Recipient Organization
AUBURN UNIVERSITY
108 M. WHITE SMITH HALL
AUBURN,AL 36849
Performing Department
Fisheries & Allied Aquaculture
Non Technical Summary
Both climate change and agricultural land use may change the amount and quality of water available to stream ecosystems. The interaction of these factors has not been considered for stream fishes in Alabama, but a recent catastrophic drought (2016) provides an opportunity to do so. The proposed research promises to not only document the effects of both drought and agricultural land use on a gradient of stream sites (headwater to mainstem), but also to investigate a possible mechanism of fish recovery. The role of spawning mode on stream fish recovery following drought has not been investigated. It has been suggested that fishes that provide parental care might have higher recruitment success post-drought. However, increased sedimentation, often the result of a high percentage of agricultural land use in a watershed, may actually decrease reproductive success of fishes with reproductive modes that include parental care. Differential recruitment may be responsible for the invasion of non-natives into stream ecosystems, and be partially responsible for the loss of sensitive species as well. The results of this research will aid in the prediction of where and which species may suffer extirpation. In addition, areas susceptible to faunal homogenization may be identified. Management recommendations may include the addition of riparian zones or the establishment of minimum flows for Alabama stream systems to balance any observed effects of agricultural land use and drought.Agricultural land use affects both the quantity and quality of water that runs off into streams and rivers. Water quantity is altered as forests are cleared and evapotranspiration and water run off rates significantly change. Increased variability in water availability has been shown to dramatically alter fish assemblages (Freeman et al., 2001, King et al. 2008). In addition, sensitive species may be lost, and the distribution and abundance of non-native fishes may increase (Helms et al., 2005). Water quality is altered as sedimentation in run off dramatically increases, especially if riparian buffer zones are absent. The negative effects of excessive sedimentation on aquatic communities have been well documented (Walser and Bart, 1999; Sutherland et al., 2008).How the effects of agricultural land use and other environmental perturbation, such as global climate change, affect fish assemblages is poorly understood. Peterson and Kwak (1999) suggest that land use may have greater effects on fish species than climate change. The effects of global climate change on fishes will include not only the obvious shifts in distribution due to thermal tolerances, but the change in the amount and distribution of freshwater may have catastrophic effects on the persistence of fish species and assemblage structure (Carpenter et al., 1992; Poff et al., 1995). Fishes with omnivorous diets and short life spans are the most likely to persist or invade disturbed habitats, and indeed, homogenization of stream habitats has already been documented and includes such species (Scott and Helfman, 2001; Helms et al., 2005). Fishes such as redbreast sunfish (Lepomis auritus) often invade or increase in numbers in stream habitats that are physically degraded or have altered hydrology, usually at the expense of other, native or endemic species (Helms et al., 2005, 2009; Johnston and Macenia, 2009). Altered hydrology is one of the biggest stressors on aquatic systems, and can include natural drought as well as changes due to agricultural or other land use changes. Although it has been well-documented that fishes in unaltered systems can rapidly recover following short-term drought (Matthews and Marsh-Matthews, 2003; Magalhães et al., 2007), the effects of prolonged drought on species and assemblage persistence is less clear, especially if the natural hydrology is altered, making water less available overall (Douglas et al., 2003).It has been suggested that species recovery may be related to spawning and recruitment success. Many spawning modes are subject to tremendous impact from sedimentation, and this factor may contribute to spawning failure, even in species that would ordinarily rebound quickly. Ensign et al. (1997) suggested that recovery rates of fishes following a fish kill in a riverine system were related to aspects of parental investment. Magalhães et al. (2007) traced fish assemblage change to water flow related recruitment success and failure in Mediterranean streams. To date, research linking fish recovery success following severe drought to recruitment success due to spawning mode, including aspects of parental care, are lacking. How land use, and in particular agricultural land use, plays a role in the recovery of fish assemblages following such disturbances in poorly known. Research that focuses on linking these factors will shed light on the mechanism of species loss and persistence following drought and in flow-altered systems.The objectives of this project are to investigate the effects of catastrophic drought (2016) on fish assemblage structure and persistence as it relates to agricultural land use in Alabama. Specifically, what roles do spawning mode and agricultural land use play in successful recruitment following drought? The proposed research will be designed to examine these factors across a gradient of stream drying, from complete to partial. Adult fishes (2010) should reflect successful recruitment during the drought years or recolonization; absences may be due to local extirpation. Recolonization should be reflected by reduced numbers, and it should be possible to trace these species to source populations. Additionally, the fish assemblages chosen for study include species with various spawning modes, from simple broadcasting to derived parental care.
