Fine-Scale Selection of Habitat by the Lesser Prairie-Chicken

Proper management of grasslands and shrublands requires an understanding of the factors that influence the persistence of organisms. We compare differences in vegetation between sites occupied by the lesser prairie-chicken (Tympanuchus pallidicinctus) and random sites to investigate composition of grasses and forbs and the importance of cover. We observed that birds selected habitat, at least in part, based on composition of grasses and forbs. There was generally a larger variance in diversity of plants for random sites compared to sites associated with presence of lesser prairie-chicken. The role of vegetative cover in selection of habitat is important for avoidance of predators, but use of cover also is a means of thermoregulation. Risksensitive behavior is a trade-off between avoiding predation and suitable microclimate. We report evidence that the lesser prairie-chicken consistently seeks to limit the risk of predation and selects locales with a favorable microclimate; birds select sites more or less exposed depending on apparent temperature. We infer that selection of habitat by the lesser prairie-chicken is the result of composition of species of plants, avoidance of predators, and thermoregulation, with the lekking mating system of this bird also playing a role. This declining species might face increasing threats as some practices of land management alter structure of vegetation and reduce shrub cover. RESUMEN—El manejo adecuado de los pastizales y matorrales requiere una comprensión de los factores que influyen en la persistencia de los organismos. Se comparan las diferencias de vegetación entre sitios ocupados por el pollo de la pradera menor (Tympanuchus pallidicinctus) y sitios al azar para investigar la composición de pastos y hierbas y la importancia de la cubierta. Observamos que las aves seleccionaron el hábitat, al menos en parte, basándose en la composición de pastos y hierbas. En general hubo una variación mayor en la diversidad de la vegetación de los sitios al azar en comparación con los sitios asociados con la presencia del pollo de la pradera menor. El papel de la cubierta vegetal en la selección de hábitat es importante para evitar a los depredadores, pero el uso de la cubierta es también un medio de termorregulación. El comportamiento sensible al riesgo es una concesión mutua entre evitar la depredación y conseguir un microclima adecuado. Se presenta evidencia de que el pollo de la pradera menor constantemente trata de limitar el riesgo de depredación y de seleccionar lugares con un microclima favorable; las aves seleccionan los sitios más o menos expuestos dependiendo de la temperatura ambiental. Se infiere que la selección del hábitat del pollo de la pradera menor es el resultado de la composición de especies de plantas, de evitar a los depredadores, y de la termorregulación, con el sistema de apareamiento lek de esta ave también jugando un papel importante. Esta especie en declive puede encontrar amenazas crecientes debido a que algunas prácticas de manejo de la tierra alteran la estructura de la vegetación y reducen la cobertura de arbustos. Temperate grasslands and shrublands are among the most endangered ecosystems and have the lowest rate of protection of all the Earth’s biomes (Brennan and Kuvlesky, 2005; Basurto and Hadley, 2006). Among these grasslands and shrublands, the prairies of the Great Plains in the United States have been severely altered and diminished (Samson and Knopf, 1994; Samson et al., 2004), harming a myriad of species that depend on them. This includes elimination of prairie-specialist mammals (Benedict et al., 1996) and substantial overall declines in populations of many avian species (Askins et al., 2007). It is important to understand what factors affect an organism’s ability to survive and reproduce. This understanding requires study of selection of habitat at different levels as well as assessment of various hypotheses about why an organism selects or avoids a particular habitat. In terms of vegetation, animals select a habitat on the basis of taxonomic composition (floristics) and structural features (physiognomy), although their relative importance is a matter of debate (Rotenberry, 1985; Mac Nally, 1990; Müller et al., 2010). Beyond vegetation, terrestrial species in open habitats, including desert and prairie, often use features of the landscape such as rocks and mounds for cover. The various factors that determine selection of habitat are frequently in flux due to human activities. Agriculture, development of energy, suppression of fire, and livestock operations have fragmented and converted the native structure of prairies. One of the icons and umbrella-species of the prairie of the southcentral United States is the lesser prairiechicken (Tympanuchus pallidicinctus), an endemic grouse that has experienced a precipitous decline in population size in the past decades, chiefly because of large-scale conversions and loss of habitat (Woodward et al., 2001; Fuhlendorf et al., 2002; Samson et al., 2004; Pruett et al., 2009a; C. A. Hagen and K. M. Giesen, http://bna.birds. cornell.edu.ezproxy.lib.ou.edu/bna/species/364). Besides conversion of habitat due to development and agriculture, the extent of heterogeneous habitat with native grasses and forbs, and cover of intermixed shrubs, in which lesser