Skip to main content

Full text of "DTIC ADA1028969: Landform-Vegetation Relationships in the Northern Chihuahuan Desert,"

See other formats


A0>A102 696 ARMY ENGINEER TOPOGRAPHIC LABS FORT BELVOIR VA F/6 8/6 

LANDFORM-VEGETATION RELATIONSHIPS IN THE NORTHERN CHIHUAHUAN DE—ETCTU) 
MAR 61 M B SATTERWHITEi J EHLEN 

UNCLASSIFIED ETL-R020 NL 




















ADA102896 


'"3 

jO 




IWCLA SS IFI S B - 

security classification of this page (Wh«n Dmta Entarmd) 


REPORT DOCUMENTATION PAGE 

READ INSTRUCTIONS 

BEFORE COMPLETING FORM 

1. REPORT number j2. GOVT ACCESSION NO. 

ETL R020 . filCl S'/hr 

3. RECIPIENT'S CAT ALOO NUMBER 

*:-TITLE (and Subiltia) 

;ANDF0RM-]|{£GETATI0N pLATIONSHIPS IN THE NORTHERN 
' CHIHUAHUAN DESERT " . 

S. type OF REPORT ft PERIOD COVERED 

Paper 

- 

6. PERFORMING ORG. REPORT number 

7. AUTHORfa; ~ ^ 

Melvin B.y^Satterwhite and Judy Ehlen 

B, CONTRACT OR GRANT NUMBERr*; 

9. PERFORMING ORGANIZATION NAME AND ADDRESS 

10. PROGRAM ELEMENT. PROJECT. TASK 
AREA ft WORK UNIT NUMBERS 

/ ■ 

11. CONTROLLING OFFICE NAME and ADDRESS .. 

US Army Engineer Topographic Laboratories ' 

Ft. Blevoir, VA 22060 

12. ftCPORT DATE 

; 29 Marctr*l‘981 ' 

13. NUMBER OF PAGES 

26 

u. MONITORING AGENCY NAME ft AODRESS('lf dff/aranf Irom CantrotUng Otttca) 

mw' 

15. security CLASS, (of thia raport) 

15a. DECLASSIFICATION/DOWNGRADING 
SCHEDULE 


1$. distribution statement ro/(AU RaporO 


Approved for public release; distribution unlimited 


Dt? 


17. distribution statement (of tha mbatrmct antarad In Block 20, It ditfarant from R> 


W 




te. SUPPLEMENTARY NOTES 


19- KEY WORDS (Continum on r*v«r«« midm it ndcmmmmry and tdantlfy by block numbor) 


remote sensing 
air photo interpretation 
geobotanical studies 
Chihuahuan Desert 
arid climate 


land forms 
plant communities 
grassland 
shrublands 
soil texture 


soil depth 


rC/Mbnja aa rmrarmm m{^ It n acmaa a rr aad Idontltr br block numbmr) 

The description and monitoring of environmental resources in arid regions 
can be a formidabele undertaking requiring substantial resources. These effort^ 
can be expedited by using remote sensing techniques. There remains, however, 
a need for correlating features that are readily extracted fro^ the imagery 
with anticipated soil and vegetation conditions on the ground! Landform featur^ 
provide a basis for the assessment of soil and vegetation conditions and 
these features can be readily identified from aerial photogr^hy and, to a 
certain extent, from Landsat imagery. The purpose of this ^udy was to evaluat^ 


DO /; 


FORM 
AN n 


1473 


EDITION OF I NOV 95 IS OBSOLETE 


UNCLASSIFIED 


security CLASSIFICATtON OF THIS PAGE Data Entart 












Block j^20 cont 


landform features as a basis for the assessment of soil and vegetation 
conditions. The study was conducted on 650,000 hectares in the northern 
Chihuahuan Desert (south-central New Mexico and western Texas). 

Landform conditions and plant communities were identified from an analysis] 
of stereo panchromatic aerial photography, and were evaluated in detail by 
intensive field investigations. Soil conditions of the various landforms identj 
fied from the imagery were established by the laboratory analysis of field sam-| 
pies. The distribution of the plant communities was closely correlated to landf 
form conditions and the edaphic factors affecting plant-available soil-water, 
soil texture, soil depth, infiltration, and slope.^t;^he most frequent landform-j 
soil-vegetation relationships in the study area wer^ (1) sand dunes and sand- 
covered alluvial fans with mesquite/broom snakeweed/grass and sand sage/grass; 
(2) washes, lower alluvial fans, and playas with tarbush/grass and burro grass/] 
tobosa grass on deep clay, silty clay, and clay loan soils; (3) dissected lime¬ 
stone hills, mesa (limestone), and upper allucial fans with vreosote and grama 
grass/partheniura on shallow clay, sandy clay loam, clay loam, and silty loam 
soils; and (4) mesa (limestone), intermedaite and lower alluvial fans, and was! 
with grama grass/creosote/mesquite on deep loam, silty loam, clay, clay loam, 
and sandy clay loam soils. Establishment of landform-plant community-soils 
relationships such as these can facilitate environmental resource inventories 
and environmental monitoring activities. 



