Next Article in Journal
Comparison of Hydraulic Travel Time and Attenuation Inversions, Thermal Tracer Tomography and Geostatistical Inversion for Aquifer Characterization: A Numerical Study
Previous Article in Journal
Investigation of Rain-Fed Horticulture Productivity in the Namangan Region, Uzbekistan
Previous Article in Special Issue
Using Multiscale Environmental and Spatial Analyses to Understand Natural and Anthropogenic Influence on Fish Communities in Four Canadian Rivers
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Water
Volume 15
Issue 13
10.3390/w15132400
water-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Assessing Impacts of Human Stressors on Stream Fish Habitats across the Mississippi River Basin

by
Jared A. Ross
1,*,
Dana M. Infante
1,
Arthur R. Cooper
1,
Joanna B. Whittier
2 and
Wesley M. Daniel
3
1
Department of Fisheries and Wildlife, College of Agriculture & Natural Resources, Michigan State University, East Lansing, MI 48824, USA
2
School of Natural Resources, University of Missouri, Columbia, MO 65211, USA
3
U.S. Geological Survey, Wetland and Aquatic Research Center, Gainesville, FL 32653, USA
*
Author to whom correspondence should be addressed.
Water 2023, 15(13), 2400; https://doi.org/10.3390/w15132400
Submission received: 7 April 2023 / Revised: 2 June 2023 / Accepted: 6 June 2023 / Published: 29 June 2023
(This article belongs to the Special Issue Disentangling Influences of Natural and Human Factors on Aquatic Ecosystems)
Browse Figures
Versions Notes

Abstract

:
Effective conservation of stream fishes and their habitats is complicated by the fact that human stressors alter the way in which natural factors such as stream size, catchment geology, and regional climate influence stream ecosystems. Consequently, efforts to assess the condition of stream fishes and their habitats must not only attempt to characterize the effects of human stressors but must account for the effects of natural influences as well. This study is an assessment of all stream fish habitats in the Mississippi River basin, USA. The basin supports over 400 stream fish species, drains a land area of 3.2 M km2, and includes a myriad of human stressors such as intensive agriculture, urbanization, nutrient loading, and habitat fragmentation by dams and road/stream crossings. To effectively characterize types and levels of human stressors specifically impacting the basin’s stream fish species, our assessment approach first accounted for the influence of natural landscape conditions on species abundances with multiple steps, including stratifying our analyses by region and stream size and quantitatively modeling the influences of natural factors on stream fishes. We next quantified individual fish species responses to explicit human stressors for different measures of land use, fragmentation, and water quality, including summaries of measures in local vs. catchment extents. Results showed that many species had negative threshold responses to human stressors and that impacts varied by species, by region, and by the spatial extents in which stressors were summarized. Our spatially explicit results indicated the degree of stream reach impairment for specific stressor categories, for individual species, and for entire assemblages, all of which are types of information that can aid decision makers in achieving specific conservation goals in the region.

1. Introduction

Human-induced stressors within watersheds have degraded stream habitats worldwide, contributing to declines in biodiversity and increased numbers of threatened fish species globally [1,2,3]. Differing types of landscape-scale stressors within catchments—including agriculture, urban landscapes, impervious surfaces, and dams and other barriers—can affect fishes by degrading stream habitats via multiple direct and indirect modes of habitat alteration. For example, buildings, roads, and other paved surfaces in urban environments indirectly affect stream habitats by altering flow and thermal regimes [4]. Similarly, increased nutrient loading to streams can occur from widespread application of fertilizers on agricultural lands or through other sources of pollution from urban landscapes [5,6]. Further, receiving waterbodies and coastal areas downstream of catchments with heavy nutrient loading are increasingly becoming eutrophic, leading to degraded habitat and altered biological communities [7,8]. Widespread construction of road culverts and dams has substantially decreased both lateral and longitudinal connectivity throughout stream networks, affecting both migratory and non-migratory fishes by limiting access to the habitats required to complete their life cycles [9,10,11]. Collectively, these varied stressors can degrade stream habitat conditions and threaten fishes, and understanding and addressing these threats are critical for effectively protecting, conserving, and restoring stream habitats and the fishes they support.
Accurate fish habitat assessments attempting to account for the effects of human stressors are challenging to conduct in part because a multitude of natural landscape factors also influence numbers and types of stream fish species and their required habitats. This challenge can be exacerbated when assessments are conducted over large regions, where variation in natural factors can be more pronounced than within smaller regions. For example, climatic characteristics, stream size, differences in geology, and network position are widely accepted landscape-scale filters broadly driving stream fish occurrence across watersheds [12,13,14], whereas more localized factors such as stream reach gradient and discrete inputs of groundwater further restrict or enable specific types of stream fish communities to occur [15]. These relationships, in conjunction with the fact that human stressors often vary along similar gradients, highlight the importance of appropriately accounting for natural influences in ecological assessments [16].
The Mississippi River basin supports a diverse assemblage of stream fish species with exceptionally high total richness and great variation within the basin [17], largely because of the notable variability in natural landscape factors occurring throughout the region. The Rocky and Appalachian Mountain ranges bound the basin from east to west, with climates varying from humid to arid depending on location. Glacial ice sheets once blanketed northern portions of the basin, contributing to differences in geological features and subsequent amounts of groundwater discharge to streams. Species present in the basin include ecologically noteworthy species of concern such as Lake Sturgeon (Acipenser fulvescens), Shovelnose Sturgeon (Scaphirhynchus platorynchus), Paddlefish (Polyodon spathula), and Blue Sucker (Cycleptus elongatus) (all of which are migratory), [18,19], as well as recreationally and commercially targeted species such as bass (Micropterus spp.), crappie (Pomoxis spp.), and catfish (Ictalurus spp.) [20,21]. Human activities across the basin have also led to changes in stream habitats and consequently the fish communities they support. For example, less diverse communities and a higher proportion of tolerant stream fishes have been observed in lower Mississippi River habitats impacted by highly urban land uses [22], whereas agricultural inputs of nutrients and sediments have led to fish species loss and fish kills in highly polluted waters [5]. Nutrient enrichment caused largely by agricultural and urban land uses across the watershed also contribute to a recurring hypoxic zone in the Gulf of Mexico where many aquatic species cannot persist, and nutrient loading of both nitrogen and phosphorus has remained steady [23]. Dams and road/stream crossings additionally fragment stream systems throughout the basin, thus blocking up- and downstream movement of migratory fishes, including many threatened and endangered species and those of greatest conservation need [19,24].
Given the diversity of fish and their habitats throughout the Mississippi River basin, as well as the multitude of human stressors within stream catchments, understanding the specific levels of anthropogenic activities within catchments and how those stressors degrade stream habitats and fish communities is critical for developing specific management actions to protect, enhance, or restore stream conditions. However, capturing the true impact of human stressors requires that the influence of important natural factors driving stream fish communities must also be accounted for. To meet these needs, we first account for the influence of important natural drivers of stream fish communities using multiple steps. With natural influences considered, we next quantify stream fish species responses to a variety of anthropogenic landscape-scale stressors including measures of urbanization, agriculture, stream fragmentation by large dams and roads, and nutrient inputs from a variety of sources. Achieving these objectives provides a dynamic range of information that can be tailored to help address specific management questions. To highlight the utility of this approach, we also provide an example application of the results focused on total nitrogen loading, given the management interest in reducing nitrogen loading to Mississippi River basin streams and ultimately the Gulf of Mexico.

