Printer Friendly

Changes in native vegetation following different disturbances in the lower Rio Grande Valley, Texas.

ABSTRACT. -- Permanent plots established in 1981 were censused again in 1986 at the Palmview tract of the Rio Grande Valley National Wildlife Refuge in southern Texas. Successional changes from bareground to communities dominated by dry-land willow (Baccharis neglecta) and huisache (Acacia smallii) were documented. In formerly grazed woodlands, overstory honey mesquite (Prosopis glandulosa) trees grew an average of 2.5 meters in height over the five-year period. Understory honey mesquite trees were dying, and cover of herbaceous vegetation declined. Cover of woody species such as coma (Bumelia celastrina), granjeno (Celtis pallida), brasil (Condalia hookeri), and Texas ebony (Pithecellobium flexicaule) were increasing beneath honey mesquite. Key words: grazing; mesquite; old fields; Prosopis; succession; Texas.


Few ecological studies have been conducted in the plant communities of southern Texas and little is known about successional changes in this vegetation. In an area 175 kilometers to the north of the Rio Grande River near Alice, Texas, Archer et al. (1988) demonstrated that honey mesquite (Prosopis glandulosa) invades grasslands, and the trees serve as foci for bird-disseminated seeds of other woody species. The resulting shrub clusters represent an intermediate stage, eventually coalescing to form closed-canopy woodlands on more mesic sites. On the San Antonio River terraces (350 kilometers north of the Rio Grande River), Van Auken and Bush (1985) observed that the number of woody species, total density, and total basal area increased with stand age and that the community changes with succession from a low-nitrogen, high-light habitat dominated by huisache (Acacia smallii) to one of high-nitrogen, low-light characterized by Texas sugarberry (Celtis laevigata).

In 1981, the U. S. Fish and Wildlife Service began a woody plant inventory program by establishing nine permanent plots in the Palmview tract of the Rio Grande Valley National Wildlife Refuge (RGVNWR) to obtain data on secondary succession in thorn woodlands of the lower Rio Grande Valley. In this paper, we compare the vegetation from the time of establishing the plots in 1981 to resampling in 1986. The Palmview tract offered an opportunity to document changes following different situations such as fire, removal of cattle, seasonal flooding, abandonment of a gas well, and old field succession.


Soils of the lower Rio Grande Valley have been formed by alluvial deposits through flooding of the Rio Grande. Palmview is five kilometers west of Mission on the second major terrace north of the floodplain. Soil textures vary from fine sandy loam to saline clay, and soils are a mixture of upland (Hidalgo and Raymondville) and terrace (Harlingen and Runn) types (Jacobs, 1981).

The site is in the Matamoros District of the Tamaulipan Biotic Province (Blair, 1950). The potential natural vegetation of the Province has been classified as Prosopis-Acacia-Andropogon-Setaria savanna (Kuchler, 1964). This site, however, may have been a mesic floodplain with perhaps a gallery forest. At present the area has islands of subtropical thorn woodland surrounded by large agricultural fields and urban development.

A portion of the Palmview tract was cultivated until its acquisition by the U.S. Fish and Wildlife Service in 1981. The rest was grazed by cattle. The grazed portion included pasture, woodland, and a seasonal wetland. No grazing or cultivation of crops occurred after 1981. A gas well was also abandoned at that time. A man-caused fire burned a portion of the former pasture on 30 January 1982.


The Palmview tract was typed as grazed woodland, pasture, cultivated field, or wetland in January 1981. Seven plots were located using a stratified random sampling procedure in the main vegetation types. Three plots (4, 5, and 6) were grazed woodlands. Plot 7 was located at the edge between woodland and pasture; the pasture portion was burned by the previously mentioned fire. Two plots (8 and 9) were in formerly cultivated fields, and plot 3 was located in a seasonal wetland. Two additional plots were established in disturbed sites; one (plot 2) was established on a manmade levee at the edge of a woodland and another (plot 1) was located on the abandoned gas well site.

