Great Basin Naturalist
Volume 57 Number 2 Article 8
5-7-1997
Classification and ordination of Coleogyne communities in southern Nevada
Simon A. Lei University of Nevada, Las Vegas
Lawrence R. Walker University of Nevada, Las Vegas
Follow this and additional works at: https://scholarsarchive.byu.edu/gbn
Recommended Citation Lei, Simon A. and Walker, Lawrence R. (1997) "Classification and ordination of Coleogyne communities in southern Nevada," Great Basin Naturalist: Vol. 57 : No. 2 , Article 8. Available at: https://scholarsarchive.byu.edu/gbn/vol57/iss2/8
This Article is brought to you for free and open access by the Western North American Naturalist Publications at BYU ScholarsArchive. It has been accepted for inclusion in Great Basin Naturalist by an authorized editor of BYU ScholarsArchive. For more information, please contact [email protected], [email protected]. Creat Basin Naturalist 57(2), iC 1997, pp. 155-162
CLASSIFICATION AND ORDINATION OF COLEOGYNE COMMUNITIES IN SOUTHERN NEVADA
Simon A. LeiI and Lawrence R. WalkerI
ABsTRAct-Woody plant community com{Xlsition was analyzed throughout the range ofColeogyne ramosissirrw. in the Spring and Sheep Mountain ranges of southern Nevada. The lower Coleogyne elevational boundary was analyzed in detau in Lucl,. Strike Canyon, on the eastern edge ofthe Spring Mountains. nvINSPAN (Z-way indicator species analy sis) identified 4 primary species and stand groups from the 2 mountain ranges (extensive survey), and 4 primary species and stand groups in the detailed study at the lower Coleogyn6 ecotone (intensive survey). Analysis of DECORANA (detrended correspondence analysis) results indicated that elevation and soil depth were the environmental facto~ most signi£i.cantly associated with distribution of species and stand groups in the extensive survey. Elevation was the only sig~ nificant physical factor associated with distribution of species and stand groups in the intensive survey. Five vegetation woes from lhe 2 mountain ranges were identi6ed based on their dominant species in 15 extensive transects. Coleogyne were subdivided into pure stands and upper and lower ecotones for further investigation ofspecies distribution and envi ronmental factors. Selected environmental factors appear to play an important role in structuring the Mojave Desert veg etation zones in southern Nevada
&'1 words: Coleogyne ramosissima Torr., cltMsifirotion, on:unation, vegetation zones, Lucky Strike Canyon, Spring Moumoinl, Sheep Range.
Three dominant vegetation types character Distribution of Coleogyne in Utah may thus ize southern Nevada: Lan'ea tridentata Cov. depend largely upon abiotic factors. However, Ambrosia dumosa Payne (valley floors and lower little is known about the factors that deter mountain slopes), Coleogyne ramomsima Torr. mine the pattern of Coleogyne distribution in (mid-elevation slopes), and Pinw; monophylla southern Nevada. Torr. & Frem.-juniperw; osteosperma Little In this study we determined the spatial (upper elevation slopes). The Coleogyne vege arrangement of Larrea-Ambrosia, Coleogyne, tation zone, ranging from 1200 to ISS0 m in and Pinus-juniperw; vegetation along an ele elevation in southern Nevada, is characterized vational gradient and the percent cover for all by a closely spaced matrix ofColeogyne with a woody perennial species. We also classified scattered distribution of other shrub species. vegetation using TWINSPAN and determined Previous studies have shown that air tempera the relationship between the distribution of tures in the COleOgytl£ community in Utah range desert plant species groups and environmental from -24· to 47·C (Korthuis 1988). Korthuis factors based on correlations with axis values (1988) proposed that the upper limit of Coleo from DECORANA. gyne distribution may be set by low air tem peratures and that cold air draining from adja STUDY AREA cent mountain slopes may limit Coleogyne establishment on basin floors. Alternatively, the This study was located in the Spring (36·O'N, lower limit of Coleogyne may be determined lIS·30'W) and Sheep (35·50'N, lIS·3S'W) by low soil moisture in Utah (Bowns 1973). Mountain ranges, approximately 6S Ian north Annual precipitation ranging from 180 to over west ofLas Vegas, Nevada. Precipitation patterns 270 mIll appears to be required for develop ofsouthern Nevada include summer storms and ment ofColeogyne stands in the Mojave Desert winter rains. Summer storms generally occur (Hunter and McAuliffe 1994). Shallow soils are in July and August and can sometimes be local typical of Coleogyne communities and may and intense. Winter rain is widespread and may partially determine its abundance and distri last up to several days. Snow is frequent at high bution in Utah (Callison and Brotherson 1985). elevations, particularly in the Pinus-juniperus
lDepllrtment of B",logieal Seieoces, U"ivenityorNevada-Las Vep.s. Las Vegas, NY 89154-4000l.
