BIG BEND The Impact of Human Use Upon the Chisos Basin and Adjacent Lands NPS Scientific Monograph No. 4

Definition and ecology

This section will present the ecology of the Chisos Basin with examples of impact that presently occur and affect the vegetation. It should be reemphasized that certain of the impacts have been in effect for many years as presented in Chapter 2. However, neither the total ecology nor all examples of impact are included. Those examples included were obvious, at least partially understood, measurable, and thought to demonstrate impact. The science of human impact ecology, exclusive of pollution and grazing studies, is relatively new. In this study, methods which I felt a priori would demonstrate impact were used. Several of these methods proved to be inadequate, but will be presented with their criticisms. Many new sampling techniques and methods of impact recognition must be developed before adequate experimentation can ensue. Another difficulty in this study is that cumulative data from major basic studies on the soil, vegetation, and vegetation dynamics are not presently available.

Herein, impact will be considered any human activity which directly alters or interferes with the vegetation or natural factors acting upon the vegetation. Not included are those indirect factors such as increased carbon dioxide, carbon monoxide, pesticides, light, temperature, and wind currents which could indeed be included by definition and are real in effect.

The generalized pattern of impact or disturbance upon vegetation is that of removing or contributing to the reduction or alteration of existing vegetation. This generally results in bare areas or in a less complete vegetation type, with cover, number of plants, and number of plants species all being reduced. This backward step, frequently called a disclimax, is essentially the opposite of succession, which leads ultimately to the most stable, mesic, and complex vegetation, the climax. The degree of vegetation removal determines the magnitude of the backward step. Unfortunately, in the basin most impact results in completely denuded sites, frequently accompanied by upheaved soil. In the upheaved condition an appreciably greater time is required for revegetation because soil microorganisms, air, temperature, nutrients, and moisture capacity have been greatly altered. Also, the lack of vegetation cover and roots results in decreased organic material and increased light penetration.

The impacted site, depending upon the desires of man, may be left bare, covered by structures, may undergo continual compaction, or be left for "nature" to revegetate. With the exception of the first, the alternatives leave a new site to be conquered by "nature." The site can now be referred to as a disturbed area into which only a few species of plants are capable of invading. These invaders must be capable of rapid growth, have great tolerance for high light intensity, high temperature, low moisture, low humidity, low nutrients, and other harsh conditions. If the species is successful, the site conditions are ameliorated, allowing a succession of less tolerant species to invade.

The vegetational cycle in the Chisos Mountains seems to be initial invasion by several herbaceous plants. These plants are then followed by succulents, shrubs, and finally trees. Table 7 presents a few of the native species found to be involved in succession in Evergreen Woodland-Chihuahuan Desert Formations. The length of this first stage is not known, but based upon the plants present in the main waterline scar produced in 1952. it lasts at least 15-18 years. This is because grasses are still dominant in the scar. The only shrubs-succulents which are just beginning to enter the scar are Dasylirion leiophyllum, Mimosa biuncifera, and Opuntia engelmannii, This is in the Chihuahuan Desert Formation section of the scar.

Table 7. Important native invaders of disturbed areas in the Chisos Basin.

Frequently accompanying these native herbaceous plants into the area are exotic species introduced from the Old World which have finally made their way into the basin. These include: Sorghum halpense (Johnsongrass), Cynodon dactylon (Bermudagrass), Bromus unioloides (rescuegrass), Salsola kali (Russianthistle), Tribulus terrestris (puncturevine), Marrubium vulgare (common horehound), Peganum harmala (harmal peganum), Echinochloa crusgalli (barnyardgrass), and Eragrostis barrelieri (Mediterranean lovegrass). Others which are native but of a ""weedy" nature are Solanum elaeagnifolium (silverleaf nightshade), Cenchrus incertus (coast sandbur), and Alternanthera peploides (mat chaff-flower).

One of the most important aspects of the cycle is that once the shrubs begin invasion, they may remain as individuals for 60-100 years; such is the case for Larrea tridentata, Prosopis glandulosa, and Acacia constricta (Shreve and Hinckley 1937). There is strong evidence that L. tridentata (Went 1955) and P. glandulosa (Sankhal et al. 1965) are capable of producing inhibitory substances which prevent establishment of other species for long periods of time. Abdul-Wahab and Rice (1967) present similar evidence for such inhibitory activity by Sorghum halpense, an abundant species in impacted areas in the basin.

