efforts have often come to naught (Rogler, 1960). A good general guide to Indian ricegrass including stand establishment and cultivars was that prepared by Ogle (2000).

Bureau of Land Management, Foothills Allotment. Alamosa County, Colorado. Mid-June, hard-dough grain stage.

141. Striking crucifer of San Luis Valley grassland- Sanddune or California wallflower (Erysimum capitatum) on a local degraded area of Indian ricegrass semidesert grassland range at northern end of San Luis Valley. This is perhaps the most conspicuous--and strikingly so--forb of this range plant community. Large size, especially as to height, coupled with bright yellow petals sets this species off in an area of minimally diverse grasslands and savannahs. Hermann (1966, p. 12) suspected that this species might be a variety of E. asperum, a species grazed by all livestock species but that one of "negligible to slight" forage value.

Alamosa County, Colorado. Mid-June, peak standing crop and full-bloom stage.

142. Shoot! Another plant of the same crucifer species- Another plant of sanddune wallflower provided closer-up shots of the sexual shoot of this conspicuous forb on Indian ricegrass-dominated semidesert grassland in the upper San Luis Valley. The indeterminate pattern of flowering (upward and outward sequence of blooming) was obvious with the siliques, crucifer fruits that are prominently elongated and considerably longer as, perhaps four to five times, than wide (Smith, 1977, ps. 66, 307). There are numerous Erysimum species throughout the Western Range Region, but E. capitatum is one of the more spectacular ones.

Alamosa County, Colorado. Mid-June, peak standing crop and full-bloom stage.