\ 




i 


oC 



UNCLASSIFIED 


SECURITY CLASSIFICATION OF THIS PAGEri*Ti»n Data Enteratf) 












LANDFORM-VEGETATION RELATIONSHIPS 
IN THE NORTHERN CHIHUAHUAN DESERT 


Melvin B. Satterwhite and Judy Ehlen 
Center for Remote Sensing, Research Institute 
U. S. Army Engineer Topographic Laboratories 
Fort Belvoir, Virginia 22060 

March 19 81 


ABSTRACT 


The description and monitoring of environmental resources in arid regions 
can be a formidable undertaking requiring substantial resources. These efforts 
can be expedited by using remote sensing techniques. There remains, however, 
a need for correlating features that are readily extracted from the imagery 
with anticipated soil and vegetation conditions on the ground. Landform fea¬ 
tures provide a basis for the assessment of soil and vegetation conditions and 
these features can be readily identified from aerial photography and, to a 
certain extent, from Landsat imagery. The purpose of this study was to eval¬ 
uate landform features as a basis for the assessment of soil and vegetation 
conditions. The study was conducted on 650,000 hectares in the northern 
Chihuahuan Desert (south-central New Mexico and western Texas). 

Landform conditions and plant communities were identified from an analysis 
of stereo panchromatic aerial photography, and were evaluated in detail by 
intensive field investigations. Soil conditions of the various landforms identi¬ 
fied from the imagery were established by the laboratory analysis of field sam¬ 
ples . The distribution of the plant communities was closely correlated to land- 
form conditions and the edaphic factors affecting plant-available soil water; 
soil texture, soil depth, infiltration, and slope. The most frequent landform- 
soil-vegetation relationships in the study area were: (l) sand dunes and sand- 
covered alluvial fans with mesquite/broom snakeweed/grass and sand sage/grass; 

(2) washes, lower alluvial fans, and playas with tarbush/grass and burro grass/ 
tobosa grass on deep clay, silty clay, and clay loam soils; (3) dissected lime¬ 
stone hills, mesa (limestone), and upper alluvial fans with creosote and grama 
grass/parthenium on shallow clay, sandy clay loam, clay loam, and silty loam 
soils; and (4) mesa (limestone), intermediate and lower alluvial fans, and washes 
with grama grass/creosote/mesquite on deep loam, silty'loam, clay, clay loam, 
and sandy clay loam soils. Establishment of landform-plant community-soils 
relationships such as these can facilitate environmental resource inventories and 
environmental monitoring activities. 


81 8 14 094 




INTRODUCTION 


Relationships between vegetation, soils, and landforms can be readily 
determined by using stereoscopic aerial photography and air photo inter¬ 
pretation techniques augmented by minimum field efforts. Although land- 
forms can be readily characterized on most photography, and general soil 
conditions can be identified by phcrographic interpretation or by association 
with specific landforms, plant sfticies often cannot be directly identified 
on aerial photography. Inferential determinations of plant communities, 
however, can be made on the basis of photo tone and photo texture dif¬ 
ferences. In arid lands where vegetation is closely adjusted to various 
edaphic conditions (soil texture, soil depth, soil water, salinity, and 
groundwater), the photo tone/photo texture relationship, used in conjunction 
with landform data, can facilitate predictions of plant community conditions. 

The purpose of this study was to identify and describe the relationships 
between vegetation, landform, and soil conditions using manual photo analysis 
and interpretation techniques together with phytosociological techniques and 
to determine the extent to which one or more of these three factors can be 
used as an indicator of the other factors. 

This study was conducted in south-central New Mexico and western Texas 
where the vegetation is largely dominated by shrub species (figure 1). These 
shrublands are thought to be disclimax communities that developed from the 
true climax desert grassland communities in response to man-induced changes 
in environmental conditions. The shrub and grass species, now adjusted to 
the new environmental conditions, from identifiable plant communities. 



Figure 1: Location of the Study Area 



2 









PROCEDURES 



t. 

1 







The aerial photography used was panchromatic 22.9-x 22,9-cm (9-x 9-in.) 