ISS 156 GREAT BASIN NATURALIST [Volume 57
woodlands and upper Goleogyne ecotones. Pre diameters were calculated by computing the cipitation is positively correlated with increase average of the longest and shortest dimension in elevation (Rowlands ct a!. 1977). of the plant cover. Elevation and aspect of Southern Nevada is an area of temperature each plot were measured. ';Ye visually estimated extremes, \vith a mean minimum winter tem percentage of rock, soil, and vegetation cover perature of _10 0 C and summer temperatures and assigned each plot to 1 of6 cover categories of above 47°C. Temperature means and ex (absent, 1-5%, 6-25%, 26-50%, 51-75%, and tremes are negatively correlated with increas 76-100%). We cla"ilied topography ofeach plot ing elevation (Rowlands et a!. 1977). Relative as slope, terrace, or dry vvash, repositioning humidity is low « 20% is common in the sum when cliffs, rocky areas, and streambeds were mer months), resulting in exceedingly high encountered. New plots were positioned hori evaporation. zontally approximately 20 Tn from the transect at an identical elevation to avoid these habitats. METHODS Ground surface was characterized as desert pavement, loose rocks, sand, and sand with Broad patterns of vegetation zonation were boulders. examined vl"ith 12 elevational transects located Using TWINSPAN (Hill 1979a), we classi on the Spring Mountains and 3 on the Sheep fied sampled plots into vegetation types and Range in summer 1993. The transects con generated groups ofspecies based on vegetation tained 9-16 circular (100-m2, 5.65-m radius) similarities among the sampled stands. Each plots spaced at a fixed elevational interval of stand group represented a vegetation zone 65 m. Average number ofplots per transect was from the lowest to the highest elevational 12. Transects were located east, west, and south groups in the part of the Spring and Sheep of the Spring Mountains and east and west of Mountain ranges sampled. Stand group 1 was the Sheep Range. Each transect included 2 characterized by Larrea-Ambrosia shrublands, plots in the Larrea-Ambrosia shrubland just while staud group 4 was characterized by Pinus below the Coleogyne shruhland (mean elevation Juniperus woodlands. Using TW1NSPAN analy 1155 + 40 m; Lei and Walker 1997), extended sis, we divided the sampled plots into 2 groups, throughout the entire Coleogyne zone (mean each of which was then divided again, and so elevation 1560 + 45 m; Lei and Walker 1997), on. The dichotomy was terminated if there and included 2 plots in the Pinus-Juniperus were 4 or fewer plots in a group (Kent and woodland just above the Coleogyne zone (mean Coker 1992). Kent and Coker (1992) further elevation 1870 + 50 m; Lei and Walker 1997). proposed that the size of the eigenvalue gen A portion ofthe lower ecntonal area ofCole erated at each dichotomy by TW1NSPAN anal ogyne was examined in detail at the location of ysis reflects the importance of each compo one of the original transects on the eastern side nent in explaining the total variation within the of the Spring Mountains. Six replicate 100-m2 data set. The eigenvalue, ranging from 0 to 1, circular plots were located at each of 6 eleva was largest at the initial dichotomy and became tions at 30-m elevational intervals in Lucky smaller with each successive dichotomy. One Strike Canyon between 1160 and 1310 m ele sampled plot from Red Rock Canyon of