Restoration of Saltcedar Infested Flood Plains on the Bosque del Apache National Wildlife Refuge1

JOHN P. TAYLOR and KIRK C. MCDANIEL2

Abstract. Vegetation development bordering the Middle Rio Grande, as with most major southwestern USA tributaries, has historically undergone rapid and dynamic change. The introduction of saltcedar (or Tamarisk, genus Tamarix) and other exotic species into this environment within the 20th century has contributed to this process. These plants are now an integral component of the riparian vegetation mix. Manpower, logistics and financial resources constrain the degree to which desired riparian habitat can be restored from saltcedar thickets on the Bosque del Apache National Wildlife Refuge near Socorro, NM. Saltcedar clearing is accomplished using a combination of herbicide, burning, and mechanical control techniques costing from $750 to $1300/ha. Soil salinity and depth to water are the principle physical features limiting revegetation efforts. Cottonwood and black willow plantings and natural regeneration after timed irrigations have produced diverse habitats that support a wide array of faunal species in areas previously occupied by homogeneous saltcedar.

Nomenclature: Saltcedar, Tamarix ramosissima Ledeb. #3 TAARA; cottonwood, Populus fremontii S. Wats.; black willow, Salix nigra Marsh. #SAXNI.

Additional index words: Phreatophyte control, riparian restoration, Rio Grande, wildlife, imazapyr, Populus fremontii.

Received for publication June 27, 1997 and in revised form.

2Biologist, U.S. Fish and Wildlife Service, Bosque del Apache National Wildlife Refuge, P. O. Box 1246, Socorro, NM 87801; Prof.., Dep. of Animal and Range Sci., New Mex. St. Univ., Las Cruces, NM 88003.

3Letters following this symbol are a WSSA-approved computer code from Composite List of Weeds, Revised 1989. Available from WSSA, 309 W. Clark St., Champaign, IL 61820.

 

INTRODUCTION

Historically, the Rio Grande's narrow riparian corridor, in a otherwise arid environment, has been a rich vegetative mosaic maintained through periodic flooding that fostered patchy habitats with extensive vertical structure. Avian species are the hallmark of this environment and reflect the diverse flora. As with many southwestern river systems, however, habitat degradation has occurred through altered river hydrography and constricted flood plains resulting from irrigation and flood control developments, and the introduction of exotic flora. Resource managers are frequently charged with restoring these riparian habitats with little knowledge of benefits, methodology, costs, and manpower requirements.

In this paper we discuss efforts by the Bosque del Apache National Wildlife Refuge (NWR) near Socorro, NM towards rehabilitating saltcedar-infested flood plains and restoring diverse riparian habitats. Efforts to manage saltcedar on the refuge began in the 1940s and continues today. Restoring saltcedar thickets is consistent with the Refuge's mission of providing suitable habitats to support a wide array of fauna.

 

RESTORATION TECHNIQUES

Riparian restoration efforts in the southwestern USA began in the 1960s largely supported by federal agencies including the U.S. Bureau of Reclamation and the U.S. Fish and Wildlife Service to compensate for river construction and operation projects (Busch et al. 1992). These efforts have varied in scale and have centered on larger river systems including the lower Colorado River in Arizona and California, the Rio Grande in central New Mexico, and the Salt River in Arizona. Native species revegetation frequently involving preliminary saltcedar control has shown varying degrees of success on impacted sites.

Several methods of saltcedar control have been employed over the last 50 years. An early method of mechanical removal involved chaining followed by maintenance mowing, but no lasting control was achieved using this method (Great Western Research 1989). Annual fall mowing to reduce the statue of saltcedar and to provide a grassland appearance continues today along portions of the Rio Grande, usually within the confines of a levee. Through the late 1940s and early 1950s, a variety of implements pulled by bulldozers and tractors were used for saltcedar control in New Mexico. These included tool bars, root planes, root knives, plows, and saws (Anonymous 1951). In 1960, a root plow pulled by a bulldozer was developed to clear saltcedar stands. The method probably evolved from previously described implements, but was standardized and adopted by river management agencies including the U.S. Bureau of Reclamation. The plow involves cutting the root crown 30 to 45 cm below the surface in dry soil and warm weather (Horton 1960). More than 90% saltcedar control was recorded using this method in early trials and the technique remains a reliable tool for saltcedar control today. More recently, improved control has been achieved by raking cut stems and roots into piles before burning to prevent sprouting from adventious buds (Bosque del Apache NWR unpublished report).

