AIRBORNE
Monitoring Native Riparian Forest Vegetation
Color infrared film aids aerial change evaluation in lower Rio Grande
By Robert I. Lonard, F. W. Judd, H.H. Everitt, D.E. Escobar, M.R. Davis, M.M. Crawford, and M.D. Desai

The Rio Grande River is significant for its role in the history of the United States and Mexico. It represents the second longest river system in the U.S. Historically, the Rio Grande flowed freely through arid regions of the southwestern United States and Mexico. Extensive riparian forests and woodland communities apparently occupied the lower reach of the Rio Grande in Texas. Construction of dams and reservoirs along the lower Rio Grande for flood control and for agricultural and municipal uses has resulted in the clearing of this riparian forest vegetation. Since the 1920s between 90% and 99% of the riparian vegetation on the United States side of the Rio Grande in the lower Rio Grande Valley has been removed .
    Surprisingly, the riparian vegetation in Texas is poorly known. Riparian vegetation along the lower Rio Grande has not been studied intensively, although qualitative information consistent with the procedures and the times has been documented. The studies emphasize the floristic composition and provide anecdotal or subjective descriptions of the riparian vegetation in the lower Rio Grande Valley. Four studies have also been conducted which provide quantitative data on the riparian vegetation of the lower Rio Grande, but each of these studies examines limited areas of the river from Falcon Dam in Starr County to the Sabal Palm Sanctuary in Cameron County. The ecological characteristics of these communities and native forests as a whole are not well understood. Specifically, information is scarce on the relative abundance of species, diversity relationships, structural features, and gradient analysis.
    The lower portion of the Rio Grande has been identified as a region where wildlife habitat is vanishing and is in critical need of protection. Knowledge of the composition of the extant, native flood plain communities is crucial for the establishment and management of wildlife habitats including those habitats for species that have been identified as threatened or endangered. To preserve existing riparian forest remnants and to establish a wildlife corridor along the river, federal and state governments are purchasing land. It will be difficult for wildlife managers to reestablish riparian vegetation if they do not know the composition of existing woodland communities.
    Soci-economic issues associated with the North American Free Trade Agreement (NAFTA) have concentrated attention on the abatement of water pollution and the existing condition of native plant communities adjacent to the Rio Grande. Thus, to provide this quantitative description of current conditions in the more xeric segment of the lower Rio Grande, a quantitative study of the lower Rio Grande linked with remote sensing technology was deemed useful. The objectives were to 1) provide a quantitative description of the vegetative composition of the riparian zone, 2) determine if dominant species exhibited characteristic image response to color infrared (CIR) film, 3) ascertain annual changes in vegetation patterns through remote sensing technology, and 4) clarify patterns of variation.
    The Rio Grande in southern Texas forms the international boundary between the United States and Mexico. The flood plain is less than 1 kilometer wide in Starr County where the river leaves the impoundment of Falcon Dam, and broadens to between 4 and 6 km wide in much of Hidalgo County.