the vation. The lowest elevation included 6 plots Spring Mountains was eliminated due to the just below the lower elevational limit of Cole existence ofnumerous rare (riparian) species. ogyne. The remaining 5 elevations represented We used DECORANA (Hill 1979b) to gen increasing levels of Coleogyne density, but the erate species and stand ordination scores. Coleogyne shrubland extends to 1600 m in These scores were based on the percent cover Lucky Strike Canyon (Lei and Walker 1997). value of each woody species. Each point on Data collected in these «intensive" plots were the diagram corresponds to a species, and dis also collected in the "extensive" plots that cov tances bctv.reen points on the graph are approx ered the entire elevational range ofColeogyne. imations of their degree of similarity (Kent Within each plot we recorded the presence and Coker 1992). Distances between points on of all woody perennial species (>10 em tall), the diagram increased as species distributions including subshrubs. Subshrnbs are plants that diverged and as species occupied different have suffrutescent stems at the base with herba vegetation zones. Similarly, stand ordination ceous stems making up the canopy. Canopy scores were also generated by DECORANA, 1997] COLEOGYNE GRADIENT ANALYSIS IN SOUTHERN NEVADA 157 and the interpretation resembled species ordi RESULTS nation scores. Distances between points in creased as stand distributions diverged. Stand Four major species groups from the 56 woody scores could be matched with several environ taxa were identi6ed by TWINSPAN analysis mental variables to detect vegetation and plot for the extensive transects in the Spring and variation in relation to the environment. Envi Sheep mountains (FIg. I). Species in group A are ronmental variables used in our study included typical of Lan'ea-Ambrosia stands found at the elevation, topography, soil depth, plot aspect, lower limit of Coleogyne; they include Lar type of ground surface, and percent soil and "ea fiidentata, Ambrosia dumasa, Ephedra neva rock cover. Stand aspect could not be correlated densis, Yucca schidigera, and Aeamptopapptls because all stands occurred on northwest-fac shockleyi. These species tend to occupy slopes ing slopes in Lud:y Strike Canyon. Similarily, and terraces at low elevations. Species in group percent rock cover could not be correlated B are typical of nearly monospeciJic stands of because all stands had a rock cover of over Caleagyne and include Ytlcca brevifolia, Pmnus 75%. The 1st and 2nd ordination axes were fasicuwta, and Thamnosma montana. Species orthogonal and indicative of different sources in group C, typical of the upper Coleogyne ofenvironmental variation. ecotone, are represented by Atriplex eaneseens, Classi6cation and ordination techniques Gutierrezia sarothme, and Chrysothamnus nau were applied to a total of180 sampled plots on seasus. Although Gtltierrezia sarothrae exists in the 15 elevational transects from the Spring Caleogyne stands, it is particularly abundant at and Sheep Mountain ranges (extensive plots) the upper Caleagyne ecotone. Species in group and 36 plots at the lower elevational limit of D are found in the pure Pinus-Juniperus wood· Coleogyne in Lucky Strike Canyon (intensive lands, with Artemisia tridentata as the major plots). understory species. Pinus-Junipems are the 2
SPRING AND SHEEP MOUNT~IN RANGES
1.0 o. , 0.' 0.' 0.'