Chemical control was used experimentally and later in control maintenance programs during the 1940s and 1950s using 2,4-D ([2,4-dichlorophenoxy] acetic acid), 2,4,5-T ([2,4,5-trichlorophenoxy] acetic acid), and Silvex (2-(2,4,5-trichlorophenoxy)propionic acid) (Bosque del Apache NWR, unpublished report, Anonymous 1951, Busch et al. 1992). These herbicides were generally not effective in killing saltcedar roots, and 2,4,5,-T and Silvex were banned in 1983 by the Environmental Protection Agency. In the 1980s, new herbicides were found effective to varying degrees for controlling saltcedar. These include triclopyr ([3,5,6-trichloro-2-pyridinyl) oxy] acetic acid) which has been most effective as a bark penetrant (Neill 1988), and imazapyr (" - 2 - [4,5-di-hydro-4-methyl-4-(1-methylethyl)-5-oxo-1 H-imidazol-2-y]-3-pyridenecarboxylic acid) which is most effective for foliar application (Taylor 1987).

Prolonged flooding has also been used as a means of saltcedar control (Warren and Turner 1975, Widemann and Cross 1978). Seventeen mo of prolonged flooding combined with some form of mechanical control, i.e., chaining, or plowing, has provided mortality rates near 80%, whereas flooding alone for 28 mo has killed 99% of saltcedar plants. Finally, DeLoach (1989) identified saltcedar as a candidate for biological control. The U.S. Department of Agriculture has approved two species for release, the leaf beetle (Diorhabda elongata) and the mealybug (Trabutina mannipara), pending an approved biological assessment.

Riparian revegetation techniques have evolved along two lines since the 1970s. The first technique has centered on complete restoration of riparian communities using supplemental irrigation. Anderson and Ohmart (1982) pioneered this technique along the lower Colorado River by installing a drip irrigation system to aid in the establishment of planted materials. Planting prescriptions were based on site capabilities, plant adaptations to soil texture, salinity and depth to water table, and wildlife habitat response models. This restoration effort remains a model of excellence today as saltcedar thickets have been transformed into native communities maximizing habitat patchiness and vertical flora structure. High cost is the primary drawback to restoration using supplemental irrigation. Cost for projects in Arizona and California have ranged from $864/ha to $1,535/ha which includes preliminary saltcedar control (Anderson and Ohmart 1982, Anderson 1988a).

A second planting technique utilizes dormant cottonwood and black willow poles augered to the water table to establish forested areas. This technique was first developed for southwestern riparian areas in New Mexico by Swenson and Mullins (1985). In many areas however, few native stands of cottonwood and willow are available nearby for harvesting poles in large numbers (Busch et al. 1992). Superior stock collections have been gathered by the USDA Plant Materials Center in Los Luna, NM, for eventual release to government agencies for restoration purposes and to nurseries for further propagation (Fenchel et al. 1987 ). Advances have recently been made propagating coyote willow (Salix exigua), seepwillow (Baccharis glutinosa), false indigo (Amorpha fruticosa L. #3 AMHFR), and New Mexico olive (Forestiera neomexicana Gray) for use as augered pole plantings (Fenchel et al. 1996). These native shrubs complement several tree species already developed for riparian restoration.

 

RIPARIAN RESTORATION ON THE BOSQUE DEL

APACHE NATIONAL WILDLIFE REFUGE

Between 13 and 18 April, 1986, a wildfire consumed 735 ha of riparian habitat on the Bosque del Apache NWR. At the time, this was the largest wildlife ever to occur on the refuge, and resulted in the destruction of 180 ha of native cottonwood and willow forest habitat. A funding plan to rehabilitate the burned area was prepared and forwarded to U.S. Fish and Wildlife Service reviewing officials in June 1986 for submission to the U.S. Department of Interior for fire rehabilitation funding. This original proposal, based on experimental restoration work accomplished on the lower Colorado River using drip irrigation (Anderson and Ohmart 1982), totaled almost $1.4 million (Bosque del Apache NWR, unpublished report 1986). The proposal was rejected as to costly and refuge officials were charged with developing a less costly, yet achievable proposal based on experimental cottonwood and willow pole plantings (Swenson and Mullins 1985).

