Remote mapping of saltcedar in the Rio Grande system of west Texas.
The Rio Grande is one of the longest river systems in the United States. The river extends 3,040 km from its source in the San Juan Mountains of Colorado to the mouth at the Gulf of Mexico on the United States-Mexico border in extreme south Texas (Gilpin 1949). Approximately two-thirds (2,020 km) of the Rio Grande is the effective border between Texas and Mexico (Davis 2002). Construction of dams and reservoirs along the lower Rio Grande for flood control and for agricultural and municipal uses have resulted in losses of much of the natural vegetation (Lonard et al. 2000).
Today, extensive areas along the Rio Grande system in Texas have been invaded by exotic, invasive plant species which have ultimately displaced much of the original native vegetation and contributed to water shortages in the river. Saltcedar (Tamarix sp.), waterhyacinth (Eichhornia crassipes [Mort.] Solms), and hydrilla (Hydrilla verticillata [L. F.] Royle) are major exotic species that have invaded the Rio Grande system of Texas (Davis 2002).
Eight species of saltcedar (Tamarix sp.) have been introduced into the United States from Europe, Asia, and Africa for ornamentals, wind-breaks, and erosion prevention of streambanks (Baum 1967). At least five species of saltcedar are found in Texas (Hatch et al. 1990). Two deciduous saltcedar species (Tamarix ramosissima Ledeb. and Tamarix chinensis Laur.) are invaders of riparian sites of the southwestern United States (including Texas) and northern Mexico. These two very similar species form dense, low thickets that displace native vegetation, impede water flow, increase sedimentation, use excessive water, and increase soil salinity (Horton & Campbell 1974; Deloach 1990). Saltcedar communities are also much less valuable for wildlife than are the native riparian communities they displace (Kerpez & Smith 1989; Deloach 1990). Research on herbarium specimens and growing plants of Tamarix chinensis and Tamarix ramosissima has shown that it is difficult to distinguish between the two species (Horton 1977). Molecular research on these two species indicates that some populations are genetically indistinguishable and that there is some evidence of hybridization among several species of saltcedar (Gaskin & Schall 2003). Although Tamarix chinensis, Tamarix ramosissima, and possible hybrids occur in west Texas, the saltcedar taxon that causes a nuisance in this area is generally referred to as Tamarix chinensis.
Riparian zones and other wildland areas are often too large and inaccessible to determine their characteristics by ground surveys. Remote sensing techniques offer potentially timely, cost-effective means of obtaining reliable data for these areas (Tueller 1982). The value of remote sensing for distinguishing among plant species and communities is well established (Carter 1982; Driscoll et al. 1997). Aerial photography, airborne electronic imagery (videography and digital), and satellite imagery have been used to remotely detect weedy species over large and inaccessible areas (Gausman et al. 1977; Tueller 1989; Anderson et al. 1993; Everitt et al. 1995; Lass & Callihan 1997; Ramsey et al. 2002).
Over the past several years remote sensing, geographic information system (GIS), and global positioning system (GPS) technologies have been integrated for detecting and mapping the distribution of noxious plant species (Dewey et al. 1991; Anderson et al. 1993; Everitt et al. 1996). Remote observations in georeferenced formats help to assess the extent of infestations, develop management strategies, and evaluate control measures on noxious plant populations.
Several studies have been conducted using remote sensing techniques to distinguish saltcedar. Everitt & Deloach (1990) described the spectral light reflectance characteristics of saltcedar and demonstrated the application of normal color aerial photography for distinguishing infestations in Texas riparian areas. Everitt et al. (1996) used normal color aerial videography integrated with GPS and GIS technologies to detect and map saltcedar infestations on three river systems in the southwestern United States. More recently, airborne multispectral digital imagery has been used to map saltcedar and other riparian vegetation along the middle Rio Grande River in New Mexico (Akasheh et al. 2004). Anderson et al. (2004) used airborne hyperspectral imagery to assess biocontrol of saltcedar in Nevada and reported limited success in differentiating this species from other riparian vegetation.
The objectives of this study were to use aerial photography and videography, GPS, and GIS technologies for detecting and mapping the distribution of saltcedar infestations along the Rio Grande system in west Texas.
MATERIALS AND METHODS
This study was conducted along the Rio Grande system on the Texas-Mexico border in west Texas. Aerial photography, airborne videography, and ground truth were conducted for this study.