Further discussions of the ecology and effects of the specific forms of impact will be presented under each example. In several cases the sequence deviates from the generalized pattern and must be considered separately. The sequence of examples is not intended to be a rank of their degree of impact.

Forms of present impact

Roads

One of the earliest forms of basin impact, other than grazing, was that of road construction. The first roads in 1934 were dirt, but these were later paved in 1948 and 1958. The two types of roads undoubtedly vary in effect, but information on the extent of impact is not available. Besides the original denuding of the vegetation, impairment of natural drainage by blockage or rerouting is pertinent. The road fill, if taken from a distant area, provides an opportunity for importation of new species and new denuded sites to be occupied. The fill soil can vary so greatly that local native vegetation is extremely slow to invade. Road cuts, made to provide local fill, cause extensive erosion sites and in conjunction with the road shoulder provide miles of continual disturbance. These areas provide constant avenues for new species' invasion. Since the roads have been paved, surface rain runoff has greatly increased on some shoulders, resulting in increased vegetation growth.

The unusually dense thickets of Aloysia lycioides that have developed east of the road to the north and south of the Campfire Circle indicate impairment of drainage here. Road tiles are absent or nonfunctional in two instances. Soil analyses (Tables 2 and 3) from the A. lycioides stand south of the circle demonstrate a damming effect, since the texture (total sand, silt, clay) of the smaller particles is higher with respect to gravel and rock. The soil displays this characteristic as well as higher carbon and phosphorus values, typical of transported soils. The sites which provide the nearest control values for comparison are sites 5 and 7.

Closely correlated with drainage is excess runoff. The most unusual example is the area to the west or below the Ranger Station near the Trailhead. All the water from the Upper Basin parking area and above flows into this region, resulting in excess moisture and transported soils which support extensive populations of A. lycioides. Much of this runoff would normally flow to the ravine behind the Service Station or remain in the downfall area. The Lower Basin exhibits other such examples in and around the group campground circle and along the north and northwest margin of the main campground. Also increasing at an alarming rate along the road to the east of the main campground is Sorghum halpense. On the road fill and ravine floor to the north of the campground are large, well-established clones.

The effect of road cuts and their erosion sites are best exemplified along the switchbacks into the basin below Panther Pass, at the entrance into the Upper Basin complex, behind the lodge, and along the road to the pumphouse below the main campground. The latter two are in need of immediate control.

The role of road margins as disturbed areas was investigated along the Upper Basin to Lower Basin road. Every 50 paces, from the junction to near the main campground, 14 line-intercept transects perpendicular to and alternating from side-to-side of the road were sampled. Coverage, to the nearest inch, was recorded for each species touched by the line from the edge of the asphalt to the line's end. This end point was determined to be where the vegetation remained stable in composition when compared to that traversed. Usually the transects were 16-20 ft long, with the first 4-6 ft being the more barren shoulder or drainage path, and that beyond the shoulder, the natural vegetation.

The results indicate that the shoulder had a total vegetation cover of 28.6%, whereas that beyond the shoulder had 82.4%. The latter area is less disturbed by erosion or denuding by blading. The following species contributed the respective percentages to the shoulder cover: Bothriochloa barbinodis, 26; Bouteloua curtipendula, 19; Viguiera stenoloba, 17; Larrea tridentata, 10; Bouteloua gracilis and B. eriopoda, 8; Parthenium incanum, 6; Xanthocephalum microcephalum, 4; Setaria macrostachya, 2; and Sphaeralcea angustifolia (narrowleaf globemallow) 1. It is of interest to note the absence, addition, and shift in percent of those beyond the shoulder: Xanthocephalum microcephalum, 21; Acacia constricta, 20; Viguiera stenoloba, 13; Bouteloua curtipendula, 10; B. gracilis and B. eriopoda, 8; Rhus aromatica, Opuntia engelmannii, and Trixis californica (American trixis), 3.