A second proposal was developed incorporating pole planting with irrigation developments along an abandoned Rio Grande channel in the area. The revised proposal totaled nearly $335,000 and was approved for funding over a 5-yr period from 1987 to 1991. The scope of this project was unique in that techniques untested on a large scale, were used for saltcedar control and revegetation with native species. Today, the project forms a portion of the largest riparian restoration program in the southwestern USA.

Project area. The Bosque del Apache NWR encompasses 23,162 ha of which 3,440 ha are floodplain habitats along a 20 km length of the Rio Grande. The Rio Grande Valley is 5 to 7 km wide through the refuge and is 5.2 km wide at the project site. Low mountain ranges rise 2,000 m to the west and 1,600 m to the east, with valley floor elevations averaging 1,470 m. Woody riparian communities consist of mixed saltcedar/bosque and homogenous saltcedar thickets. Native species include cottonwood, black willow (Salix nigra Marsh #SAXNI), coyote willow, New Mexico olive, false indigo, seepwillow, screwbean mesquite (Prosopis pubescens Benth.), wolfberry (Lycium andersonii Gray), and fourwing saltbush (Atriplex canescens [Pursh] Nuttall #ATXCA). Understory herbage in cottonwood dominated areas include common lambsquarter (Chenopodium album L. #CHEAL), narrowleaf globemallow (Sphaeralcea angustifolia (Cav.) G. Don #SPHAN), white sweet clover (Melilotus alba Medik.. #MEUAL), jimsonweed (Datura stramonium DC. ex Dunal #DATIN), Virginia groundcherry (Physalis virginiana Mill. #PHYLC), silverleaf nightshade (Solanum elaeagnifolium Cav. #SOLEL), western ragweed (Ambrosia psilostachya DC. #AMBPS), horseweed (Conyza canadensis (L.) Crong. #ERICA), and trailing fleabane (Erigeron flagellaris Gray) (Ellis et al. 1994). Homogenous saltcedar communities are generally devoid of herbaceous growth but some saltgrass (Distichlis spicata [L.] Greene #DISSP) occurs in remnant meadows.

The site chosen for restoration is bisected by a Rio Grande channel abandoned in 1942 during the last great flood when the river shifted about 1 km to the east. The area is bordered to the east by a low spoil levee resulting from construction of the Rio Grande low flow channel in the 1960s and on the west by an old levee constructed in the early 1940s prior to the flood. East-west cross dikes were constructed perpendicular to the repaired abandoned channel separating the project site into three management areas totaling 159 ha. Two areas, unit 28 and 29, are 61 ha while the remaining area, unit 30, is 37 ha. Flood irrigation developments divert water from the main Refuge canal to the repaired abandoned channel on the north end of unit 28 with continued flow to units 29 and 30. Water control structures were placed in cross dikes to provide irrigation capabilities and to create wetlands in lower elevations in units 28 and 29.

 

Saltcedar control. Work related to saltcedar control was staggered by unit and began in 1987 on unit 28; 1988 on unit 29; and 1991 on unit 30. Both contractor and refuge resources were used with records kept on all costs corresponding to contract equipment and labor prices (Table 1). Prior to clearing, saltcedar canopy closure averaged 70% (" 4%) and was not different among units (est. from 1987 aerial photos, 2.54 cm = 152 m; 30 random 5-cm measurements per unit). Saltcedar was generally shorter with fewer stems in unit 28 (2.4 " 0.15 m ht; 4,580 " stems/ha) than units 29 and 30 (ave. 3.4 " 0.13 m ht; 7000 " 585 stems/ha; measured within 90 random 2.4 m2 circular plots per unit).