2. Materials and Methods

2.1. Study Area

The Mississippi River has the largest watershed in North America and the fourth largest watershed in the world at more than 3.2 million km2. The basin drains all or portions of 31 U.S. states, 2 Canadian provinces, and 9 aggregated level III ecoregions [25]. Ecoregions are a compilation of areas with distinctive landform and climate characteristics, and natural habitats within an ecoregion are thought to respond similarly to human stressors as well as management actions [25]. The mainstem Mississippi River is formed by several large tributaries including the Missouri River, Illinois River, Ohio River, Red River, and Arkansas River. The region is highly diverse in natural landscape characteristics with generally increasing precipitation from west to east and increasing temperature from north to south. Natural land cover includes shrub and grasslands in the west and central portions of the basin, while forested landscapes dominate in the east.
Anthropogenic activities are also prevalent in the basin. Large, urbanized areas include the cities of Denver, Colorado, Minneapolis, Minnesota, Chicago, Illinois, Indianapolis, Indiana, Pittsburgh, Pennsylvania, and St. Louis, Missouri, among others. Agricultural land uses such as cultivated crops, pasture, and livestock farming are widespread throughout the watershed and dominate the landscape all through central portions of the basin. The hydrology and connectivity of the Mississippi River are highly altered by roads and dams, as more than 1.15 million road/stream crossings [26] and more than 26,000 large dams (i.e., ≥1.8 m high) occur throughout the basin [10,27]. The number and locations of smaller dams are generally unknown across the entire basin but suspected to be in the millions [28,29].

2.2. Spatial Units and Landscape Information

We used the National Hydrography Dataset Plus Version 2 (NHDPlusV2, U.S. Geological Survey and U.S. EPA, Washington, DC, USA) for this study, which is a 1:100,000 scale dataset representing all streams in the conterminous U.S. [30]. The finest spatial unit within the NHDPlusV2 is a stream reach, which is generally defined as a section of stream spanning headwaters to confluences, confluences to confluences, or confluences to river mouths (Figure 1).
Each stream reach also has a local catchment representing the land area draining directly to the downstream end of that reach. Furthermore, we created local buffers extending 90 m on each side of the stream to broadly characterize riparian influences on stream habitat. These units allow for summarization of landscape characteristics (e.g., land use/land cover) within local catchments and buffers, which can then be aggregated to summarize upstream network catchments and buffers representing landscape factors throughout the entire network [31], providing information in four spatial extents for analysis (Figure 1).
Six landscape factors were used to account for natural variation in analyses (Table S1). We calculated network catchment area by aggregating all upstream local catchments and summing their local catchment areas [31]. Reach elevation and reach slope (i.e., gradient) were included with the NHDPlusV2. Average annual precipitation and average annual air temperature data were acquired and summarized from the Parameter-elevation Regressions on Independent Slopes Model (PRISM) spanning 1995–2015 [32]. Air temperature was summarized within local catchments, while precipitation was summarized within network catchments. We used an estimate of the percent of stream baseflow contributed from groundwater [33] and summarized this information within network catchments.
A total of 38 anthropogenic landscape-scale stressor variables were used in analyses and occurred in three categories: land use, water quality, and fragmentation (Table S1). For land use, the percentages of impervious surfaces, three classes of urban land uses (i.e., open + low intensity, medium intensity, and high intensity urban), cultivated crops, and pasture/hay classes from the 2011 National Land Cover Dataset (NLCD 2011) [34] were summarized in local and network catchments as well as local and network buffers for the entire Mississippi River basin. Each land-use stressor variable summarized within a single spatial extent was considered a unique stressor (e.g., % of local catchment impervious surface vs. % of network catchment impervious surface). Water quality variables were only summarized within network catchments and included data from SPARROW [35] to characterize nitrogen and phosphorus loads from various sources (e.g., farm, non-farm, manure from confined operations), while estimated nitrogen and phosphorus fertilizer use was also summarized [36]. Water quality stressors were summarized as yields (kg/km2/yr) by aggregating the total upstream loads and dividing this by the network catchment area [31]. Fragmentation variables included three large dam metrics and two road/stream crossing metrics; upstream and downstream mainstem dam densities (#/100 km) and upstream degree of regulation, defined as the percentage of estimated annual stream flow volume stored in all upstream reservoirs [10]. TIGER 2011 roads were intersected with streams from the NHDPlusV2 and summarized as a density by stream length at the local reach and upstream network extents to represent potential sources of fragmentation due to road/stream crossings.

2.3. Stream Fish Data

Fish assemblage data from streams were collected between 1990 and 2015, representing relatively current conditions. Data acquisition was leveraged from a previous project [37] and originated from various museum, university, and state and federal agency collections. We only considered data that were collected using single-pass electrofishing methods and targeted the entire fish community as opposed to targeted sampling of game fish, for example. Data additionally had to include spatial coordinates associated with each sample site that was used to link samples to the NHDPlusV2 (Figure 2).
Geographic accuracy of sites was verified manually when the recorded stream names did not match NHDPlus stream names and when points fell further than 50 m from a stream reach. If a sample site could not be verified, it was removed from the broader dataset. Relative abundances for fish species were calculated by dividing the number of individuals captured for each species by the total number of individuals sampled.

2.4. Accounting for Natural Variation

Prior to evaluating human stressor influences on stream fish species, we took several steps to account for natural influences. These included stratifying analyses by aggregated level III ecoregions (Figure 2; [25]), as fishes are known to respond differently to similar stressors in different regions because of broad-scale climatic and physiographic differences [38,39]. Within each ecoregion, we also further stratified sample sites into creeks (drainage area < 100 km2) and rivers (drainage area > 100 km2) to account for influences of stream size on fish species presence and abundance (Figure 2; [40,41,42]). As an additional step, we further accounted for the influences of natural factors on stream fish abundances by developing boosted regression tree models (BRT, Figure 2; [43]) for each fish species with at least 40 total sites. We attempted to model species using sites having the least disturbed conditions from a set of natural landscape variables and retained models that adequately explained variability in the fish species response (i.e., those that converge to a solution). The remaining models were used to predict the expected abundance to all sites [39], and the residuals (i.e., difference between observed and expected) were rescaled from 0 to 100 and used as the response variable in subsequent analyses [16]. We identified the least disturbed sites as those that had a very low or low risk of habitat degradation based on scores from the 2015 national assessment of stream fish habitat conducted for the National Fish Habitat Partnership (NFHP, [37]). The six natural landscape variables used in the models included network catchment area, contribution of groundwater to baseflow, average annual precipitation and air temperature, stream reach elevation, and stream reach slope (described above). We developed models for numerous species in each ecoregion in each size strata, except for the Northern Plains (NPL) ecoregion, where data from creeks and rivers were combined for analysis. We initially set tree complexity at 5, learning rate at 0.01, and a bag fraction at 0.5. If necessary, we iteratively reduced the learning rate in half in order to achieve a final BRT with a minimum of 2000 trees. The BRT and all subsequent analyses were performed in R.

