Filters
Date Range
Contacts
Tag Schemes
Tags (with Type=Subject )
|
Key elements of the 2015 national assessment of stream fish habitats follow the 2010 assessment, including: 1) the idea that distributions and numbers fishes reflect the quality of habitat in which they live; and 2) human landscape factors pose a risk to the condition of stream habitat, and indirectly, to fishes. The 2015 inland stream assessments for the contiguous United States, Alaska, and Hawaii all followed five broad steps (Figure 1) that are described in detail below for the inland stream assessment for Alaska. Note that analytical details for the Alaska assessment differed in southeast Alaska as compared to the remainder of the state (referred to as greater Alaska) due to differences in the resolution of...
Accounting for natural variation With the exception of differences in spatial units, assessments for greater Alaska and southeast Alaska were conducted similarly across regions. Because stream fish assemblage data were not available for the state, no steps were taken to account for natural variation in stream habitats for either southeast or greater Alaska. This represents an important need for future work.
Integrating data into a spatial framework After acquiring data, variables were attributed to a national stream coverage for use in assessment following Wang et al. (2011). The National Hydrography Dataset Version 1 (NHDV1) is a 1:100,000 scale representation of streams from throughout the conterminous United States. The NHDV1 identifies stream reaches as sections of streams occurring between confluences (Figure 2). We attributed all data to stream reaches (i.e., fish data, fragmentation metrics by dams) or to local catchments and 90m buffers draining to stream reaches (i.e., human land uses, mining activities, impervious surfaces, etc.). Local catchments (watersheds) and buffers are the land areas draining directly...
Identifying disturbances to fish habitat The approach for identifying disturbances to fish habitat was based on the assumption that greater intensities and types of human landscape disturbances would most likely lead to more disturbed stream fish habitat (e.g., Danz et al. 2007, Esselman et al. 2011). Twenty-two human landscape variables were identified for the Alaska assessment, with 21 variables used in the southeast and 19 in greater Alaska. We grouped variables into six sub-indices representing specific types of disturbances including: urban land use, agricultural land use, stream fragmentation, point source pollution, infrastructure, and active mines. Each sub-index of disturbance was represented by 2 to...
Integrating data into a spatial framework Greater Alaska For most of Alaska excluding the southeast portion of the state, watershed boundaries for individual stream reaches were unavailable, and the highest resolution spatial units available for assessment were 12-digit USGS hydrological units (HUC-12s). Greater Alaska includes 12,824 HUC-12s that partially follow watershed boundaries; however, boundaries are also intended to capture roughly similarly-sized regions vs. entire upstream landscape areas draining to streams (Figure 11). After acquiring data, variables were attributed to HUC-12s for the greater Alaska assessment. Southeast Alaska For the southeast portion of Alaska, watersheds were delineated from...
Assembling Response Data The assessment uses available fish and shellfish species presence/absence as indicators of the effects of anthropogenic (human caused) stressors on the estuarine habitats where fish and shellfish live, feed, and reproduce. Fish data were obtained from state and federal trawl survey programs, including each of the five coastal states as well as the U.S. Environmental Protection Agency’s (EPA) Environmental Monitoring and Assessment Program (EMAP) and National Coastal Assessment (NCA). Fish trawl nets are pulled through the water at specified sampling locations for a set period of time to determine the abundance and diversity of fish in the area. Environmental data like water temperature,...
Data on stream fishes were provided for use in the 2015 assessment by the Hawaii Division of Aquatic Resources. Data were collected from 1992 to 2010, and assemblages were sampled using standardized visual surveys (Higashi and Nishimoto 2007). Fish data indicated presence or absence of nine native taxa in stream reaches including five fluvial fish species, two shrimp species, a gastropod, and two species of native flagtails (treated as a single taxonomic group analytically) that periodically enter the stream from the nearshore coastal environment (Table 6). Fish presence-absence data were available for 403 perennial stream reaches throughout the five main Hawaiian Islands. Many different human landscape factors...
Stream fish data providers for 2015 national assessment of stream fish habitats.
Developing Sub-Indices of Disturbance Variables within each of the four disturbance categories (Land Use/Land Cover, Alteration of River Flows, Sources of Pollution, and Estuary Eutrophication) were combined to create a sub-index of disturbance associated with each category. To calculate the sub-indices, a percent rank was calculated for variable scores in each estuary. Percent rank scores were inverted (i.e. 1-percent rank) where necessary to maintain consistency of interpretation, with low scores representing the highest degree of disturbance. Within each disturbance category, the average of each of the variable percent rank values was calculated. Finally, the percent rank of this average was calculated – this...
Creating cumulative habitat condition scores Greater Alaska All six sub-indices of disturbance scores in each HUC-12 were summed together to yield a cumulative habitat condition index (CHCI) score for each HUC-12. The maximum value for the CHCI was 6, indicating that a HUC-12 was in the worst condition class for each sub-index of disturbance, while the minimum value of the CHCI was 0, indicating that a HUC-12 was in the best condition class for each sub-index of disturbance. We followed methods applied for the conterminous US and created condition classes using Jenk’s natural breaks. With the exception of the HUC-12s that received a CHCI score of <0.001, which were given a priory assignment of “very low” risk of...
Assembling Data We used datasets of habitat stressors available at a national scale and measured within estuaries and their associated watersheds. Included datasets represent anthropogenic stressors likely to affect fish habitat within estuaries based on evidence from the published habitat ecology literature. Although many important factors were included, not all data were available at sufficient spatial resolution or geographic breadth for meaningful analysis. Some important datasets that were investigated but determined to be insufficient for inclusion in the current version of the National Estuary Assessment included: historic habitat extent/habitat loss; storm/wastewater discharges; sediment contaminants; biogenic...
Screening Responses to Individual Stressors We used hierarchical models to develop the northern Gulf of Mexico Regional Estuary Assessment. This approach is a compromise between pooling all data from each estuary together into a single model, and modeling each estuary separately. For hierarchical models, the intercept and/or slope parameters can vary among different groups in the model. The assessment defines groups as estuaries and states (FL, AL, MS, LA, and TX). “Random effects” in hierarchical models account for group-level differences that are not accounted for by the available predictor variables. For example, some variation between estuaries is due to different morphologies and physical features that are...
Creating cumulative habitat condition scores To create the cumulative habitat condition index (CHCI) for streams of the conterminous United States, associations between multiple fish metrics and multiple human landscape factors were synthesized into a single number using the following scoring process. 5a. For each significant association between a fish metric and a human landscape factor, we evaluated the shape of the relationship to identify two key points. The “ threshold point” is the level of a landscape factor associated with a decrease in abundance of a particular fish metric (change in condition class between 5 and 4 in Figure 4), and the “plateau point” is the level of a landscape factor beyond which increasing...
Assembling data Data on stream fishes were provided for use in the 2015 assessment from many federal and state agencies and organizations from around the country. For a list of data providers, see Table 2. Due to the cooperation and support of multiple data providers, the 2015 assessment used stream fish assemblage data from 39,405 stream reaches as compared to 26,468 stream reaches in 2010 assessment. Data now reflects abundances of different fish species found in streams throughout the conterminous United States. Besides fish data, many different human (anthropogenic) landscape factors were assembled and used to characterize habitat condition. These factors include: urban and agricultural land use; intensity...
Assembling data Many different human landscape factors were assembled and used to characterize condition of Alaska stream fish habitat (Table 3). Factors include: urban and agricultural land uses; density of point sources of pollution in catchments; measures of stream network fragmentation, including densities of dams and culverts; density of infrastructure (road length, pipelines, etc.) and locations of mines. The availability of some landscape factors varied between southeast Alaska and greater Alaska (see Table 3). Some important threats to fish and fish habitats could not be incorporated into the assessment due to data availability limitations (one example includes forest harvest information across the entire...
After acquiring data, variables were attributed to a stream coverage for use in assessment following Wang et al. (2011). The Hawaii Fish Habitat Partnership (HFHP) stream layer (Tingley et al. in prep) is a modified version of the 1:24,000 National Hydrography Dataset that consists of 11,436 intermittent and perennial stream reaches across the five largest Hawaiian Islands (Hawai’i, Maui, Moloka’i, O’ahu, Kaua’i). The HFHP stream layer distinguishes stream reaches as sections of streams occurring between confluences, headwaters and confluences, and confluences and pour points, following the definition of stream reaches in the contiguous United States. Each reach has an associated local catchment (watershed), upstream...
Integrating Data into a Spatial Framework This part of the assessment examined 33 estuaries across the northern Gulf of Mexico (Florida, Alabama, Mississippi, Louisiana, and Texas). These estuaries were cataloged in a spatial framework used to quantify and analyze the effects of stressors on the estuaries. The framework was composed of a series of estuaries, linked to associated shoreline buffer units, Estuarine Drainage Areas (EDAs), and watershed basins. The spatial framework was developed based on estuary boundaries used for the National Estuary Assessment. A number of changes were made for the northern Gulf of Mexico region to improve the way estuaries and their coastal catchments were defined in the framework...
The approach for identifying disturbances to fish habitat was based on the assumption that greater intensities and types of human disturbances would most likely lead to more disturbed stream fish habitat (e.g., Danz et al. 2007, Esselman 2011). Twenty human landscape variables were identified for the Hawaii assessment (Table 2). We grouped variables into seven categories representing specific types of disturbances including: agricultural land use, urban land use, former plantation lands, point source pollution, density of ditches, stream fragmentation, and 303d listed streams. Disturbance sub-indices were then created for each category of disturbance variables for each spatial scale (i.e., local catchments, network...
To create the cumulative habitat condition index (CHCI) for streams of Hawaii, we first standardized and summed disturbance sub-indices described above within each of the three spatial scales to create three habitat condition indices (HCI) for each stream reach. Based on the assumption that urban land use can have excessive negative effects on stream habitat compared to other disturbances and at the request of the HFHP, the urban sub-index was upweighted in each spatial scale by a factor of 2. The CHCI was developed by summing HCI scores across spatial scales for each stream reach and standardizing from 0 (best condition) to 1 (worst condition). We followed methods applied for the contiguous U.S. and created condition...
Identifying disturbances to fish habitat The first step in identifying disturbances to fish habitat involved summarizing stream fish species data into a set of metrics that could be potential indicators of stream habitat condition. Examples of metrics include summaries of fish species by their feeding preferences, reproductive strategies, or tolerance to stressors. While many potential indicators were generated, an analytical process was used to identify a subset of metrics that were the most effective indicators of habitat condition in each of nine large ecoregions (Stoddard et al. 2008). Next, each of the key fish metrics was tested against each of the human landscape factors summarized in watersheds and stream...
|
