Using packrat middens to assess grazing effects on vegetation change

Research on grazing effects usually compares the same sites through time or grazed and ungrazed sites over the same time period. Both approaches are complicated in arid environments where grazing can have a long undocumented history and landscapes can be spatially heterogenous. This work employs both approaches simultaneously by comparing grazed and ungrazed samples through both time and space using fossil plant macrofossils and pollen from packrat middens. A series of 27 middens, spanning from 995 yr BP to the present, were collected from Glen Canyon in southeastern Utah, USA. These middens detail vegetation change just prior to, and following, the historical introduction of domesticated grazers and also compares assemblages from nearby ungrazable mesas. Pre-grazing middens, and modern middens from ungrazed areas, record more native grasses, native herbs, and native shrubs such as Rhus trilobata, Amelanchier utahensis, and Shepherdia rotundifolia than modern middens from grazed areas. Ordinations demonstrate that site-to-site variability is more important than any temporal changes, making selection of comparable grazed versus ungrazed study treatments difficult. But within similar sites, the changes through time show that grazing lowered the number of taxa recorded, and lessened the pre-existing site differences, homogenizing the resultant plant associations in this desert grassland.