The absence of Bothriochloa barbinodis, Setaria macrostachya, Larrea tridentata, Parthenium incanum, and Sphaeralcea angustifolia is significant since the latter three are most common at lower elevations in the park. The first two frequent disturbed areas in the basin. The species in the natural vegetation do not deviate significantly from site 5 (Table 4) nearby. The natural vegetation cover percentage is also very similar to the site which had a total plant cover of 92%. The major deviation is the reduction in grass, probably due to the greater growth of shrubs along these roads from greater runoff.

A major factor contributing to the high importance of desert species along the road is the desert origin of the road fill. It has many desert characteristics. including its color. Larrea tridentata size and distribution also substantiate the soil's desert origin, with most being on road fill and from 9-15 years of age, near the 1958 improvement date.

Other evidence to substantiate the desert origin of road fill is the old road scar to the northeast of the present Upper to Lower Basin junction. Table 8 presents the Importance Percentages (IP—average of the sum of the species relative density and relative cover) of the species sampled across this road scar. The transect was a series of contiguous linear quadrats 10 ft long and 3 ft wide. Quadrats 4, 5, and 6 are the old scar, with the remainder being natural vegetation. Soil data (Tables 2 and 3) correlate most closely with the other desert vegetation soils in having high pH and calcium carbonate and low carbon and phosphorus.

It is significant that the species in the second group in Table 8, composed of those occurring in the scar and natural vegetation. are all invaders of disturbed areas in the basin. Both Agave lecheguilla and Dasylirion leiophyllum may have been planted in the scar as their clone sizes are quite large for the available growth time, 11 years. The planting of such desert species by the National Park Service seems to be a general practice, even in the basin.

Table 8. Importance percentages of the species in an old road scar in the Chisos Basin.

The line-intercept method used along the road is an adequate method to obtain qualitative differences, but is rather insensitive to quantitative differences in the vegetation. It was learned from the sampling of this section of roadside vegetation, in conjunction with sites 5 and 8, that the entire northern basin region is composed of a highly disturbed vegetation type. This disturbed condition decreased the sensitivity of the method for comparing impacted vegetation with nonimpacted or natural vegetation. Contiguous plots, which could give added density data, would be a recommended alternative method rather than the line-intercept.

Trails

The most widespread form of basin impact is that of the trail system, which traverses many steep slopes and vegetation types. The effects are similar to road impact as the native vegetation is removed, creating bare areas and impairing natural drainage. The trails, however, lack a complex fill since they are carved from the natural slope. This may impact root damage to nearby trees and shrubs, but does not necessarily result in immediate physical removal or death. The carving process creates, on both sides of the trail, disturbed margins of a xeric nature. The upslope surface is generally hard and steep, while the downslope side is soft and steep. These erosive margins provide many continuous miles of varied habitats for new species.

The rockiness, tree roots, steep terrain, and rainfall runoff require extensive use of drainbars to stabilize such trails. The drainbars in turn create many different vegetative microclimates, where many fine soils and organic materials accumulate and increase their size with every downpour of rain. Because the trails have a soft surface, their maintenance must be greater than that of the paved roads. The method of upkeep can very rapidly influence the trails' overall impact. Important to the trails are the organisms which use them; in this case, man, his horses, and native wildlife,

Throughout the trail system, the organisms' effect upon the natural drainage systems can be observed. This is most readily seen in the way that the trails diagonally dissect the slopes, intercept and concentrate the runoff, and direct its flow to drainbars or eventual ravines. Because of the steepness of the terrain, the rapid flow carries much soil, which can erode the slopes and expose many plant roots. Likewise, great amounts of organic material are transported from their natural sites to lower elevations. Some of the more outstanding examples are on the upper portions of the Window Trail. Here, the drainbars exceed the trail in width and are frequently taken unintentionally by hikers, resulting in cross-country hiking to return to the trail. With such constant agitation in this more xeric trail section, very few plants are inhabiting the bars, although along their margins Acacia roemeriana (Roemer acacia). A. constricta, and Xanthocephalum spp. are invading. On lower Laguna Meadow Trail the drainbars are dominated by Xanthocephalum spp. while along the upper section of the trail, Artemesia ludoviciana forms a nearly continuous border; both are unnatural conditions.