1:50,000 scale stereo aerial photography laid as three 102-x 152-cm uncon¬ 
trolled photomosaics. Landform units were identified and described from the 
three-dimensional characteristics of shape, relief, slope, arrangement, and 
orientation; the drainage characteristics of pattern, density (spacing of 
drainageways), gulley cross sections, gulley gradient (uniformity and steep¬ 
ness), and degree of incision; and the bedrock characteristics of rock layer¬ 
ing, dip, type, and structure. Four major landform categories (mountains/ 
hills, alluvial fans, basin areas, and washes) were divided into 14 landform 
units. The percentage of the study area occupied by each landform unit was 
determined using a grid sampling technique. Soil samples were collected at 
selected sites and were analyzed in the laboratory for textural and moisture 
conditions. 

Plant communities were identified from phytosociological data collected 
at 298 sites. These were representative of the plant community and associated 
conditions found in the land cover mapping units. Species composition, species 
percent ground cover, and ancilliary data representative of the plant community 
were collected using a 20-x 20-m quadrant. Perennial grass species and woody 
plants provided the basis for the plant community descriptions because of their 
constant presence and reoccurring nature. 

The tabular comparison method was used to cluster the phytosociological 
data and to assist in the identification of plant communities (Mueller-Dombois 
and Ellenberg, 1974; Kuchler, 1967). The phytosociological data were partially 
analyzed using the automated clustering technique developed by Lieth and Moore (1971). 

Three major land use categories (grassland, shrubland, and forestland) and 
22 land cover mapping units were identified from the phytosociological data. 

For each species, the ground cover class, absolute frequency, relative fre¬ 
quency, and an important coefficient [(cover x frequency)/2], were determined. 

The procedures followed for characterizing each community and for assigning 
botanical names to the plant communities are discussed in Satterwhite and Ehlen 
(1980). Coefficients of similarity were calculated between all the plant commu¬ 
nities . 

Probable plant communities were mapped on the aerial photography by eval¬ 
uating the photo tone and photo texture differences and plant physiognomic char¬ 
acteristics. Relationships between the plant communities and the photo mapping 
(land cover) units were identified by evaluating these same characteristics at 
each sample site and from numerous random observations made en route between 
sample sites. Consequently, an association can be made between the plant commu¬ 
nities described from the phytosociological data and the photo patterns mapped from 
the aerial photography. 

The relationships between the plant communities and specific landform units 
were determined by using rhe data collected in the field together with data collec¬ 
ted by simultaneous sampling of the landform and land cover maps. Data from the 
maps were collected at 6,033 sample points using the grid point sampling techniques. 
The frequency of each relationship was calculated as a percentage of the total sam¬ 
ple . 







RESULTS 


Landform and land cover units were mapped on three photo mosaics; for 
convenience, only one photo mosaic is presented here. All three photo mosa¬ 
ics and their accompanying landform and land cover maps are reported else¬ 
where (Satterwhite and Ehlen, 1980). Geographic names are shown on figure 2. 


Landforms 

The relative percentages of the major landforms and their subunits with¬ 
in the study area are shown in Table I, Figure 3 illustrates the spatial 
distribution of the landform subunits on one of the photo mosaics. 


Table I: 

Landform 

Units and Their Percentages of the 

Study 

Area 

Major 

Map 


Percentage of 

Landform Unit 

Symbol 

Landform 

the 

Study A.rea 

Mountains/hills 

A1 

Mesa 


15.8 


A2 

Highly dissected hills 


16.1 


A3 

Rugged, sharp-crested mountains 


2.3 

Alluvial Fans 

B1 

Primary, high elevation fans 


10.8 


B2 

Secondary, high elevation fans 


1.2 


B3 

Mottled, intermediate elevation 

fans 

6.8 


B4 

Dark-toned, lowest elevation fans 

3.6 


B5 

Fans covered with aeolian sand 


3.5 


B6 

High elevation, anomolous fans 


0.7 

Basin Areas 

Cl 

Light-toned, speckled sand dunes 

30.2 


C2 

Dark-toned, rough-textured sand 

dunes 

2.6 


C3 

Low, smooth areas 


2.5 


C4 

Small, dark-toned depressions 


1.5 

Washes 

D 

— 


2.0 


Total; 99.6 


4 










- * 5 ^ ^ 

• ■ ’fe 












r - * / -* ^ V^ 



.„..,&0r9grond« - * . *. •<'.>• —•^•'J *"’* 

'•«ono« Cofl^.;^ <r . ) V ' '• 

^ 4 =^' i-\X' 


.iS^* 







^-■4* 



.-■ :-HiQy'W*otio?-^.t^^.- ^ 

^^■^' t'4* ”' :::vA. 