0.5 eIGENVALUE
0.' 0.3
0.2
O. ,
0.0 Loo SAOO opec CORA A.1'C'A ARPU ARTR ""GUT l....nrr FOffJ DYCO IATR Y.RP;\ Em! OPP.A "'.ESP LYAN CHL! SAM! PRFA. OPBA CHNA 2RAN JUOS PIMO CEGR RHTR M1I.TO CACO PSFR opro COIl' THMO )'l]BA GUSA EPVI CELl:: STPA KRPA 'NY! YUBR STPI FAPA OPAC 'YSA >SCO MIPR 10007
1 '-__-'I L--J LI -'1 • B c D
SPECIES CROUPS
Fig. 1. Dendrogram of species groups identified by TWINSPAN from 15 elevationaJ transects in the Spring and Sheep Mountain ranges..Main groups were arranged by elevation from the lowest (group A) to the highest (group D). Species are arranged alphabeticaLLy within groups. See Appendh 1 for species abbreviations. 158 GREAT BASIN NATURALIST [Volume 57 co-dominant genera in these woodlands, with with Coleogyne as the dominant species and JurlipenlS osteosperma more abundant at lower Grayia spinosa as a common associated species. elevations and Pinus monophyUa at higher ele Hymenoclea salsow exists in or near the edges vations. Associated species including Fallugia of washes witbin the Coleogyne-dominated pamdoxa, Ephedra viridis, Cercocarpus ledi stands in this group. foli.lS, and Forsellesia nevadensis are also pre Analysis of the 180 extensive and 36 inten sent, but with significantly fewer individuals. sive plots using DECORANA resulted in sig Four main species groups from 26 woody nificant segregation of stand groups from taxa were identified by TWINSPAN analysis TWINSPAN along axis 1, but not along axis 2 for the intensive plots at the lower boundary (Fig. 3A). Pearsons correlation analysis indi of the Coleogyne shrubland in Lucky Strike cated that axis 1 of the stand ordination is sig Canyon (Fig. 2). Species in group A are typical nWcantly correlated (P < 0.001) with eleva of Larrea-Ambrosia stands dominated by Lar tion, soil depth, topography, and percent soil rea tridentatll and Yucca schidigem. Species in cover in the Spring and Sheep Mountain ranges group B, typical of the lower half of the lower ('rable 1). Axis 2 is signillcantly correlated with Coleogyne ecotone and dominated by Ambmsia elevation, topography, and type of ground sur dumosa, are generally short shrubs. Species face (Table 1). In Lucky Strike Canyon hath in group C are typical of the upper half of axis 1 (r = 0.90, P < 0.001) and axis 2 (r = the lower ecotone, with Yucca brevifolia and 0.40, P < 0.001) of the stand ordination are Acamptopappus shockleyi as the most abun significantly correlated with elevation (Table dant species. Species in group D are typical of 2). DECORANA also showed a significant stand the nearly monospecific Coleogyne stands, group segregation along axis 1, bnt not along
LUCKY STRIKE CANYON
1.0
0.9
0.8
0.7
0.6
0.5 EIGENVALUE 0.4
0.3
0.2
0.1
0.0 OPEC ENVI LATR PAPA AMDU KRPA A ACSH CORA ECPO • SAME ERP'1\: OPRA QPAC CELA LYAN GRSP HYSA PSFR PSCO EPNE MATO OPBA YUSC MESP YUER
1-1- ---', 'L -" 1'- -', L' ---',
A B C o SPECIES GROUPS
Fip;. 2. Dcndrogr.tm of specie... groups identified hy lWlNSPAN from 6 elevations throughout the lower Coloogyne ecotone in Lucky Strike Ca.nyon. Major groups were arranged by elevation from the lowest (group A) to the highest (grollP D). Species are arranged alphabetically within groups. See Appendix 1 for species abbreviations. 1997] COJ,EOGYNE GRADIENT ANALYSIS IN SOUTHERN NEVADA 159
'00 ~--~---~--~--~ TABLE L Results of Pearsons correlation analysis by A, SPRING AND SHEEP MOUNTAIN RANGES matching the 1st and 2nd axis of stand ordination scores STANDS acquired from DECORANA to various physical factors in 400 the Sprlllg and Sheep Mountain ranges (Fig. lA). r is the o GROUP 1 coefficient oflinear correlation. • • GROUP ::; 300 • L1 GROUP 3 Factor 1st axis 2nd axis • GROIlP 4- ~levSlliorl -0.79*** -0.24*** 200 Soil depth -0.54*** _0.03 NS Percent soil cover -0.35*** -0. IONS Topography -0.33*** -0.24** '00 o Ground surface O.lONS --O.20NS Plot aspect O.lONS O.03 NS .. Percent rock cover 0.03 NS O.WNS o_'---:-::-----:c':'+--::::----:c a 200 400 600 800 ~'p < 0.01 AXIS 1 n~p < O.OOJ Ns = non.~ignificallt