Sampling was conducted in plots 0.1 hectare in diameter; the center of each plot was marked permanently with a steel post. Two groups of vegetation were sampled in each plot. The first group consisted of larger woody species--one stem at least five centimeters diameter at 1.4 meters height (dbh). Each individual in this group was identified to species, numbered, and marked with a stainless steel tag. Information recorded for each individual included azimuth and distance from the center of the plot, dbh of the largest stem, plant height, and the number of stems at least five centimeters dbh. The second group sampled consisted of woody vegetation less than five centimeters dbh and dominant herbaceous species. Individual locations were plotted on circular graphs. Canopy cover was estimated ocularly for each species with the aid of the graphs. Density estimates for each plot were obtained from photographs of a 2.0 X 0.5-meter density board located 17.8 meters from plot center. Photographs were taken from the center of plots in the four cardinal directions. In addition a species list was made for each plot. Plots were sampled in 1981 and again during March 1986.


Old fields. -- Cultivated fields (bare soil) were colonized by dry-land willow (Baccharis neglecta), which formed an overstory with a canopy cover of 65 to 80 percent and was one to five meters tall. Exotic Washingtonia palms (Washingtonia sp.) invaded plot 8 (they were present nearby on the tract in 1981). Understory grass cover was 60 to 80 percent and was dominated by exotics, primarily buffelgrass (Cenchrus ciliaris) and bermudagrass (Cynodon dactylon).

Former pasture and succession after fire. -- This disturbance type was represented by the southern half of plot 7. Photographs show all herbaceous vegetation and woody plant seedlings removed by the fire on 30 January 1982. Fire did not enter the woodland. By 1986, large shrub and tree canopy cover in the burned pasture was 30 to 50 percent and was dominated by dry-land willow, huisache, retama (Parkinsonia aculeta), and honey mesquite. Tree heights were up to six meters, and the dbh of the largest stems were 8.5 cm. Dry-land willow plants had 10 or more stems, and the total basal area of one of these plants was often 100 square centimeters at 1.4 meters height. The understory in 1986 was typically a dense mat of Texas virgin's bower (Clematis drummondii) and bermudagrass. The portion of the pasture that burned in 1982 had a denser cover of woody vegetation than the unburned portion.

Abandoned gas well. -- This once-denuded site had 25 percent overstory canopy cover of honey mesquite and Washingtonia palms in 1986. The major understory species were dry-land willow, honey mesquite, buffelgrass, and whorled dropseed (Sporobolus pyramidatus), the last-listed an indicator of saline soil (Raymondville clay loam).

Man-made levee. -- By 1986, overstory large shrub and tree cover was 5 percent and was primarily retama and honey mesquite. The understory was dominated by dry-land willow, retama, buffelgrass, and bermudagrass.

Seasonal wetland. -- Plant cover in 1988 was 95 percent and one to four meters tall, and was dominated by dry-land willow and cat-tail (Typha domingensis). Abundance of these species did not change much over the five years. The northeastern portion of the plot, however, changed from cat-tail to dry-land willow and an island of dry-land willow in the western part changed to cat-tail. Photographs showed dry-land willow to be taller and denser in the southern and eastern parts of the plot in 1986 relative to 1981. The one huisache tree measured on the plot in 1981 had died and rotted. Honey mesquite seedlings found scattered near the plot center in 1981 were no longer there in 1986, perhaps killed by flooding.

Woodland. -- Plot 6 and the northern half of plot 7 were similar. Cattle grazed there until 1981, and a narrow road once passed through plot 6. The overstory in 1986 was dominated by honey mesquite with a canopy cover of about 70 percent and a density of about 230 trees per hectare (Table 1). The height of the dominant tree layer varied from 5.9 to 8.8 m, and the average five-year height growth was three (1.5 to 5.0) meters. At least 20 percent of the crown of all honey mesquite trees was composed of dead material, and about half of them had at least 60 percent dead material. The dbh of the largest stem of these overstory trees varied from 5.3 to 30.9 centimeters, and average increase in dbh of this stem was three (one to six) centimeters.

The understory brush had a canopy cover of about 20 percent. Plant heights were up to six meters, and dbhs were up to 18 centimeters. The dominant understory species were honey mesquite, granjeno (Celtis pallida), brasil (Condalia hookeri), mist flower (Eupatorium odoratum), huisache, coma (Bumelia celastrina), ivy treebine (Cissus incisa), snakeeyes (Phaulothamnus spinescens), coyotillo (Karwinskia humboldtiana), and guayacan (Guaiacum angustifolium). Thirteen of 28 honey mesquite trees with crowns below the dominant overstory had died over the five-year period. A comparison of photographs between the two measurements showed an increase in shrubs obscuring the density boards.

