Aquatic habitats are among the most imperiled habitats on dryland military installations, yet they harbor a disproportionately high amount of biodiversity given the small land area they cover. Aquatic invertebrates (insects and allied taxa) constitute a major part of this biodiversity and form a critical part of the food web that sustains aquatic, riparian, and terrestrial organisms, including Federally Threatened or Endangered species. Biodiversity in dryland aquatic habitats is strongly influenced by spatial and temporal variability, which presents challenges for predicting how management decisions on military lands could affect landscape-scale patterns of aquatic invertebrate biodiversity. This project integrates mathematical modeling, invertebrate sampling, and statistical estimation to understand and measure biodiversity of aquatic invertebrates on dryland military bases. 

The interplay of niche-based and neutral models was a central theme of this project. The intent was to use both of these methods to determine optimal management techniques across aridland ecosystems. The researchers describe the development of the core models to be used for this purpose. They also emphasize the importance of using trait analysis to make general inferences that cross taxonomic boundaries.

The effort assembled an extensive dataset of invertebrate biodiversity patterns through space and time, replicated across three military installations spanning three different physiogeographic provinces: White Sands Missile Range in New Mexico, Fort Huachuca in Arizona, and NAWS China Lake in California. Thus far, researchers have identified an extraordinary diversity of aquatic invertebrates with the sampling, approaching 400 taxa. This was a high amount of biodiversity for a type of habitat that is relatively rare in the landscape. Reseacrhers have also discovered major range extensions of known taxa and possible new species on these military lands.

This effort developed both neutral and niche-based models that are useful for understanding how species occurrences and patterns of biodiversity are distributed across variable landscapes, such as the aridland stream networks present on many large military installations in the western US. 

From a modeling perspective, the niche-based models show the importance of stream hydrology and other abiotic factors in determining the distribution and detectability of species. A surprising number of taxa appear to be detectable in one season or microhabitat but not in others. This presents substantial challenges to installation resource managers, who need reliable information on species distributions and abundances to make informed management decisions. Niche-based methods can also allow the determination of the temporal and environmental conditions under which species are likely to be present and detectable, but given the number of physical variables involved, this was not a simple task to execute.

Conversely, the neutral models revealed some surprising findings about patterns of alpha diversity (within sites) and beta diversity (among sites) across the landscape. First, reseacrhers have found a strong effect of habitat capacity (the size of a specific site, in terms of the number of individuals it can harbor) on beta diversity. This pattern was important because it means that differences in habitat capacity among sites can distort measured biodiversity patterns. It was especially relevant to aridland aquatic habitats, because these habitats can differ in capacity by several orders of magnitude (from small isolated seeps to large river systems). This finding was of great importance to general biodiversity studies, and it has been published as a stand-alone paper in addition to implementing it into the larger project. Second, the neutral models showed that neutral biodiversity processes were more important under extreme hydrological conditions (high water and drought) versus more mesic conditions. Thus, the interplay between niche processes (which depend on the traits of individual species) and neutral processes (which are driven by stochastic processes of immigration and local extinction) appear to be mediated by hydrology. 

While the data collection and model development has revealed findings relevent to general ecology, especially in aridland aquatic ecosystems, general recommendations can be made pertinent to the management of aridland aquatic ecosystems, especially as they pertain to the three military installations studied here.

1. The hydroperiod of a site strongly determines the species pool of that site. In nearly all the modeling efforts, hydroperiod was a strong explanatory variable. Thus, sites that are perennial have a unique character and species pool that sets them apart from intermittent sites, and vice-versa. Care should be taken to ensure that perennial sites retain this character and do not become intermittent due to groundwater pumping or diversions. Conversely, naturally-intermittent sites can sometimes harbor a unique fauna, and it may not be advisable to change them to a perennial hydroperiod via artificial dams or impoundments. Hydroperiod of sites can be monitored with electrical resistance sensors such as those deployed in the study.
2. The dispersal ability of taxa strongly influences their distribution on the landscape. Taxa with little or no mobility contribute to unique community structures, much in the same way that low-dispersal taxa often have unique genetic structures across the landscape (as revealed in SERDP RC-1724). Thus, taxa that are known to be poor dispersers, including some hemipteran bugs such as belostomatids and naucorids, should be managed with particular care. Trait databases such as the one generated by Schriever et al. (2015) are a valuable management tool for determining the dispersal ability of particular species.
3. In aridland and desert ecosystems, much dispersal occurs overland via aerial dispersal, as opposed to within the stream network. For this reason, springs and off-channel habitats that are seemingly disjunct from other habitats may play an important role in maintaining the overall biodiversity of a particular region.
4. The habitat capacity of a site plays an important role in biodiversity. The neutral modeling revealed that large-sized habitats -- such as larger ponds, or more contiguous reaches of stream -- operate in fundamentally different ways than smaller habitats. In general, larger sites harbor greater biodiversity, but the results caution that smaller sites may be important for maintaining biodiversity at a broader landscape level. Thus, aridland aquatic habitats should be managed as a catchment-level matrix of diverse aquatic habitats, rather than as disjunct points on the landscape.