•^-'v.vVv- -.. 

2.. ,>' -j-^'^ •-."K; .=?'.^59 



5 -■Xo*p'»" 


Sinol: 








tOISON 






Figure 2; Geographic Names in the Study Area 


5 


























The mountains/hills landform comprises 34 percent of the study area and was 
identified by its three-dimensional shape in the stereo images. These areas 
are the highest and most rugged parts of the study area. Drainage patterns are 
dendritic and a high degree of structural control is often exhibited locally. 

The drainageways are usually incised, and density is generally high. Three 
subunits were identified in this landform category: (1) large, flat-topped 
mesas (Al) that are composed of gently dipping limestone with minor sandstone 
and shale; (2) large, highly dissected rock masses (A2), primarily interbedded 
limestones or limestones and shales, along with many relatively small, isolated 
rock bodies (inselbergs); and (3) large areas of rugged, sharp-crested mountains 
(A3) composed of coarse-grained, intrusive, igneous rocks, which also include 
some inselbergs. 

The alluvial fans comprise 27 percent of the study area. They were iden¬ 
tified by their distinctive fan shape, distributary drainage pattern, moderate 
slope (1 to 5 percent), and topographic position between the mountains and hills 
and the low, flat basin areas. Relief and drainage characteristics on the allu¬ 
vial fans very with topographic position and the degree of fan development. 

Fans occur as individuals and as broad aprons (bajadas). The alluvial fans were 
separated into six discrete subunits. 

Flat, with little relief, and occupying the lowest elevations, the basin 
areas comprise 37 percent of the study area and are the largest landform units. 
Coppice sand dunes are common in this landform. Few drainageways can be seen, 
although some channel-like depressions were identified on the air photos. The 
gentle regional slope of the basin areas is to the north in the Tularosa Basin, 
and to the south in the Hueco Bolson (figure 2). The boundary between these 
basins is difficult to identify on the air photos. The basin areas were divided 
into four discrete units. 

The washes (the major stream channels upstream, from where the drainage pat¬ 
tern becomes distributary or where the major channels enter the lowest alluvial 
fans) comprise 2 percent of the study area. Washes include the immediate over¬ 
flow areas associated with the channels. Distinct, steep-sided channels were 
occasionally noted on the alluvial fans. Most of the drainageways in these areas, 
however, lacked distinct channels and were identified by tonal pattern. This 
pattern consists of very dark and very light tones caused by the contrast between 
dense vegetative cover and bare soil. Washes in the mountains/hills landform have 
either U-shaped or V-shaped cross sections, whereas in the alluvial fan complex, 
the cross sections are box-shaped. 

Soils 

Soil texture and soil depth data, summarized in Table II, were collected in 
the field for most of the landform subunits. The soil conditions form four groups 
defined by textural and depth differences: (1) sand, loamy sand, and sandy loam 
soils more than 30 cm deep; (2) clay, clay loam, loam, gravelly loam, and grav¬ 
elly clay loam soils less than 15 cm deep; (3) clay loam, clay, sandy clay, and 
sandy clay loam soil 15- to 30- cm deep; and (4) clay loam, silty clay loam, 
clay, and sandy clay soils more than 30 cm deep. Group 1 was associated with the 



Table II: Lan^lform, Soil Texture, and Soil Depth Relationships 



Sandy clay loam, loamy sand 30- to 45- cm or more 








basin area landforms; group 2, with the uppermost alluvial fan unit and 
the moutains/hills units; group 3, with the lower and sand-covered allu¬ 
vial fans and the washes; and group 4, with the mesa and the upper and 
sand-covered alluvial fan units. 


Plant Communities 

The spatial distribution of the land cover units on one photo mosa¬ 
ic is shown in figure 4. The statistics describing the areal extent of 
each of the 22 land cover units are presented in Table III. 

Grasslands and grass-shrublands occupy 37 percent of the study area, 
are widely distributed, and occur on most landform units. Five discrete 
grassland communities were identified; (1) Bouteloua eriopoda - Bouteloua 
curtipendula , (2) Bouteloua curtipendula - Bouteloua uniflora , (3) Scleropogon 
brevifolius - Hilaria mutica , (4) Sporobolus cryptandrus - Sporobolus 
flexuosus , and (5) Sporobolus giganteus . These communities were mapped as 
undifferentiated grassland (10) because of their small size and the dif¬ 
ficulty in differentiating them at the 1:50,000 photographic scale. The 
shrub species. Acacia constricta , Artemisia filifolia , Fl ourensia cernua , 
Larrea tridentata , Parthenium incanum , or Fr osopis glandulos a. can occur 
as scattered individuals with 1 to 5 percent cover in the grassland areas. 