Conventional color photography and videography were acquired simultaneously of the Rio Grande from Lajitas near Big Bend National Park to near E1 Paso in west Texas on 10 December 2002. Imagery was obtained at an altitude above ground level of 3,050 m. Kodak Aerochrome conventional color (0.40 to 0.70 [micro]m) type 2448 film was used with a Fairchild type K-37 large format (23 cm by 23 cm) mapping camera. The camera aperture setting was f8 at 1/250 sec. Conventional color video was acquired with a Canon mini-digital video camera (model GL-1) with a zoom lens (4.2 to 84mm) and a super-VHS recorder.
A Cessna (model 404) airplane, equipped with a camera port in the floor, was used to obtain the aerial photography and videography. The cameras were maintained in nadir position during image acquisition. Imagery was acquired between 1130 and 1300 hours Central Standard Time under sunny conditions.
An Omnistar (model 3000L) differential GPS and Horita (model GPT-50) real-time GPS video/digital captioner/interphaser were integrated with the video system. The GPS acquired the latitude-longitude coordinate data of the aircraft location over the scene of interest, while the video interphaser transferred and superimposed the GPS data at the bottom of the video scene. The accuracy of the GPS was approximately [+ or -]20 m from the center coordinates of each video scene. Location coordinates of saltcedar were obtained from each video scene and entered into the computer manually. Before the GPS data were obtained from the video scenes, population levels of saltcedar were assigned to each photographic image. Population levels were assigned to the photographic images because they had better spatial resolution than the video scenes. Population levels of saltcedar were accomplished by breaking down the width of stands that grow in corridors along the Rio Grande. Population levels of saltcedar were assigned to each photograph using the following criteria: > 120 m wide, dense; 60 to 120 m wide, moderate; and < 60 m wide, light. The length of the corridor was not considered since most were greater than 0.75 km long. Each video scene of saltcedar covered a linear distance along the river of approximately 2,100 m, whereas the photographs covered a linear distance of approximately 2,500 m. Population density or percent cover of the population levels was not quantitatively estimated. Personal computer MapInfo-GIS software (MapInfo, Inc. 1998) was used to generate regional and detailed maps along the Rio Grande. MapInfo uses StreetWorks which is a street display mapping product that provides coverage of U.S. streets, highways, city and town boundaries, area landmarks, point locations, and water features. StreetWorks is based on U.S. Census Bureau TIGER (Topologically Integrated Geographic Encoding and Referencing) 1995 data that includes street-level detail to the local level. The TIGER map-based system was constructed using USGS 1:100,000 scale digital line graph maps. These maps were produced to geographically map saltcedar infestations along the Rio Grande using the airborne video survey GPS data.
Ground truth surveys were conducted at sites where aerial photography and videography were obtained. In some instances, ground surveys were done of some sites prior to acquiring aerial imagery. Observational data recorded were plant species and cover. Ground surveys were made only on the U.S. side of the border.
RESULTS AND DISCUSSION
Figures 1a and 1b show a black-and-white photographic print of a normal color photograph and a normal color analog video image, respectively, of a saltcedar infestation on the Rio Grande north of Candelaria in west Texas. The photograph is a portion of a 23 cm photograph (1:10,000 scale), whereas the video image (3.0 m pixel size) was extracted from a slightly larger video scene. The arrows on the two images point to the orange-brown tonal response of a dense stand of saltcedar. Bare soil and sparsely vegetated areas have white to various light gray tones, shrubs have a dark gray or black image response, and water has light green to dark green tones. Although the video image has coarser resolution than the photograph, saltcedar can be easily distinguished. The GPS latitude-longitude coordinates of the area are displayed at the top of the video image. The distinct image response of saltcedar was due to its yellow-orange to orange-brown late fall foliage color prior to leaf drop. Saltcedar could be readily distinguished in all the normal color photography and videography obtained along the Rio Grande. Saltcedar has higher visible reflectance during this phenological stage that facilitates it detection on normal color photography and videography (Everitt & Deloach 1990; Everitt et al. 1996). Figure 2 shows a GIS map of the four-county area of west Texas where the aerial survey was conducted. The Rio Grande forms the left boundary of the map. The GPS latitude-longitude data provided on the aerial videographic imagery from the December 2002 survey of the Rio Grande have been integrated with the GIS to georeference infestations of saltcedar along the river. Areas with stars represent the densest populations of saltcedar, those with dots were moderate populations, and those represented by triangles have light populations. Many of the population symbols are stacked on each other because of the small scale of the map. Results indicate that approximately 460 river-km of saltcedar occurred along the Rio Grande study area. The densest populations occurred in the eastern portion of Hudspeth County and the western part of Presidio County where many of the corridors were > 120 m wide. Ground surveys of the individual sites indicated that saltcedar stands with > 25% cover could be distinguished on the imagery; however, those stands with > 50% cover had a more pronounced image response. Many small saltcedar plants and some isolated larger plants with sparse canopies could not be distinguished on the imagery.
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
The 2002 survey map (Figure 2) of saltcedar distribution was similar to a 1994 survey map of the same general area (Everitt et al. 1996). Total river-km of saltcedar was not computed in the 1994 survey, but a qualitative comparison between the two maps is very similar. However, some of the saltcedar density levels differed between the two surveys. This was partially due to changes in plant populations over the eight-year interval between surveys, but was primarily attributed to using different criteria for assignment of population levels in the 1994 survey and acquisition of imagery at a different altitude. The 1994 imagery was obtained at altitudes ranging from 1,050 to 1,500 m, whereas the 2002 survey was obtained at an altitude of 3,050 m. The higher altitude imagery of the 2002 survey provided a much greater horizontal width of coverage of the Rio Grande floodplain and the detection of more saltcedar populations than in the 1994 survey.
Results from this study have shown that airborne remote sensing, GPS, and GIS technologies are valuable tools for detecting and mapping saltcedar along the Rio Grande system of west Texas. These findings indicate that approximately 460 river-km of the Rio Grande from Lajitas to near E1 Paso are infested by saltcedar. This estimate is probably an underestimation of the actual number of river-km since many small saltcedar plants and some individual larger plants could not be distinguished on the aerial imagery.
The integration of airborne videography with GPS technology can serve as a permanent geographically located image database to monitor future contraction or spread of saltcedar over time. The GIS database can be used to record attribute information for areas of interest. The joint use of these technologies provides important information on the distribution of saltcedar in the Rio Grande system along the Texas-Mexico border. It is anticipated that these technologies can be used for a variety of other natural resource management applications.
The authors thank Fred Gomez and Jim Forward for their assistance in conducting ground surveys, Buck Cavazos for preparing illustrations, and Juan Ramos for his assistance in obtaining the aerial imagery. Thanks are extended to personnel of the Big Bend State Natural Area, Presidio, Texas, and Dr. Harry Miller and family, Candelaria, Texas, for allowing us to conduct ground surveys on their properties. The authors also thank personnel from the Bureau of Reclamation and International Boundary and Water Commission for providing ground truth information.
Akasheh, O., C. Neale & F. Farag. 2004. Spatial mapping of riparian vegetation in the middle Rio Grande River, New Mexico. Proceedings 19th Biennial Workshop on Color Photography, Videography, and Airborne Imaging for Resource Assessment. Amer. Soc. Photogrammetry and Remote Sensing, Bethesda, MD. CD-ROM.
Anderson G. L., J. H. Everitt, A. J. Richardson & D. E. Escobar. 1993. Using satellite data to map false broomweed (Ericameria austrotexana) infestations on south Texas rangelands. Weed Tech., 7:865-871.
Anderson, G. L., J. Knight, R. I. Carruthers & C. J. Deloach. 2004. Monitoring the dynamics of saltcedar biological control in Lovelock, Nevada. Proceeding 19th Biennial Workshop on Color Photography, Videography, and Airborne Imaging for Resource Assessment. Amer. Soc. Photogrammetry and Remote Sensing, Bethesda, MD. CD-ROM.
Baum, B. R. 1967. Introduced and naturalized tamarisks in the United States and Canada. Baileya, 15:19-25.
Carter, V. 1982. Application of remote sensing to wetlands. Pp. 284-300, in C. J. Johannsen and J. L. Sanders (eds.), Remote Sensing in Resource Management. Soil Conservation Soc. Amer., Ankeny, IA, 665 pp.
Davis, R. 2002. Rio Grande no mas. Texas Parks and Wildlife, 60(7):90-103.
Dewey, S. A., K. P. Price & D. Ramsey. 1991. Satellite remote sensing of Dyer's woad (Isatis tinctoria). Weed Tech., 5:479-484.
Deloach, C. J. 1990. Prospects for biological control of saltcedar (Tamarix spp.) in riparian habitats of the southwestern United States. Proc. VII Int. Sympos. Biol. Contr. Weeds. March 6-11, 1988, Rome, Italy. pp. 307-314.
Driscoll, R. S., J. H. Everitt, R. Haas & P. T. Tueller. 1997. Ranges and range management. Pp. 441-474, in W. R. Philipson (ed.), Manual of Photographic Interpretation. Amer. Soc. Photogrammetry and Remote Sensing, Bethesda, MD, 689 pp.
Everitt, J. H. & C. J. Deloach. 1990. Remote sensing of Chinese tamarisk (Tamarix chinensis) and associated vegetation. Weed Sci., 38:273-278.
Everitt, J. H., D. E. Escobar & M. R. Davis. 1995. Using remote sensing for detecting and mapping noxious plants. Weed Abst., 44:639-649.
Everitt, J. H., D. E. Escobar, M. A. Alaniz, M. R. Davis & J. V. Richerson. 1996. Using spatial information technologies to map Chinese tamarisk. Weed Sci., 44:194-201.
Gaskin, J. F. & B. A. Schaal. 2003. Molecular phylogenetic investigation of U. S. invasive Tamarix. Systematic Botany, 28(1):86-95.
Gausman, H. W., R. M. Menges, D. E. Escobar, J. H. Everitt & R. L. Bowen. 1977. Pubescence affects spectra and imagery of silverleaf sunflower (Helianthus argophyllus). Weed Sci., 25:437-440
Gilpin, L. 1949. The Rio Grande: River of destiny. Duell, Sloan, and Pearce, New York, 241 pp.
Hatch, S. L., K. N. Gandhi & L. E. Brown. 1990. Checklist of the vascular plants of Texas. MP-1655. Texas Agricultural Experiment Station, Texas A & M Univ., College Station, Texas, 158 pp.
Horton, J. S. 1977. The development and perpetuation of the permanent tamarisk type in the phreatophyte zone of the southwest. USDA For. Serv. Paper RM-137, 3 pp.
Horton, J. S. & C. J. Campbell. 1974. Management of phreatophyte and riparian vegetation for maximum multiple use values. USDA For. Serv. Paper RM-117, 23 pp.
Kerpez, T. A. & N. S. Smith. 1989. Saltcedar control for wildlife habitat improvements in the southwestern United States. USDI, U. S. Fish and Wildl. Serv. Resour. Publ. 169, 16 pp.
Lass, L. W. & R. H. Callihan. 1997. The effect of phenological stage on detectability of yellow hawkweed (Heiracium pratense) and oxeye daisy (Chrysanthemum leucanthemum) with remote multispectral digital imagery. Weed Tech., 11:248-256.
Lonard, R. I., F. W. Judd, J. H. Everitt, D. E. Escobar, M. R. Davis, M. M. Crawford & M. D. Desai. 2000. Evaluation of color-infrared photography for distinguishing annual changes in riparian forest vegetation of the lower Rio Grande in Texas. Forest Ecology and Management, 128:75-81.
MapInfo, Inc. 1998. MapInfo Professional V5.0. MapInfo Inc. Troy, New York.
Ramsey, E. W., G. A. Nelson, S. K. Sapkota, E. B. Seeger & K. D. Martella. 2002. Mapping Chinese tallow with color-infrared photography. Photogramm. Eng. and Remote Sens., 68:251-255.
Tueller, P. T. 1982. Remote sensing for range management. Pp. 125-140, in C. J. Johannsen and J. L. Sanders (eds.), Remote Sensing for Resource Management. Soil Conserv. Soc. Amer., Ankeny, IA, 665 pp.
Tueller, P. T. 1989. Remote sensing technology for rangeland management applications. J. Range Manage., 42:442-453.
JHE at: firstname.lastname@example.org
J. H. Everitt, C. Yang, M. A. Alaniz and M. R. Davis
USDA-ARS, Integrated Farming and Natural Resources Research
2413 E. Hwy 83, Weslaco, Texas 78596
|Printer friendly Cite/link Email Feedback|
|Author:||Everitt, J.H.; Yang, C.; Alaniz, M.A.; Davis, M.R.|
|Publication:||The Texas Journal of Science|
|Date:||Feb 1, 2006|
|Previous Article:||A 10 million year old ash deposit in the Ogallala Formation of the Texas Panhandle.|
|Next Article:||The vascular flora of Howard County, Texas.|