Plots 4 and 5 were similar in appearance and appeared to be a later successional stage than plots 6 or 7. Overstory canopy cover was also about 70 percent and dominated by honey mesquite in 1986. Trees such as Texas ebony (Pithecellobium flexicaule), coma and huisache entered the upper canopy to a minor extent. The density of trees in the overstory was about 250 to 380 per hectare. Tree height varied from 5.3 to 9.6 meters, and average five-year growth was 2.5 (one to four) meters--Table 1. Diameter at breast height of the largest stem varied from 5.4 to 37.8 centimeters, and average five-year dbh increase was about three (zero to nine) centimeters for all overstory trees. Vigor of honey mesquite trees was similar to that on plots 6 and 7. Other species had less dead material in their upper crowns.

Understory shrub cover was 20 to 40 percent, with heights of up to 5.5 meters, and dbh of the largest stem up to 14.1 centimeters. The species in the understory shrub layer did not appear to have changed much over the five years. Photographs showed shrub biomass higher in 1986, and density boards not visible at 17.8 meters. They also showed less herbaceous vegetation in 1986 (cover less than 10 percent). Major understory species included brasil, granjeno, honey mesquite, coma, elbowbush (Forestiera angustifolia), Berlandier wolfberry (Lycium berlandieri), Texas ebony, snake-eyes, and the exotic guinea grass (Panicum maximum).


The development of woody assemblages at this site seems to fit the Relay Floristic (Egler, 1954) or Facilitation (Connell and Slayter, 1977) models of plant succession, although the data from this study are not sufficient to demonstrate theories because only a few seres were present. Theory suggests that each successive type establishes itself because the preceding type modifies the site in a way favorable to its successor (Clements, 1916). Species may be killed in competition with later species, although this had not yet occurred with honey mesquite at this site. Archer et al. (1988) suggested that succession on their study site fit the Facilitation Model. Early successional communities seem to change from low-nitrogen, high-light to high-nitrogen and low-light as suggested by Tilman (1982) and supported by Van Auken and Bush (1985). Tilman, in his "resource-ratio" theory, suggested that species with a lower requirement for a limiting resource (that is, light) may displace others (that is, Texas ebony may replace honey mesquite).

Later successional species were not present initially. The environment at this site has been modified by humans and dormant seeds of later successional species may not be present. Seed dispersal from other areas by birds and mammals may be necessary. Van Auken and Bush (1985) observed that mature forest species, except Texas sugarberry, first occurred about 30 years after disturbance in their study site. Several of the understory brush species in plots 4, 5, and 6, such as brasil, coma, snake-eyes, coyotillo, guayacan, elbowbush, Berlandier wolfberry, and Texas ebony, are thought to be 'mature' woodland species because they are common on the Santa Ana National Wildlife Refuge, which has remained undisturbed for more than 40 years, but former disclimax (flooding) or climax species are unknown.

Dry-land willow, buffelgrass, and bermudagrass were dominant species after five years of old-field succession. Dry-land willow may prepare the site for colonization by pioneer tree species such as honey mesquite, huisache, and retama on more mesic sites. The tall dry-land willow plants may provide shade and retention of surface soil moisture for seedlings of other species. The exotic buffelgrass and bermudagrass probably inhibit establishment of other species by successfully competing for soil moisture, light, growing space, and the former possibly by allelopathy (Akhtar et al., 1978; Hussain et al., 1982). Van Auken and Bush (1988) observed that little bluestem (Schizachyrium scoparium) reduced biomass of honey mesquite in a greenhouse experiment.

Grazing and fire may reduce the vigor of grasses and provide the seed bed necessary for establishment of woody species, as long as the intensity and frequency are low enough to permit establishment of woody species (Vora, 1989). Fire may remove some of the inhibitory allelopathic effects. Woody plants were far more prevalent in the pasture (burned and unburned portions) than in the formerly cultivated fields, and the burned portion of the pasture had a denser cover of woody vegetation. Fire tends to be restricted to grassy fields in this area as live fuel moisture in woodlands is usually too high for it to burn except possibly under extreme conditions.