Shrublands, the major physiognomic group, occupy 5E percent of the 
study area. The major shrub species are Acacia constricta , Artemisia 
filifolia , Flourensia cernua , L. tridentata , Part henium incanum , and Prosopis 
glandulosa . Less frequent shrub species, Chi lops is chi_Hensis, Falliqua 
paradoxa , and Thelesperma longipes can form dense stands along some drain- 
ageways. These species were rarely observed in the upland areas. The 
grasses commonly forming the understory in these shrub communities are B. 
eriopoda , gracilis , fi. mutica , Muhlenbergia spp. , Scler ' pogon brev ¬ 
ifolius , or Sporobolus flexuosus . 

The distribution of forestlands is very limited. For all practical 
purposes, the forested areas of any size (more than 5 hectares) are in the 
Sacramento and Organ Mountains in the northeastern and northwestern parts 
of the study area (figure 2). The major tree species, J uniperus monosperma , 
Pinus edulis , and Quercus u n dulata are limited to these areas and to 
isolated draws along the utero Mesa escarpment. The associated shrub species 
in the forestlands are Agave sp., Ce rcoparpus montanu s, Chrysothamnus sp., 
Molina sp., and X anthocephalu m Sarothrae . Acacia sp., Celtis sp., Populus 
sp., Prosopis glandulosa , Tama rix ramosissima , and Ulmus sp. are associated 
with some cattle watering tanks (ponds). Fros opis g landulosa occurs as a 
small tree in these areas, but its normal stature is a short- to medium¬ 
sized shrub. Chilopsis ch i liensis occasionally occurs as dense stands of 
tall shrubs and small trees in major drainageways. 


10 


























Table III: Land Cover Mapping Units and Ihelr Percentages of the Study Area 


Physiognomic Group Percent 

Group of Area Land Cover Unit 


Grassland 


Shrubland 


Grassland (10) 

Grass - Larrea trldentata (11) 
Grass - Flourensla cemua - Larrea 
trldentata (12) 

Grass - Acacia constricts (13) 
Grass - Artemisia filifolla (14) 
Grass - Prosopls elandulosa (15) 
Grass - Parthenlum Incanum (16) 


Larrea trldentata (20) 

Larrea trldentata - Grass (21) 

Larrea trldentata - Parthenlum Incanum 

- Grass (22) 

Larrea trldentata - Prosopls elandulosa 

- Xanthocephalum Sarothrae (23) 

Larrea trldentata - Flourensla cemua 

- Grass (25) 

Acacia constricts - Grass (30) 

Acacia constricts - Larrea trldentata 

- Grass (31) 

Flourensla cemua - Grass (40) 
Flourensla cemua - Larrea trldentata 

- Grass (41) 

Prosopls glandulosa - Xanthocephalum 
Sarothrae - Atrlplex canescens - 
Grass (50) 

ProsOTls ^andulosa - Larrea trldentata 


Prosopls elandulosa 
(51) 

Prosopls glandulosa 
folia (52) 
Artemisia filifolla 
Artemisia filifolla 
dulosa (61) 


- Artemisia fill- 

- Grass (60) 

- Prosopls glan- 


Percent 
of Area 


Forestland 


Other 


Junlperus monosperma - Quercus 
undulata (70) 


Total 100.1 


100.1 


12 









Landform/Soils/Plant Community Relationships 


The relationships between landform and land cover show that landform 
can be used as an indicator of both vegetation and soil conditions. Data 
describing the associations between the plant communities and landform u- 
nits are summarized in Table IV. The value given at the intersection of 
a row and a column is the percentage of the study area occupied by a partic¬ 
ular landform/plant community association. The Chi square test (95 percent 
level of confidence) confirmed what the frequency table shows: there is a 
relation between the plant communities and landform units. Figure 5 il¬ 
lustrates the association of the major and minor plant communities with a 
landform unit. 

Mountains/Hills Plant Communities . The plant communities associated 
with the mesa landform are primarily grassland communities (lOA and lOB); 
the Bouteloua eripoda - Bouteloua curtipendula grassland (lOA) is the 
most common. Soils are silty clay loam, loam, and gravelly clay loam in 
texture and are generally more than 30 cm deep. In some swales, the soil 
depth is 60 cm or more to a petrocalcic horizon or bedrock. The mesa 
soils in some areas, however, are less than 15 cm to bedrock or a petro¬ 
calcic horizon and contain substantial amounts of gravel. 

