Tailoring rangelands to enhance wildlife habitat: The Welder Wildlife Refuge Experience

D. LYNN DRAWE, Welder Wildlife Refuge, Sinton, Texas

The Welder Wildlife Refuge is located in a transitional zone between the Gulf Prairies and Marshes and the South Texas Plains. The flora of the Coastal Bend is complex, containing almost 1,400 species of flowering plants and ferns including 218 species of grasses. Plants of the area are mostly of tropical or sub tropical origin. The 7,800-acre refuge contains 63 of the 130 species of woody plants listed by Jones (1975).

Vegetational changes

The general vegetation of the region around the Welder Refuge is part of a grassland climax that occurs along the Gulf of Mexico. Although generally considered a grassland, the area today is actually a brush-grass complex, basically a grassland with brush of various densities covering almost all sites except the deep sands of the Nueces-Sarita series.

Differences of opinion exist as to the relationship of brush and grassland in the past. Many writers have stated that there was a considerable encroachment of brush on the native prairies during the latter part of the 19th century, while others state that the present species have not appreciably increased their range, but that their density has increased in recent years (Drawe et al. 1978). Certainly, the vegetation of the Welder Wildlife Refuge and surrounding areas has changed considerably since Newell (1838) described it as open prairie. More recently, an early settler of the Sinton area described the general area around Sinton as a “bald prairie” (Williges 1959). However, references to berries of trees and shrubs, pricklypear tunas, and other food items attracting the Indians from the Refugio Mission to the Aransas River indicate that there was considerable woody vegetation on or near the Refuge 200 years ago (Drawe et al. 1978).

There is good evidence that considerable change has taken place in the woody vegetation of the Welder Wildlife Refuge in the last 43 years. O.W. Maley, Foreman on the Welder Ranch for many years, stated that there had been a great increase in brush north of the semipermanent lakes on the Refuge since he first saw the refuge area in 1937. Comparison of aerial photos taken in 1939 and 1992 shows that brush has increased throughout the refuge. Mr. Maley stated that the brush had increased in spite of a vigorous brush control program practiced by Rob Welder before his death in 1953. W.C. Glazener, former Director of the Refuge, stated: “Huisache, in particular, has become more widely distributed and increased in density on numerous sites of the Refuge since I first visited here in October 1954. Much of the increase in density and height has been obvious in the past decade (1965-1975).” Frank Rooke, Jr., longtime owner of a large ranch adjoining the refuge, recalled in 1960: “When I was a small boy (in the decade of the 1920s) and looked south-westward from our house on the north bank of the Aransas River all I could see between us and Sinton was large live oak trees.”

Seral stages, as measured by range condition class, have changed since the Refuge was established. Range technicians placed the area in poor condition during the mid-1950s. During the 1960s most of the area was in fair condition. Through the 1980s and 1990s the area has remained in high-fair to low-good condition. The upward trend has been attributed to light to moderate grazing by domestic livestock, periodic deferment of range units, and more than a decade of above-average rainfall. The decline in the late 1980’s has been attributed to a three-year drought ending in 1990.

Historical brush management

The brush management history of the 7,800-acre refuge has been recorded with some accuracy. Rob Welder used heavy machinery to ‘control’ the chaparral along the ‘break’ between the clay uplands and the Aransas River bottomland in the 1940s. Using a rolling chopper designed and built to his specifications, he treated approximately 2,000 acres of the refuge.

After the refuge was established in 1954, brush manipulation and studies of the effects of various procedures became commonplace because of the obvious spread of honey mesquite and huisache and the increase in density of most chaparral species. These management and research efforts have added much to the knowledge of brush management in south Texas.

Thad Box, the first graduate student to receive a degree on a Welder Fellowship, initiated a comparison of five methods of mechanical brush control in 1963. The 20-acre plots were replicated at three locations on clay soils. A follow-up study was conducted by Mutz et al. (1977). Their data show that soil disturbing mechanical practices such as root plowing or root plowing plus raking cause long-term changes in both the woody and herbaceous communities that must be considered by the manager/landowner. Huisache, mesquite, and pricklypear cactus increased on root plowed treatments at the expense of more desirable (for wildlife) woody species. Herbage scores and rate of return of desirable grasses indicate a slower recovery of the herbaceous community on root plowed treatments.

Mechanical top removal increases deer preference values for a variety of south Texas woody species (Box and Powell 1965). The rate of regrowth of some species such as huisache is very rapid, but the increase in preference values is short-lived. Thus, a brush management program using top removal on a portion of a range each year would provide a continuing supply of high-value, readily available browse.

Until recently, little information was available concerning the amount of brush to remove. Naderman and Dahl (1978) reported preliminary results of a study of various sizes and shapes of brush openings made by roller chopping or mowing and their effects on deer use and deer density. Square and rectangular openings ranging from 10-80 acres in size were created. Deer utilized only the edges of larger openings, and were more likely to use the entire opening if it was rectangular. A 20-acre rectangular opening seemed to be the preferred size and shape. The overall impact on the deer population of opening up more area was an increase in deer on the study ranch.