Saltcedar in unit 28 was sprayed in September 1987 by fixed-wing aircraft with imazapyr applied at 1.12 kg/ha in a 140 L/ha solution with a drift control agent and nonionic surfactant added at 0.25% v/v. The next summer, green leaf material was absent and stems were desiccated so it was decided to chain the standing debris before broadcast burning. Two D-7 class bulldozers dragging a heavy gauge ship chain about 1 m above the surface laid down the standing vegetation in preparation for broadcast burning. Conditions during the burn in September 1988 averaged 40% relative humidity, 25EC air temperature, and 8 km/h wind speed from the south. Fuel moisture content was less than 10% and the fire consumed over 90% of the woody debris. One year later saltcedar resprouts were common over most of the burned area. We are not certain, but the high number of resprouts may partially have been the result of incomplete herbicide activity and the fire being conducted too soon after spraying. To control resprouts, the entire 61-ha area was root plowed to a 45-cm depth in autumn 1989 using procedures described by Horton (1960). Roots and other woody debris were then stacked in piles using bulldozers equipped with front-mounted brush blades. After burning and burying the piles, the entire area was smoothed by dragging a rail iron behind a bulldozer in preparation for revegetation. Saltcedar resprouts were still common after root plowing thus, in August 1990 larger plants were cut and immediately treated with a cut-stump formulation of imazapyr (Chopper 7) by ground crews using backpack pump sprayers. In August 1991, remaining resprouts were treated with a foliar application of imazapyr (Arsenal 7) applied at 1% v/v in water with a 0.25% nonionic surfactant added to a tank sprayer mounted on an all-terrain vehicle. Estimated cost for saltcedar control on unit 28 was $1,030/ha which included 2 years of spot herbicide applications on resprouts. In 1995, 6 yr after treatments began, an average of 72 live saltcedar resprouts/ha were recorded after a complete count over the unit.

Saltcedar in unit 29 was initially treated with imazapyr applied at 0.84 kg/ha in a 140 L/ha solution by fixed-wing aircraft in September 1988. Trees were 100% defoliated the next summer, and a broadcast burn was attempted without prior chaining in September 1989. This burn was incomplete, thus additional aerial vegetation clearing was required. Standing debris was raked at ground surface and windrowed using a hydraulic 6.4 m wide root rake pulled behind a D-7 bulldozer in autumn 1989. Rake teeth on the implement were 1.2 m in length, spaced 38 cm apart. Windrowed material was consolidated into piles for burning using a 1.9 m3 capacity scoop articulating loader adapted with a brush rake. As in unit 28, follow-up mechanical control was required over the entire area to reduce resprouts. During summer 1990, unit 29 was root plowed and raked, and debris was stacked into piles. These piles were then burned and the entire area smoothed by dragging a rail iron over the surface. Resprouts were treated in August 1991 with a 1% v/v imazapyr foliar application using an all-terrain vehicle. Total cost for this operation was $1,292/ha and an average of 63 resprouts/ha were counted across the unit in 1995.

From control experiences gained treating units 28 and 29, it was decided to forego the aerial herbicide application on unit 30 and use mechanical methods as initial treatments. Saltcedar aerial growth was first pushed down with a front mounted dirt blade attached to a D-7 bulldozer. Debris was stacked with the articulating loader, brush was raked, and piles were burned. Root plowing and root raking followed aerial clearing using equipment and implements previously described. Roots were stacked in piles, burned, and the entire area was smoothed for revegetation using the rail iron dragged over the surface. All work was accomplished over a 6 month period during spring and summer 1991. Few saltcedar resprouts were found following these control activities and maintenance spot herbicide applications were not required. Control cost for 37 ha in unit 30 was $750/ha, and saltcedar density averaged 15 resprouts/ha in 1995.

 

Revegetation. In 1988, 5 observation wells were hand drilled in unit 28 using a bucket auger and cased in 10 cm PVC pipe, and 4 additional wells were installed in units 29 and 30. Monitored monthly for 2 years, the wells indicated water table fluctuations of <0.6m, which is ideal for pole planting survival (Swenson and Mullins 1985). Water tables ranged from 1.2 to 4.3 m over the entire project area, but water was generally nearer the surface in the northern portion of the area (unit 28). After saltcedar control work was complete, a 0.2-ha grid system was mapped across each unit and elevational information was gathered using standard field surveying techniques to determine irrigation zones. Soil samples were taken at the center of each grid, 38 cm below the surface and 38 cm above the water table to determine salinity (electrical conductivity, EC) and soil texture. From this information a series of contour maps showing salinity levels, depth to water, and elevation were prepared. These maps formed data layers from which grid planting prescriptions were based.