The plant community most commonly associated with the highly dissected 
hills is the Grass - Parthenium incanum community (16), which occurs pri¬ 
marily on this lanform unit. Grassland communities are limited, with 
Bouteloua curtipendula - Bouteloua uniflora (lOB) the most common. Grass- 
Acacia constricta (13), Acacia constricta - Grass (30), and Acacia constricta- 
Larrea tridentata - Grass (31) communities occur alm-'st exclusively on 
this landform unit. The Juniperus monosperma - Quercus undulata community (70) 
occurs on this landform unit along the southern flanks of the Sacramento 
Mountains. Soils are shallow (less than 15 cm) and their textures are clay 
and clay loam. Grass species in the Acacia constricta communities are H. 
mutica and Scleropogon brevifolius , which occur on clay soils 15- to 30- cm 
deep. 


Bouteloua curtipendul a - Bouteloua uniflora (lOB) and Juniperus mono ¬ 
sperma - Quercus undulata (70) communities are commonly found on the rugged, 
sharp-crested mountains. The latter community occurs in the Organ Mountains. 
Larrea tridentata - Parthenium incanum - Grass (22) is a minor associated 
community. Soils are shallow (less than 15 cm), and their textures are grav¬ 
elly loam and sandy. 

Alluvial Fan Plant Communities . Larrea tridentata is either the dominant 
or the major associate species in plant communities on the alluvial fans. The 
major plant communities on the upper alluvial fans (B1 and B2) are Larrea 
tridentata (20) and Larrea tridentata - Grass (21). Soils on these fans are 
less than 15 cm to bedrock or a petrocalcic horizon. Soil textures are clay, 
clay loam, and loam. Some soils contain substantial amounts of gravel. 


13 










PCN = Plant Community Number .0 = Frequencies <0.1% U = Frequency >1.0 percent 

0.0 = No Observations 










PLANT COMMUNITIES 



Aausnbajj %0Z+ Asuanbaij %0Z 0l 
tajtiunujiuoo iofaiAi saiiiunuiuiOQ joui^ 


15 


Figure 5: Plant Cominunity Associations with Specific Landforms. The Width of each Landfortn Unit is Proportional to its 
Relative Percentage of the Study Area. The Frequencies of Major and Minor Plant Communities Reflect the 
Relative Percentage that that Community Occurs on the Landform. For example, the Bouteloua Grassland (lOA) 
Occurs on More Than 20% of Landform A1 and is the Major Plant Community. 












The major plant communities on the mid-elevation alluvial fans (B3) are 
Larrea tridentata (20) and Larrea tridentata - Prosopis glandulosa - Grass (23). 

The lower alluvial fans are dominated by Flourensia cernua - Larrea tridentata 
- Grass (41), Larrea tridentata - Flourensia cernua - Grass (25), Grass - Flour - 
ensia cernua - Larrea tridentata (12) and Scleropogon brevifolius - Hilaria mu - 
tica (lOC) in addition to the two _L. tridentata communities, 20 and 21. The 
Flourensia cernua - Grass community (40) is found only on the lowest alluvial 
fan unit. This community covers large areas of these fans but its frequency is 
low in regard to the total study area. The grass species commonly found on 
these landforms are _H- mutica , j1. Porteri , or Scleropogon brevifolius . Soil 
textures are clay loam, silty clay loam, clay, and sandy clay; soil depth is 
more than 30 cm. 

The major plant communities found on the sand-covered alluvial fans (B5) 
are the Larrea tridentata - Prosopis glandulosa ~ Xanthocephalum Sarothrae (23) 
and the Prosopis glandulosa - Larrea tridentata (51) communities. The presence 
of Prosopis glandulosa -dominated communities indicates a change in soil condi¬ 
tions compared to the other alluvial fans. Where substantial amounts of sand 
occur, the L. tridentata shrubs, and other shrubs incapable of tolerating par¬ 
tial or complete burial, are eliminated and_P. glandulosa is becoming estab¬ 
lished. glandulosa is capable of adventitious root growth and rapid estab¬ 

lishment on active fan areas. The ^rrea tridentata - Prosopis glandulosa - 
Xanthocephalum Sarothrae community (23) occupies sites where aeolian deposition 
is active, i.e., between the upper alluvial fan communities dominated by L. 
tridentata (20 and 21) and the P. glandulosa communities (50) of the basin areas. 
Soil textures on the B5 landforms are sandy clay loam and loamy sand and soil 
depths are 30- to 45- cm or more. 

The alluvial fans on the Otero Mesa (B6) are vegetated almost exclusively 
with Bouteloua spp. grassland communities, particularly Bouteloua eriopoda- 
Bouteloua curtipendula (lOA). Two isolated areas of Artemisia fjlifolia - Grass 
community (60) were found. Soil textures are gravelly loam and gravelly clay 
and soil depth is more than 45 cm. 

