Integration of geospatial and cattle nutrition information to estimate paddock grazing capacity in Northern US prairie Rebecca Phillips a, * , Ofer Beeri b , Eric Scholljegerdes a , David Bjergaard c , John Hendrickson a a USDA-ARS, Northern Great Plains Research Laboratory, Highway 6 South, P.O. Box 459, Mandan, ND 58554, United States b University of North Dakota, Space Studies Department, P.O. Box 9008, Grand Forks, ND 58202, United States c The John Hopkins University, Baltimore, MD, United States article info Article history: Received 6 March 2008 Received in revised form 9 January 2009 Accepted 14 January 2009 Available online 13 February 2009 Keywords: Landsat thematic mapper Advanced spaceborne thermal emission and reflection radiometer ASTER Livestock Grassland Remote sensing Grazing capacity abstract Spatiotemporal variability in forage quantity and quality requires that regular assessment is needed of the capacity for grasslands to support livestock nutritional requirements. Current methods for estimating grazing capacity are typically production-based and lack the forage quality data necessary to match nutrients in forage with livestock requirements in real time. This paper describes a method for estimating short-term grazing capacity for small (1–20 ha) paddocks using cattle nutrition and high spatial resolu- tion forage data in Geographic Information Systems (GIS) for mixed-grass prairie. We define grazing capacity as the number of days a specific paddock will support the nutritional requirements of beef cattle. We integrate previously published methods for estimating cattle nutritional requirements, forage quality (crude protein) and forage quantity (phytomass) to estimate grazing capacity based on current standing- crop. The model utilizes high-resolution (<30-m) satellite imagery or field data to estimate short-term grazing capacity for small paddocks. Three versions of the model were evaluated on one paddock under cattle use in 2007. One version was parameterized using data collected on June 22 from the Landsat The- matic Mapper (TM), one version was parameterized using data collected June 23 from the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER), and one version was parameterized using data collected June 20 from field clippings. TM and ASTER versions underestimated grazing capacity by four days while the field version overestimated grazing capacity by one day. Results suggest integra- tion of cattle nutrition and forage data in GIS could assist with stocking rate adjustments, but additional trials are needed. Published by Elsevier Ltd. 1. Introduction Producers need real-time estimates of grazing capacity for mul- tiple paddocks because the capacity for Northern US prairie grass- lands to meet livestock nutritional requirements changes seasonally and annually (Vallentine, 2001; Diaz-Solis et al., 2006; Grigera et al., 2007). Stocking rates should be dynamic to preserve the sustainable balance between livestock production and grass- land health (Diaz-Solis et al., 2006). Livestock performance is di- rectly influenced by the quantity and quality of forage, so these variables are typically used as indicators of grazing capacity. How- ever, determination of forage quality and quantity in the field typ- ically requires intensive surveys that are time-consuming and expensive, so managers often estimate grazing capacity based on historical land-use and visual inspections (Vallentine, 2001). Forage quantity and quality can be estimated remotely for North Dakota grasslands at multiple locations with <20% error (Phillips et al., 2006; Beeri et al., 2007), so incorporation of forage data into existing livestock performance models is a logical next-step to- wards building decision support systems. Linking these dynamic forage data with cattle nutrition requirements in a Geographic Information Systems (GIS) framework would help optimize re- source utilization by matching real-time forage availability to live- stock nutritional needs (Moen, 1984; Hobbs et al., 1985) and support adaptive and sustainable grassland management (Hunt et al., 2003). Landsat Thematic Mapper (TM) and Advanced Very High Reso- lution Radiometer (AVHRR) data have been applied toward moni- toring grassland trend and condition in Western Australia (Bastin et al., 1998; Edirisinghe et al., 2000; Wallace et al., 2004), the Bra- zilian Amazon (Asner et al., 2004; Numata et al., 2007), the South- western, USA (Qi et al., 2000), and Western China (Zha et al., 2003). In many cases, the focus is on delineation of vegetation from bare soil or on qualitative differences in spectral vegetation index values (Pickup et al., 1993; Qi et al., 1994), which is practical for long-term monitoring. Some have used Moderate Resolution Imaging Spect- roradiometer (MODIS) and AVHRR data in models of radiation 0308-521X/$ - see front matter Published by Elsevier Ltd. doi:10.1016/j.agsy.2009.01.002 * Corresponding author. Tel.: +701 667 3002; fax: +701 667 3054. E-mail address: rebecca.phillips@ars.usda.gov (R. Phillips). Agricultural Systems 100 (2009) 72–79 Contents lists available at ScienceDirect Agricultural Systems journal homepage: www.elsevier.com/locate/agsy