Advances in Understanding Landscape Influences on Freshwater Habitats and Biological Assemblages

<i>Abstract.</i>—Linking landscape features, both natural and human-altered, to aquatic ecosystem structure and function is a fundamental objective in landscape ecology and freshwater science, but this process is data- and resource-intensive. Quantifying how landscape stressors influence aquatic communities requires balancing logistic and financial constraints with effectively sampling the landscape to capture gradients of interest. There are a variety of ways to balance these constraints, such as using existing data, handpicked site selection, or a statistical site-selection scheme. Poor sampling design reduces statistical power; however, we do not know how differences in site-selection designs influence our ability to measure ecological responses to landscape gradients. We quantified how the distribution of sample sites across landscape gradients affected the measured responses of stream fish assemblages to these gradients at different sample sizes. Specifically, we used randomization tests to compare the variability in the responses of fish assemblage structure (species richness and composition) to catchment area and land use (agricultural land) with manipulated distributions (random, highly skewed, and uniform) of sites across these landscape gradients. Assemblage composition was more sensitive than species richness to sampling design, and we observed less variability in the detected response of assemblage composition when samples were distributed uniformly across landscape gradients, especially when sample sizes were small. Although strong responses to environmental gradients, such as species richness to catchment area, are robust to sampling distributions, large sample size and a uniform distribution of samples might help elucidate more subtle responses to environmental gradients.

Advances in Understanding Landscape Influences on Freshwater Habitats and Biological Assemblages

<i>Abstract.</i>—Lakes, like islands, have been model systems for testing important concepts in landscape ecology. Lake assemblage and community composition, generally, and the occurrence of invaders, specifically, are controlled by a range of factors across scales. Here, I use the example of Rainbow Smelt <i>Osmerus mordax </i>invasions in inland lakes to illustrate common problems in both predictive and explanatory models of invasive species distributions across landscapes. Using variables related to dispersal and regional- and lake-scale environment, I fitted a series of boosted regression tree models to examine the factors that explain Rainbow Smelt invasion success. These models illustrate the potential effects of extrapolation and nonequilibrium conditions, the role of human activities, and the difficulty of understanding the importance of biotic interactions in the spread of invasives. Understanding the factors controlling invasions should inform management and conservation of inland lake ecosystems. For this to be effective, a mechanistic framework is needed to untie correlations in potential driving factors. Emerging data sets with fine spatial grain and broad spatial extent will support the transition from correlative models to mechanistic understanding of aquatic invasions.

Advances in Understanding Landscape Influences on Freshwater Habitats and Biological Assemblages

<i>Abstract.</i>—Waterfalls, caused by the abrupt changes of elevation in streams, are natural barriers that influence the distribution and dispersion of aquatic species. The resulting habitat fragmentation has contributed to species specialization as well as barriers that inhibit passage of nonnative species upstream. In Hawai‘i, it is assumed that nonnative species are unable to pass waterfall barriers, yet they are present above some waterfalls, possibly facilitated by human introduction. In this study, we used a landscape approach to identify likely human introductions and examine the ability of nonnative stream fauna to bypass waterfalls. We identified the human activities associated with the high likelihood of species introduction. We found that when a local catchment has a population density >4.24 people/km2 or road length density is >0.01 km/km2, the presence of nonnative species in the stream is likely a result of human introduction. After filtering human facilitated introduction, we also assessed the potential waterfall climbing ability of 14 nonnative taxa. We found that 12 out of the 14 taxa were absent upstream of waterfalls, indicative of their inability to traverse waterfalls. Only two species, Tahitian prawn (also known as monkey river shrimp) <i>Macrobrachium lar </i>and American bullfrog <i>Rana catesbiana</i>, seem able to pass waterfalls. This study highlights the role that people play in facilitating species introductions in otherwise inaccessible habitats. Without human interference, waterfalls can be considered effective barriers to nonnative species and can be instrumental in supporting nonnative species eradication and control strategies.

Advances in Understanding Landscape Influences on Freshwater Habitats and Biological Assemblages

<i>Abstract.</i>—Over the past decade, numerous studies have identified correlative relationships between aquatic biota and human activities at landscape scales. In addition to demonstrating the pervasive effects of these activities on aquatic biota, these findings have encouraged researchers to suggest that predictive relationships between human activities and aquatic biota could be used to enhance diagnostic power of biological assessments, predict future changes in species distributions, and inform land-use planning. However, to achieve these important goals, descriptions of human activities will need to become more detailed than the simple land use/land cover classifications frequently used. Our purpose is to highlight four sources of human activity data (existing geographic information system layers, census data, remotely sensed images, and visual landscape surveys) that can be used to increase the level of detail with which the human environment is described. Strengths and weaknesses of each data source are discussed and methods for adapting those data to aquatic studies are described by drawing on experiences from studies in the agricultural landscapes of southern Manitoba and southwestern Ontario, Canada. Based on the observations and lessons learned from our previous experiences, we make recommendations for how researchers can identify and apply the data sources that best meet their needs. We also discuss challenges and possible solutions for applying the described data sources as well as for improving data availability in the future. Moreover, we encourage aquatic researchers to allot more time to detailed description of human activities because we believe this to be an effective approach to improving our ability to predict the effects of human activity and thus better assist decision makers in protecting aquatic ecosystems.