prairie-chicken evolved, has been reduced with traditional practices of rangeland management (Fuhlendorf et al., 2006). Excessive grazing renders nesting cover insufficient (Riley et al., 1992), and suppression of fire has allowed encroachment by trees, a major contributor to reduction of the size of populations of the lesser prairie-chicken (Fuhlendorf et al., 2002). Collisions with low fences, which proliferate in the prairie following settlement by humans, also have become a major source of mortality (Patten et al., 2005a; Wolfe et al., 2007). This lekking species requires open areas when displaying and choosing mates; there also is evidence that the lesser prairie-chicken selects habitats with a favorable microclimate, one associated with higher survival of adults (Patten et al., 2005b) and broods (Bell et al., 2010). Our effort goes beyond traditional studies of use of habitat in that we examine selection of microhabitat and how it affects mortality from predation. The specific predator-cover tactics of species of grouse are considered major determinants of annual rates of mortality, mostly independent of fecundity and density (Bergerud and Gratson, 1988). Grasses, forbs, and shrubs are integral parts of the habitat of the lesser prairie-chicken; we wished to determine how floristics and physiognomy at this fine scale influenced selection of habitat. We generated predictions based on the view that use of cover is predicated on a mix of avoidance of predators, thermoregulation, and reproductive requirements (Table 1). To avoid aerial predators, extent of cover at occupied sites always ought to exceed that available at random sites in that same habitat. We expect mammalian predation to occur chiefly with the use of scent at night, but cover affects dispersion of scent (Bergerud and Gratson, 1988; Conover and Borgo, 2009; Conover et al., 2010). To thermoregulate, extent of cover used should be higher when temperature is high (i.e., cover provides shade) but lower when temperature is low (i.e., birds seek solar radiation for warmth to avoid hypothermia). During the reproductive season, cover should be sacrificed when lekking or searching for suitable nest sites (March–June) but prioritized otherwise. Only if predators, temperature, and reproduction jointly drive cover-seeking behavior did we expect to find evidence in support of all three predictions, which would imply that selective pressures on the species converge sometimes (e.g., exposure and associated predation should be lowest in autumn) but conflict at others, leading to tradeoffs under certain climatic conditions. Our objectives are to describe selection of habitat by lesser prairie-chicken by determining associations between choice of site and flora, physical exposure in different seasons and microclimate, and whether mortalities from predation change with amount of exposure. MATERIALS AND METHODS—Our two study areas were located in the mixed and shortgrass prairies of the southwestern Great Plains, in Beaver, Harper, and Ellis counties in northwestern Oklahoma, and Roosevelt County in eastern New Mexico. Natural vegetation was characterized by a community of sand shinnery oak (Quercus havardii) or sand sagebrush (Artemisia filifolia; Dhillion et al., 1994; Peterson and Boyd, 1998) on predominantly sandy soil or sandy clay loam. The study area in Oklahoma (56,175 ha) supported native prairie (59%) dominated by sand sagebrush and plums (Prunus) and fields of the Conservation Reserve Program (21%) dominated by Old World bluestems (Bothriochloa), lovegrass (Eragrostis), or native-mix, with numerous forbs and grasses (Appendix 1). There was a natural gradient with increased sand sagebrush in the east (western Oklahoma) and increased shinnery oak in the west (eastern New Mexico). The study area in New Mexico (42,150 ha) included the New Mexico Department of Game and Fish’s North Bluit and Milnesand Prairie-Chicken Areas and portions of private ranches. The habitat in this area was analyzed through geographic information system for planning conservation of the lesser prairie-chicken by Johnson et al. (2006). The dominant sand-shinnery-oak communities were fragmented by cultivation, and some sites on one private ranch were treated with tebuthiuron, an herbicide designed to thin or kill stands of shinnery oak and other shrubs (Patten and Kelly, 2010). Other common shrubs and subshrubs at the study site in New Mexico included honey mesquite (Prosopis glandulosa), tree cholla (Cylindropuntia imbricata), broom snakeweed (Gutierrezia sarothrae), and soapweed yucca (Yucca glauca). Common grasses were sand bluestem (Andropogon hallii), little bluestem (Schizachyrium scoparium), blue grama (Bouteloua gracilis), black grama (Bouteloua eriopoda), sideoats grama (Bouteloua curtipendula), sand dropseed (Sporobolus cryptandrus), and purple threeawn (Aristida purpurea). Common forbs were western ragweed (Ambrosia psilostachya), annual buckwheat (Eriogonum annuum), and camphorweed (Heterotheca subaxillaris; Bell et al., 2010). Approximately half of each study area was subjected to grazing by cattle. Climate at both study sites was semiarid continental with hot summers and cold, dry winters with a frost-free growing period extending from mid-April to late October (Wright, 2003). Mean (–1 SD) annual precipitation (New Mexico, 45.4 – 15.5 cm; 136 vol. 58, no. 2 The Southwestern Naturalist