Prescribed burning research initiated in 1974 on the Welder Wildlife Refuge has been designed to determine (1) the influence on the ecosystem of the removal of accumulations of mulch, (2) the possibility of the suppression of undesirable brush species, and (3) the influence of fire on wildlife populations and wildlife habitat. These were some of the first fires on the refuge since the cessation of burning in the early 1900s. Prior to burning, data were collected on brush cover, herbaceous cover and composition, and herbaceous production. Many of these initial management burns have been repeatedly burned over the years.

On a mesquite grassland we studied fire in combination with oiling and low energy grubbing of huisache (Bontrager et al. 1979). Huisache was removed initially by oiling or grubbing. Fire was then superimposed over oiling and grubbing, resulting in a combination of treatments. The combination of oiling and/or grubbing and burning was particularly effective in suppression of huisache. Both oiling and grubbing were effective and economical in stands of huisache trees up to 150 trees per acre. However, when the number of trees exceeded 150 per acre, costs were prohibitive. In stands of huisache with greater than 150 trees per acre burning was not an effective tool. There was not enough fine fuel on the ground to carry the fire through a dense stand. In dense stands therefore, it may be desirable to roller chop or chain and defer grazing to allow herbaceous fuel to build up so a fire will carry through the area.

A series of three 50-acre blocks was selected in chaparral grassland in 1975 to assess the effects of repeated fires: (1) burn as often as possible, (2) burn every 3 years, or (3) burn every 5 years. Repeated burning effectively suppressed woody vegetation. The 50-acre plot burned as often as possible was burned four times. It was possible to burn this plot every year or every other year, depending on the amount of fine fuel buildup. In drier years annual burning was not possible. It may take more than two years to build up sufficient fuel for a fire if the area is not deferred from grazing. (Our plots received moderate grazing pressure.) Fuel loads on the areas burned every 3 years or every 5 years were adequate.

The three repeat-burn study plots were in chaparral grassland with a 40 to 45% woody cover. After the second prescribed burn there was almost a total elimination of woody cover. One extremely hot reclamation fire burned through most smaller chaparral mottes. On the windward side of a motte the fire destroyed 100% of the cover. On the leeward side of a motte, depending upon wind speed during the fire, there have been protected areas where some woody plants avoid topkill. In larger mottes, i.e., those greater than 20 feet in diameter, woody plants survived in the center and leeward side of the motte. Fine fuel was nonexistent within mottes. Only crown fires have burned completely across large mottes. The whole motte survived in cooler maintenance burns, even in the case of some of the smaller mottes less than 20 feet in diameter. During maintenance or reclamation burns, individual plants between the clumps were top-killed.

Essentially all chaparral species resprout. These species may regain their original size in 18 to 22 months (Hamilton and Scifres 1981). On the area burned as often as possible, burning every two years reduced 40% cover to zero. Prior to the second burn the cover was essentially 40% again.

On the area burned every three years, fine fuel accumulations were adequate and canopy cover reduction of chaparral was as great as the area burned as often as possible. After the third burn on both areas it appeared that there was a loss of many individual woody plants, and vigor of individual chaparral plants appeared lower.

On the area burned every five years, reductions in brush cover comparable to the burn-as-often-as-possible treatment and the burn-every-three-years treatment were obtained. However, the long period between burns allowed the chaparral to regrow to its former height and density and possibly increase in density. Burning every five years may be a good recommendation for the rancher/land manager interested in preserving most of his brush for wildlife cover. If the primary interest is in suppressing brush it will be necessary to burn more often.

Individual plant response of the chaparral species was monitored. Following the first fire on the burn-every-third-year plot, 100 plants of each species of chaparral were examined three months after the fire and again one year after the fire. Initially, the plant that appeared to be most susceptible to fire was agarito. Agarito burns “explosively” and usually burns completely down to ground level in a hot fire, thus possibly sustaining more physiological damage than other species. Three months after the fire essentially 100% of the agarito appeared to be dead. About 20% of the Mexican persimmon appeared to be killed. Regrowth was not suppressed on the other 10 or 12 chaparral species. One year after the burn, all species including agarito and Mexican persimmon were regrowing vigorously. Following the third burn on the burn-every-third-year plot, there appeared to be a reduction in the amount of cover of all chaparral species. Some mottes were almost totally top-killed after three fires. Many decadent stumps of all chaparral species were apparent. Weedy herbaceous vegetation initially grows in the mottes that are totally top-killed by the fire.

Fire was an effective tool in removing excess mulch accumulations that occur during wet cycles under conservative grazing. One of the problems to the range resource manager/wildlife habitat manager in the Coastal Prairie is that during wet years heavy accumulations of mulch can occur. During the 1974 burns, fine fuel loads from 4,500 pounds per acre to 16,000 pounds per acre were removed (average 5,500 pounds per acre). Greater amounts of fine fuel form a mat that may not be penetrated by light and actually acts to suppress plant growth. We found an increase in herbage production in the year following the initial burn.

