nena masthead
SENA Home Staff & Editors For Readers For Authors

Satellite-derived Temperature Data for Monitoring Water Status in a Floodplain Forest of the Upper Sabine River, Texas
Mary Grace T. Lemon, Scott T. Allen, Brandon L. Edwards, Sammy L. King, and Richard F. Keim

Southeastern Naturalist, Volume 16, Special Issue 9 (2016): 90–102

Full-text pdf (Accessible only to subscribers.To subscribe click here.)


Access Journal Content

Open access browsing of table of contents and abstract pages. Full text pdfs available for download for subscribers.

Issue-in-Progress: Vol. 22 (2) ... early view

Current Issue: Vol. 21 (4)
SENA 21(3)

All Regular Issues


Special Issues






JSTOR logoClarivate logoWeb of science logoBioOne logo EbscoHOST logoProQuest logo

Southeastern Naturalist M.G.T. Lemon, S.T. Allen, B.L. Edwards, S.L. King, and R.F. Keim 2016 90 Vol. 15, Special Issue 9 Satellite-derived Temperature Data for Monitoring Water Status in a Floodplain Forest of the Upper Sabine River, Texas Mary Grace T. Lemon1,*, Scott T. Allen1, Brandon L. Edwards1, Sammy L. King2, and Richard F. Keim1 Abstract - Decreased water availability due to hydrologic modifications, groundwater withdrawal, and climate change threaten bottomland hardwood (BLH) forest communities. We used satellite-derived (MODIS) land-surface temperature (LST) data to investigate spatial heterogeneity of canopy temperature (an indicator of plant-water status) in a floodplain forest of the upper Sabine River for 2008–2014. High LST pixels were generally further from the river and at higher topographic locations, indicating lower water-availability. Increasing rainfall-derived soil moisture corresponded with decreased heterogeneity of LST between pixels but there was weaker association between Sabine River stage and heterogeneity. Stronger dependence of LST convergence on rainfall rather than river flow suggests that some regions are less hydrologically connected to the river, and vegetation may rely on local precipitation and other contributions to the riparian aquifer to replenish soil moisture. Observed LST variations associated with hydrology encourage further investigation of the utility of this approach for monitoring forest stress, especially with considerations of climate change and continued river management. Introduction Bottomland hardwood (BLH) forests are high-value, high-productivity wetland ecosystems, (Tockner and Stanford 2002), but many are under threat of ecosystem conversion due to current water-management strategies and changing climatic conditions (Graf 2001, Olson and Dinerstein 1998). Many of these forests were converted to agriculture in the latter half of the 20th century (Turner et al. 1981), and those that remain have become degraded due to hydrologic modifications to large rivers, among other causes (King et al. 2012). Hydrologic modifications that reduce water availability to floodplain vegetation during the growing season (e.g., dams, levees, and channelization; Hupp et al. 2009, Wharton et al. 1982) may be particularly detrimental because wetland trees tend to be poorly adapted to dry conditions. Wetland trees tend to have shallow roots (Kozlowski 1997), especially those developed during flooding, which may not be ideal for accessing soil water (Burke and Chambers 2003). In general, trees that are well adapted to one stress are vulnerable to other stressors (Niinemets 2010). Predicted increases in drought frequency (Georgakakos and Zhang 2011, Hay et al. 2011, Orlowsky and Seneviratne 1School of Renewable Natural Resources, Louisiana State University Agricultural Center, Baton Rouge, LA, 70803. 