Advances in Understanding Landscape Influences on Freshwater Habitats and Biological Assemblages

<i>Abstract.</i>—Scale influences the detection of relationships between landscape alterations and stream fishes and therefore has strong implications for conservation planning and implementation. This is especially true in riverscapes because terrestrial landscapes drain into riverscapes in a manner that can be measured at multiple scales. Two commonly employed scales in riverscape ecology and conservation include local catchment (i.e., the area of land draining directly into a segment of stream between two confluences) and network catchment (i.e., the total area of upstream land). We used a multispecies extension of species distribution modeling (i.e., gradient forest) to describe relationships between landscape alterations (measured at local catchment and network catchment scales) and stream fish occurrence patterns in portions of the Mississippi and Tennessee River basins in western Tennessee, USA. Landscape alterations included seven urban or agricultural classes, and densities of roads, road crossings, dams, human population, mines, and confined animal feeding operations. At the network catchment scale, the most influential landscape alterations affecting fish distributions were cultivated crops and pasture/hay land uses, but at the local catchment scale, open-space development, human population density, and road density were most important for describing multispecies fish distributions. Despite these differences, gradient forest model performance measured as explained variation at the species level was consistent between local catchment and network catchment scales. Furthermore, when predictions for unsampled stream segments were mapped across the region, both scales produced consistent patterns in fish assemblages affected by low, medium, and high development or cultivated crops. Our results provide direction to conservation practitioners by identifying regions where limited resources might be allocated to increase efficiencies within two highly altered and taxonomically diverse riverscapes. The framework described here provides a case study for application of new statistical innovations to address conservation challenges and can be used in other landscapes and riverscapes to identify locations where management efforts might be best allocated.

Advances in Understanding Landscape Influences on Freshwater Habitats and Biological Assemblages

<i>Abstract.</i>—Anthropogenic activities including urbanization, agriculture, and dams degrade stream habitats and are a dominant reason for global biodiversity declines in fluvial fish assemblages. Declining diversity trends have been well documented in many regions of the world; however, fishes vary regionally in response to anthropogenic land use, resulting from complex relationships between landscape variables and mechanisms controlling stream fish assemblages. To test for differences in regional fish response to anthropogenic land use, we conducted our study across five freshwater ecoregions in the temperate mesic portion of the United States and evaluated data characterizing stream fish assemblages from 10,522 locations across all study freshwater ecoregions. Fishes were summarized by metrics describing assemblage structure, trophic groupings of species, levels of tolerance to anthropogenic stressors, and life history characteristics, with seven metrics used for analyses. Natural and anthropogenic landscape variables were assessed across freshwater ecoregions, and we tested for regionally specific influences of percent catchment urbanization, percent catchment agriculture, and catchment densities of dams and stream-road crossings on stream fishes. We used cascade multivariate regression trees to quantify variance explained in fish metrics by these landscape variables after controlling for influences of natural landscape variables, including catchment area, catchment lithology, and elevation of study sites. Results indicated differences in dominant influences by freshwater ecoregion, as well as differences in the levels of anthropogenic land use influencing fishes within and across freshwater ecoregions. For example, urban land use was the most influential anthropogenic land use in both Appalachian Piedmont and Chesapeake Bay freshwater ecoregions, with fish assemblage metrics showing responses at 10% and 1% catchment urban land use, respectively. In contrast, dam density in the network catchment was the most influential anthropogenic variable on fish assemblage metrics in both the Laurentian Great Lakes and Middle Missouri freshwater ecoregions. Also, large amounts of agriculture in the catchment was the most influential anthropogenic land use on fish assemblage metrics in the Upper Mississippi freshwater ecoregion. Knowledge of regional differences in the top contributing anthropogenic landscape variables and the levels at which fish assemblages respond to these variables lends insight into mechanisms controlling stream fish assemblages by freshwater ecoregions and can aid in development of region-specific conservation strategies to prevent biodiversity loss from current and future anthropogenic land use.