Scope of the Study
Two study sites were selected located in the riparian zone of the Rio Grande in Hildalgo and Starr counties. The first site is located 8 km southwest of Mission at Bensten-Rio Grande Valley State Park. The 237 hectare park is one of the few public natural preserves adjacent to the Rio Grande in the southern portion of Texas. Vegetation in upland sites at the park is dominated by Prosopis glandulosa (mesquite) and Opuntia lindheimeri (pricklypear).
    The Starr County site is located at Saline–o, 10 km downstream from Falcon Dam. The plant community in adjacent upland sites is dominated by P. glandulosa, Celtis pallida (granjeno), and O. lindheimeri. Identified as the Chihuahuan Thorn Forest, this region is known locally as the Falcon Thorn Forest and is ranked among national ecosystems for its unique and valuable plant community.
    Using Kodak aerochrome type 2443 film, aerial images were taken using a Hasselblad 70mm camera outfitted with an orange filter and an aperture setting of f5.8 at 1/500 second. Equipped with an 80mm lens, images were acquired at an altitude of approximately 460 meters (m) above the ground, providing a ground scale resolution of about 1:5,600. Aerial images were acquired a year apart on June 6, 1997 and June 12, 1997.
    The line intercept technique was used to provide ground truth information and to quantify vegetation abundance at the two study sites. Three replicate transects were placed perpendicular to the margin of the river at each site. Transects, established along a compass line 10m apart, began at the edge of the river and ended at the crest of the first terrace. Transects were 30m long at Bentsen-Rio Grande Valley State Park, and 40m long at Saline–o. Transects were divided into 10m intervals and readings were taken along the total length of each interval. Each species intercepted by the line was rated individually and scored with separation into strata, i.e., tree, shrub, and ground layers. Records of species, frequency, and foliage cover were recorded. Importance values were assigned to determine which species were dominant at each site.
    These investigations showed that the dominant species in the tree layer within Bentsen-Rio Grande Valley State Park's riparian forest were Celtis laevigata (sugar hackberry), which contributed 62.4% of the relative cover; Salix nigra (black willow); and Acacia smallii (huisache) which accounted for 16.7% and 12.6% of the relative cover, respectively. In the shrub zone, only four species were identified, with the dominant shrub being Celtis pallida (granjeno). The introduced grass, Panicum maximum (Guineagrass) was the prevailing ground cover, representing 98.6% of the total cover. Another 2% of ground cover was identified as Paspalum lividum, and was located along the edge of the river.
    A more diverse tree, shrub, and ground cover population was identified in the Saline–o forest. The dominant species was P. glandulosa, which contributed 36.1% of the total riparian forest cover. Other important species included Parkinsonia aculeata (retama), Fraxinus berlandieriana (Mexican ash) and C. laevigata. Five species of shrubs were identified, including the dominant Celtis pallida. Thirty-five species were present in the ground layer, led by the dominant species C. ciliaris, which accounted for 43% of the total cover.
    Scattered, robust colonies of Paspalum virgatum (talquezal), which contributed only 1.7% of the relative cover, were present on the edge of the river. The last documented collection of this species was from the vicinity of Brownsville, Texas, in 1940. The richness in specie populations is attributed to the saturated soils that result from the periodic release of water from Falcon Dam.
    The vegetation of two mature riparian forest communities in the lower Rio Grande Valley was ascertained. Aerial CIR photography proved useful in discriminating certain tree species, but due to tonal similarities it was difficult to identify each species in the riparian forest canopy. These limitations established that the technique cannot be used to quantify composition in the tree layer. The technique was also not useful in mapping the distribution of shrubs and herbaceous species in the under growth. Based on the findings, a primary benefit appears to be the ability to monitor the senescence and death of trees in the riparian zone in successive years.

About the Authors:
Robert I. Lonard is a professor of biology at the University of Texas-Pan American in Edinburg, Texas. His research interests include the flora and vegetation of southern Texas. Dr. Frank W. Judd is professor and chairman of the Biology Department at the University of Texas-Pan American. His interests range from physiological and population ecology to community ecology. Jim Everitt, David Ewcobar, and RenŽ Davis are range scientist, remote sensing specialist, and airplane pilot/photographer, respectively at the USDA-Agricultural Research Service Center in Weslaco, Texas. They are members of a team conducting research on the development and application of remote sensing technology for natural resource management. Dr. Melba Crawford is a professor in the Center for Space Research at the University of Texas at Austin. Her research interests include the development of algorithms to analyze remotely sensed digital imagery, including image segmentation, statistical image reconstruction, data fusion, and image classification. Dr. Mita Desai is an associate professor in the Department of Electrical Engineering at the University of Texas-San Antonio. Her research expertise includes investigations into wavelet-based techniques and Markov random fields methods for the analysis of image texture.

Back