600 ,-~--~-~-~-- B, SPRING AND SHI':EP MOUNTAIN RANGES and soil depth in the Spring and Sheep Moun SPUOIES '00 tain ranges. Four major species groups were o GROUI' A also distributed along the lower Co!eogyne eco 400 THIiO•• OPEA • GROUP B o E!'NE tone in Lucky Strike Canyon and were domi '"U o GROUP C liESP 0 N 300 nated by Larrea tridentata, Ambrosia dumosa, s.o.""o@ AIWU • GROUP D ~ CORA LYAiP x PIIIO JOOS 0 J.\.'l'R Yucca brevifalia, and Coleogyne ramosissirna, < 200 • A1<"rR • 0 •• >,"PA \'\lER YUSC respectively, determined by elevation. EPVj "" • GIlSP '00 CIlN .YUBA 0 In southern Nevada, Coleogyne shrublands o ENVI OUSAO 0 HYSA establish at mid-elevations with well-drained OPEC -. 0 o SAl)() colluvial slopes between elevations of 1300 I'RI'A. -100 and 2100 m. Co!eogyne stands rarely establish -200 o 200 400 600 800 above 2000 m unless they are situated on south AXIS 1 facing slopes; Coleogyne sometimes establishes below 1300 m on north-facing slopes (Lei and Fig. 3. Ordination of stand (A) and species (B) groups determined by TWINSPAN from the Spring and Sheep Walker 1997). Callison and Brotherson (1985) Mountain ranges across the entire elevational range of noted that Coleogyne is the most abundant Coleogyne. Some uncommon species and overlapping species in its community and may contribute points were eliminated for ease ofinterpretation and visu over 75% of the total vegetation cover, while alization in the species ordination diagram. See Appendix 1 for species abbreviations. associated species may contribute less than 15%. Our data confirmed that the Coleogyne plant community is nearly monospecific, with axis 2 in Lucky Strike Canyon (Fig. 4A). Corre other species comprising less than 28% of the lations of stand groups vary with each environ total number of species among the 15 transects mental factor. Figures 3B and 4B illustrate the on 2 mountain ranges. Our data show that distribution ofTWINSPAN species groups and Coleogyne is a dominant shrub in its vegeta individual species along DECORANA axes 1 tion type in terms of height and cover. Many and 2 in the Spring and Sheep Mountain ranges shrub species occur around the periphery of and Lucky Strike Canyon, respectively, reflect Coleogyne canopies because Coleogyne roots ing differences in the OCCUlTence of various do not release chemical toxins to repress the plant species over different elevations and vege growth and development ofother nearby species tation zones sampled in southern Nevada. (Bowns and West 1976), which is in contrast to Larrea roots (Mahall and Callaway 1992). Cacti DISCUSSION are sparsely distributed in the Co!eagyne slnub lands; many cacti are vulnerable to prolonged Four major species groups that were domi winter freezing temperatures (Larson 1977). nated by Ambrosia dumosa, Coleogyne ramo The greatest woody plant diversity is found in sissima, Gutierrezia sarothrae, and Pinus mono~ or near edges of washes within the Larrea phylla, respectively, were found to be distrib Ambrosia community and tbe lower Co!eogyne uted along gradients determined by elevation community. 160 CREAT BASIN NATURALIST [Volume 57
TAli!.¥: 2. Results of Pcarsons correlation analysis by matching the 1st and 2nd axis of stand ordination -"cores "'" A. LUCKY STRIKE CANYON. STANDS o GROUP 1 acquired from DECOltA A to various physiC"oil fa-doN; in 250 the lower Coleogyne elcvahonal bounuary at Lucky Sh-ike • GR
500 The lower Coleogyne boundary generally B. LUCKY STmKE CANYON. SPECIES contains relatively high species richness. This ~'o url0 "'" m. vegetation zone consists of floras from both 0 o GROUP A ?5tfl O. iIll!:3P L",na-Ambrosia and Coleogyne shrublands. "'" "" • GR Mountain ranges (extensive survey). Decreased and in structuring vegetation zones in south air and soil temperatures, increased precipita ern Nevada. Relationships between the distri tion, and increased soil moisture are associated bution of sample stands and various physical with increases in elevation for the extensive factors are purely correlative. Experimental, transects. Factors such as topography and per functional, and ecosystem approaches are cent soil cover show significant, but weak, cor required to further understand relationships relations with stand ordination scores, indicat between distributions ofplaut communities and ing some