Texas virgin's bower probably also reduces colonization by woody species by forming a dense, smothering mat allowing little light penetration. Portions of new accretions from the Rio Grande on the nearby Santa Ana National Wildlife Refuge that have not been disturbed for 40 years remain covered with a dense mat of this species. Fire may be needed to remove it.

The seeds of the three colonizing leguminous trees, honey mesquite, huisache, and retama, probably require animal dissemination. Plants of these species proliferate in areas where cattle have grazed, especially once cattle are taken off the site (personal observation). Livestock transport seeds, scarify them, deposit them in a nutrient-rich media (dung), and reduce herbaceous competition by grazing (Archer et al., 1988).

Periodic flooding will drown seedlings and large trees of these species depending on the duration, depth, and time of year as in most bottomland hardwood stands (Wharton et al., 1982). Most of these trees cannot live under prolonged flooding during the growing season, and seasonal flooding has maintained plot 3 in a mixture of dry-land willow and cat-tail. Honey mesquite seedlings present in 1981, and established presumably during a dry period, were gone in 1986. Decomposition is rapid as evidenced by the complete disappearance of a huisache tree on plot 3 in five years. The effects on vegetation of regular flooding of the Rio Grande before dams and diversions are unknown.

The successional role of exotic Washingtonia palms is undocumented. Presumably they are an early successional species and will die out when enveloped by later successional trees.

The Palmview tract did not have the intermediary early seral stages to demonstrate if honey mesquite trees serve as foci for cluster development of woody species as suggested by Archer et al. (1988), although there was no reason to doubt it. Several woody species, such granjeno, brasil, coma, ivy treebine, snake-eyes, coyotillo, guayacan, elbowbush, Berlandler wolfberry, and Texas ebony, were growing under the honey mesquite canopy on plots 4, 5, 6, and 7.

The older woodland stands (plots 4, 5, and 6), which are known to have been in existence for at least 20 years, were still dominated by honey mesquite. The dominant honey mesquite trees had not yet realized their growth potential and were still growing an average of 0.5 to 0.6 meters per year. Dominant trees seemed to be shading out intermediate and suppressed honey mesquite trees as indicated by the loss of 13 of 28 such trees over five years on plot 6. Honey mesquite did not seem to be replacing itself, and in time will presumably be reduced by competition from tree species presently in the understory. Light intensities under adult tree canopies may be insufficient for honey mesquite seedlings, but this alone may not explain lack of seedling establishment (Bush and Van Auken, 1987).

Huisache and retama may play a role similar to honey mesquite, although Archer et al. (1988) concluded that Acacia farnesiana played no role in cluster development; they suggested that its canopy architecture was unattractive to perching birds. All three leguminous trees probably facilitate survival and growth of other species by enhancing soil nutrient levels (especially nitrogen and carbon), water-holding capacity and infiltratation, and mineralization potential (Van Auken and Bush, 1985; Bush and Van Auken, 1986; Archer et al., 1988) and by providing shade for tolerant species.

These data are the first repeated measurements of individual trees showing plant community change in the lower Rio Grande Valley. They begin to provide growth rates needed to determine time frames for succession.
TABLE 1. Five-year changes in mean density, stem area, and heights of
honey mesquite trees. Standard deviations are given in parenthesis.

 Plot and year of measurement
 Early succession (old fields)
 Plot 1 Plot 7
Variable 1981 1986 [DELTA] 1981 1986 [DELTA]

Density (a) 40 160 +120 180 170 -10
Stem area (b) 63 68 -5 285 311 -26
 (44) (84) (78) (82) (369) (290)
Height (c) 4.9 4.9 0 4.1 5.4 +1.3
 (1.0) (0.6) 1.2 (0.3) (2.8) (2.2)

 Plot and year of measurement
 Later succession (formerly grazed woodlands)
 Plot 6 Plot 4
Variable 1981 1986 [DELTA] 1981 1986 [DELTA]

Density (a) 430 400 -30 520 600 +80
Stem area (b) 148 245 -97 192 184 -8
 (22) (206) (241) (31) (212) (132)
Height (c) 3.9 6.4 +2.5 4.9 5.6 +0.7
 (0.2) (1.8) (2.2) (0.2) (1.4) (1.7)