2US Geological Survey, Louisiana Cooperative Fish and Wildlife Unit, Louisiana State University Agricultural Center, Baton Rouge, LA, 70803. *Corresponding author - Manuscript Editor: Jerry Cook Proceedings of the 6th Big Thicket Science Conference: Watersheds and Waterflow 2016 Southeastern Naturalist 15(Special Issue 9):90–102 Southeastern Naturalist 91 M.G.T. Lemon, S.T. Allen, B.L. Edwards, S.L. King, and R.F. Keim 2016 Vol. 15, Special Issue 9 2012) and anthropogenic water use (Brown et al. 2014) will further decrease water levels in rivers, and thus, water availability for BLH forests, potentially converting hydrological conditions to those favored by species associated with drier habitats (Gee et al. 2014, Kroes and Brinson 2004, Shankman et al. 2012) Water available to floodplain vegetation is controlled by river discharge, precipitation, and shallow-groundwater flow, but relative contributions to water accessible by vegetation are still poorly understood (e.g., Kaplan and Muñoz-Carpena 2011, Krause et al. 2007). The physiological effects of flooding have been well documented in BLH forests (reviewed by Kozlowski 1997), but few studies have focused on the importance of water availability and deficit during low-flow periods, when floodplains become hydrologically disconnected from their associated rivers. Recent research suggests that even in floodplains typically defined by annual water excess, periods of limited availability can result in reduced water-use by trees (Allen et al. 2014, Bosch et al. 2013), an indicator of stress and reduced productivity (Lu et al. 2004). Identification of areas within the floodplain where water availability to vegetation is limited can provide valuable information about floodplain hydrology that can lead to improved understanding of BLH ecology and management. Decreased water availability reduces tree hydraulic-conductance (Meinzer 2001), limiting transpiration and causing an increase in canopy temperature (Jones 1998, Monteith 1965). Many researchers have employed a variety of methods for efficiently determining plant stress (Akhtar et al. 2013, Jackson et al. 1981) in crops and forests (e.g., Luvall and Holbo 1989, Nemani and Running 1989, Sun and Mahrt 1994). In particular, remotely sensed canopy temperature is useful as an index of water stress (Berni et al. 2009, Moran and Jackson 1991, Nagler et al. 2003). Areas with comparatively high land-surface temperature (LST) are assumed to have less available water. In addition, diel fluctuation in LST is greater in moisture-limited areas (e.g., Carlson 1986, Carlson et al. 1981, Gauthier and Tabbagh 1994), and accordingly used to identify areas of moisture stress (Tramutoli et al. 2001). In this study, we used satellite-derived (MODIS) LST data to investigate spatial heterogeneity of water stress in a floodplain forest of the upper Sabine River. The first objective was to determine whether the spatial patterns of diel LST fluctuations were comparable to spatial patterns identified by daytime LST anomaly. The second objective was to identify inter-annual spatial patterns in LST during the growing season, and determine whether years defined by water scarcity result in greater spatiotemporal variability of LST. The location of the study site in eastern Texas, is characterized by periods of extended drought, including during the study period. We hypothesized that the LST would be relatively spatially homogeneous (LST anomaly is convergent between groups) during relatively wet years, but that it would exhibit more spatial variability in drought years (LST anomaly is divergent between groups) as water sources become localized within the flo odplain. Field Site Description The study area includes 2 sections of BLH forests that lie adjacent to the Sabine River: Old Sabine Bottoms Wildlife Management Area (OSB WMA) and Southeastern Naturalist M.G.T. Lemon, S.T. Allen, B.L. Edwards, S.L. King, and R.F. Keim 2016 92 Vol. 15, Special Issue 9 the neighboring Little Sandy National Wildlife Refuge (LS NWR) (Fig. 1a, b). Dominant tree species include Planera aquatica J.F. Gmel. (Water Elm), Fraxinus pennsylvanica Marsh. (Green Ash), Celtis laevigata Willd. (Sugarberry), Ulmus crassifolia Nutt. (Cedar Elm), Quercus nigra L. (Water Oak), Quercus phellos L. (Willow Oak), Quercus lyrata Walter (Overcup Oak), and several Carya (hickory) species (Alden 1998). Soils are dominated