Plantings were made from January to April 1990 and 1991 in units 28, and 29, respectively; and unit 30 in 1992 and 1993. Field crews were provided grid sheets showing the planting prescription based on guidelines developed for optimum survival and growth (Table 2). In retrospect, tolerable salinity levels may have been too high for some species and the Refuge has recently adjusted the original planting guideline given in Table 2. A maximum of 49 cottonwood, black willows, or shrub seedlings were planted in each grid, spaced 6.4 m apart depending on site suitability. For units 28 and 29, propagated and Refuge-cut stock was supplied by the Los Lunas Plant Material Center, NM (Fenchel et al. 1987). Tree poles were 2 to 3 m long with 5 to 7 cm butt diameter. For unit 30, Refuge-grown cottonwood and black willow tree poles were obtained from areas with 3 to 7 year old saplings. A 2 to 3 person Refuge crew cutting saplings near the surface harvested about 200 to 300 poles per d. The ends of these poles were soaked in water for 10 d before planting in augured holes that penetrated the water table and left 2 to 3 apical branches above the surface on each pole.

On average, a 3 person crew planted 150 to 180 poles per d using a Texoma production auger (30.5 cm dia.) drilling to 4.3 m depth. About 100-125 poles per d were planted using a McMillan auger (30.5 cm dia.) mounted on a front-end loader bucket drilling to a 3.6 m depth. A bobcat landscaping machine fitted with an attachable auger (23 cm dia.) drilling to 2.4 m depths was the most rapid planting method, as a 2 person crew planted 200-250 poles per d. Shrubs supplied by the Los Lunas Plant Materials Center were planted in units 28 and 29 but not unit 30 using propagated seedlings (wolfberry, New Mexico olive, silver buffaloberry (Shepherdis argentea (Pursh) Nutt.), screwbean mesquite, skunkbush sumac (Rhus trilobata Nutt.), and fourwing saltbush) in standard book containers with a minimum 20 cm of root development. Holes were drilled to the water table and refilled before planting shrubs to aid root penetration (Anderson 1988b). After hand planting, seedlings were watered from a tank truck and mulched with standard roofing felt as a weed control measure. Seedlings were later flood irrigated about every other month the first growing season and annually thereafter in late May.

Costs for planting 32 ha in unit 28 with 5,500 cottonwood and willow poles and 1,500 shrub seedlings averaged $7.75 per planting. In unit 29, 4,200 cottonwood and willow poles and 2,500 shrub seedlings planted over 28 ha and cost $7.30 per planting. Costs in unit 30 were lower than other units because poles were obtained on the refuge and no shrub seedlings were planted. Crew efficiency also increased, allowing 5,500 cottonwood and black willow poles to be planted on 23 ha for about $3.75 per plant.

Cottonwood survival exceeded 80% 4 yr after planting units 28 and 30, but frost damage shortly after leaf out in April killed many first-year poles in unit 29 (53% alive after 4 yr). Cottonwood growth was slowed somewhat when poles were planted in areas where water table depth exceeded 4.3 m and soil salinity approached 3.0 dS/m, but overall annual growth was linear and equal across units (Figure 2; height = 1.35 + 0.68 (yr), F = 0.871, df = 2,6, n = 1032). By comparison, deep tillage to 3 m and drip irrigation for 165 d resulted in 100% cottonwood and black willow survival when plantings were made on a dredge spoil site on the lower Colorado River in Arizona (Ohmart and Anderson 1982). Under drip irrigation, cottonwoods grew about 10 to 15 mm/d during June and July from sites ranging from the Rio Grande near Presidio, Texas to the Kern River in California (Anderson 1989). Black willows established easily in units 28 and 29 (> 80% survival), but the deeper water table and girdling damage by rabbits to nearly 90% of the first yr trees lowered survival in unit 30 (25% alive after 4 yr). Black willow height increased about 0.75 m/yr through the first 4 yr (Figure 3, height = 1.35 + 0.68 (yr), F = 0.97, df = 2,6, n = 517).

Planted shrub survival, with the exception of New Mexico olive, was disappointing but was more than offset by natural regeneration (Table 3). Native species including coyote willow, seepwillow, numerous herbs, and to a certain extent cottonwood and black willow, responded to irrigations, resulting in nonplanted species comprising nearly 98% of the woody composition in 1995. Considering these results, the refuge now attempts to mimic natural flooding processes using controlled water level manipulations to facilitate regeneration of native flora where possible. Although saltcedar seedlings are recruited using this technique, experience has shown they form a minor component of the overall flora assemblage.

Discussion

Several studies have attempted to compare native riparian versus saltcedar habitat values for wildlife but such assessments are difficult, particularly for breeding species (Ellis et al. 1993, Ellis et al. 1994, Thompson et al. 1994). By nature, riparian habitats are linear and influenced by zones of transition (edge) from one habitat type to another. Faunal use of edge habitats can be markedly different than more interior homogeneous habitats (Hink and Ohmart 1984), further restricting interpretations from wildlife surveys. Some general conclusions can be drawn, however. Native riparian communities with rich canopy structures and abundant decadent trees which support nest cavities harbor greater species diversity and numbers than saltcedar monocultures (Anderson and Ohmart 1982 and 1984, Sedgwick and Knopf 1986, Ohmart et al. 1988, Busch et al. 1992, and Ellis 1995). While this statement is usually true in a given area, Hunter et al. (1988) cautions against applying avian use data from one river system to another when assessing habitat values. For example, Anderson and Ohmart (1984) reported lower avian diversity and density in saltcedar habitats than native habitats on the lower Colorado River, yet Raitt and Desantro (1980) found lower diversity but higher bird densities in saltcedar habitats along portions of the middle Rio Grande.

The restoration effort on the 159-ha project area has been instrumental in attracting a diverse fauna. The reactivation of an abandoned river channel through the project area, which in turn provided a water source for wetland development and natural plant regeneration, was perhaps the single most important factor influencing faunal response. Initial site disturbance resulted in abundant seed-producing pioneer herbs that attracted fauna with an affinity to open grassland or shrub habitats. While floral structure and habitat patchiness is still in the early stages of development, a noticeable shift is occurring in the project area toward a wider variety of birds species having various forage guilds (Franzreb 1981, Ellis 1995). Over a 5-yr monitoring period (1992-1996), avian richness nearly doubled (Table 4) (Stuart and Farley 1993, Bosque del Apache NWR unpublished biomonitoring program reports). Compared with other riparian communities, including cottonwood gallery forest and mixed cottonwood forest, the project area now harbors the highest avian species richness of any Refuge habitat. Reptile and amphibian species richness is also higher and small mammal diversity on the project area is similar to other nearby riparian habitats on the Refuge (Table 4, Stuart et al. 1992 and 1993, Stuart and Farley 1994, Bosque del Apache NWR unpublished report, Ellis et al 1993 and 1994). Developing flora and progressively more mesic microhabitats reflected a general shift from fauna typical of dryer, less densely vegetated habitats to species representing more mesic fringes.

In terms of faunal response, there is no question that restoration on the project area has been worthwhile. The high cost of this effort however, cannot be overlooked (Table 1). The least expensive mechanical saltcedar control cost were $752/ha in 1993 dollars which is similar to mechanical clearing on the Cibola NWR in Arizona at $775/ha in 1982 dollars (Anderson and Ohmart 1982). Revegetation cost were nearly twice as high when both seedlings and poles were planted from nursery stock compared to use of locally harvested plant materials. Planted areas do not capture the full scope of restoration work which occurred during the project, however. By providing irrigation capabilities, wetlands were developed in areas unsuitable for planting, and enabled recruitment of native flora through natural regeneration.

 

LITERATURE CITED

Anderson, B. W. 1988a. Revegetation of two 25 acre plots on the Kern River Preserve. Nature Conservancy and California Department of Fish and Game, Sacramento, California, 42 pp.

Anderson, B. W. 1988b. Deep tillage aids tree establishment in riparian revegetation projects in arid southwest. Restoration and Management Notes 6:84-87.

Anderson, B. W. 1989. Research as an integral part of revegetation projects. Pages 413-419 in R.E. Warner and K.M. Hendrix, eds. Proceedings of the California Riparian Systems Conference, Berkeley, Ca.

Anderson, B. W. and R. D. Ohmart. 1982. Revegetation and wildlife enhancement along the lower Colorado River. U.S. Dept. of the Interior, Bureau of Reclamation. Contract No. 7-07-30-V0009. 215 pp.

Anderson, B. W. and R. D. Ohmart. 1984. Final Report-vegetation management study for the enhancement of wildlife along the Lower Colorado River. U.S. Dept. of the Interior, Bureau of Reclamation. Contract No. 7-07-30-V0009. 529 pp.

Anonymous. 1951. Report to saltcedar interagency council task force for New Mexico. Albuquerque, NM.

Busch D., Herbranson, L., Johns, E., Pinkney, F., and D. Sisneros. 1992. Vegetation management study: lower Colorado River. Phase I report to the U.S.D.I.-U.S. Bureau of Reclamation, lower Colorado region. 103 pp.

Campbell, H. W. and S. P. Christman. 1982. Field techniques for herpetofaunal community analysis. Pp 193-200 in Herpetological communities. N.J. Scott, Jr. (ed.). U.S. Dept. Interior, U.S. Fish and Wildlife Service. Wildlife Research Report 13. 239 pp.

Deloach, J. C. 1989. Saltcedar, a weed of western North American riparian areas: a review of its taxonomy, biology, harmful and beneficial values, and its potential for biological control. Final Report for Bureau of Reclamation, Lower Colorado Region, Boulder City, Nevada, 296 pp.

Ellis, L. M., C. S. Crawford, and M. C. Molles. 1993. The effects of annual flooding on the Rio Grande riparian forests: Bosque del Apache National Wildlife Refuge, San Antonio, New Mexico. Progress report submitted to U.S. Fish and Wildlife Service, Albuquerque, N.M. 60 pp.

Ellis, L. M., C. S. Crawford, and M. C. Molles. 1994. The effects of annual flooding on the Rio Grande riparian forests: Bosque del Apache National Wildlife Refuge, San Antonio, New Mexico. Progress report submitted to U.S. Fish and Wildlife Service, Albuquerque, N.M. 91 pp.

Ellis, L. M. 1995. Bird use of saltcedar and cottonwood vegetation in the Middle Rio Grande valley of New Mexico, U.S.A. J. Arid Envir. 30:339-349.

Emlen, J.T. 1971. Population densitites of birds derived from transect counts. Auk 88:323-342.Farley, G. H., L. M. Ellis, J. N. Stuart, and N. J. Scott. 1994. Avian species richness in different-aged stands of riparian forest along the Middle Rio Grande, New Mexico. Conservation Biology. 8:1098-1108.

Fenchel, G., W. Oaks, and E. A. Swenson. 1987. Selecting desirable woody vegetation for environmental mitigation and controlling wind erosion and undesirable plants in the Rio Grande and Pecos river valleys of New Mexico. 5 year interim report (1983-87). USDA-SCS-Plant Materials Center, Los Lunas, N.M. 49 pp.

Fenchel, G. A., D. Dreesen, and J. Fraser. 1996. 1996 interagency riparian report. USDA-NRCS-Plant Materials Center, Los Lunas, N.M. 42 p.

Franzreb, K. E. 1981. A comparative analysis of territorial mapping and variable-strip transect censusing methods. Studies in Avian Biology 6:164-169.

Great Western Research, Inc. 1989. Economic analysis of harmful and beneficial aspects of saltcedar. Final Report for Bureau of Reclamation, Lower Colorado Region, Boulder City, Nevada, 259 pp.

Hink, V. C. and R. D. Ohmart. 1984. Middle Rio Grande biological survey. Final Rep. to the U.S. Army Corps of Engineers, Albuquerque, NM. 193 pp.

Horton, J. S. 1960. Use of a root plow in clearing tamarisk stands. U.S.D.A. Forest Service Res. Note 50. Ft. Collins, CO. 6 pp.

Hunter, W. C., R. D. Ohmart, and B. W. Anderson. 1988. Use of exotic saltcedar (Tamarix chinensis) by birds in arid riparian systems. Condor 90:113-123.

Neill, W. M. 1988. Control of tamarisk at Desert Springs. Desert Protective Council. 6 pp.

Ohmart, R. D., Anderson, B. W., and W. C. Hunter. 1988. The ecology of the lower Colorado River from Davis dam to the Mexico-United States international boundary: a community profile. U.S.D.I.-U.S. Fish and Wildlife Service Biological Report 85(7.19). 296 pp.

Raitt, R. J. and M. C. Delasantro. 1980. Avifauna census, Elephant Butte and Caballo reservoirs, New Mexico. Final Report to the U.S.D.I.-U.S. Water and Power Resources Services, Rio Grande Project. Contract No. 9-07-54-V0507. 226 pp.

Ralph, C. J., G. R. Geupel, P. Pyle, T. E. Martin, and D. F. DeSante. 1993. Handbook of field methods for monitoring landbirds. Gen. Tech. Rep. PSW-GTR-144. Albany, Ca. 41 pp.

Reynolds, R.T., J.M. Scott, and R.A. Nussbaum. 1980. A variable circular-plot method for estimating bird numbers. Condor 82:309-313.

Sedgwick, D. and F. Knopf. 1986. Cavity-nesting birds and the cavity-tree resource in plains cottonwood bottomlands. J. Wildl. Manage. 50:247-252.

Stuart, J. N. and G. H. Farley. 1994. Use of riparian revegetation sites by terrestrial vertebrates in the Rio Grande Valley, New Mexico. Final Report submitted to U.S. Fish and Wildlife Service, Fort Collins, Colorado. 58 pp.

Stuart, J. N., N. J. Scott, Jr. and G. H. Farley. 1992. Use of riparian revegetation sites along the Rio Grande by terrestrial vertebrates. Report submitted to U.S. Fish and Wildlife Service, Fort Collins, Colorado. 85 pp.

Stuart et al. 1993. Use of riparian revegetation sites by terrestrial vertebrates in the Rio Grande Valley, New Mexico. Report submitted to U.S. Fish and Wildlife Service, Fort Collins, Colorado. 173 pp.

Swenson, E. A., and C. L. Mullins. 1985. Revegetating riparian trees in southwestern floodplains. Pages 135-138 in R. R. Johnson, C. D. Ziebell, D. R. Patton, P. F. Ffolliott, and R. H. Hamre, eds. Riparian ecosystems and their management: reconciling conflicting uses. U.S. For. Serv., Gen. Tech. Rep. RM-120. 523 pp.

Swenson, E. A., and M. Houck. 1989. Selection and development of plants for use in southwestern riparian zones. Page 68 in Issues and technology in the management of impacted wildlife-proceedings of a national symposium. Glenwood Springs, CO.

Taylor, J. P. 1987. Imazapyr (arsenal) use and performance on saltcedar (Tamarix pentandra), willow (Salix and Baccharis spp.) and Phragmites (Phragmites communis) at the Bosque del Apache NWR. U.S. Fish and Wildlife Service. Albuquerque, NM. 15 pp.

Thompson, B. C., D. A. Leal, and R. A. Myer. 1994. Bird community composition and habitat importance in the Rio Grande system of New Mexico with emphasis on neotropical migrant birds. US.F.W.S. and N.B.S. Cooperative Agreement 14-16-0009-1592, No. 11. 151 pp.

Warren, D. K., and R. M. Turner. 1975. Saltcedar (Tamarix chinensis) seed production, seedling establishment, and response to inundation. J. Ariz. Acad. Sci. 10:135-144.

Wiedemann, H. T., and B. T. Cross. 1978. Water inundation for control of saltcedar along the periphery of lakes. Page 229 in B. Truelove, ed. Proceedings Southern Weed Science Society 31st Annual Meeting. Auburn University Printing Service, Auburn, Ala. 439 pp.