Animal Health Component
10%
Research Effort Categories
Basic
90%
Applied
10%
Developmental
(N/A)
Goals / Objectives
The objectives of this project are to investigate the effects of catastrophic drought (2016) on fish assemblage structure and persistence as it relates to agricultural land use in Alabama. Specifically, what roles do spawning mode and agricultural land use play in successful recruitment following drought? The proposed research will be designed to examine these factors across a gradient of stream drying, from complete to partial, as well as varying percentages of agricultural land use and type. Adult fishes (2017) should reflect successful recruitment during the drought years or recolonization; absences may be due to local extirpation. Recolonization should be reflected by reduced numbers, and it should be possible to trace these species to source populations. Additionally, the fish assemblages chosen for study include species with various spawning modes, from simple broadcasting to derived parental care. This will allow an evaluation of the effect of spawning mode, the severity of drought and land use effects on fish species and assemblage recovery, as well as the interaction of these factors
Project Methods
This work will take place across a wide geographic area in many watershedsthroughout state of Alabama. These watersheds will be selected based on the availabilityof historic data on fish assemblage structure and representation of a range of thephysiographic regions, percentage in agricultural land use, and drainage basins of Alabama. Collection MethodsStreams will be sampled according to the same protocol used in past studiesconducted by the same lab. Site coordinates were logged using a hand-held GPS unit(Garmin GPSmap 60CSx). Fishes will be collected using a backpack electrofisher and a 3m x 1.75m seine. A stream reach of approximately 15 times the wetted width wassampled at each site. All available mesohabitats are typically sampled roughlyproportional to their area at the site. Samples take roughly an hour to complete. Fishescollected are anesthetized using tricaine mesylate (MS-222) and fixed in a 10% formalinsolution. Collections are sorted and identified in the lab and stored in a 50% isopropanolsolution. Collections will be deposited in the Auburn University Museum of NaturalHistory. Fish collection records were managed in Microsoft Excel 2013. Most sites aresampled during two consecutive summers in order to validate our methods and detection. Data AnalysisHydrologic data for analysis will be retrieved from the United States GeologicSurvey (USGS) National Water Information System. Active stream gages will be usedwhen available. Gages in adjacent, similar watersheds can be used as proxies wheregages were not available in the study watershed or the record from the gage was onlyavailable for a small period of the time range investigated by this study.Change in discharge over time will be investigated for each watershed usingsimple linear regression of the mean annual discharge in the open source statisticsprogram R. Annual mean, median, and standard deviation of discharge were calculated inMicrosoft Excel from daily mean discharges for the water years (October 1 to September30) that coincided with fish collections used in this study. The Richards-Baker FlashinessIndex (R-B Index) was used instead as a measure of the flow stability in the watersheds studied. This index compares the pathlength of the of the annual discharge hydrograph tothe sum of the mean daily discharges. The higher the index score, themore extreme the hydrologic fluctuations.Land cover data will be taken from National Land Cover Dataset (NLCD). Datasetsfrom the 2001, 2006 and 2011 NLCD reports were used to measure changes to the landdraining to the sample sites. Arc GIS 10.3.1 was used to measure the proportion of eachland cover type in the watersheds draining to the sample sites with a focus on forest,cropland, and development land cover types.To validate the sampling methods, fish assemblages from sites will be sampledtwice during the contemporary period are compared between the two years using theMorisita Similarity Index (Im). The Morisita Index is more conservative than othersimilarity indices and incorporates species abundances so that is less biased by naturallyrare species. Values for the Morisita Index range from 0 to 1 with 1representing completely identical assemblages and 0 indicative of no overlap in theassemblages. Past studies have considered values ≤ 0.4 to have low similarity and valuesof ≥ 0.7 to have high similarity (Johnston & Maceina 2008, Matthews et al. 1988, Phillips& Johnston 2004).The correlations between fish and environmental data will be analyzed usingmultivariate, constrained ordination techniques using the community ecology analysispackage, 'vegan 2.0-10', in the program R. Fish assemblage 215 data from each watershedwill be ordinated using partial redundancy analysis (pRDA). This ordination methodcombines multiple regression (MR) with principal component analysis (PCA) and is auseful tool to analyze the response of a community to environmental variables. Theordination in the PCA portion of the analysis is constrained by environmental variables,better explaining the variance caused by these variables. Partial RDA partitions thevariance from selected environmental variables allowing them to be accounted for likecovariates in linear regression.