In the Mojave Desert, spring ecosystems are important hotspots of biodiversity which host a disproportionate density of plants and wildlife compared to the surrounding landscape. Yet we understand little about the diversity at desert springs, and even less about how spring communities will respond to threats such as climate change, human use, and non-native species. Here we present a new long-term vegetation monitoring protocol for 17 springs across six national parks in the Mojave Desert, and assess the initial results from four springs. We also discuss our assessment of three survey methods considered for this protocol, since it was particularly difficult to select standardized methods to apply across distinctive springs. We address two monitoring questions: 1. What are the status and trends in plant community composition and species richness at select desert springs? 2. How is species diversity and richness associated with human-related and natural disturbances? We measure species composition and disturbance using line-point intercept (LPI) along transects (number of transects is a function of spring area). We estimate species richness and overstory composition in “belt transects” along permanent transects. Pilot data (two springs, 2018) and the first year of data (four springs, 2019) were collected in Lake Mead National Recreation Area (LAKE).
Results/Conclusions
After comparing methods, LPI was the best choice for estimating species richness and diversity. Measuring vegetation in Daubenmire quadrats along transects required more sampling effort than LPI (for the same power) by over 20%, and due to the clumped nature of the the vegetation, quadrats appeared inappropriate. A fully random quadrat method, while avoiding the challenges of relocating sampling locations, required 45% more sampling effort than LPI. Therefore, we opted to intensively measure permanent transects, and accepted the limits to sample size, spatial extent, and the scope of inference that necessarily come from this design.
Springs in the initial season at LAKE, by vegetation cover, had a low occurrence of invasive plants (< 5% total cover), low occurrence of rare plants (< 5% total cover), and tended to be shrub dominated (20-50% total cover). Repeat measurements (to be collected every 3 years) will give the ability to detect trends in community dynamics at springs.
This long-term monitoring protocol provides much needed knowledge about an important ecological indicator in the Mojave Desert. Further, this protocol provides a good balance of time invested to information provided and can be adapted for monitoring vegetation at other groundwater-fed systems.