Several sampling approaches were used to demonstrate other effects of trails upon the vegetation. The first approach involved two transects sampled across the lower Laguna Meadow Trail approximately 200 and 100 yards north of the Juniper Flat short-cut trail. A line was placed perpendicular to the trail and five contiguous, linear quadrats (10 x 3 ft) were sampled for cover and number of species. Two quadrats occurred above and two below the trail and one encompassed the trail. At these points the trail tread was 8-10 ft wide and located on a western exposure, with a slope of 15-18°. Table 9 presents the IP for each species sampled.

Table 9. Importance percentages of the species in two Laguna Meadow Trail transects.

It is of interest that the average total vegetation cover above the trail approaches twice that below, as well as exceeding this multiple for average total individuals. The species complex above the trail is more diverse with nearly double the species, several being woody plants with woodland affinities. Examples of this are: Rhus aromatica, Cercocarpus montanus, and the two ferns, Pellaea intermedia (creeping cliffbrake) and Cheilanthes eatoni. Below both trails the herbaceous, weedy Xanthocephalum microcephalum is important, a sign of great disturbance.

A second procedure was used to determine the effect of trails upon tree reproduction. A total census recorded the kind and number of trees and seedlings within 10 ft of both trail margins. To obtain data more sensitive to the impact, the zone was divided into two 5-ft zones of 0-5 ft and 5-10 ft from the trail. These zones were maintained both above and below the trail. A tree was defined as possessing a diameter at breast height (dbh) of 3 inches or greater, whereas any of lesser diameter were classified as seedlings. In the case of multi-trunked Quercus spp. and Juniperus pinchoti, only one of the upright trunks had to have the prerequisite dbh to be classified as a tree. An attempt was also made to determine and record the generic name of the dead trees within each zone. Only those which appeared near a probable root site were counted in order to avoid mistakenly counting misplaced trees.

A census was taken along the trail from the water-barrel drive to Juniper Flat, on a northern exposure and from 5500 to 5600 ft in elevation. Table 10 indicates that a greater number of trees, living and dead, and seedlings occurred in the less disturbed 5-10 ft zone of the upper side. This is not the case in the below-the-trail census, where the values did not differ significantly.

Reproduction is much greater on the above-trail-side, reaching a near 2 seedlings: 1 tree ratio in both zones, On the other hand, the below trail reproduction is less than a 1,5 seedlings: 1 tree ratio. As the leading tree, Pinus cembroides seems to be the most affected by reduced reproduction, especially below the trail. Juniperus deppeana maintains a greater than 2 seedlings: 1 tree ratio on both sides of the trail. Quercus grisea, Juniperus pinchoti, and J. flaccida are reproductively poor on both sites. Quercus emoryi and Q. gravesii are reproducing well, especially on the above trail sites, indicating, along with the presence of Pinus cembroides and Juniperus deppeana trees, that conditions are more conducive for growth above the trail.

A second census along the more heavily impacted trail from the Campfire Circle to the corral (near Ranger Station) presents different trends. This trail is also on a northern exposure, with an elevation between 5100 and 5300 ft. The distance traversed is about half of that covered by the former census and the sample size about one-fifth. Again, the number of trees and seedlings was greater above the trail, but the difference was not significant (Table 11). No outstanding statement of differences can be made concerning the two zones of the trail sides.

The leading tree, Pinus cembroides, again had reduced reproduction especially below the trail. With the exception of Quercus grisea and Prosopis glandulosa, the reproduction of species is greatest above the trail. The general trend of reproduction is toward more xeric species such as Quercus grisea, Juniperus pinchoti, and Prosopis glandulosa, whereas the former census favored the more mesic species Juniperus deppeana, Quercus emoryi. and Q. gravesii.

A census on a new section of trail above Stipa Flat (1967) was made and compared with a similar distance of old trail immediately below (250 paces), with only the above trail side sampled in each case. Table 12 presents the results obtained for two trail sections, the first used for 2-3 years and the other for 10 years. The two sections are extremely similar in species counts, with the major shift being in seedlings. It is of interest that both trees and seedlings are more numerous in the 0-5 ft zone than in the 5- to 10-ft zone. This finding is different from the lower elevational trails sampled. The number of dead trees comprises a much lower percentage of the populations on those trail sections at higher elevations.

Table 10. Tree and seedling census along trail from water barrel to Juniper Flat.

Table 11. Tree and seedling census along trail from corral to Campfire Creek.

