Novel ecosystems are composed of species that have not historically co-occurred within a geographic extent. Species interactions may lead novel ecosystems to function differently from those they have replaced with regard to ecosystem services like fire regimes, water availability, and nutrient cycling. Both species interactions and anthropogenic factors (e.g. the deliberate introduction of the species and climatological shifts) are major drivers of ongoing novel ecosystem formation. These impacts will continue and so identifying, evaluating, and understanding novel ecosystems is critical to any kind of land management.
Because novelty can only be recognized in comparison to a baseline, historic data are necessary. In order to identify changes in species distributions and potential novel ecosystems on larger scales, we used two publicly available data sets: the Bureau of Land Management’s (BLM) Soil-Vegetation Inventory Method (SVIM) data set covering 1977-1983 and the BLM Assessment, Inventory, and Monitoring (AIM) data set covering 2011-2018, both of which are spatially explicit and extend across much of the BLM-managed land in the American west. We confronted compatibility challenges by limiting our focus to discrete geographic areas of high-density overlap between data sets, deriving equivalent presence information from the data from cover values, and avoiding analyses dependent on sample design information.
Results/Conclusions
By comparing the SVIM and AIM data, we were able to detect range changes of species of management concern and identify the occurrence of novel ecosystems over the 40-year time span of the two datasets. Regional shifts in distribution differed in magnitude and by species, as would be expected from spatial variation in driving factors and biotic potential. Additionally, results corroborate the expansion of systems dominated by annual grasses, such as Bromus tectorum, and associated decreases in deep-rooted perennial grasses.
Methodological differences between the two datasets, as well as a lack of documentation for the SVIM data contributed to uncertainty in estimating changes in species distribution and abundance. Additionally, using cover to detect presence may underestimate species occurrence when species fall below the cover threshold for likely detection. However, rare species may be unlikely to have significant impact on ecological function.
Detecting trends in large-scale ecosystem changes and assessing ecosystem function require long time series, which in turn requires either extended monitoring or the use of historic data sets. We demonstrate the viability of the latter for answering these questions to inform management decisions.