Relationships between benthic macroinvertebrate assemblages and the physical environment in sandy SE streams, with emphasis on stream geomorphology, hydrology and watershed disturbance
Type of DegreeDissertation
MetadataShow full item record
Biodiversity is predicted to decline over the next 100 years, and stream ecosystems are predicted to be impacted by landuse change. This predicted decline highlights the need to understand sand-bed streams and biotic response to important agents of disturbance in these understudied ecosystems when compared to gravel/cobble, upland streams. In this dissertation, I report on 4 different studies: Chapter 2) I develop and validate models for predicting discharge in sand-bed streams; in Chapter 3) I develop and validate the hydrogeomorphic reference condition; in Chapter 4) I investigate the direct and indirect effects of landuse on relevant system components with structural equation modeling within the theoretical context of the landcover cascade; and in Chapter 5) I conducted an experiment to explore the competing hypotheses that macroinvertebrate sand-bed stream assemblages are either deterministically or stochastically assembled in the face of disturbance intensity. Below I present abridged abstracts for each of the 4 studies presented in this dissertation. Chapter 2: Manning’s equation is often used to predict discharge when stream gauging becomes logistically untenable. In this chapter, I developed and validated a regional equation that predicts discharge from easily measured channel morphology variables in sand-bed streams of the Southeastern Plains without specific estimation of Manning’s n M . The results of this study suggest that the equation developed should be favored over all other equations presented in the literature to predict in-channel discharge in sand-bed, Southeastern Plains streams. iiChapter 3: Stream reference condition in sand-bed streams of the Sand Hills ecoregion of the Southeastern, US has not been defined. In this chapter, I develop and validate the hydrogeomorphic reference condition (HGM). The results of this study suggest that the HGM is a physico-chemical and biologically relevant reference condition. Further, the HGM stream class (reference/non-reference) can be determined from easily measured channel morphology variables. The results of this chapter should have regional implications for managing and preserving biodiversity in the SE Sand Hills, and, potentially, for sand-bed streams in general. Chapter 4: The landcover cascade (LCC) was developed in SE Blue Ridge streams. These systems are inherently different from lowland, sand-bed streams of the SE Sand Hills. The major question addressed in this chapter is whether the indirect effect of land cover is translated through relevant, hierarchical system components to instream biota in a similar manner as the LCC, and if direct effects of hierarchical system components are similar to those shown in the Blue Ridge. The results of this study suggest that landcover is translated differently to instream biota. Specifically, hydrologic disturbance increased when silviculture increased and decreased when restored long leaf pine decreased. Hydrology was the most important direct system component and indirect cascade to biota resulting from land use change. Chapter 5: How communities are assembled in the face of disturbance is an on-going question of interest to both applied and basic ecologist. Empirical data suggest 2 contrasting patterns of assembly under differing levels of disturbance intensity along a stochastic/deterministic gradient. In order to rectify this discrepancy and to explore how stream communities are assembled in the iiiface of disturbance intensity, I conducted a randomized complete block field experiment with 5 levels of disturbance and an un-manipulated stream-bed control. The results of this study suggest that streams communities are assembled deterministically at intermediate levels of disturbance. Further, this study is the first to show the appropriate conditions for deterministic filtering are satisfied, and to show that intermediate levels of disturbance show traits that are similar to the stream-bed controls suggesting stream macroinvertebrates are deterministically assembled at intermediate disturbance.