Advances in Understanding Landscape Influences on Freshwater Habitats and Biological Assemblages

<i>Abstract.</i>—Surrounding land use and cover can have profound effects on the physical, chemical, and biological properties of stream ecosystems. For this reason, changes in land use and cover throughout catchments often have strong effects on stream ecosystems that are particularly interesting to researchers. Additionally, natural physical and climatic, or physiographic, characteristics are important for determining natural land cover and constraining human land use and are also strongly related to stream habitat and biota. Because the physiographic template differs among catchments and is an important mediator of catchment processes, it is important to account for natural physiographic differences among catchments to understand the relationship between land use/cover and stream biota. In this paper, we develop and assess the usefulness of a regional framework, land use/cover distinguished physiographic regions (LDPRs), which is designed for understanding relationships between land use/cover and stream biota while accounting for the physiographic template. We classified hydrologic units into LDPRs based on physiographic predictors of land use and cover for the eastern and western United States through the use of multivariate regression tree analysis. Next, we used case study data to assess the usefulness of LDPRs by determining if the relationships between fish assemblage function and land use/cover varied among classes using hierarchical logistic regression models. Eight physiographic characteristics determined land cover patterns for both the eastern and western United States and were used to classify hydrologic units into LDPR classes. Five commonly used biotic metrics describing trophic, reproductive, and taxonomic groupings of fish species responded in varying ways to agriculture and urban land use across LDPRs in the upper Mississippi River basin. Our findings suggest that physiographic differences among hydrologic units result in different pathways by which land use and cover affects stream fish assemblages and that LDPRs are useful for stratifying hydrologic units to investigate those different processes. Unlike other commonly used regional frameworks, the rationale and methods used to develop LDPRs properly account for the often-confounded relationship between physiography and land use/cover when relating land cover to stream biota. Therefore, we recommend the use and refinement of LDPRs or similarly developed regional frameworks so that the varying processes by which human land use results in stream degradation can be better understood.

Advances in Understanding Landscape Influences on Freshwater Habitats and Biological Assemblages

<i>Abstract.</i>—Currently, much is known about influences of landscape attributes, including timber harvest practices, on large wood dynamics in streams. Comparatively, much less is known about influences of catchment attributes on Southeast Alaska streams, in part because of a historical lack of consistent catchment-scale data available for the region. As in forested regions elsewhere, large wood is an important resource to stream habitats and fishes in Southeast Alaska. Therefore, the purpose of this study was to characterize catchment-scale influences, including various timber harvest practices, on large wood in Southeast Alaska stream sites. We delineated local catchment boundaries for all stream reaches in the region and summarized landscape influences, including natural and harvest-related attributes in local and network catchments. Relative amounts of variation in four large wood habitat variables explained by natural versus harvest practice-related landscape attributes were evaluated to compare different influences on 28 randomly selected study sites. We used those results to predict variation in large wood variables from our sites through use of both natural and harvest practice-related catchment attributes to identify those that may be most influential to large wood. Natural characteristics, including catchment area, deciduous forests, forested wetlands, and catchment slope, all had significant influences on large wood variables, as did various measures of contemporary and historical timber harvest practices. We found that large wood length was positively related to conventional harvest after 1990, suggesting the potential effectiveness of contemporary logging regulations in protecting large wood characteristics. In our study, both natural and timber harvest practice-related attributes had measureable influences on stream habitat, underscoring the importance of considering catchment-scale attributes, including riparian management schemes, for managing Southeast Alaska streams.

Advances in Understanding Landscape Influences on Freshwater Habitats and Biological Assemblages

<i>Abstract.</i>—Habitat fragmentation, land use practices, and flow impediments modify the natural course of rivers, disrupting connectivity and subsequently affecting dispersal and gene flow in aquatic organisms. Many of the relationships between the physical river network and the genetic structure of populations are not well understood. Riverscape genetics is a developing field that uses population genetic metrics to assess genetic structure within the context of the environmental variables that drive functional connectivity in a river network. Here, we applied an effective distance network approach to characterize the effects of hydrology in shaping neutral genetic population structure of fall-run Chinook Salmon <i>Oncorhynchus tshawytscha </i>within a small, coastal Oregon catchment. We evaluated whether gene flow was limited by (1) site-specific features occurring within spawning habitat, using a dissimilarity matrix, and (2) the cumulative effect of the environment accrued while traveling en route between reaches. We found that Chinook Salmon that spawned at higher elevations (site specific effects) after traversing steeper gradients (en-route effects) were more genetically distinct from individuals that traversed gradual gradients and spawned at lower elevations. This effect (isolation by resistance) was distinguishable from isolation by distance, which was not detected among spawning groups. Our study enhanced interpretation of habitat heterogeneity in constraining gene flow and spatial genetic structure among reaches within a small, coastal catchment. Given that smaller catchments may hold life history 36 variation that is important to long-term population persistence, there is need to understand these relationships that maintain genetic diversity.