Tree reproduction again exhibits a shift with respect to dominant trees and seedlings on the two previous trail samples. In both sections Quercus grisea, Juniperus deppeana, and Pinus cembroides are dominant trees, whereas none of their seedlings is dominant. The leading seedlings are Q. emoryi and Q. gravesii, both adapted to mesic conditions, especially the latter. It is interesting that the more xeric Q. emoryi is the most common seedling in the lower section, Quercus grisea and J. deppeana both have reduced reproduction. The success and high number of Q. emoryi and Q. gravesii are due most probably to the same factors which are at work in Boot Canyon, the primary factor being a shift to a series of more moist years in the climatic cycle. The size of most Q. gravesii (3-6 inches) seedlings is similar to the size of those in Boot Canyon. The Q. emoryi seedlings are larger (12-30 inches), suggesting an earlier response to the moisture cycle.

A topic relevant to trail ecology and its effect upon the vegetation in the Chisos Mountains is that of horse use and trail maintenance. While I was hiking the various trails to and from areas of concentrated investigation, a syndrome of utmost importance to this study of human impact became evident. Although a means of quantifying the problem could not be developed, once the problem was recognized, its ramifications were further observed on many occasions. To consider the problem from its origin is difficult because both horse behavior and trail maintenance practices are involved. The sheer topography and rocky terrain of the mountains are also important factors.

Table 12. Comparative census of trees and seedlings along new and old sections of trail.

Observations indicate that the horses consistently prefer to use the outer or downslope margin of the trail tread. Several hypotheses could be proposed for such behavior, all of which focus on the fact that the horse selects the smoothest course on the trail. Because of physical factors such as downslope slippage of soil, erosion, and trail site, the smoothest course is on the outer portion of the trail. The trail maintenance philosophy of the National Park Service supports a smooth trail, thus unknowingly making it easier for the horse and more difficult for the trail crew's work in maintaining the trail system.

With continued usage and natural washing, the outer lip of the trail becomes sunken or begins to slip downhill, exposing the underlying rocks. Since the condition is philosophically undesirable, the crew corrects the condition by adding soil. Understandably, since little soil is available to fill the depression, material from the inner margin of the trail tread, which is higher, is picked and raked into the area. The uprooted rocks are pushed over the outer lip to tumble downhill and mar the downslope vegetation. On the many occasions when the trail tread yields too little soil, the bank on the inner trail margin is picked, yielding sufficient soil to fill the outer margins and the rocks are again pushed downhill. The maintenance results in a wider and smoother trail tread until the horses' sharp hoofs and physical factors again expose the more shallow rocks on the inner tread surface and promote the return of soft soils and usage impact to the outer lip, and the horse-maintenance syndrome is perpetuated. Seemingly, the syndrome is a never-ending cycle, leading only to wider and wider trails, higher and higher inner eroding banks, and longer and longer talus-like slopes downhill. All three conditions contribute to less natural vegetation, but promote more encroachment by disturbed area species. This disturbance is furthered also by the common occurrence of horses treading across natural vegetation at switchbacks.

The condition is not confined to small segments of trail, but can be seen from the basin Ranger Station to Boot Canyon via Juniper Flat or Laguna Meadow, along the main trail to the Lower Basin and the Window trail. A quick glance at an air photo of the more xeric, less wooded Laguna Meadow Trail will show long segments of the trail nearly half as wide as the paved roads. Such wide trails also occur above Juniper Flat, but are obscured by the overtopping woodland vegetation. Although the trail system has been in use for 10-11 years from the Chisos Basin to Boot Canyon via Juniper Flat, and is 29-33 years old from the basin to Laguna Meadow, with significant horse use for only 22 years on the latter sector, the trails and nearby vegetation are already severely damaged. Only time, the amount of future horse use, and types of maintenance practices will be important factors in determining the width of these trails by 2000 A.D., a mere 30 years hence. A major reevaluation of trail use and maintenance must be made in this xeric, steep, rocky environment. Well-defined programs, including maintenance experimentation and well-trained and informed maintenance crews, must be implemented to cope with the natural conditions of the syndrome. Such programs might include motorized wheelbarrows to return rock and soil trapped on drain bars below to areas of maintenance above,