Cattle Grazing Effects on Spring Ecosystems in California's Oak Woodland
This study was designed to examine effects of cattle grazing on undeveloped, cold-water spring water quality, vegetation cover and composition, insect family richness, and channel morphology.
As early as 1991, regulatory agency interest in nonpoint source pollution surfaced as a major issue for management of grazing animals on California rangelands. In 1994, the Range Management Advisory Committee (RMAC) to the State Board of Forestry identified water quality as the major environmental issue facing the California range livestock industry. More recently, coho and steelhead salmon populations for much of the north coast were added to the federal endangered species list. Subsequently, the United States Environmental Protection Agency (EPA) announced that total maximum daily load (TMDL) water quality standards for nonpoint source pollution will be developed for 18 critical north coast salmon rivers. The EPA's decision to develop TMDLs for nonpoint source pollutants on these 18 rivers by the year 2007 likely will set a national precedent. Loading of sediment and heat due to grazing activities has been identified as a cause for concern for water quality and habitat degradation on most of these 18 waterways.
These recent events have focused immediate, intense attention on the impacts of livestock production on water quality and endangered fisheries habitat, as well as upon the ability of watershed management practices to minimize these impacts. Limited, specific, science-based information is available concerning "background" water quality, the water quality impacts of livestock production, and the effectiveness of water quality best management practices on California's rangeland watersheds.
Thus in 1992, funded by the Integrated Hardwood Range Management Program, I began research focused on the response of perennial springs and their associated ephemeral creek to cattle grazing. In 1996, Dr. Ken Tate (UC Davis) and I began a fruitful collaboration and expanded my spring-based plot research to examine 3 small, but complete, watersheds.
Three spring creeks from each of 3 watersheds (Campbell, Schubert, and Forbes) at Sierra Foothill Research and Extension Center, Marysville, were located and randomly assigned one of 3 grazing treatment levels. Mean annual residual dry matter (RDM) levels were ungrazed (~2500 kg/ha RDM), lightly grazed (~1500 kg/ha RDM), or moderately grazed (~1000 kg/ha RDM). Permanent line-point transects were established parallel to spring-creek flow for vegetative cover and plant community composition estimation. Analyses of variance (ANOVA) and community ordination (TWINSPAN [Two-way Indicator Species Analysis], a polythetic, divisive classification computer application) were used to assess effects of time and grazing treatment on plant cover and community composition and stability of plants. Permanent line-point transects also were established perpendicular to spring-creek flow to measure fluctuations in channel morphology. Total change in cross-sectional channel depth was estimated annually. Aquatic insects frequently are studied to evaluate the ecological integrity of streams. Reduced community richness may indicate organic pollution and/or habitat degradation. Therefore, insect emergence traps were placed at springheads during four seasonal periods in 1996 to assess insect family richness over time and grazing treatments. Trapped insects were identified to family. ANOVA were used to assess species richness over time and among grazing treatments, and was limited to families with aquatic genera; wholly terrestrial families were excluded. Spring water was monitored to assess potential livestock effects on spring-creek sediment and nutrient loading. Water samples were collected from spring-creek effluent post-livestock grazing to estimate nitrate, orthophosphate, dissolved oxygen, temperature, and pH.
Total herbaceous cover varied significantly over time (year) but not grazing treatment (Fig. 1). Plant communities persisted and were stable regardless of grazing intensity. Temporal variation in total cover apparently was associated with rainfall from the previous year (Fig. 1). Channel morphology in ungrazed plots varied more over time than in grazed plots (Fig. 2). Although a trampling effect from cattle clearly affected the morphology of grazed springs, our method did not demonstrate this quantitatively. Moderately grazed springs exhibited reduced insect family richness compared to ungrazed and lightly grazed sites (Fig. 3). Insect family richness was significantly lower in February than July, September, and April sampling dates. No significant differences among grazing treatments were found for cumulative 5-year means of nitrate, orthophosphate, dissolved O2, temperature, and pH (Table 1).
While some riparian habitats have been characterized as fragile, we argue that these systems have a "built-in" buffering ability, probably resulting from abundant water and nutrients supplied from surrounding annual grassland and oak woodlands. Indeed, we now think of these systems as mini-wetlands, fortuitously situated at watershed lowlands. These ideas are directing our future research agenda: to test whether or not these systems act as nutrient and sediment filters at the terrestrial-aquatic interface.
Table 1. Water quality characteristics (mean and standard error) of moderately grazed (MG), lightly grazed (LG), and ungrazed (UG) springs, Marysville, California.
|Disolved oxygen (mg/l)||4.85||1.62||7.02||10.96||5.32||1.90|
prepared and edited by Richard B. Standiford, Justin Vreeland, and Bill Tietje