Basin Area Plant Communities . The most prevalent plant community on the 
light-toned, speckled sand dunes is the Prosopis glandulosa - Xanthocephalum 
Sarothrae - Atriplex canescens - Grass community (50). Other communities 
dominated by Prosopis glandulosa , Artemisia filifolia or the Sporobolus spp. 
grasses. Grass - Prosopis glandulosa (15), Grass - Artemisia filifolia (14), 

Prosopis glandulosa - A rtemisia filifolia (52), and Artemisia filifolia - 
Prosopis glandulosa (61), occur on limited areas. Soils in these dune areas 
are sandy loams, loamy sands and sands. Soil depth is normally more than 2 m on the 
dunes, but is less than 30 cm in the interdunal areas. 

The major plant community associated with the large, dark-toned sand dunes 
(C2) is the Artemisia filifolia - Grass community (60). Minor plant communities 
are the Sporobolus cryptandrus - Sporobolus flexuosus grassland (lOD), Grass - 
Prosopis glandulosa (15), and Prosopis glandulosa - Xanthocephalum Sarothrae - 
Atriplex canescens - Grass (50). The sand and loamy sand soils in these areas 
are more than 2 m deep. 


16 








The plant communities in the low, smooth areas (C3) are mostly Sporobolus 
- Sporobolus flexuosus grassland (lOD) and the Artemisia filifolia 
- Grass shrubland (60). Sporobolus cryptandrus and flexuosus are the domi¬ 
nant grass species in both communities. Soils are similar in texture to those 
in the Cl and C2 landforms, but are 1.0 meter or more in depth. 

The major plant community in the small depressions (C4) is Larrea triden - 
tata - Flourensia cernua - Grass (25). The minor communities associated with 
this landform are the Sporobolus cryptandrus - Sporobolus flexuosus grassland 
(lOD) and the Artemisia filifolia - Grass community (60). In the almost circu¬ 
lar depressions, or playas, the grass communities are dominated by Scleropogon 
brevifolius and _H. mutica . The soils in the depressions contain more silt 
and clay than the other basin area soils and are more than 1.0 meter in depth. 

Vi'ash Plant Communities . Along the major drainageways, several plant commu¬ 
nities were encountered which had common species, but which differed in dominant 
species. The major plant community is the Scleropogon brevifolius - Hilaria 
mutica grassland (lOD). The Sporobolus giganteus community (lOE) is also pres¬ 
ent on small areas of the upper reaches. Minor plant communities are Grass - 
Flourensia cernua - Larrea tridentata (12), Flourensia cernua - Larrea tridentata- 


Grass (41), and Larrea tridentata (20). Soils are more than 30 cm deep and are 
clay loam, silty clay loam, clay, and sandy clay in texture. 


DISCUSSION 

Five major plant communities that covered 94 percent of the study area were 
associated with specific landform-soil conditions that in turn accounted for 80 
percent of the landform areas: (1) mesa with the grassland community Bouteloua 
eriopoda - Bouteloua curtipendu la on clay loam, clay, and sandy clay soils more 
than 30cm deep; (2) the interbedded limestone and shale member of the highly 
dissected hills landform with the Bouteloua curtipendula - Parthenium incanum 


community on loamy and silty loam soils less than 15 cm deep; (3) the upper and 
intermediate alluvial fans with the Larrea tridentata and the Larrea tridentata- 


Grass communities on sandy clay loam and clay loam soils more than 15 cm deep; 

(4) the light-toned, speckled sand dunes (coppice dunes) with the Prosopis gland - 
ulosa - Xanthocephalum Sarothrae - Atriplex canescens - Grass community on loamy 
sand, sandy loam and sandy clay soils more than 30 cm deep; and (5) the dark- 
toned, rough-textured sand dunes with the Artemisia filifolia - Sporobolus spp. 
community on sandy, sandy loam, and loamy sand soils more than 65 cm deep. 

The field and laboratory data show that soil depth and soil textural condi¬ 
tions vary with landform; each of the twenty-two plant communities is associated 
more frequently with one particular landform unit. The variation in these edaphic 
factors determines the moisture retention characteristics of a soil (Satterwhite, 
1979) and the potential soil-water-holding capacity is the factor controlling 
plant community distribution. These relationships permit landforms to be used as 
an indicators of plant communities and soil conditions. 