2.5. Identifying Multiple Breakpoint Responses of Fishes to Human Stressors

We identified multiple breakpoints in evaluating negative responses (i.e., declining abundances) of stream fishes to increasing levels of human stressors. The first breakpoint is an ecological threshold, which occurs when a substantial change in a biological response variable is observed after a comparatively minor increase in an environmental stressor [1,44]. Threshold responses, as compared to linear responses, are assumed to be indicative of a sensitivity of the biological response variable to the stressor under consideration [1]. To identify thresholds in species abundances to increasing levels of human stressors, we used two complimentary analytical techniques ([41]; Figure 2). We conducted both piecewise linear regression analyses [45] in conjunction with a change-point analysis using indicator species analysis [41,46]. Threshold responses of individual species were identified from individual stressors for every ecoregion/stream size stratum, except for the NPL region where creeks and rivers were combined because of sample size limitations. We considered a threshold to be significant if it had a p-value from piecewise linear regression ≤0.10 and a p-value from the change-point with indicator species analysis ≤0.05. As statistical significance is directly related to sample size, we used these criteria to ensure we captured trends in species having fewer sample sites (i.e., rarer species). Scatterplots of threshold results were visually investigated to ensure a negative response, and threshold values had to overlap within 10% of their range along the human stressor data gradient. The second breakpoint, described as the plateau breakpoint, was identified visually as the point where the fish response no longer decreases with increasing human stressors. Resulting threshold values from the piecewise linear regression analysis were retained and the plateau breakpoints were averaged across all species for each stressor to be used in subsequent scoring.

2.6. Risk of Habitat Degradation Indices for Land Use, Fragmentation, and Water Quality

The threshold values identified above were used to score stream reaches into one of five possible risk of habitat degradation classes separately for human land use, fragmentation, and water quality (Figure 2). These classes include very low, low, moderate, high, and very high risk of habitat degradation. All reaches with a landscape stressor value smaller than the lowest threshold response for that stressor were scored very low risk of habitat degradation for that stressor. Reaches with a stressor value between the lowest and highest fish species threshold responses were scored as having a low risk of habitat degradation. Reaches having a stressor greater than the average plateau point value were scored at a very high risk of habitat degradation. The remaining stressor values were divided into two equal parts, with the lower half scored as moderate risk of habitat degradation while the upper half scored at high risk. The cumulative index depicting the risk of habitat degradation for any particular reach was identified as the lowest score across all stressors for the land-use, fragmentation, and water quality indices. Each cumulative index was summarized and mapped individually as well as in combination to support management decision making.

2.7. Identifying Areas beyond Ecological Breakpoints: An Example Application Focused on Total Nitrogen Yield

Managers in the region are concerned with not only human impacts to stream habitats, but also with impacts to the condition of their receiving waterbodies. Because excess nitrogen has been found to be a primary stressor on stream fish communities in agriculturally dominated catchments [47] and because of the overwhelming influence of excess nitrogen delivered through the Mississippi River on the hypoxic zone in the Gulf of Mexico [7], we integrated a subset of results into an example application focused on total nitrogen yield to highlight the utility of this approach and its flexibility to help address specific management challenges. For those catchments with a water quality score driven by total nitrogen yield, we subtracted the total nitrogen yield breakpoint values from each stream’s current total nitrogen yield to determine areas across the landscape where stream fishes are currently limited by total nitrogen. Finally, we mapped both the locations and threshold exceedance amounts to help prioritize where management actions intended to reduce nitrogen input to streams may have the biggest impact for improving stream fish habitat.

3. Results

3.1. Fish Species Response to Anthropogenic Stressors

In total, we had abundance data for 441 individual stream fish species (Table S2) and documented 3010 significant negative responses of these species to human stressors (Table S3). Of these, 173 species (39%) had a significant negative response to at least one land-use, fragmentation, or water quality variable (Table 1).
Greater than 35% of species had negative threshold responses to anthropogenic stressors in the NAP (38%) and UMW (37%), whereas the XER and SAP ecoregions had 34% and 33% of species respond negatively. The NPL, CPL, SPL, TPL, and WMT ecoregions all had fewer than 30% of species respond negatively (ranging from 9% to 26%). Specific land-use stressors had a negative impact on 36% of species basin-wide, whereas regionally, the NAP had the highest percentage of species with a negative response at 34%. The TPL, WMT, SAP, UMW, and XER ecoregions all had >20% of species respond negatively to land-use stressors (ranging from 24% to 29%). The remaining ecoregions had fewer than 20% of species with threshold responses. Fewer than 30% of species showed negative responses to fragmentation variables basin-wide and regionally. The XER, TPL, UMW, and NAP ecoregions all had greater than 20% of species showing negative responses (ranging from 21–28%). More than 32% of species responded negatively to water quality stressors basin-wide; however, the NAP was the only ecoregion with greater than 25% of species to negatively respond.

3.2. Influence of Spatial Extent on Threshold Responses to Land-Use Variables

The impact of variables summarized within network catchments was greatest within all stressor types, with agricultural variables accounting for 41%, impervious surfaces at 32%, and urban land use comprising 38% of urban thresholds. Similarly, the impact of stressors summarized within network buffers was consistently greater than that of stressor variables summarized at the local catchment and local buffer extents (Figure 3).
Agricultural and impervious surface stressors within network buffers each accounted for 31% of thresholds for each stressor type, while urban variables summarized for network buffers accounted for 29% of all urban thresholds. Agricultural variables summarized at local catchments and local buffers comprised 14% and 15% of all agricultural thresholds. Similarly, thresholds from impervious surfaces within local catchment and buffer extents each accounted for 19% of impervious thresholds, whereas the impact of urban land use was slightly greater in local catchments and slightly lower for local buffers, accounting for 20% and 13% of urban thresholds, respectively.

3.3. Human Land-Use Index

Within the study region, 13% of total reach length was classified as very low risk of habitat degradation, including large areas of the NPL ecoregion in the northwestern portion of the basin (Figure 4).
The remaining human land-use conditions were generally evenly distributed across the low (20%), moderate (22%), high (20%), and very high (25%) risk of habitat degradation classes. Ecoregions within the basin exhibited variability in habitat condition, resulting in differences in percentages of streams assigned to habitat degradation classes across the study region. For instance, the TPL and UMW ecoregions had high percentages of stream length classified as high or very high risk of habitat degradation class at 69% and 61%, respectively, whereas the percentage of stream length classified as very low or low risk ranged from 56% to 92% among the NPL, WMT, and XER ecoregions. The TPL had the lowest percentage of stream length classified as very low or low risk of habitat degradation at ~7%.

3.4. Fragmentation Index

The habitat condition of streams based on the fragmentation index for the Mississippi River basin resulted in most of the stream length being classified as either very low (25%) or low (58%) risk of habitat degradation from dams and road/stream crossings. A relatively small percentage of stream length was classified as moderate (13%), high (2%), or very high (2%) risk of degradation (Figure 5).
Fragmentation index results by ecoregion resulted in the NAP having the greatest percentage of total stream length in the high or very high risk classes at 10%, while XER had the least stream length in the very low or low risk classes at 56%.

3.5. Water Quality Index

Water quality stressor impacts on stream habitat condition varied throughout the entire Mississippi River basin, with the percentage of stream length classified as very low or low risk totaling 43% predominately occurring in the northwestern portion of the basin within the NPL ecoregion (Figure 6).
Approximately 31% of stream length in the basin was classified as high or very high risk of degradation, with these streams being frequently located across the TPL ecoregion including major sections in the states of Iowa, Nebraska, and southern Minnesota. Additionally, the habitat condition of much of the mainstem of the lower Mississippi River was at high and very high risk of degradation. Across individual ecoregions, the CPL, SPL, and TPL ecoregions had the greatest percentages of stream length in the high or very high risk of habitat degradation classes, at 37%, 42%, and 52%, respectively. No other ecoregions had >30% combined stream length in the high and very high-risk classes. Further, the TPL ecoregion only had 2% of the total stream length considered to be at very low or low risk of degradation, with the SPL ecoregion having the second lowest percentage of stream length in these categories at 17%. The greatest percentage of stream length in the very low or low risk of habitat degradation classes occurred in the XER ecoregion at 87%, while the NPL (75%) and SAP (67%) ecoregions had the next greatest stream lengths in the two lowest risk classes.

3.6. Integrating Land-Use, Fragmentation, and Water Quality Indices

By combining and mapping results using all three risk indices, streams with low or very low risk of habitat degradation from all stressors could be identified. For instance, headwaters of the Mississippi River basin meeting these criteria primarily occurred in the northwestern portion of the basin, with smaller areas occurring in the upper Midwest, as well as in the southern and southeastern portions of the basin (Figure 7).
Streams with a low risk of habitat degradation for land uses and fragmentation but a high risk for nutrient loading were spread across the basin with the highest density occurring in the corn belt states of Illinois, Indiana, Iowa, Kansas, and Nebraska. Other portions of the basin meeting these criteria included southern Oklahoma as well as southwestern Montana (Figure 8).

3.7. Identifying Areas That Exceed Fish Breakpoint Responses: Total Nitrogen Yield as an Example Application

We found total nitrogen yield to be a significant anthropogenic stressor across the Mississippi River basin, with 43% of streams currently characterized as having surpassed thresholds and negatively impacting 62 stream fishes (Table 2 and Table S3 and Figure 9).
Regionally, the SPL ecoregion had the greatest percentage of stream catchments exceeding the total nitrogen yield threshold at 60%, followed by the SAP (58%) and TPL ecoregions (55%; Table 2). A relatively lower percentage of streams exceeded total nitrogen yield thresholds in the UMW, CPL, and NAP ecoregions (ranging from 35% to 43% of streams), whereas the percentages in the WMT and XER ecoregions were much lower (15% and 4% of streams, respectively). Streams in the NPL ecoregion were largely characterized as not currently exceeding total nitrogen thresholds (0.55% of streams; Table 2); however, the average exceedance value of the few streams in these cases was by far the greatest across all ecoregions at 8466 kg/km2/yr. Of the remaining streams currently exceeding thresholds for total nitrogen yield, the TPL ecoregion contained those with the next greatest average exceedance value of 5335 kg/km2/yr, followed by the SPL (3101 kg/km2/yr), UMW (2867 kg/km2/yr), CPL (2303 kg/km2/yr), NAP (1653 kg/km2/yr), and SAP (1470 kg/km2/yr) ecoregions. The average threshold exceedance values of streams in the WMT and XER ecoregions were an order of magnitude lower at 513 and 230 kg/km2/yr, respectively.

4. Discussion

Anthropogenic landscape stressors across the Mississippi River basin are prevalent and influence stream fish communities in a myriad of ways. Here we illustrated the utility of our assessment approach to help understand how specific types and levels of anthropogenic activities within catchments and buffers negatively impact fish communities and their habitats across large spatial extents by quantifying specific stream fish responses to individual stressors. Our results showed that a substantial proportion of the fish species studied respond to the human disturbances assessed, and that these impacts vary regionally among stressor variable types. Quantifying individual species responses to specific human stressors provides the opportunity to identify those that are most impactful to individual species and fish communities of management interest. Additionally, species-specific results can be tailored into a variety of metrics, indices, and products most relevant for supporting decisions that managers are tasked with making.

4.1. Species-Specific Threshold Responses to Individual Stressors at Multiple Spatial Extents

Stream habitat assessments across large regions have generally quantified human impacts on aquatic communities as represented by composite variables such as multi-metric indices or metrics characterizing groups of organisms by their tolerance to disturbances, feeding preferences, or reproductive guilds [37,39,48,49,50]. Our approach allows for similarly meaningful assessments to be made with the added benefit of having the capability to simultaneously assess impacts to individual species. This contrasts with other landscape-scale habitat assessments that rely on responses of groups of species, which is an approach that does not fully provide managers the flexibility to tease apart impacts to the specific species that they are most interested in [37,39]. Our bottom-up approach allows for identification of stressor variables most impactful to individual species while retaining the ability to conduct community-wide assessments.

4.2. Quantifying Stressor Impacts on Stream Fish Species across Multiple Spatial Extents

Regardless of the type of land-use variable tested, a far greater percentage of threshold responses among fish species occurred at the network-based spatial extents when compared to local-based summaries of stressor variables. Local-scale predictor variables have been found to have relatively greater importance in explaining fish assemblages in generally undisturbed areas when compared to network-wide variables, given their direct effects on stream fish species [15]. However, as human disturbance becomes more prevalent within catchments, thus directly altering important hydrologic characteristics (e.g., groundwater input, flow extremes), water temperatures, water quality conditions, instream cover, and other factors, network-based summaries of landscape characteristics become increasingly meaningful [15,51,52]. This further supports the fundamental need to account for natural influences prior to assessing human stressor effects on stream fishes.

4.3. Characterizing Current Stream Habitat Condition by Grouping Stressor Influences

Degradation of stream habitat occurs differentially depending on many factors such as the type, location, and intensity of stressors occurring throughout catchments [38,39]. This study illustrated the utility of assessing their impacts individually as well as within meaningful groupings of variables. For instance, a manager interested in prioritizing areas to implement agricultural practices to reduce nutrient loading will have a greater likelihood of success by utilizing parameters developed specifically from water quality variables [53] than one combined with measures of stream fragmentation, for example. In contrast, levels of nutrients in streams may be less impactful to migratory species requiring access to specific habitats to complete their life cycle [54]. Further, combining indices from various groupings of stressors to determine where catchments are currently at low risk of degradation from land use, water quality, and fragmentation can be a beginning point for identifying areas that would benefit most from protection, while catchments with a low risk of degradation from land use and fragmentation but a high risk from water quality may warrant investments in providing incentives to promote practices that reduce runoff and nutrient loss [55].

4.4. Identifying Limiting Stressors to Maximize Conservation Opportunities

Efforts to reduce the impacts of excess nutrients delivered to the Gulf of Mexico through the Mississippi River have been a major focus of research, best management practice implementation, and state and federal policy for decades [56], and the benefits of targeting nutrient reduction for improving downstream receiving waterbodies also directly translate to tributaries delivering such nutrients [53]. However, prioritizing specific locations where implementation of nutrient reduction strategies will have the most impact across the Mississippi River basin is challenging and varies with individual state priorities [57]. Identifying where nitrogen yields are most limiting to stream fishes and the extent to which catchments are having an impact can be a useful consideration when identifying priority watersheds that require the least amount of reduction to achieve meaningful improvement while additionally helping to assess whether realistic benefits can be achieved in areas targeted for nutrient reduction efforts. Further, the ability to assess the spatial scale at which certain types of stressors most impact fishes allows conservation actions and priorities to be more effectively implemented, a benefit that is not possible with metric-based assessments [37].