 Plot and year of measurement
 Later succession (formerly grazed woodlands)
 Plot 5
Variable 1981 1986 [DELTA]

Density (a) 430 420 -10
Stem area (b) 183 253 -71
 (27) (207) (283)
Height (c) 4.7 6.8 2.1
 (0.2) (1.6) (2.3)

(a) Trees per hectare with one stem at least five centimeters in dbh.
(b) Mean cross-sectional area ([cm.sup.2]) of largest stem at 1.4 meters
height of trees with one stem greater than five centimeters dbh.
(c) Heights (m) of trees with one stem greater than five centimeters


Special thanks to R. Schumacher and N. Gilbertson for initiating this project and completing the original plot measurements, and to Z. Labus, R. Starman, V. Pulaski, J. Rodriguez, and G. Bryant for data collection, and to S.Archer, F. Gehlbach, J. Everitt, R. Lonard, F. Johnson, T. Fulbright, D. Van Auken, and the staff of the Rio Grande Valley National Wildlife Refuge for editorial review.


Akhtar, N., H. H. Naqvi, and F. Hussain. 1978. Biochemical inhibition exhibited by Cenchrus ciliaris Linn. and Chrysopogon aucheri (Bioss) Stapf. Pakistan J. For., 28:194-200.

Archer, S., C. Scifres, C. R. Bassham, and R. Maggio. 1988. Autogenic succession in a subtropical savana: conversion of grassland to thorn woodland. Ecol. Monogr., 58:111-127.

Blair, W. F. 1950. The biotic provinces of Texas. Texas J. Sci., 2:95-117.

Bush, J. K., and O. W. Van Auken. 1986. Changes in nitrogen, carbon, and other surface soil properties during secondary succession. J. Soil Sci. Soc. Amer., 50:1597-1601.

____. 1987. Light requirements for growth of Prosopis glandulosa seedlings. Southwestern Nat., 32:469-473.

Clements, F. E. 1916. Plant succession: an analysis of the development of vegetation. Publ. Carnegie Inst. Washington, 242:1-512.

Connell, J. H., and R. O. Slatyer. 1977. Mechanisms of succession in natural communities and their role in community stability and organization. Amer. Nat., 3:1119-1194.

Egler, F. E. 1954. Vegetation science concepts. I. Initial floristic composition, a factor in old-field vegetation development. Vegetatio, 4:412-417.

Hussain, F., H. H. Naqvi, and I. Iiahi. 1982. Interference exhibited by Cenchrus ciliaris and Bothriochloa pertusa. Bull. Torrey Bot. Club, 109:513-523.

Jacobs, J. L. 1981. Soil survey of Hidalgo county, Texas. U.S.D.A., Soil Conserv. Serv. 171 pp.

Kuchler, A. W. 1964. The potential natural vegetation of the conterminous United States. Spec. Publ. Amer. Geogr. Soc., 36:vi + 1-40 116 + map.

Tilman, D. 1982. Resource competetion and community structure. Princeton Univ. Press, 296 pp.

Van Auken, O. W., and J. K. Bush. 1985. Secondary succession on terraces of the San Antonio River. Bull. Torrey Botanical Club 112:158-166.

____. 1988. Competition between Schizachyrium scoparium and Prosopis glandulosa. Amer. J. Bot., 75:782-789.

Vora, R. V. 1989. Fire in an old field adjacent to a sabal palm grove in south Texas. Texas J. Sci., 41:107-108.

Wharton, C. H., W. M. Kitchens, E. C. Pendleton, and T. W. Sipe. 1982. The ecology of bottomland hardwood swamps of the southeast: a community by profile. U. S. Fish Wildlife Serv., FWS/OBS-81/37, 133 pp.


Rio Grande Valley National Wildlife Refuge, Alamo, Texas 78516

Current address of author: U. S. Forest Service, 1170 4th Avenue South, Park Falls, Wisconsin 54552.
COPYRIGHT 1990 Texas Academy of Science
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1990 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Vora, Robin S.; Messerly, John F.
Publication:The Texas Journal of Science
Geographic Code:1U7TX
Date:May 1, 1990
Previous Article:A record of Bootherium bombifrons (Artiodactyla: Bovidae) from Hunt County, Texas.
Next Article:Fish assemblage structure in an intermittent Texas stream.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters