Sandhill Crane (Antigone canadensis) overwintering resource use and selection in west Texas using GIS and remote sensing



Journal Title

Journal ISSN

Volume Title



I hypothesized that Sandhill Crane (Antigone canadensis, hereafter SACR) resource selection patterns in west Texas may change across late fall, winter, and early spring as a function of changes in the temporal distribution of available saline lakes, playa wetlands, other waterbodies, and the temporal dynamics of grain crops. Cranes (n=27) were trapped, banded, and tagged with platform transmitter terminal (PTT) global positioning system (GPS) devices during 2014-15 and 2015-16 winters. All water surface areas were delineated through Random Forests supervised classification of Landsat 8 imagery using the Google Earth Engine platform for 2014 and 2015. We used a grid with 2.5 x 2.5 km cells to calculate the differences in water surface area and crop types in 2014 and 2015 in each cell and used crane GPS locations to evaluate whether crane resource selection was determined by changes in water surface area and crop type. In addition, the influence of crop productivity on SACR resource use was assessed using time series of vegetation indices in a subset of the study area. Total water surface areas differed between years and in 2014 between spring/summer (34,458.70 ha), fall (45,931.11 ha), and winter (27,211.82 ha), and in 2015 between spring/summer (113,804.54 ha), fall (78,233.50 ha), and winter (55,781.60 ha). Overall differences in water surface availability affected SACR resource use and selection. As total winter surface water acreage increased from 2014 (27,211.82 ha) to 2015 (55,781.60 ha), so, too, did the average home range sizes according to 90% kernel density estimates for 2014 (1,495,035.53 ha) to 2015 (3,259,732.21 ha). As expected, SACRs showed selection for larger water surface areas in specific saline and playa wetlands for roosting and resting rather than choosing random water surface areas. However, cranes also showed increasing selection of playa wetlands towards the end of winter periods, and greater selectivity for playa wetlands in the 2014-15 winter. Distances of crane locations to playa wetlands and saline lakes also differed based upon the times of the day. Cranes increased selection for sorghum (Sorghum bicolor) in both winters from the first to final month in the 2014 winter and the 2015 winter. A landscape-scale study of cranes near Baileyboro Lake showed these select post-harvested winter wheat (Triticum aestivum), corn (Zea mays), and sorghum crops that are close to the lake, but that they are able to spend more energy in travel efforts if plenty of post-harvested sorghum and corn are available at greater distances. Data presented in this work shows how resource use and selection is influenced by changes in nutritional requirements as the birds prepare for spring migration and the oncoming breeding season and interannual differences in the availability of certain crop types. Future studies in overwintering SACR behavior should further investigate the confounding effects of changes in saline lakes and playa wetlands water surface area and crop productivity on SACR distribution and resource selection during winter using complete time series of Landsat data.



Sandhill crane, Resource selection, Overwintering, West Texas, GIS, Remote sensing