4.5. Limitations and Next Steps

Our study provides a wealth of information for stream conservation and management efforts across the Mississippi River basin by first disentangling natural influences on stream fish assemblages before quantifying species-specific responses to a suite of human landscape stressors. These results can help guide decisions being made as they relate to mitigating impacts from differing types of stressors on stream habitat, while allowing the flexibility to aggregate results in meaningful ways to address explicit management goals. However, one limitation of using a landscape approach is the inability to validate predictions effectively at a site-specific spatial scale, in part owing to heterogeneous landscape conditions across large regions as well as local factors that could amplify or mitigate disturbances. Despite these limitations, this approach provides a relative estimate of the condition for all sites over large regions, and the utility to management is that the information can be used to identify locations of interest and further supplement our estimates with local information to better understand conditions. Previous studies have used similar landscape approaches to provide resource managers with additional information to help inform decisions [58,59]. Additional useful applications could include compiling results for a group of climate-sensitive species that might inform decision makers tasked with addressing impacts of a changing climate, while another example could include focusing on results specific to game fish to help state natural resource agencies allocate resources to improve recreational angling opportunities. Further, incorporating newly developed, large-scale datasets (e.g., subsurface tile drainage locations, forest disturbance frequencies) as they become available can add more utility to such an approach [60,61], while extending the assessment nationally with updated datasets would provide additional yet complimentary information in conjunction with existing national stream habitat assessment efforts.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/w15132400/s1, Table S1: List of natural and anthropogenic variables used in analyses; Table S2: List of Mississippi River basin stream fish species with abundance data available for use in analyses; Table S3: List of initial and plateau breakpoints identified for Mississippi River basin stream fish species organized by ecoregion and stream size.

Author Contributions

Conceptualization, J.A.R., D.M.I. and W.M.D.; methodology, J.A.R., W.M.D., D.M.I., A.R.C. and J.B.W.; formal analysis, J.A.R. and A.R.C.; data curation, J.A.R.; writing—original draft preparation, J.A.R.; writing—review and editing, J.A.R., A.R.C., D.M.I., J.B.W. and W.M.D.; visualization, J.A.R. and A.R.C.; supervision, D.M.I.; project administration, D.M.I.; funding acquisition, D.M.I. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the U.S. Fish and Wildlife Service, Eastern Tallgrass Prairie and Big Rivers Landscape Conservation Cooperative (F17AC00981), as well as Michigan State University AgBioResearch. We also leveraged support from the U.S. Geological Survey Northeast Climate Adaptation Science Center (G12AC20418), U.S. Fish and Wildlife Service (F13AC00565), and U.S. Geological Survey (G17AC00185).

Data Availability Statement

Data presented in this study are either referenced and available online or available upon reasonable request from the corresponding author. Locations of certain species used in the study are considered sensitive by the agencies providing these data and thus are not publicly available.