influence of these parameters on the environmental factors in southern Nevada. final groupings of stand aud species identified Moreover. extensive studies of these physical by TWINSPAN aualysis. Soil depth is a signif factors in ColeogylW shrublauds across its entire icaut physical factor limiting Coleogyne distri geographical range in the southwestern deserts bution. Coleogyne communities exbibit rela are required to determine ecological require tively shallow soil depth due to caliche layers ments of Coleogyne and the specific environ 30-50 em beueath the soil surface (West 1983). ment it occupies. The area 0-10 em below the soil surface con tains few ColeogylW roots; only roots ofaunuals ACKNOWLEDGMENTS were identified in this zone (Bowns and West 1976). Shrubs can trap aeolian materials, aud We express our gratitude to Yin-Chin Lei greater root activity and weathering cause a de and Steven Lei for collecting vegetation and pression ofthe petrocalcic layer directly under environmental data. Anna Sala assisted with shrubs (West 1983). Root biomass is primarily the multivariate analyses (TWINSPAN and located between 10 and 30 em and is nega DECORANA). We also thank the Motor Pool tively correlated with soil depth in Coleogyne and Department of Biological Sciences of the communities (Bowns aud West 1976). Soil depth University of Nevada-Las Vegas for providing is shallowest in ColeogylW communities but logistical support. deepest in Pinus-Juniperus woodlands. Shallow soils result in low root:shoot ratio and limited LITERATURE CITED root development in the ColeogylW communi ties (West 1983). BEATLEY, J. C. 1976. Vascular plants of the Nevada Test DECORANA results ofaxis 1 ofstand ordi Site and central-southern Nevada: ecological and geographical distributions. National Technical Infor nation scores showed a significant correlation mation Service, U.S. Department of Commerce, between the distribution of sample plots and Springfield, VA. elevation at the lower Coleogyne ecotone of BOWNS, J. E. 1973. An autecological study of blackbrush Lucky Strike Cauyon. DECORANA axis 2 did (Coleogyne ramosissima Torr.) in southwestern Utah. not correlate significantly with auy physical fac Unpublished dissertation, Utah State University, Logan. tors except elevation. Habitat aud topography BOWNS, J. E., AND N. E. WEST. 1976. Blackbrush (Coleog did not chauge significantly to enhance hetero yne ramosissima Torr.) on southern Utah rangelands. geneous environmental conditions throughout Utah Agricultural Experiment Station, Research the lower Coleogyne elevational boundary. Report 27, Department of Range Science, Utah State Hence, elevation is the most essential inde University, Logan. pendent variable to detect staud and species CALLISON, J., AND J. D. BROTHERSON. 1985. Habitat rela tionship of the blackbrush community (Coleogyne groupings on axis 1 in Lucky Strike Canyon ramosissima) ofsouthern Utah. Great Basin Natural and the Spring and Sheep Mountain ranges in ist 45: 321-326. southern Nevada (Tables 1,2). HILL, M. O. 1979a. TWINSPAN: a Fortran program for Certain physical factors appear to play an arranging multivariate data in an ordered two-way essential role in limiting ColeogylW distribution table by classification of individuals and attributes. Ecology and Systematics, Cornell University, Ithaca, to a well-defined elevational band between NY approximately 1050 aud 2150 m in the Spring ___. 1979b. DECORANA: a Fortran program for de and Sheep Mountain ranges near Las Vegas, trended correspondence analysis and reciprocal aver Nevada. Studying au elevational gradient anal aging. Ecology and Systematics, Cornell University, ysis of the Mojave Desert plant communities Ithaca, NY. would contribute to au understanding ofwhich HUNTER, K. L., AND J. R. McAULIFFE. 1994. Elevational shifts of Coleogyne rarrwsissima in the Mojave Desert physical factors are most important in deter during the little Ice Age. Quaternary Research 42: mining the current distribution of Coleogyne 216-221. 162 GREAT BASIN NATURALIST [Volume 57 Kl~l\~r> M., AND P. COKER 1992. Vegetntjon description and Amelam;hier utahensis Koehne T AMUT analysis: a [>l'