by vertisols (shrink–swell clays), with bands of coarser soils that outline former channels of the Sabine. There have been substantial hydrological modifications in the Sabine watershed throughout the 20th century, mainly reservoir construction, leading to reduced flood-peak stages and Figure 1. (a) Location and grouping of study pixels within the study area. Darker pixels are cool in temperature while brighter pixels are warm. (b) Approximate location of the study area within East Texas and along the Sabine Riverwithin East Texas and along the Sabine River. (c) Topographic map of the study area derived from USGS National Elevation Dataset with shading indicating 1-m changes in elevation within in the floodplain. Lowest elevations are shaded black. Stars indicate point locations of observed mortality during the 2011 drought. Southeastern Naturalist 93 M.G.T. Lemon, S.T. Allen, B.L. Edwards, S.L. King, and R.F. Keim 2016 Vol. 15, Special Issue 9 lower variability in river flows. Two large reservoirs, Lake Fork, on a large tributary of the Sabine and Lake Tawakoni on the main stem Sabine, are directly upstream from the study site. Sabine River modifications and increased frequency of drought conditions over the past decade have presumably caused water stress in some floodplain trees, as demonstrated by large tree-mortality events (Christopher Farrell, Texas Parks and Wildlife Department, OSBWMA, Lindale, TX, pers. comm.) concurrent with the severe drought conditions in 2011 (Figs. 1 and 2). In addition, regeneration of hydric-adapted tree species has been minimal in these locations (March et al. 2012). Methods Data acquisition and processing We used MODIS (1 km ×1 km resolution) MOD11A2 8-d average LST data from 2008–2014 (USGS 2014). Although a finer spatial scale is preferable, other thermal sensors do not provide the temporal resolution required for this analysis. The development and processing of this MODIS product includes corrections for emissivity Figure 2. Mean land-surface temperature (LST) calculated across all pixels from 1 May 2008 to 1 November 2014. Southeastern Naturalist M.G.T. Lemon, S.T. Allen, B.L. Edwards, S.L. King, and R.F. Keim 2016 94 Vol. 15, Special Issue 9 variation and removal of periods with clouds (Wan et al. 2002). To minimize thermal contamination, we removed pixels within the study-area boundary if they contained more than 20% permanent surface-water, any upland area, or recent clear-cut forest. This yielded a total of 30 pixels (Fig. 1). We included only imagery captured from 1 May through 1 November to restrict the analysis to the growing season. We acquired weather data from the National Climate Data Center (GHCND:USC00414020), which were collected ~2 km away from the study site in Hawkins, TX. River stage was measured at USGS gage 08019200 on the Sabine River near Mineola, TX, during the study period. Analyses For each pixel and each 8-d LST composite, we subtracted day LST from night LST to calculate the magnitude of mean diel fluctuation for each pixel across the study period. We performed a principal component analysis (PCA) on these data (rows: pixels; columns: 8-d mean diel fluctuation). We ordered pixels by mean diel fluctuation according to first principal component scores, which we compared against mean daytime LST anomaly (mean deviation of each pixel from the median LST of all pixels for each 8-d period) to test the hypothesis that daytime LST anomaly arises because of diel fluctuation, as would be expected in waterlimited canopies. We compared the spatial mean LST anomaly for the warmest 10, middle 10, and coolest 10 pixels (based on first principal component scores) through time and against river stage and recent rainfall. We estimated the effect of rainfall on soil moisture using the antecedent precipitation index (API) as APIi = R + k × APIi-1, where R is rainfall occurring on day i and k is 0.9 (Linsley et al. 1949). All analyses were performed in MATLAB (Mathworks, Inc., Natick, MA). Results Overbank flooding only occurred 5 times throughout the 7-y study period (Fig. 3a). Pulses were short in duration except in 2008, fall of 2009, and during late winter of 2010. In some growing seasons, there were only a few, small, peak flows well below bankfull, and in 2011, the river remained at low base-flow for the entire growing season. Rainfall-controlled soil moisture (API) was highest during the growing season of 2009 and lowest during 2011 (Fig. 3b). Peak soil moisture was typically higher during the late growing season; however, there were short pulses of elevated soil moisture near the beginning of the growing seasons in 2008, 2010, and 2014. The combined results from river flow and API distinctly showed greater water scarcity in 2011 than other years. The wettest period occurred from the beginning of the 2009 growing season through the start of the 2010 growing season. The first eigenvector in the PCA analysis accounted for 90.5% of the variance of diel LST variation, and was thus an appropriate variable for differentiating pixels. Rank by diel variation was also correlated with mean LST anomaly (Fig. 4). Pixels ranked by the first component of diel variation showed consistent behavior throughout the study period, and the difference in LST anomaly between the Southeastern Naturalist 95 M.G.T. Lemon, S.T. Allen, B.L. Edwards, S.L. King, and R.F. Keim 2016 Vol. 15, Special Issue 9 Figure 3. (a) 8-day mean Sabine River hydrograph at USGS gage 08018500 near Mineola, TX, from 1 November 2007 to 1 November 2014. Asterisks indicate overbank events at the study location (>5.18 m at gage 08018500) and (b) 8-day mean antecedent precipitation index (API) for soil moisture from 1 November 2007 to 1 November 2014. Figure 4. Land-surface temperature (LST) anomaly for each pixel sorted by the weight of the first eigenvector of the diel variation from 1 May 2008 to 1 November 2014. Color scale has been set from -2 to 2 ºC in order to aid visualization but ranged from approximately -4 to 4 ºC. Southeastern Naturalist M.G.T. Lemon, S.T. Allen, B.L. Edwards, S.L. King, and R.F. Keim 2016 96 Vol. 15, Special Issue 9 warmest 10 pixels and the remaining pixels was greater than the differences among the 2 other pixel groups (Fig. 5). Similarly, LST anomaly of the warmest pixels was more variable than cooler pixels. The largest divergence (highest variability) in LST anomaly was during the droughty 2011 growing season, with a mean difference between the warmest and coolest groups reaching 1.60 °C, the largest of all years (Fig. 5). Among-year trends in LST anomaly (Fig. 5) match among-year trends in LST (Fig. 2). Both LST anomaly for the warmest pixels and raw LST values reflect the severity of drought in 2011 compared to all other years; pixel mean LST reached 36.5 ± 1.7 °C during the growing season (Fig. 2). Increasing rainfall-controlled soil moisture corresponded with convergence of LST anomaly between the warmest and coolest pixels (Fig. 6). For river stages at or below baseflow (~0.5 m), LST heterogeneity was high; however, above baseflow there was an increasing trend toward homogeneity in terms of LST (Fig. 7). For stages near or above bankfull (~5.2 m), there was convergence of LST; however, the sample size of surface-flooding events during the growing season across the study period was extremely small (Fig. 7). The spatial distribution of the first eigenvector weights across the floodplain generally corresponded to relative disconnection from the river in terms of topographic position and distance (Fig. 1a). Discussion The results indicate that heterogeneity of floodplain temperature generally increased in drought years, which supports our hypotheses. Convergence to a more homogeneous condition was related to rainfall and stage of the Sabine River; Figure 5. Land-surface temperature (LST) anomaly for 3 groups of pixels from 1 May 2008 to 1 November 2014. The pixels were sorted in ascending order of the weight of the first eigenvector of the diel variation with the low LST making up the first 10 pixels, the median LST making up the middle 10 pixels, and the high LST group making up the last 10 pixels. Numbers below data along x axis indicate mean difference of actual daytime LST between low and high groups (ºC). Southeastern Naturalist 97 M.G.T. Lemon, S.T. Allen, B.L. Edwards, S.L. King, and R.F. Keim 2016 Vol. 15, Special Issue 9 however, pixel temperature was loosely organized by floodplain topography—the warmest pixels were generally at higher topographic positions, which were loosely related to distance from the river. The relative lack of overbank flooding events at the study site during the growing season could explain why the relationship between growing season LST and river stage is not stronger, but the river still seemingly affected the floodplain forest. This dichotomy remains unexplained for our study floodplain, but is likely related to topographic variability controlling water availability from multiple sources that were not well quantified by the simple hydrologic analyses we conducted for this work. Although our analysis relied on the correlation between temperature and water stress, leaf-level processes link carbon- and water-exchange; thus, increasing temperature anomalies are likely coupled to reduced photosynthetic capability. Further research is needed to match canopy temperature to specific stress thresholds. Given that there was substantial tree-mortality in 2011 (March et al. 2012), the high LST observed at that time was certainly indicative of high-stress events. However, mortality is usually associated with multiple-year droughts (e.g., Nepstad et al. 2007), so maximum temperatures alone are not necessarily useful indicators. Figure 6. Spatial variation in 8-d mean land-surface temperature (LST) anomaly (ºC) as a function of 8-d mean, rainfall-derived soil moisture (API). Southeastern Naturalist M.G.T. Lemon, S.T. Allen, B.L. Edwards, S.L. King, and R.F. Keim 2016 98 Vol. 15, Special Issue 9 Dams and resulting hydrologic changes generally reduce peak-flow frequency and magnitude while increasing low-flow frequency and magnitude (Brandt 2000, Graf 2001, Williams and Wolman 1984). Similarly, Magilligan and Nislow (2005) found that dams can increase low-magnitude high-frequency hydrograph variations and decrease the average length of high pulses downstream. This type of pre-dam– post-dam analysis has not been completed for the upper Sabine River. Although Phillips (2001) found no evidence of a reduction in peak annual discharge of the Sabine River at the lowermost gaging station (USGS no. 08030500) near Ruliff, TX, despite multiple upstream dams, local effects on floodplain forests have been documented in the lower Sabine River (Alldredge and Moore 2012) and are likely also present on our study site due to its proximity to 2 large upstream reservoirs. This conclusion is supported by the low frequency of over-bank flooding events that were not sufficient to routinely cause floodplain-wide alleviation of drought stress (Fig 3a). Alluvial aquifers are replenished during surface-flooding events in floodplains (Winter 1999), and changes in the frequency, timing, and magnitude of these events can reduce water availability to floodplain ecosystems. Additionally, short-duration, Figure 7. Spatial variation in 8-d mean LST anomaly (ºC) as a function of Sabine River stage. Asterisks indicate time intervals where stages at or above 5.18 m occurred. Southeastern Naturalist 99 M.G.T. Lemon, S.T. Allen, B.L. Edwards, S.L. King, and R.F. Keim 2016 Vol. 15, Special Issue 9 high-flow pulses may not cause long-distance propagation of the floodplain watertable as compared to long-duration, high-flow pulses, even when considering sub-bankfull flows (Jung et al. 2004). Therefore, the modifications upstream of our study site may be affecting subsurface hydrology and water availability to vegetation, particularly in floodplain areas with low subsurface hydraulic connectivity to the river (e.g., distant and microtopographically higher in elevation). Demonstrated dependence of LST on rainfall instead suggests that vegetation relies on local precipitation and other contributions to the riparian aquifer (Kroes and Brinson 2004) to replenish soil moisture during times of low flows resulting from increased regional demand for water during drought periods. Increased drought frequency is predicted for eastern Texas over the next century (Orlowsky and Seneviratne 2012); thus, further mortality events may be observed during severe drought years in areas of low connectivity to the river. In this study, decreasing connectivity in terms of elevation and distance from the river was generally associated with decreased water availability as interpreted by high LST anomaly; however, in other floodplains, spatially diverse geomorphic features and complex groundwater connections may result in more complex canopy- temperature patterns. The coarse spatial resolution of the study limits inferences about smaller floodplain features (i.e., sloughs, back-swamp ponding). Higher spatial-resolution (less than 1 km) thermal imagery is required to investigate canopy-temperature patterns related to smaller-scale connectivity, and therefore, could not be addressed by our study. Remote sensing allows inference with greater spatial and temporal frequency than field-based methods. In addition to field measurements, this tool enables identification of specific mechanisms and development of detailed ecological conclusions without limiting spatial extent. To increase the value of both techniques for purposes of research and monitoring, coupled ground-based and remote-sensing data-collection and analysis is needed. Conclusions LST (a surrogate for plant water-status) in the floodplain of the upper Sabine River is most spatially variable during times of drought. Dependence of LST on rainfall and spatial patterns suggests that vegetation in some parts of the floodplain is less hydrologically connected to the river than in others. Areas less connected to the river and its influence had greater temperature variations, indicating greater vulnerability to climatological drought. Remotely sensed LST shows promise as a tool for better understanding the spatial distribution of water stress within floodplains, which is expected to increase in relevance during times of increased water scarcity due to water-management projects and climate change. Acknowledgments We thank Christopher Farrell from TPWD for information about the site, particularly in determining bankfull stages. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the US Government. Southeastern Naturalist M.G.T. Lemon, S.T. Allen, B.L. Edwards, S.L. King, and R.F. Keim 2016 100 Vol. 15, Special Issue 9 Literature Cited Akhtar, F., B. Tischbein, and U.K. Awan. 2013. Optimizing deficit-irrigation scheduling under shallow-groundwater conditions in lower reaches of Amu Darya River Basin. Water Resource Management 27:3165–3178. Alden, S.M. 1998. Woody-plant species diversity of the Old Sabine Bottom Wildlife Management Area, Smith Country, Texas. M.Sc. Thesis. University of Texas at Tyler, Tyler, TX. Alldredge, B., and G. Moore. 2012. Assessment of riparian-vegetation sensitivity to river hydrology downstream of a major Texas dam. River Research and Applications 30:230–244. Allen, S.T., K.W. Kraus, J.W. Cochran, B. Edwards, R.F. Kiem, and S.L. King. 2014. Hydrological controls over water-use in a forested-floodplain wetland. Abstracts, American Geophysical Union Fall meeting. San Francisco, CA. 15–19 December 2014. Berni, J.A.J., P.J. Zarco-Tejada, G. Sepulcre-Canto, E. Fereres, and F. Villalobos. 2009. Mapping canopy-conductance and CWSI in olive orchards using high-resolution thermal remote-sensing imagery. Remote Sensing of Environment 113:2380–2388. Bosch, D.D., L.K. Marshall, and R. Teskey. 2014. Forest transpiration from sap-flux density measurements in a Southeastern Coastal Plain riparian buffer system. Agricultural and Forest Meteorology 187:72–82. Brandt, S.A. 2000. Classification of geomorphological effects downstream of dams. Catena 40:375–401. Brown, J.H., J.R. Burger, W.R. Burnside, M. Chang, A.D. Davidson, T.S. Fristoe, M.J. Hamilton, S.T. Hammond, A. Kodric-Brown, N. Mercado-Silva, J.C. Nekola, and J.G. Okie. 2014. Macroecology meets macroeconomics: Resource scarcity and global sustainability. Ecological Engineering 65:24–32. Burke, M.K., and J.L. Chambers. 2003. Root dynamics in bottomland-hardwood forests of the Southeastern United States Coastal Plain. Plant and Soil 250:141–153. Carlson, T.N. 1986. Regional-scale estimates of surface-moisture availability and thermal inertia using remote thermal measurements. Remote Sensing Review 1:197–247. Carlson, T.N., J.K. Dodd, J.F. Benjamin, and J.N. Cooper. 1981. Satellite estimation of the surface energy-balance, moisture availability, and thermal inertia. Journal of Applied Meteorology 20:67–87. Gauthier, F., and A. Tabbagh. 1994. The use of airborne thermal remote-sensing for soil mapping: A case study in the Limousin region (France). International Journal of Remote Sensing 15:1981–1989. Gee, H.K.W., S.L. King, and R.F. Keim. 2014. Tree growth and recruitment in a leveed floodplain forest in the Mississippi River Alluvial Valley, USA. Forest Ecology and Management 334:85–95. Georgakakos, A., and F. Zhang. 2011. Climate-change scenario assessment for ACF, OOA, SO, ACT, TN, and OSSS Basins in Georgia. Georgia Water Resources Institute (GWRI) Technical Report sponsored by NOAA, USGS, and the Georgia EPD, Georgia Institute of Technology, Atlanta, GA. 229 pp. Graf, W.L., 2001. Damage control: Restoring the physical integrity of America’s rivers. Annals of the Association of American Geography 91:1–27. Hay, L.E., S.L. Markstrom, and C. Ward-Garrison. 2011. Watershed-scale response to climate change through the 21st century for selected basins across the United States. Earth Interactions 15:1–37. Southeastern Naturalist 101 M.G.T. Lemon, S.T. Allen, B.L. Edwards, S.L. King, and R.F. Keim 2016 Vol. 15, Special Issue 9 Hupp, C.R., A.R. Pierce, and G.B. Noe. 2009. Floodplain geomorphic process and environmental impacts of human alteration along coastal-plain rivers, USA. Wetlands 29:413–429. Jackson, T.J. 1986. Soil-water modeling and remote sensing. Transactions on Geoscience Remote Sensing GE-24:37–46. Jones, H.G. 1998. Stomatal control of photosynthesis and transpiration. Journal of Experimental Botany 49(Special Issue):387–398. Jung, M., T.P. Burt, and P.D. Bates. 2004. Toward a conceptual model of floodplain watertable response. Water Resources Research 40(12):1–13. Kaplan, D., and R. Muñoz-Carpena. 2011. Complementary effects of surface water and groundwater on soil-moisture dynamics in a degraded coastal floodplain-forest. Journal of Hydrology 398:221–334. King, S.L., L.L. Battaglia, C.R. Hupp, R.F. Keim, and B.G. Lockaby. 2012. Floodplain Wetlands of the Southeastern Coastal Plain. Pp. 253–266, In D. Batzer, and A.H. Baldwin (Eds.). Wetland Habitats of North America: Ecology and Conservation Concerns. University of California Press, Berkeley, CA. Kozlowski, T.T. 1997. Responses of woody plants to flooding and salinity. Tree Physiology 17:490–490. Kroes, D.E., and M.M. Brinson. 2004. Occurrence of riverine wetlands on floodplains along a climatic gradient. Wetlands 24:167–177. Krause, S., A. Bronstert, and E. Zehe. 2007. Groundwater–surface-water interactions in a North German lowland floodplain: Implications for the river discharge dynamics and riparian water balance. Journal of Hydrology 347:404–417. Linsley, R.K., M.A. Kohler, and J.L.H. Paulhus. 1949. Applied Hydrology. McGraw-Hill Book Company, Inc., New York, NY. 689 pp. Lu, P., L. Urban, and Z. Ping. 2004. Granier’s thermal-dissipation probe (TDP) method for measuring sap-flow in trees: Theory and practice. Acta Botanica Sininca 46:631–646. Luvall, J.C., and H.R. Holbo. 1989. Measurements of short-term thermal responses of coniferous-forest canopies using thermal-scanner data. Remote Sensing of the Environment 27:1–10. Magilligan, F.J., and K.H. Nislow. 2005. Changes in hydrologic regime by dams. Geomorphology 71:61–78. March, R., G. Moore, and C. Edgar. 2012. A remote-sensing method for mapping the extent of tree mortality in Texas following the 2011 exceptional drought using publicly available datasets. Abstracts, American Geophysical Union, 3–7 December 2012, San Francisco, CA. Available online at meeting-at-a-glance/. Accessed 5 April 2015. Meinzer, F.C., M.J. Clearwater, and G. Goldstein. 2001. Water transport in trees: Current perspectives, new insights, and some controversies. Environmental and Experimental Botany 45:239–262. Monteith, J.L. 1965. Evaporation and environment. Symposia of the Society for Experimental Biology 19:205–224. Moran, M., and R. Jackson. 1991. Assessing the spatial distribution of evapotranspiration using remotely sensed inputs. Journal of Environmental Quality 20:725–737. Nagler, P.L., E.P. Glenn, and T.L. Thompson. 2003. Comparison of transpiration rates among Saltcedar, Cottonwood, and willow trees by sap-flow and canopy-temperature methods. Agricultural and Forest Meteorology 116:73–89. Nemani, R.R., and S.W. Running. 1989. Estimation of regional surface-resistance to evapotranspiration from NDVI and thermal-IR AVHRR data. Journal of Applied Meteorology 28:276–284. Southeastern Naturalist M.G.T. Lemon, S.T. Allen, B.L. Edwards, S.L. King, and R.F. Keim 2016 102 Vol. 15, Special Issue 9 Nepstad, D.C., I.M. Tohver, D. Ray, P. Moutinho, and G. Cardinot. 2007. Mortality of large trees and lianas following experimental drought in an Amazon forest. Ecology 88:2259–2269. Niinemets, Ü. 2010. Responses of forest trees to single and multiple environmental stresses from seedlings to mature plants: Past stress history, stress interactions, tolerance, and acclimation. Forest Ecology and Management 260:1623–1639. Olson, D.M., and E. Dinerstein. 1998. The global 200: A representation approach to conserving the earth’s most biologically valuable ecoregions. Conservation Biology 12:502–515. Orlowsky, B., and S.I. Seneviratne. 2013. Elusive drought: Uncertainty in observed trends and short- and long-term CMIP5 projections. Hydrology and Earth System Science 17:1765–1781. Phillips, J.D. 2001. Sedimentation in bottomland hardwoods downstream of an East Texas dam. Environmental Geology 40:860–868. Shankman, D., C.W. Lafon, and B.D. Keim. 2012. Western range-boundaries of floodplain trees in the southeastern United States. Geographical Review 102:35–52. Sun, J., and L. Mahrt. 1994. Spatial distribution of surface fluxes estimated from remotely sensed variables. Journal of Applied Meteorology 33:1341–1353. Tockner, K., and J.A. Stanford. 2002 Riverine flood plains: Present state and future trends. Environmental Conservation 29:308–330. Tramutoli, V., P. Claps, M. Marella, N. Pergola, C. Pietrapertosa, and C. Sileo. 2001. Hydrological implications of remotely sensed thermal inertia. Pp. 7–11, In M. Owe, K. Brubaker, J. Ritchie, and A. Rango (Eds.). Remote Sensing and Hydrology 2000. International Association of Hydrological Sciences, Oxfordshire, UK. 610 pp. Turner, R.E., S.W. Forsythe, and N.J. Craig. 1981. Bottomland hardwood-forest land resources of the southeastern United States. Pp. 13–28, In J.R. Clark and J. Benforado (Eds.). Wetlands of Bottomland Hardwood Forests. Elsevier, New York, NY. 401 pp. US Geographical Survey (USGS). 2014. MODIS MOD11A2 data product: 2008–2014. Available online at Accessed 27 February 2015. Wan, Z., Y. Zhang, Q. Zhang, and Z. Li. 2002. Validation of the land-surface–temperature products retrieved from Terra Moderate Resolution Imaging Spectroradiometer data. Remote Sensing of the Environment 83:163–180. Wharton, C.H., W.M. Kitchens, E.C. Pendleton, and T.W. Sipe. 1982. The ecology of bottomland-hardwood swamps of the southeast: A community profile. Technical Report. Biological Service Program, FWS/OBS-81/37, US Fish and Wildlife Service, Washington, DC. 133 pp. Winter, T.C. 1999. Relation of streams, lakes, and wetlands to groundwater-flow systems. Hydrogeology Journal 7:28–45.