Acknowledgments

We thank Kyle Herreman for assistance with GIS and data management, Hao Yu for providing statistical advice, Jana Stewart and Brad Potter for data insights and input on preliminary results, and Gwen White for assistance initiating this project. We would also like to thank Amy Benson and four anonymous reviewers for comments and suggestions that improved upon our original submission. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Allan, J.D. Landscapes and riverscapes: The influence of land use on stream ecosystems. Annu. Rev. Ecol. Evol. Syst. 2004, 35, 257–284. [Google Scholar] [CrossRef] [Green Version]
  2. Dudgeon, D.; Arthington, A.H.; Gessner, M.O.; Kawabata, Z.-I.; Knowler, D.J.; Leveque, C.; Naiman, R.J.; Prieur-Richard, A.-H.; Soto, D.; Stiassny, M.L.J.; et al. Freshwater biodiversity: Importance, threats, status and conservation challenges. Biol. Rev. 2006, 81, 163–182. [Google Scholar] [CrossRef] [PubMed]
  3. Jelks, H.L.; Walsh, S.J.; Burkhead, N.M.; Contreras-Balderas, S.; Diaz-Pardo, E.; Hendrickson, D.A.; Lyons, J.; Mandrak, N.E.; McCormick, F.; Nelson, J.S.; et al. Conservation status of imperiled North American freshwater and diadromous fishes. Fisheries 2008, 33, 372–407. [Google Scholar] [CrossRef]
  4. Paul, M.J.; Meyer, J.L. Streams in the urban landscape. Annu. Rev. Ecol. Syst. 2001, 32, 333–365. [Google Scholar] [CrossRef]
  5. Carpenter, S.R.; Caraco, N.F.; Correll, D.L.; Howarth, R.W.; Sharpley, A.N.; Smith, V.H. Nonpoint pollution of surface waters with phosphorus and nitrogen. Ecol. Appl. 1998, 8, 559–568. [Google Scholar] [CrossRef]
  6. Mitsch, W.J.; Day, J.W.; Gilliam, J.W.; Groffman, P.M.; Hey, D.L.; Randall, G.W.; Wang, N.M. Reducing nitrogen loading to the Gulf of Mexico from the Mississippi River Basin: Strategies to counter a persistent ecological problem. Bioscience 2001, 51, 373–388. [Google Scholar] [CrossRef]
  7. Scavia, D.; Bricker, S.B. Coastal eutrophication assessment in the United States. Biogeochemistry 2006, 79, 187–208. [Google Scholar] [CrossRef] [Green Version]
  8. Watson, S.B.; Miller, C.; Arhonditsis, G.; Boyer, G.L.; Carmichael, W.; Charlton, M.N.; Confesor, R.; Depew, D.C.; Hook, T.O.; Ludsin, S.A.; et al. The re-eutrophication of Lake Erie: Harmful algal blooms and hypoxia. Harmful Algae 2016, 56, 44–66. [Google Scholar] [CrossRef]
  9. Cooke, S.J.; Paukert, C.; Hogan, Z. Endangered river fish: Factors hindering conservation and restoration. Endanger. Species Res. 2012, 17, 179–191. [Google Scholar] [CrossRef] [Green Version]
  10. Cooper, A.R.; Infante, D.M.; Daniel, W.M.; Wehrly, K.E.; Wang, L.; Brenden, T.O. Assessment of dam effects on streams and fish assemblages of the conterminous USA. Sci. Total Environ. 2017, 586, 879–889. [Google Scholar] [CrossRef]
  11. Warren, M.L.; Pardew, M.G. Road crossings as barriers to small-stream fish movement. Trans. Am. Fish. Soc. 1998, 127, 637–644. [Google Scholar] [CrossRef]
  12. Poff, N.L. Landscape filters and species traits: Towards mechanistic understanding and prediction in stream ecology. J. N. Am. Benthol. Soc. 1997, 16, 391–409. [Google Scholar] [CrossRef]
  13. Zorn, T.G.; Seelbach, P.W.; Wiley, M.J. Distributions of stream fishes and their relationship to stream size and hydrology in Michigan’s lower peninsula. Trans. Am. Fish. Soc. 2002, 131, 70–85. [Google Scholar] [CrossRef]
  14. Yu, H.; Cooper, A.R.; Infante, D.M. Improving species distribution model predictive accuracy using species abundance: Application with boosted regression trees. Ecol. Model. 2020, 432, 109202. [Google Scholar] [CrossRef]
  15. Wang, L.Z.; Lyons, J.; Rasmussen, P.; Seelbach, P.; Simon, T.; Wiley, M.; Kanehl, P.; Baker, E.; Niemela, S.; Stewart, P.M. Watershed, reach, and riparian influences on stream fish assemblages in the Northern Lakes and Forest Ecoregion, USA. Can. J. Fish. Aquat. Sci. 2003, 60, 491–505. [Google Scholar] [CrossRef]
  16. Stoddard, J.L.; Herlihy, A.T.; Peck, D.V.; Hughes, R.M.; Whittier, T.R.; Tarquinio, E. A process for creating multimetric indices for large-scale aquatic surveys. J. N. Am. Benthol. Soc. 2008, 27, 878–891. [Google Scholar] [CrossRef]
  17. Bertuzzo, E.; Muneepeerakul, R.; Lynch, H.J.; Fagan, W.F.; Rodriguez-Iturbe, I.; Rinaldo, A. On the geographic range of freshwater fish in river basins. Water Resour. Res. 2009, 45, 1–11. [Google Scholar] [CrossRef]
  18. Pracheil, B.M.; Lyons, J.; Hamann, E.J.; Short, P.H.; McIntyre, P.B. Lifelong population connectivity between large rivers and their tributaries: A case study of shovelnose sturgeon from the Mississippi and Wisconsin rivers. Ecol. Freshw. Fish 2019, 28, 20–32. [Google Scholar] [CrossRef] [Green Version]
  19. Cooper, A.R.; Infante, D.M.; O’Hanley, J.R.; Yu, H.; Neeson, T.M.; Brumm, K.J. Prioritizing native migratory fish passage restoration while limiting the spread of invasive species: A case study in the Upper Mississippi River. Sci. Total Environ. 2021, 791, 148317. [Google Scholar] [CrossRef]
  20. Garvey, J.; Ickes, B.; Zigler, S. Challenges in merging fisheries research and management: The Upper Mississippi River experience. Hydrobiologia 2010, 640, 125–144. [Google Scholar] [CrossRef]
  21. Alford, J.B.; Walker, M.R. Managing the flood pulse for optimal fisheries production in the Atchafalaya River basin, Louisiana (USA). River Res. Appl. 2013, 29, 279–296. [Google Scholar] [CrossRef]
  22. Chen, Y.; Herzog, K.; Shrestha, S.; Grigas, D.; Farrelly, J.; Laskodi, C.; Skoog, M. Urban land use, water quality, and biological conditions in the Lower Mississippi River basin bayous. Fisheries 2015, 40, 334–335. [Google Scholar] [CrossRef]
  23. Oelsner, G.P.; Stets, E.G. Recent trends in nutrient and sediment loading to coastal areas of the conterminous US: Insights and global context. Sci. Total Environ. 2019, 654, 1225–1240. [Google Scholar] [CrossRef] [PubMed]
  24. Hoover, J.J.; Bailey, P.; Januchowski-Hartley, S.R.; Lyons, J.; Pracheil, B.; Zigler, S. Anthropogenic obstructions to paddlefish movement and migration. In Paddlefish: Ecological, Aquacultural, and Regulatory Challenges of Managing a Global Resource; Schooley, J.D., Scarnecchia, D.L., Eds.; American Fisheries Society: Bethesda, MD, USA, 2019; Symposium 88; pp. 67–101. [Google Scholar]
  25. U.S. Environmental Protection Agency. Wadeable Streams Assessment: A Collaborative Survey of the Nation’s Streams; EPA 841-B-06-002; Office of Research and Development and Office of Water: Washington, DC, USA, 2006.
  26. U.S. Geological Survey. Anthropogenic Disturbances and Natural Variables in the Conterminous United States Linked to Catchments and Buffers of the National Hydrography Dataset Plus Version 2.1; U.S. Geological Survey: Reston, VA, USA, 2022. [CrossRef]
  27. U.S. Geological Survey. Dam Metrics Representing Stream Fragmentation and Flow Alteration for the Conterminous United States Linked to the NHDPLUSV1; U.S. Geological Survey: Reston, VA, USA, 2017. [CrossRef]
  28. L.R. Johnston Associates. A Status Report on the Nation’s Floodplain Management Activity: An Interim Report; TV-72105A; Federal Emergency Management Agency: Washington, DC, USA, 1989.
  29. Graff, W.L. Landscapes, commodities, and ecosystems: The relationship between policy and science for American rivers. In Sustaining Our Water Resources; Water Science and Technology Board, Ed.; National Academy Press: Washington, DC, USA, 1993; pp. 11–42. [Google Scholar]
  30. U.S. Environmental Protection Agency & U.S. Geological Survey National Hydrography Dataset Plus V2 (NHDPlusV2). Available online: https://www.epa.gov/waterdata/nhdplus-national-data (accessed on 24 August 2016).
  31. Tsang, Y.-P.; Wieferich, D.; Fung, K.; Infante, D.M.; Cooper, A.R. An approach for aggregating upstream catchment information to support research and management of fluvial systems across large landscapes. SpringerPlus 2014, 3, 589. [Google Scholar] [CrossRef] [Green Version]
  32. PRISM Climate Group, Oregon State University. Available online: https://prism.oregonstate.edu (accessed on 20 February 2017).