The manager can create a plant community desirable for wildlife or livestock by selecting the time of the burn. Late winter burning reduces forb populations and benefits grasses, whereas early winter burning is followed by normal forb populations (Hansmire et al. 1988). In addition to reducing forbs, late winter burns may reduce the amount of other cool season plants, particularly Texas wintergrass.

Fire increased quail use of 40-acre plots on the Welder Wildlife Refuge, but the burned areas were too small and the time span too short to determine the effect on overall population densities (Wilson and Crawford 1979).

Breeding birds in a mesquite grassland were not affected in the year following winter burning (Reynolds in press). Mourning doves and great-tailed grackles increased, and black-bellied whistling ducks declined in burned treatments in the breeding season. Mean abundance of wintering birds, mourning doves, and savannah sparrows increased in burned treatments. Wintering bird species that required dense grass and litter cover declined in burned treatments. A mosaic of burned and unburned habitats was recommended to maintain a variety of bird species.

Recent brush management treatments on the Welder Refuge have included roller chopping and aerial spraying. Approximately 1500 acres in a chaparral grassland were roller chopped in 1989 in a checkerboard pattern of 10 acres chopped and 10 acres untreated. This pattern has proven to be effective in providing more grazing for cattle while improving the habitat for deer. These plots have regrown to the extent that follow-up treatment became necessary. A research project was initiated in 1995 to re-treat a portion of the roller chopped plots in a study of the effects of removal of various components of the vegetation on deer behavior (Stewart in press). Results of that study indicate that deer used plots less where the woody and forb components have been removed. Deer use was higher where either forbs, brush resprouts, or both were present.

Aerial spraying from a helicopter using Grazon PC plus Remedy was used in mesquite grassland to control mesquite. Approximately 1,500 acres were treated between 1991 and 1996. This treatment has been effective in reducing the mesquite canopy by approximately 75-85%. It appears that the life of this treatment will be 5-7 years, and follow-up prescribed burning is being applied to extend the life of the treatment.

Nolte (1995) studied the effects of aerial spraying of mesquite on grassland birds. The removal of the mesquite canopy had the effect of eliminating nesting habitat of most of the species that nest in the coastal prairie; however, all species except scissortail flycatchers were able to adapt by using other chaparral species as nesting sites.


Brush manipulation to enhance wildlife habitat has been a major research and management thrust of the Welder Wildlife Foundation. The woody vegetation of the Welder Refuge has been tailored to enhance wildlife habitat while maintaining a commercial cattle operation.

Literature Cited

Bontrager, O.E., C.J. Scifres, and D.L. Drawe. 1979. Huisache control by power grubbing. J. Range Manage. 32:185-188.

Box, T.W., and J. Powell. 1965. Brush management techniques for improved forage values in south Texas. Trans N. Amer. Wildl. Conf. 13:285-296.

Drawe, D.L., T.W. Box, and A.D. Chamrad. 1978. Plant communities of the Welder Wildlife Refuge. Welder Wildlife Found. No. 5, Ser. B/revised, Sinton, TX. 38 pp.

Hamilton, W.T., and C.J. Scifres. 1982. Cool season prescribed burning for maintenance of bufflegrass. J. Range Manage. 35:9-12.

Hansmire, J.A., D.L. Drawe, D.B. Wester, and C.M. Britton. 1988. Effect of winter burns on forbs and grasses of the Texas Coastal Prairie. Southwestern Nat. 33:333-338.

Jones, F.B. 1975. Flora of the Texas Coastal Bend. Welder Wildl. Found., Sinton, TX. Contrib. 6. Series B. 262 pp.

Mutz, J.L., C.J. Scifres, D.L. Drawe, T.W. Box, and R.E. Whitson. 1978. Range vegetation after mechanical brush treatment on the Coastal Prairie. Tx. Ag. Exp. Sta., College Station. 16 pp.

Naderman, J., and B.E. Dahl. 1978. White-tailed deer use of small openings in brush. Research Highlights – Noxious Brush and Weed Control; Range and Wildlife Management, Texas Tech Univ., Lubbock, TX. pp. 19-20.

Newell, C. 1838. History of the Revolution in Texas. Wiley and Putnam, New York.

Reynolds, M.C. 1996. Effects of winter burning on songbirds in mesquite-grassland in south Texas. M. S. Thesis. Welder Wildl Found., Sinton, TX. (in press)

Stewart, K.M. 1996. Effect of brush manipulation on the behavior of white-tailed deer. M.S. Thesis. Welder Wildl. Found., Sinton, TX (in press).

Williges, G.G. 1959. A line transect across sections 50 and 51 of the Welder Wildlife Refuge. M.A. Thesis. Texas A&I Univ., Kingsville. 82 pp.

Wilson, M.M. and J.A. Crawford. 1979. Response of bobwhites to controlled burning in south Texas. Wildl. Soc. Bull. 7:53-56.

Comments: Dale Rollins, Professor and Extension Wildlife Specialist
Updated: Mar. 18, 1997

Comments are closed.