Grantee Research Project Results
Final Report: Regional Ecological Resource Assessment of the Rio Grande Riparian Corridor: A Multidisciplinary Approach to Understanding Anthropogenic Effects on Riparian Communities in Semi-arid Environments
EPA Grant Number: R827677Title: Regional Ecological Resource Assessment of the Rio Grande Riparian Corridor: A Multidisciplinary Approach to Understanding Anthropogenic Effects on Riparian Communities in Semi-arid Environments
Investigators: Raney, Jay , Crawford, Melba , Neuenschwander, Amy , Paull, Gene , Judd, Frank , Lonard, Robert , Tremblay, Thomas , White, William , Encheva, Tatiana
Institution: The University of Texas at Austin , The University of Texas - Pan American , The University of Texas at Brownsville
EPA Project Officer: Packard, Benjamin H
Project Period: September 1, 1999 through August 31, 2002 (Extended to December 31, 2003)
Project Amount: $642,496
RFA: Regional Scale Analysis and Assessment (1999) RFA Text | Recipients Lists
Research Category: Aquatic Ecosystems , Ecological Indicators/Assessment/Restoration
Objective:
Riparian ecosystems of the Southwestern United States are among the most productive ecosystems of North America. The rapid decline of these ecosystems throughout the United States, including the Lower Rio Grande Valley, has made riparian conservation a focal issue. This was a multidisciplinary study of riparian communities along the Lower Rio Grande Valley of Texas and Mexico. The objectives of this research project were to: (1) acquire and analyze high-resolution, remotely sensed data from multiple sensors; (2) integrate existing and new field data and remotely sensed data into a geographic information system (GIS); (3) ascertain whether the native vegetation communities are maintaining themselves and identify the topographic, edaphic, and other ecological factors that perpetuate these communities; (4) interprete spatial variations in riparian habitats, including comparisons of the north and south banks of the Rio Grande; (5) analyze temporal changes at specific locations; and (6) develop a foundation for future analysis of riparian floodplain communities by linking local and remotely sensed regional data using GIS.
Summary/Accomplishments (Outputs/Outcomes):
Analysis and classification of riparian vegetation in the Lower Rio Grande Valley using remote sensing data supported by field surveys confirmed what other researchers have qualitatively suggested, that riparian vegetation has been diminished greatly since the early 1900s. Digital analysis of historical maps and aerial photographs of woodland distribution in Cameron County as part of this study revealed that in the mid-1930s, there were approximately 81,887 ha of woodlands in Cameron County (Tremblay and White, 2002). By the early to mid-1980s, only 7,337 ha of woodlands in this original area remained, indicating a loss of approximately 91 percent of this resource. This quantitative assessment of woodland loss helps confirm the earlier qualitative estimates of up to a 95 percent loss.
Today, riparian vegetation in the Lower Rio Grande Valley has a limited distribution. Based on repeated vegetation surveys at selected sites (see Figure 1), researchers at the University of Texas–Pan American (UT-PanAm) concluded that the dominant trees and shrubs along the Rio Grande appeared to be replacing themselves. In addition, they found that there were no trees at the mouth of the river and the vegetation there was similar to that found along the Laguna Madre shore of barrier islands. Mesquite (Prosopis glandulosa) was the dominant tree near the coast, where soil salinity and wind blown salt spray are greatest. It also was dominant in the western section of the river near Falcon Dam, where rainfall is least and where the Rio Grande floodplain is narrow. Sugar hackberry (Celtis laevigata) was the dominant tree species at all other sites except at the Santa Ana National Wildlife Refuge (NWR), where cedar elm (Ulmus crassifolia) and anacua (Ehretia anaqua) were the dominant trees. Granjeno (Celtis pallida) was a dominant shrub throughout the riparian corridor. The introduced Guinea grass (Panicum maximum) and buffel grass (Pennisetum ciliare) were the dominant species in the ground cover, displacing native species. The present riparian communities may be greatly influenced by human interventions such as construction of dams that have eliminated annual flooding of the Rio Grande. Blair (1950) reported that cedar elm (Ulmus crassifolia) was the dominant tree in the floodplain of the Rio Grande in the Lower Rio Grande Valley of Texas. We found cedar elm was a dominant species only at the Santa Ana NWR (Lonard and Judd, 2002). The distribution and abundance of this species may have been adversely affected by the curtailment of annual flooding of the Rio Grande. Certainly, it is no longer a widespread dominant species in the riparian zone of the lower reach of the Rio Grande.