  33. Wolock, D.M. Base-Flow Index Grid for the Conterminous United States; No. 2003-263; U.S. Geological Survey: Lawrence, KS, USA, 2003.
  34. Homer, C.; Dewitz, J.; Yang, L.; Jin, S.; Danielson, P.; Xian, G.; Coulston, J.; Herold, N.; Wickham, J.; Megown, K. Completion of the 2011 national land cover database for the conterminous United States—Representing a decade of land cover change information. Photogramm. Eng. Remote Sens. 2015, 81, 345–354. [Google Scholar]
  35. Schwarz, G.E.; Hoos, A.B.; Alexander, R.B.; Smith, R.A. The SPARROW Surface Water-Quality Model—Theory, Application and User Documentation; 6–B3; U.S. Geological Survey: Reston, VA, USA, 2006.
  36. Stewart, J.S.; Schwarz, G.E.; Brakebill, J.W.; Preston, S.D. Catchment-Level Estimates of Nitrogen and Phosphorus Agricultural Use from Commercial Fertilizer Sales for the Conterminous United States, 2012; Scientific Investigations Report 2018-5145; U.S. Geological Survey: Reston, VA, USA, 2019.
  37. Crawford, S.; Whelan, G.; Infante, D.M.; Blackhart, K.; Daniel, W.M.; Fuller, P.L.; Birdsong, T.; Wieferich, D.J.; McClees-Funinan, R.; Stedman, S.M.; et al. Through a Fish’s Eye: The Status of Fish Habitats in the United States 2015. 2016. Available online: http://assessment.fishhabitat.org (accessed on 20 February 2017).
  38. Utz, R.M.; Hilderbrand, R.H.; Raesly, R.L. Regional differences in patterns of fish species loss with changing land use. Biol. Conserv. 2010, 143, 688–699. [Google Scholar] [CrossRef]
  39. Esselman, P.C.; Infante, D.M.; Wang, L.; Cooper, A.R.; Wieferich, D.; Tsang, Y.-P.; Thornbrugh, D.J.; Taylor, W.W. Regional fish community indicators of landscape disturbance to catchments of the conterminous United States. Ecol. Indic. 2013, 26, 163–173. [Google Scholar] [CrossRef]
  40. Wang, L.; Infante, D.; Esselman, P.; Cooper, A.; Wu, D.; Taylor, W.; Beard, D.; Whelan, G.; Ostroff, A. A Hierarchical Spatial Framework and Database for the National River Fish Habitat Condition Assessment. Fisheries 2011, 36, 436–449. [Google Scholar] [CrossRef] [Green Version]
  41. Daniel, W.M.; Infante, D.M.; Hughes, R.M.; Tsang, Y.-P.; Esselman, P.C.; Wieferich, D.; Herreman, K.; Cooper, A.R.; Wang, L.; Taylor, W.W. Characterizing coal and mineral mines as a regional source of stress to stream fish assemblages. Ecol. Indic. 2015, 50, 50–61. [Google Scholar] [CrossRef]
  42. Goldstein, R.M.; Meador, M.R. Comparisons of fish species traits from small streams to large rivers. Trans. Am. Fish. Soc. 2004, 133, 971–983. [Google Scholar] [CrossRef]
  43. Elith, J.; Leathwick, J.R.; Hastie, T. A working guide to boosted regression trees. J. Anim. Ecol. 2008, 77, 802–813. [Google Scholar] [CrossRef]
  44. Groffman, P.; Baron, J.; Blett, T.; Gold, A.; Goodman, I.; Gunderson, L.; Levinson, B.; Palmer, M.; Paerl, H.; Peterson, G.; et al. Ecological thresholds: The key to successful environmental management or an important concept with no practical application? Ecosystems 2006, 9, 1–13. [Google Scholar] [CrossRef]
  45. Muggeo, V.M.R. Segmented relationships in regression models with breakpoints/changepoints estimations. CRAN-R 0.2, 2012; Volume 9. [Google Scholar]
  46. Dufrene, M.; Legendre, P. Species assemblages and indicator species: The need for a flexible asymmetrical approach. Ecol. Monogr. 1997, 67, 345–366. [Google Scholar] [CrossRef]
  47. Meador, M.R.; Frey, J.W. Relative importance of water-quality stressors in predicting fish community responses in midwestern Streams. JAWRA 2018, 54, 708–723. [Google Scholar] [CrossRef]
  48. Fore, J.D.; Sowa, S.P.; Galat, D.L.; Annis, G.M.; Diamond, D.D.; Rewa, C. Riverine threat indices to assess watershed condition and identify primary management capacity of agriculture natural resource management agencies. Environ. Manag. 2014, 53, 567–582. [Google Scholar] [CrossRef] [PubMed]
  49. Hill, R.A.; Fox, E.W.; Leibowitz, S.G.; Olsen, A.R.; Thornbrugh, D.J.; Weber, M.H. Predictive mapping of the biotic condition of conterminous US rivers and streams. Ecol. Appl. 2017, 27, 2397–2415. [Google Scholar] [CrossRef] [PubMed]
  50. Schinegger, R.; Palt, M.; Segurado, P.; Schmutz, S. Untangling the effects of multiple human stressors and their impacts on fish assemblages in European running waters. Sci. Total Environ. 2016, 573, 1079–1088. [Google Scholar] [CrossRef] [Green Version]
  51. Wang, L.; Lyons, J.; Kanehl, P.; Gatti, R. Influences of watershed land use on habitat quality and biotic integrity in Wisconsin streams. Fisheries 1997, 22, 6–12. [Google Scholar] [CrossRef]
  52. Wang, L.Z.; Lyons, J.; Kanehl, P. Impacts of urbanization on stream habitat and fish across multiple spatial scales. Environ. Manag. 2001, 28, 255–266. [Google Scholar] [CrossRef]
  53. Keitzer, S.C.; Ludsin, S.A.; Sowa, S.P.; Annis, G.; Arnold, J.G.; Daggupati, P.; Froehlich, A.M.; Herbert, M.E.; Johnson, M.-V.V.; Sasson, A.M.; et al. Thinking outside of the lake: Can controls on nutrient inputs into Lake Erie benefit stream conservation in its watershed? J. Great Lakes Res. 2016, 42, 1322–1331. [Google Scholar] [CrossRef] [Green Version]
  54. Dean, E.M.; Cooper, A.R.; Wang, L.; Daniel, W.; David, S.; Ernzen, C.; Gido, K.B.; Hale, E.; Haxton, T.J.; Kelso, W.; et al. The North American Freshwater Migratory Fish Database (NAFMFD): Characterizing the migratory life histories of freshwater fishes of Canada, the United States and Mexico. J. Biogeogr. 2022, 49, 1193–1203. [Google Scholar] [CrossRef]
  55. Fales, M.; Dell, R.; Herbert, M.E.; Sowa, S.P.; Asher, J.; O’Neil, G.; Doran, P.J.; Wickerham, B. Making the leap from science to implementation: Strategic agricultural conservation in Michigan’s Saginaw Bay watershed. J. Great Lakes Res. 2016, 42, 1372–1385. [Google Scholar] [CrossRef] [Green Version]
  56. Niebling, W.; Baker, J.; Kasuri, L.; Katz, S.; Smet, K. Challenge and response in the Mississippi River Basin. Water Policy 2014, 16, 87–116. [Google Scholar] [CrossRef]
  57. Mississippi River/Gulf of Mexico Hypoxia Task Force—Priority Watershed Selection Processes in Hypoxia Task Force States. Available online: https://www.epa.gov/ms-htf/priority-watershed-selection-processes-hypoxia-task-force-states (accessed on 12 January 2023).
  58. Sievert, N.; Paukert, C.; Whittier, J.; Daniel, W.; Infante, D.; Stewart, J. Projected stream fish community risk to climate impacts in the Northeastern and Midwestern United States. Ecol. Indic. 2022, 144, 109493. [Google Scholar] [CrossRef]
  59. Tsang, Y.-P.; Infante, D.; Wang, L.; Krueger, D.; Wieferich, D. Conserving stream fishes with changing climate: Assessing fish response to changes in habitat over a large region. Sci. Total Environ. 2021, 755, 142503. [Google Scholar] [CrossRef]
  60. Valayamkunnath, P.; Barlage, M.; Chen, F.; Gochis, D.J.; Franz, K.J. Mapping of 30-meter resolution tile-drained croplands using a geospatial modeling approach. Sci. Data 2020, 7, 257. [Google Scholar] [CrossRef]
  61. Zhou, Q.; Xian, G.; Horton, J.; Wellington, D.; Domke, G.; Auch, R.; Li, C.; Zhu, Z. Tree regrowth duration map from LCMAP collection 1.0 land cover products in the conterminous United States, 1985–2017. GISci. Remote Sens. 2022, 59, 959–974. [Google Scholar] [CrossRef]
Water 15 02400 g001 550
Figure 1. The four spatial extents used for summarizing anthropogenic stressors and natural landscape conditions, including local catchment, network catchment, local buffer, and network buffer.
Figure 1. The four spatial extents used for summarizing anthropogenic stressors and natural landscape conditions, including local catchment, network catchment, local buffer, and network buffer.
Water 15 02400 g001
Water 15 02400 g002 550
Figure 2. Flowchart depicting steps taken to account for natural variation and test fish species responses to anthropogenic stressors. Ecoregions include Coastal Plains (CPL), Northern Appalachians (NAP), Northern Plains (NPL), Southern Appalachians (SAP), Southern Plains (SPL), Temperate Plains (TPL), Upper Midwest (UMW), Western Mountains (WMT), and Xeric (XER).