Species growing on the different soils throughout the study area tend to 
reflect these edaphic differences. For example, the drought-tolerant species 
occur on the more xeric sites where soils are shallow and coarse textured. Acacia 
constricta . Fouquieria splendens , Opuntia spp., and Nolina sp. are routinely 



■vVt'iM-’iiirV' 





found on the dissected, interbedded limstones where soils are very shallow 
or non-existent. The amount of plant-available water held in these soils 
is vei 7 small. 

Other species exhibit similar edaphic relationships. L. tridentata , 
Parthenium incanum , Viguieria stenoloba , curtipendula , M. arenacea a nd 
Sporobolus Wrightii are found on the dissected limestone hills and the upper 
alluvial fans where soil depth is 3- to 30-cm and soil textures are gravelly 
clay loam and gravelly loam. These soils are capable of holding substantial 
plant-available water; once the soil water is depleted, they can remain 
droughty throughout the summer and fall growing season. Only the surface 
decimeter(s) may be wetted by the high-intensity, short-duration summer 
rains. The steep slopes of these landforms, combined with slow infiltration 
and percolation rates, prevent recharge of the soil-water reservoir during 
this period. Less intense precipitation events recharge surface and sub¬ 
surface soil horizons during the late fall and winter months. 

Even though the relationships between plant communities and edaphic 
conditions on the lower alluvial fans and washes are similar, the species 
on these landforms can be less drought tolerant. The soils in these areas 
often receive surface runoff from the upper slopes. The shrubs Atriplex 
canescens , Flourensia cemua . Prosopis glandulosa , Rhus aromatica . and 
R. mycrophylla and the grasses H. mutica , Porteri , Scleropogon brevifol- 
ius , and Panicum obtuscum are most common. 

For example, Prosopis glandulosa encroaches along the thread of small 
drainageways on the mid-elevation alluvial fans on the eastern slopes of the 
Franklin Mountains. tridentata , the dominant species on these alluvial fans, 
occurs at slightly higher elevations (less than 0.3 m) on the coarse-textured, 
shallow soils that form the interfluvs. These occurrences of the two species 
on dissimilar microhabitats on the same landform illustrate their relative 
drought tolerances and water requirements. 

Prosopis glandulosa and L. tridentata also occur in limited areas of the 
coppice dune region (Cl). glandulosa occupies the dunes where soil depth can 
be 2 m or more, and L. tridentata occurs as a rare-to-infrequent species in the 
dunal areas where soil depth to the petrocalcic horizon is generally less than 

O. 5m. These two species occupy basin sites comparable to those on the mid¬ 
elevation alluvial fans. I^. tridentata occupies the shallow, droughty soils and 

P. glandulosa occupies soils of greater depth and a potentially greater soil- 
water reservoir. 


CONCLUSIONS 

Aerial photography and conventional air photo interpretation techniques can 
provide a successful way to obtain information about soils, landform, and vege¬ 
tation. These relationships can permit the most easily determined factor on 
aerial photography, soils, landforms, or vegetation, to be used as an indicator 
of the other two factors. Because landforms are directly observable on stereo 
aerial photography, they can be used more easily than either vegetation or soil 
conditions as an indicator. 


18 




Extrapolating these relationships to other areas is a matter for continued 
investigation. Observations made adjacent to the study area show these rela¬ 
tionships are applicable to those areas where the same species occur. Problems 
can arise when extrapolating to other regions, however, because species eco- 
typical differentiation can change tolerance limits, and hence, landform and 
soil relationships, even when comparable landform units and soil conditions are 
encountered. 


REFERENCES CITED 


Kuchler, A. W., 1967, Vegetation Mapping: New York, Ronald Press Co., 472 p. 

Lieth, H., and G. W. Moore, 1971, Computerized Clustering of Species in Phyto- 
sociological Tables and Its Utilization for Field Work. IN Patil, G. P., 

E. C. Pielou, and W. E. Waters (eds). Spatial Patterns and Statistical Dis¬ 
tributions: The Pennsylvania State University Press, University Park, 
Pennsylvania, p. 403-422. 

Mueller-Dombois, D. and H. Ellenberg, 1974, Aims and Methods of Vegetation 
Ecology: New York, John Wiley and Sons, Inc., 547 p. 

Satterwhite, Melvin B., 1979, Evaluating Soil Moisture and Textural Relation¬ 
ships Using Regression Analysis: U.S. Army Engineer Topographic Laboratories, 
Ft. Belvoir, Virginia, ETL-0226, 31 p. 

Satterwhite, Melvin B., and Judy Ehlen, 1980, Photo Analysis of the Relation¬ 
ships Between Vegetation and Terrain Features in South-central New Mexico 
and Western Texas; U.S. Army Engineer Topographic Laboratories, Ft. Belvoir, 
Virginia, ETL-0245, 228 p.