Figure 1. Index Map Showing the Rio Grande and Approximate Locations of 8 of 11 Vegetation Transects Along The River. Vegetation surveys were repeated at several sites that had been surveyed in past years to determine if riparian vegetation was replacing itself. Photo of riparian vegetation was taken in Santa Anta NWR.
Using remote sensing data acquired of the Lower Rio Grande Valley, scientists at the Center for Space Research (CSR) analyzed and classified woodlands and riparian vegetation. The most recent Landsat imagery acquired between 2000 and 2002, was used to determine the current distribution of riparian woodlands. The data set was entered into the Bureau of Economic Geology (BEG) GIS for analysis. In addition to the lower-resolution multispectral (Landsat Thematic Mapper [TM]) data analyzed by CSR, high-resolution hyperspectral (HYMAP) data were acquired for selected sites and used to refine our classification of woodlands and riparian vegetation. Color-infrared photography with 1-m resolution, in conjunction with the high-resolution (4 to 7 m) spectrally calibrated hyperspectral data supported by field surveys, were used to train classification algorithms and visually evaluate resulting classes in the Santa Ana NWR and the Bentsen-Rio Grande Valley State Park. The Santa Ana NWR contains one of the largest contiguous riparian communities along the Rio Grande. The remote-sensing signatures at training sites on the high-resolution data were used for classification of medium-resolution Landsat 7 data to evaluate the utility of these sites in: (1) scaling upward from medium to high resolution data; and (2) improving the riparian classification of the medium resolution data. The Landsat 7 data have extensive areal coverage but lower spatial and spectral resolution than hyperspectral data, and lower spatial resolution than DOQs (digital orthophoto quadrangle).
Because of the large number of species representing riparian vegetation along the Rio Grande, and the difficulty in adequately differentiating the various species using remotely sensed imagery, we established five classes of vegetation communities defined by the presence of evergreen and deciduous species as well as combinations of the two. The composition of the vegetation was determined from field surveys and from interpretation of high-resolution, digital CIR aerial photographs (DOQs) acquired during winter months. This classification approach is modeled after the U.S. Fish and Wildlife Service (USFWS) National Wetlands Inventory program, in which riparian vegetation inventory and mapping conventions were developed for the Western United States. The USFWS classification is hierarchical, with the riparian system having two subsystems, lentic and lotic, subdivided into forested and scrub/shrub classes. These, in turn, have three subclasses: deciduous, evergreen, and mixed, from which we established five subclasses consisting of: (1) evergreen; (2) deciduous; (3) mixed, codominant; (4) mixed, evergreen dominant; and (5) mixed, deciduous dominant. Examples of common evergreen species identified through field surveys in the Santa Ana NWR and along other reaches of the Rio Grande include Texas ebony (Chloroleucon ebano), anacua, granjeno, la coma (Sideroxylon celastrina), huisache (Acacia minuata) (see Figure 2), and tepeguaje (Leucaena pulverulenta). Examples of deciduous species include hackberry, cedar elm, mesquite, black willow (Salix nigra), retama (Parkinsonia aculeata), Texas persimmon (Diospyros texana), and Rio Grande ash (Fraxinus berlandieriana). This last species is deciduous or semi-evergreen.
Using remote sensing data of various scales, resolution, and seasons of acquisition, and supported by the detailed field surveys, we classified riparian vegetation communities into the five classes defined by the presence of evergreen and deciduous species as well as combinations of the two as described above. We achieved relatively good results in the Santa Ana NWR (see Figure 3). Poorer results, however, were achieved in scaling upward from the hyperspectral data to Landsat 7 TM data; results degraded further when extended beyond the refuge. Although general trends in vegetation communities outside the refuge were defined, boundaries between classes were less distinct and there was a larger scattering of classes. We concluded that the best results in the evergreen and deciduous characterization were obtained using only three subclasses—–evergreen, deciduous, and mixed—–as defined by the USFWS. Five subclasses, as discussed above, could not be as consistently classified because of complex mixtures in vegetation communities.
Digital land use and climate maps were completed by The University of Texas at Brownsville (UTB). Current land use was based on maps prepared from 1995 DOQs, and historical land use was based on existing BEG land use maps based on 1960 aerial photographs. The largest land use parcel was agriculture, followed by range-pasture and urban. Observations from the Brownville-Harlingen-McAllen sector of the Lower Rio Grande Valley show that the urban-residential category increased dramatically from 1960 to 1995. There was a slight decrease in agricultural land use. Overlays of 1995 and 1960 data show an explosive growth of residential urban parcels, particularly in the McAllen-Pharr-Edinburg area. Mapping of woodland shows very little of this category left in Hidalgo County. The year 2000 U.S. Census data for the four counties of the Lower Rio Grande Valley show a combined population approaching 1,000,000. The land use maps graphically indicate how this growth has impacted natural vegetation.
Maps on climate include average annual precipitation, September precipitation, average annual temperature, January mean temperature, July mean temperature, heating degree days, and cooling degree days. The climatic maps show systematic variations in precipitation and temperature in the study area, including decreasing average rainfall and increasing average temperatures as one proceeds up the Rio Grande Valley from the Gulf of Mexico. There is evidence that the decreasing annual precipitation up the valley corresponds with a relatively lush mesic plant community in riparian areas near the coast to a more xeric assemblage farther inland.
Figure 2. Photo of Vegetation That Includes Huisache Near the Entrance to Bentsen-Rio Grande Valley State Park. Photo taken in December 2003, before leaf fall.
Figure 3. Southern Half of the Santa Ana NWR Classified Using HYMAP Data Into Five Classes: Deciduous (D), Evergreen (E), Mixed (M), Mixed-Evergreen Dominant (M-e), Mixed-Deciduous Dominant (M-d).
We found a strong correlation between riparian vegetation and soils. Along the Rio Grande in Cameron County, for instance, although 17 different soils were associated with riparian vegetation, 3 soils made up more than 60 percent of the association (Rio Grande silt loam 22 percent; Zalla loamy fine sand 21 percent, and Matamoros silty clay 18 percent). Within a 3-km-wide corridor along the Rio Grande, which includes Cameron, Hidalgo, and Starr Counties, we found a similarly strong relationship. Within the 3-km corridor, these three soils, plus Laredo silty clay loam, cover only 32 percent of the area, but they are the soils on which 61 percent of the riparian vegetation occurs.
To further investigate the relationship between soils and riparian vegetation, we analyzed the distribution of common species of trees and shrubs that were identified at the approximately 160 field sites visited by researchers from UT-PanAm. All shrub and tree species identified at the sites were entered into our GIS, and a GIS layer of the common species found at the sites was developed for analysis of soil relationships. Results indicate that most species were more common on two soils, Laredo Silty Clay Loam and the Rio Grande Silt Loam. There were fewer occurrences on clays such as the Grulla Clay and Harlingen Clay (see Table 1). In addition, we analyzed the relationship between soil salinities and 10 common species of shrubs and trees. This was accomplished by analyzing the number of occurrences of the trees and shrubs on soils with salinities (based on conductivity) ranging from 0 to 4 millimhos/cm. This analysis was based on all species found at distinct field check sites and transect locations, as reported by Lonard and Judd (2002). Soil salinity is represented as electrical conductivity in millimhos/cm at 25°C. The Natural Resource Conservation Service classifies soils as either nonsaline (0-2) or slightly saline (2-4). Among the results was that mesquite occurred more frequently in slightly saline soils than did other species. This finding is in agreement with that of Lonard and Judd (2002), who found mesquite to be the dominant species near the coast, where the effects of salinity and salt spray are most pronounced. This relationship between vegetation and soils, when correlated with other parameters such as topography, hydrology, and land use, is useful in analyzing riparian vegetation with respect to historical trends, anthropogenic effects, and optimal sites for reestablishment of riparian tracts.
Table 1. Frequency (number of occurrences at all field sites) of Common Riparian Tree and Shrub Species on Four Soils in the Lower Rio Grande Valley
Laredo Silty Clay Loam | Rio Grande Silt Loam | Grulla Clay | Harlingen Clay | |
---|---|---|---|---|
Acacia minuata (Huisache) | 9 | 6 | 2 | 2 |
Celtis lavigata (Sugar hackberry) | 15 | 8 | 7 | 2 |
Chloroleucon ebano (Texas ebony) | 15 | 4 | 1 | |
Leucaena pulverulenta (Tepeguaje) | 7 | 3 | 1 | |
Parkinsonia aculeata (Retama) | 6 | 7 | 3 | 2 |
Phaulothamnus spinescens (Snake Eyes) | 5 | 1 | 1 | 2 |
Prosopis glandulosa | 14 | 6 | 3 | 2 |
Sakix nigra (Black willow) | 2 | 2 | ||
Ulmus crassifoha (Cedar elm) | 1 | 6 | 3 | |
Zanthoxylum fagara (Colima) | 2 | 2 | 4 | |
Total occurrences | 83 | 45 | 25 | 12 |
To make comparisons between the remaining riparian vegetation in Texas and Mexico, we created a 20-km-wide buffer zone along the Rio Grande, with 10 km on the United States side and 10 km on the Mexico side (see Figure 4). By comparing the distribution and amount of riparian vegetation classified within the 20-km corridor along the Rio Grande (10 km in the United States and 10 Km in Mexico), we found that of the total woodlands mapped within this area of analysis, 74 percent occurs in the United States, and 26 percent occurs in Mexico. Compared to other types of land cover such as cropland, only small percentages of woodlands, 6 percent in the United States and 2 percent in Mexico, remain. If we assume that in the past, most of the area was vegetated with riparian woodlands and brushlands, as has been suggested by some authors, then almost 95 percent of these wooded areas have been cleared in the United States, and 98 percent in Mexico. On the United States side, this is in agreement with estimates by Jahrsdoerfer and Leslie (1988), who stated that since the early 1900s, 95 percent of the native brushland has been cleared for agriculture, urban development, and recreation, and in riparian areas, an estimated 99 percent of native brush has been destroyed.
Figure 4. Illustration Showing 20-km Buffer Zone Along the Rio Grande From the Gulf of Mexico to Falcon Dam, Within Which Analysis of Riparian Vegetation Was Analyzed in the United States and Mexico. Dark (red) areas are riparian woodlands.
Among the more optimistic aspects regarding riparian vegetation along the Lower Rio Grande Valley are the efforts of the USFWS, the Texas Parks and Wildlife Department (TPWD), the National Audubon Society, and the Nature Conservancy. These agencies have been involved in programs that actively help preserve and restore riparian habitats, ranging from the TPWD's acquisition of white-winged dove habitat, to the National Audubon Society's Sabal Palms Santuary, and the USFWS large-scale acquisitions as part of the USFWS Lower Rio Grande Valley NWR. Associated with the acquisition of land is a rigorous planting program in which a variety of evergreen and deciduous shrubs and trees are being planted to help restore riparian habitat corridors along the Rio Grande. It is hoped that the analysis of riparian distribution and dominant plant species identified and reported in this study, and their relationship to soils, hydrology, land use, salinity, topography, and other parameters, will assist in riparian restoration programs in the Lower Rio Grande Valley, and serve as a foundation for future analysis of riparian floodplain communities by linking local and remotely sensed regional data using GIS.
References:
Blair WF. The biotic Provinces of Texas. Texas Journal of Science 1950;2:93-117.
Jahrsdoerfer SE, Leslie Jr. DM. Tamaulipan brushland of the Lower Rio Grande Valley of south Texas: description, human impacts, and management options. Fish and Wildlife Service, Biological Report 1988;88(36):63.
Journal Articles on this Report : 2 Displayed | Download in RIS Format
Other project views: | All 20 publications | 3 publications in selected types | All 2 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Lonard RI, Judd FW. Riparian vegetation of the Lower Rio Grande. Southwestern Naturalist 2002;47(3):420-432. |
R827677 (2001) R827677 (2002) R827677 (2003) R827677 (Final) |
Exit |
|
Tremblay TA, White WA, Raney JA. Native woodland loss during the mid 1900s in Cameron County, Texas. Southwestern Naturalist 2005;50(4):479-482. |
R827677 (2003) R827677 (Final) |
Exit |
Supplemental Keywords:
EMAP, land management, irrigation, Lower Rio Grande Valley, agriculture, resource management, salinity, biome, videography, spectral imagery, ecosystem protection/environmental exposure and risk, geographic area, ecological effects, ecological indicators, ecology, ecosystem protection, cosystem/assessment/indicators, environmental chemistry, hydrology, regional/scaling, southwest, state, Bayesian classifiers, GIS, Rio Grande Riparian Corridor, riparian ecosystem, Texas, TX, agriculture, anthropogenic, ecological assessment, ecological exposure, environmental data, floodplain communities, land use, landscape patterns, regional scale impacts, remotely sensed data, scaling methods., RFA, Scientific Discipline, Geographic Area, Ecosystem Protection/Environmental Exposure & Risk, Hydrology, Southwest, Ecology, Environmental Chemistry, Ecosystem/Assessment/Indicators, State, Ecological Effects - Environmental Exposure & Risk, Regional/Scaling, Ecological Risk Assessment, ecological exposure, EMAP, semi-arid environments, Texas, Riparian ecosystem, floodplain communities, ecological assessment, environmental data, anthropogenic, regional scale impacts, Rio Grande Riparian Corridor, agriculture, GIS, landscape patterns, remotely sensed data, land use, scaling methodsProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.
Project Research Results
- 2003 Progress Report
- 2002 Progress Report
- 2001 Progress Report
- 2000 Progress Report
- Original Abstract
2 journal articles for this project