Figure 2. Flowchart depicting steps taken to account for natural variation and test fish species responses to anthropogenic stressors. Ecoregions include Coastal Plains (CPL), Northern Appalachians (NAP), Northern Plains (NPL), Southern Appalachians (SAP), Southern Plains (SPL), Temperate Plains (TPL), Upper Midwest (UMW), Western Mountains (WMT), and Xeric (XER).
Water 15 02400 g002
Water 15 02400 g003 550
Figure 3. Bar graph showing the percentage of threshold responses of fish species to agricultural land-use, impervious surface, and urban land-use variables summarized at the local catchment (LC), local buffer (LB), network buffer (NB), and network catchment (NC).
Figure 3. Bar graph showing the percentage of threshold responses of fish species to agricultural land-use, impervious surface, and urban land-use variables summarized at the local catchment (LC), local buffer (LB), network buffer (NB), and network catchment (NC).
Water 15 02400 g003
Water 15 02400 g004 550
Figure 4. Map and percent of stream length in each risk of habitat degradation class due to land use. Ecoregions include Coastal Plains (CPL), Northern Appalachians (NAP), Northern Plains (NPL), Southern Appalachians (SAP), Southern Plains (SPL), Temperate Plains (TPL), Upper Midwest (UMW), Western Mountains (WMT), and Xeric (XER).
Figure 4. Map and percent of stream length in each risk of habitat degradation class due to land use. Ecoregions include Coastal Plains (CPL), Northern Appalachians (NAP), Northern Plains (NPL), Southern Appalachians (SAP), Southern Plains (SPL), Temperate Plains (TPL), Upper Midwest (UMW), Western Mountains (WMT), and Xeric (XER).
Water 15 02400 g004
Water 15 02400 g005 550
Figure 5. Map and percent of stream length in each risk of habitat degradation class due to fragmentation. Ecoregions include Coastal Plains (CPL), Northern Appalachians (NAP), Northern Plains (NPL), Southern Appalachians (SAP), Southern Plains (SPL), Temperate Plains (TPL), Upper Midwest (UMW), Western Mountains (WMT), and Xeric (XER).
Figure 5. Map and percent of stream length in each risk of habitat degradation class due to fragmentation. Ecoregions include Coastal Plains (CPL), Northern Appalachians (NAP), Northern Plains (NPL), Southern Appalachians (SAP), Southern Plains (SPL), Temperate Plains (TPL), Upper Midwest (UMW), Western Mountains (WMT), and Xeric (XER).
Water 15 02400 g005
Water 15 02400 g006 550
Figure 6. Map and percent of stream length in each risk of habitat degradation class due to water quality. Ecoregions include Coastal Plains (CPL), Northern Appalachians (NAP), Northern Plains (NPL), Southern Appalachians (SAP), Southern Plains (SPL), Temperate Plains (TPL), Upper Midwest (UMW), Western Mountains (WMT), and Xeric (XER).
Figure 6. Map and percent of stream length in each risk of habitat degradation class due to water quality. Ecoregions include Coastal Plains (CPL), Northern Appalachians (NAP), Northern Plains (NPL), Southern Appalachians (SAP), Southern Plains (SPL), Temperate Plains (TPL), Upper Midwest (UMW), Western Mountains (WMT), and Xeric (XER).
Water 15 02400 g006
Water 15 02400 g007 550
Figure 7. Map showing catchments with a low or very low risk of habitat degradation from land use, fragmentation, and water quality. Ecoregions include Coastal Plains (CPL), Northern Appalachians (NAP), Northern Plains (NPL), Southern Appalachians (SAP), Southern Plains (SPL), Temperate Plains (TPL), Upper Midwest (UMW), Western Mountains (WMT), and Xeric (XER).
Figure 7. Map showing catchments with a low or very low risk of habitat degradation from land use, fragmentation, and water quality. Ecoregions include Coastal Plains (CPL), Northern Appalachians (NAP), Northern Plains (NPL), Southern Appalachians (SAP), Southern Plains (SPL), Temperate Plains (TPL), Upper Midwest (UMW), Western Mountains (WMT), and Xeric (XER).
Water 15 02400 g007
Water 15 02400 g008 550
Figure 8. Map showing catchments with a low or very low risk of habitat degradation from land use and fragmentation but high or very high risk from water quality. Ecoregions include Coastal Plains (CPL), Northern Appalachians (NAP), Northern Plains (NPL), Southern Appalachians (SAP), Southern Plains (SPL), Temperate Plains (TPL), Upper Midwest (UMW), Western Mountains (WMT), and Xeric (XER).
Figure 8. Map showing catchments with a low or very low risk of habitat degradation from land use and fragmentation but high or very high risk from water quality. Ecoregions include Coastal Plains (CPL), Northern Appalachians (NAP), Northern Plains (NPL), Southern Appalachians (SAP), Southern Plains (SPL), Temperate Plains (TPL), Upper Midwest (UMW), Western Mountains (WMT), and Xeric (XER).
Water 15 02400 g008
Water 15 02400 g009 550
Figure 9. Map showing catchments exceeding total nitrogen yield threshold. Catchments not shown either do not exceed the threshold or have a water quality score driven by a different stressor. Ecoregions include Coastal Plains (CPL), Northern Appalachians (NAP), Northern Plains (NPL), Southern Appalachians (SAP), Southern Plains (SPL), Temperate Plains (TPL), Upper Midwest (UMW), Western Mountains (WMT), and Xeric (XER).
Figure 9. Map showing catchments exceeding total nitrogen yield threshold. Catchments not shown either do not exceed the threshold or have a water quality score driven by a different stressor. Ecoregions include Coastal Plains (CPL), Northern Appalachians (NAP), Northern Plains (NPL), Southern Appalachians (SAP), Southern Plains (SPL), Temperate Plains (TPL), Upper Midwest (UMW), Western Mountains (WMT), and Xeric (XER).
Water 15 02400 g009
Table
Table 1. Percent of species with at least one negative threshold across all stressors, and specifically for the land-use, fragmentation, and water quality stressor variables summarized by ecoregion and basin-wide. Number of species tested is in parentheses.
Table 1. Percent of species with at least one negative threshold across all stressors, and specifically for the land-use, fragmentation, and water quality stressor variables summarized by ecoregion and basin-wide. Number of species tested is in parentheses.
EcoregionAll StressorsLand UseFragmentationWater Quality
CPL (246)16.713.87.312.6
NAP (125)37.634.428.028.0
NPL (80)8.86.31.33.8
SAP (349)32.727.519.822.9
SPL (132)22.019.712.118.2
TPL (217)25.823.521.220.3
UMW (149)36.927.521.521.5
WMT (50)26.024.012.024.0
XER (38)34.228.921.121.1
Basin-wide (441)39.236.127.932.4
Table
Table 2. Count and percent of catchments that are currently exceeding the identified total nitrogen yield threshold, and average, minimum, maximum, and standard deviation (kg/km2/yr) of the exceedance organized by ecoregion and basin-wide.
Table 2. Count and percent of catchments that are currently exceeding the identified total nitrogen yield threshold, and average, minimum, maximum, and standard deviation (kg/km2/yr) of the exceedance organized by ecoregion and basin-wide.
EcoregionCountPercentAverageMinMaxSD
CPL63,78141.192302.880.0115,334.073552.41
NAP727043.031652.830.0115,386.842788.02
NPL9620.558466.3121.7412,253.452827.23
SAP152,82758.011469.67<0.0115,376.592120.12
SPL110,21159.883101.040.0215,254.823419.89
TPL149,38354.995334.910.3114,396.162767.63
UMW21,45535.162867.350.1013,585.662528.54
WMT903015.36513.260.0215,201.311395.85
XER8644.17229.840.0113,950.751318.75
Basin-wide515,78342.613095.700.0115,386.843250.60
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Ross, J.A.; Infante, D.M.; Cooper, A.R.; Whittier, J.B.; Daniel, W.M. Assessing Impacts of Human Stressors on Stream Fish Habitats across the Mississippi River Basin. Water 2023, 15, 2400. https://doi.org/10.3390/w15132400

AMA Style

Ross JA, Infante DM, Cooper AR, Whittier JB, Daniel WM. Assessing Impacts of Human Stressors on Stream Fish Habitats across the Mississippi River Basin. Water. 2023; 15(13):2400. https://doi.org/10.3390/w15132400

Chicago/Turabian Style

Ross, Jared A., Dana M. Infante, Arthur R. Cooper, Joanna B. Whittier, and Wesley M. Daniel. 2023. "Assessing Impacts of Human Stressors on Stream Fish Habitats across the Mississippi River Basin" Water 15, no. 13: 2400. https://doi.org/10.3390/w15132400

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop