Louisiana Waterthrush and Benthic Macroinvertebrate Response to ...

Notes

Louisiana Waterthrush and Benthic Macroinvertebrate Response to Shale Gas Development Petra B. Wood,* Mack W. Frantz, Douglas A. Becker

P.B. Wood U.S. Geological Survey, West Virginia Cooperative Fish and Wildlife Research Unit, P.O. Box 6125, West Virginia University, Morgantown, West Virginia 26506 M.W. Frantz, D.A. Becker West Virginia Cooperative Fish and Wildlife Research Unit, Division of Forestry and Natural Resources, West Virginia University, Morgantown, West Virginia 26506 Present address of D.A. Becker: Department of Biology, Kutztown University, 245 Boehm Hall, Kutztown, Pennsylvania 19530

Abstract Because shale gas development is occurring over large landscapes and consequently is affecting many headwater streams, an understanding of its effects on headwater-stream faunal communities is needed. We examined effects of shale gas development (well pads and associated infrastructure) on Louisiana waterthrush Parkesia motacilla and benthic macroinvertebrate communities in 12 West Virginia headwater streams in 2011. Streams were classed as impacted (n ¼ 6) or unimpacted (n ¼ 6) by shale gas development. We quantified waterthrush demography (nest success, clutch size, number of fledglings, territory density), a waterthrush Habitat Suitability Index, a Rapid Bioassessment Protocol habitat index, and benthic macroinvertebrate metrics including a genus-level stream-quality index for each stream. We compared each benthic metric between impacted and unimpacted streams with a Student’s t-test that incorporated adjustments for normalizing data. Impacted streams had lower genus-level stream-quality index scores; lower overall and Ephemeroptera, Plecoptera, and Trichoptera richness; fewer intolerant taxa, more tolerant taxa, and greater density of 0–3-mm individuals (P 0.10). We then used Pearson correlation to relate waterthrush metrics to benthic metrics across the 12 streams. Territory density (no. of territories/km of stream) was greater on streams with higher genus-level stream-quality index scores; greater density of all taxa and Ephemeroptera, Plecoptera, and Trichoptera taxa; and greater biomass. Clutch size was greater on streams with higher genus-level stream-quality index scores. Nest survival analyses (n ¼ 43 nests) completed with Program MARK suggested minimal influence of benthic metrics compared with nest stage and Habitat Suitability Index score. Although our study spanned only one season, our results suggest that shale gas development affected waterthrush and benthic communities in the headwater streams we studied. Thus, these ecological effects of shale gas development warrant closer examination. Keywords: headwater streams; Marcellus shale; stream quality; waterthrush demography; West Virginia Received: September 9, 2015; Accepted: April 11, 2016; Published Online Early: May 2016; Published: December 2016 Citation: Wood PB, Frantz MW, Becker DA. 2016. Louisiana waterthrush and benthic macroinvertebrate response to shale gas development. Journal of Fish and Wildlife Management 7(2):423–433; e1944-687X. doi: 10.3996/092015JFWM-084 Copyright: All material appearing in the Journal of Fish and Wildlife Management is in the public domain and may be reproduced or copied without permission unless specifically noted with the copyright symbol &. Citation of the source, as given above, is requested. The findings and conclusions in this article are those of the author(s) and do not necessarily represent the views of the U.S. Fish and Wildlife Service. * Corresponding author: [email protected]

Journal of Fish and Wildlife Management | www.fwspubs.org

December 2016 | Volume 7 | Issue 2 | 423

Shale Gas Effects on Waterthrush and Macroinvertebrates

P.B. Wood et al.

Figure 1. Location of study site (Lewis Wetzel Wildlife Management Area [WMA]) in West Virginia in relation to the Marcellus shale region and range of the Louisiana waterthrush Parkesia motacilla based on North American Breeding Bird Survey (BBS) data from Sauer et al. (2012).

Introduction The Marcellus shale natural gas reservoir (Figure 1) underlies much of the Central Appalachian and Western Allegheny Plateau ecoregions. Both ecoregions are heavily forested with numerous headwater stream systems and consequently are of great conservation importance to forest songbirds and to species associated with forested headwater streams. In recent years, hydraulic fracturing, often called ‘‘fracking,’’ has allowed recovery of natural gas from unconventional reservoirs throughout the United States. In West Virginia and other parts of the central Appalachians, the number of wells accessing unconventional natural-gas reservoirs (e.g., Marcellus shale) has increased considerably. The Marcellus Center for Outreach and Research (2015) estimated that 111 unconventional gas wells occurred in the central Appalachians in 2005, which increased to .12,964 by 2015. Thus the region has been subjected to expanding networks of well pads, roads, and pipelines leading to forest loss, increased fragmentation of forested systems (Drohan et al. 2012), and disturbance of headwater streams (Farwell, et al. 2016). Journal of Fish and Wildlife Management | www.fwspubs.org

Little is known about the impacts of Marcellus naturalgas extraction on terrestrial or aquatic ecosystems. Although concerns regarding environmental impacts of hydraulic fracturing have centered on risks of water contamination and effects on human health, unconventional gas development also poses a potential threat to regional forests and biodiversity (Kargbo et al. 2010; Drohan et al. 2012; Kiviat 2013; Brittingham et al. 2014; Souther et al. 2014). In addition to the direct impacts of deforestation, habitat fragmentation, large-scale freshwater consumption (3–5 million gallons/well), runoff and changed hydrology from increased impervious surfaces, and noise and light disturbance, there are also potential secondary effects on forest ecosystems, including soil erosion, soil and water contamination, sedimentation of streams, increased human access and presence, and changes in biotic communities, among others (Adams et al. 2011; Baker et al. 2013; Kiviat 2013). The Louisiana waterthrush Parkesia motacilla, hereafter waterthrush, is a riparian-obligate songbird that forages primarily on benthic macroinvertebrates and nests along forested headwater streams that have welldeveloped riffle and pool complexes (Prosser and Brooks 1998). Previous studies have found waterthrush breeding December 2016 | Volume 7 | Issue 2 | 424


P.B. Wood et al.

Figure 2. Location of study streams, benthic sampling locations, and stream reaches used by the Louisiana waterthrush (LOWA) Parkesia motacilla that were impacted by shale gas development during 2011 on the Lewis Wetzel Wildlife Management Area, West Virginia, overlaid on a 2011 aerial photograph showing forest cover. Stream names ending in 0 were considered impacted at the benthic sampling location and those ending in 1 were considered not impacted by shale gas development.

density to be negatively influenced by stream acidification (Mulvihill et al. 2008) and waterthrush occupancy to be related to relative abundance and biomass of particular benthic macroinvertebrate taxa (Mattsson and Cooper 2006). Some authors have suggested that this species may serve as a biological indicator of headwater stream quality (Brooks et al. 1998; Mattsson and Cooper 2006, 2009). Thus, the study of waterthrush demography and their benthic macroinvertebrate prey could provide an indication of the effects of shale gas development on headwater stream ecosystems at multiple trophic levels. Consequently, in this study, we Journal of Fish and Wildlife Management | www.fwspubs.org

examined the effects of gas well and associated infrastructure development on benthic macroinvertebrate communities and related them to waterthrush nesting success, territory density, and habitat quality.

Study Area We studied waterthrush and benthic macroinvertebrate communities during the 2011 breeding season along ~50 km of first- and second-order tributary streams (n ¼ 12; Figure 2) on 4,541 ha of the Lewis December 2016 | Volume 7 | Issue 2 | 425


Wetzel Wildlife Management Area located in northwestern West Virginia (Figure 1). The study area lies within the Permian Hills subdivision of the Western Allegheny Plateau, an area of deeply dissected topography and relatively continuous Appalachian Oak and MixedMesophytic Forest (Woods et al. 1999). Elevations at Lewis Wetzel Wildlife Management Area are 221–480 m (mean 356 m). The study area also lies within the Marcellus shale region and is where waterthrush reach their greatest abundances within the central Appalachians (Figure 1) based on Breeding Bird Survey data (Sauer et al. 2012). Gas well development activities on Lewis Wetzel Wildlife Management Area included building of well pads, ponds, parking areas, bridges and stream crossings, construction of new and widening of existing roads and pipelines, and removal of timber associated with these activities. Within a Geographic Information System (GIS), we used a sequence of aerial and satellite imagery (2003–2011) combined with field visits to map areas of nonforest cover on the study area and to identify type of impact. In 2011, 93.9% of Lewis Wetzel Wildlife Management Area was mature forest and 1.5% was in new gas wells and infrastructure that were a result of Marcellus shale gas development. We then classified each study stream as being impacted (n ¼ 6) or unimpacted (n ¼ 6) by gas development. Each impacted stream had well pads or their infrastructure within 60 m of the stream centerline; we used a 60-m streamside buffer because it is the typical extent of waterthrush streamside habitat use (Mattsson and Cooper 2009). Streams classed as unimpacted had no shale gas–related impacts.

Methods Field sampling Benthic surveys. We sampled benthic macroinvertebrates in two riffles ~40 m apart near the base of each stream reach (Figure 2). Although sample locations best represent local conditions, the benthic community typically is affected by upgradient conditions (e.g., proximity and magnitude of stressors) including drifting colonizers from upstream sources. We paired impacted and unimpacted streams and placed benthic sample locations on second-order streams at locations with similar flow accumulation and generally similar habitat conditions within stream pairs. In each of the two riffles, we used a one-quarter-m2 kicknet during 30-s timed samples to collect two samples from the center of each of the two riffles during 6–10 May 2011. We composited the four kicknet samples to yield a 1-m2 sample following bioassessment methods outlined in West Virginia Department of Environmental Protection standard operation procedures (WVDEP 2011). We collected samples from the paired streams on the same day when possible, at a similar time of day, and during periods of normal flow. Waterthrush territory mapping and nest monitoring. Between 28 April and 2 June 2011, we mapped Journal of Fish and Wildlife Management | www.fwspubs.org

P.B. Wood et al.

waterthrush locations and nests using Global Positioning System units and detailed stream maps during intensive nest searching and monitoring, territory mapping, and banding and resighting activities on each stream. We noted locations of waterthrush behavioral activities, particularly those that would help to delineate territories, such as counter-singing and chasing of neighboring males (Bibby et al. 2000). We delineated territories at the end of the field season based on the mapped locations and territorial behaviors. We then calculated territory density as the number of territories per km of stream searched for each of the 12 streams. We initially visited nests approximately every 3–4 d, and more frequently as fledging approached (Martin and Geupel 1993). We counted the number of eggs after incubation commenced to determine clutch size. We assumed that an empty nest had fledged if the nest was active the previous day and was near its predicted fledge date. We searched for fledglings to confirm fledging success. The number of fledglings for each nest was the count of nestlings in the visit prior to fledging for successful nests and zero for failed nests. We considered a nest successful if it produced 1fledgling. Riparian habitat assessments. Riparian habitat quality was assessed using the Louisiana Waterthrush Habitat Suitability Index (hereafter, HSI; Prosser and Brooks 1998) and the Environmental Protection Agency Rapid Bioassessment Protocol habitat index for high gradient streams (Barbour et al. 1999). The waterthrush HSI evaluates instream and upland habitat in terms of both nesting and foraging requirements (Prosser and Brooks 1998). The Rapid Bioassessment Protocol assesses stream habitat quality based primarily on instream characteristics that relate to the abundance and composition of the waterthrush’s aquatic macroinvertebrate prey, and thus may indicate the quality of instream foraging habitat. We quantified each index on 50-m stream segments centered on the waterthrush nests monitored in 2011. Data analyses Benthic macroinvertebrates. In the laboratory, we washed the composited kicknet samples in a 500-lmmesh sieve, then we sorted a random 200 (620%) invertebrate fixed-count sample from the detritus using a gridded pan, and identified all subsampled organisms to the genus-level (WVDEP 2011). We counted organisms, and measured body lengths to the nearest 1 mm. We used length measurements to estimate the ash-free dry mass of benthic organisms using length–weight equations found in the literature (e.g., Benke et al. 1999; Sabo et al. 2002). We based macroinvertebrate density (individuals/m2) and biomass (lg/m2 ash-free dry mass) estimates on extrapolated values using the proportion of sorted grids required to reach the target 200 organism subsample (Ligeiro et al. 2013). We calculated macroinvertebrate density and biomass for several categories (Table 1) that might be important to foraging waterthrush. Categories included all taxa combined; all taxa in December 2016 | Volume 7 | Issue 2 | 426


Table 1. Definitions and notations of all candidate models and model sets used in Program MARK for evaluating the association of temporal, habitat, and benthic covariates with daily survival rate (DSR) of Louisiana waterthrush Parkesia motacilla nests during 2011 on the Lewis Wetzel Wildlife Management Area, West Virginia. EPA, Environmental Protection Agency; GLIMPSS, genus-level index of most probable stream status to assess the overall quality of each stream; ETA, Ephemeroptera, Plecoptera, and Trichoptera taxa. Covariate of interest

Model notation

Model Set 1: temporal covariates (n ¼ 5 models) Intercept only CONSTANT Nest stage Nest Stage Rain Rain Linear time trend T Quadratic time trend TT Model Set 2: habitat covariates (n ¼ 4 models) Intercept only CONSTANT Habitat Suitability Index (HSI) HSI EPA Rapid Bioassessment RBP Nonforest cover (%) NonForest Model Set 3: benthic covariates (n ¼ 14 models) Intercept only CONSTANT GLIMPSS GLIMPSS Biomass total Biomass T Biomass size class 0–3 mm Biomass 0 Biomass size class .3–6 mm Biomass 3 Biomass size class .6 mm Biomass 6 Density total Density T Density size class 0–3 mm Density 0 Density size class .3–6 mm Density 3 Density size class .6 mm Density 6 EPT biomass EPT Bio EPT density EPT Den Heptageniidae biomass Hep Bio Heptageniidae density Hep Den Model Set 4: covariates in top (DAICc , 2) models from sets 1–3

three size classes (3 mm, .3–6 mm, and .6 mm); only Ephemeroptera; and Plecoptera and Trichoptera (EPT) taxa because these are suggested preferred prey (Mattsson et al. 2009); and only heptageniid mayfly taxa, which are commonly fed to nestlings (Trevelline et al. 2016). We calculated a macroinvertebrate multimetric index called the genus-level index of most probable stream status (GLIMPSS) to assess the overall quality of each stream (Pond et al. 2013). The GLIMPSS index is a regionally and seasonally stratified macroinvertebrate multimetric index calculated using 8–10 macroinvertebrate metrics, depending on stratum (mountains vs. plateau, spring vs. summer). Metric values are individually scored on a 100-point scale and then averaged to yield the GLIMPSS score where the impairment threshold is set at the fifth percentile of the reference distribution within each stratum. For our data, we used the 8-metric, spring-plateau GLIMPSS (see Pond et al. 2013). We also calculated overall taxa richness, EPT richness, number of intolerant taxa, and % tolerant taxa. We selected these metrics to evaluate benthic community composition and diversity because EPT and intolerant taxa typically Journal of Fish and Wildlife Management | www.fwspubs.org

P.B. Wood et al.

require good water quality whereas tolerant taxa tend to increase in streams with poor water quality. We classed intolerant taxa as those having a tolerance value of 0–3, whereas tolerant taxa had tolerance values of 7– 10 as defined in the WVDEP Watershed Assessment Branch database (WABase; WVDEP 2011). We compared each benthic metric between impacted and unimpacted streams with a Student’s t-test that incorporated adjustments for normalizing data. We then used a Pearson correlation matrix to relate waterthrush metrics (territory density, average clutch size per stream, average number of fledglings per stream) regardless of impact status of streams to the benthic metrics (Table S1). Because GLIMPSS and territory density were calculated per stream, we calculated the average clutch size (number of eggs per completed clutch) and average number of fledglings (number of young fledged per nesting attempt) per stream to use in the correlation. For all statistical tests, we set significance at a ¼ 0.10 to account for variation that may have biological significance (Askins et al. 1990). All values presented in results are means 6 SE. We used SAS version 9.4 for data summaries and statistical analyses. Nest survival and productivity. We developed a set of a priori candidate models containing temporal, habitat, and benthic covariates that we hypothesized might influence daily survival rate (DSR) of waterthrush nests (see details below). We used Akaike’s Information Criterion for small sample sizes (AICc) to evaluate support for candidate models (Burnham and Anderson 2002) and used Program MARK (White and Burnham 1999) to estimate DSR and evaluate covariates. We included nests that reached at least the egg-laying stage in analyses, but removed 2 of 45 nests that had unknown fates. We used a multistage modeling process to evaluate candidate models similar to Cunningham and Johnson (2006). We modeled individual covariates in each of three model sets (see Table 1 and below). We retained the best-supported models from each set (those with DAICc , 2) and used these individual covariates in the fourth final model set. Model set one included temporal and rainfall covariates. Temporal covariates included nest stage (laying, incubation, brooding) and the linear and quadratic effects of time within season. Encounter histories in MARK were divided into three groups to correspond to the three nest stages (Dinsmore et al. 2002). We included nest stage because variation in nest vulnerability and parental investment across the nesting cycle may result in nonconstant survival (Dinsmore et al. 2002; Grant et al. 2005; Burhans et al. 2010). We included time within season because nest survival can vary in concert with patterns in predator activity and density across the breeding season (Dinsmore et al. 2002; Grant et al. 2005; Burhans et al. 2010). We included quadratic terms because the relationship between temporal effects and daily survival rate may be nonlinear (Grant et al. 2005). We included mean daily rainfall as a covariate because December 2016 | Volume 7 | Issue 2 | 427


Mattsson and Cooper (2009) found a strong effect of mean daily rainfall on nest survival. Intermediate levels of rainfall may result in optimal prey availability and, consequently, greater foraging efficiency and nest vigilance (Mattsson and Cooper 2009). Flooding events may result in direct loss of nests (Price and Bock 1983), reduced abundance and diversity of benthic macroinvertebrates (Bond and Downes 2003), and reduced waterthrush foraging efficiency in deeper, faster water. For each nest, we averaged daily rainfall estimates across the period in which a nest was under observation (Mattsson and Cooper 2009). Precipitation estimates were from the closest weather station to the study area located in Middlebourne, West Virginia (39830 0 N, 80854 0 W) at a distance of 20–27 km from any specific nest. Model set two included covariates related to habitat quality. We hypothesized that gas development could negatively influence waterthrush reproduction through removal and fragmentation of riparian forest cover, modifying predator assemblages and activity, or altering stream hydrology and water quality (Petit and Petit 1996; Mulvihill et al. 2008; Mattsson and Cooper 2009). Therefore, we included the percentage of nonforest cover from all sources of impact within a 100-m radius of each nest as a covariate in nest survival models. Although most studies have reported reduced nest success within 50 m of forest edges (Paton 1994), we used a 100-m radius because forest edges may negatively affect the reproductive success of groundnesting species such as the waterthrush on a scale greater than that at which it affects canopy-nesting species (Flaspohler et al. 2001). We also included HSI and Rapid Bioassessment Protocol habitat index habitatquality scores as covariates in habitat models. Model set three included covariates that described the benthic macroinvertebrate community on each stream. We hypothesized that water quality impairment and its effects on prey abundance and species composition could influence nest survival (Mulvihill et al. 2008; Mattsson and Cooper 2009) in part by reducing waterthrush foraging efficiency. Increased foraging effort at the expense of nest vigilance may increase the risk of nest predation (Martindale 1982; Martin 1992) or reduced prey abundance could increase nestling starvation. We therefore included the GLIMPSS scores to assess the overall quality of each stream and the macroinvertebrate biomass and density metrics to assess prey availability as covariates in the nest survival models. We evaluated competing models of DSR within each model set using the nest survival model in Program MARK (version 7.1, Colorado State University, Ft. Collins, Colorado, USA; White and Burnham 1999). We modeled the binomially distributed data with the user-defined, logit-link function while simultaneously considering associations with temporal, habitat, and benthic covariates. We used standard coding for analysis of data in Program MARK (Dinsmore et al. 2002; Rotella et al. Journal of Fish and Wildlife Management | www.fwspubs.org

P.B. Wood et al.

2004). For our analysis of nest stage, we assumed a 29-d nesting period (egg-laying of 5 d, incubation of 14 d, nestling of 10 d) based on chronology of nests monitored on our study area. We considered the model with the lowest AICc value to be the best-supported model given the data, and any models with DAICc 2 were considered plausible (Burnham and Anderson 2002). We assessed the relative plausibility of each model in each model set by comparing Akaike weights (wi). We used beta coefficients and their confidence intervals to infer biological importance of covariates in plausible models.

Results Of the 17 benthic macroinvertebrate metrics that we compared, 6 differed between impacted and unimpacted streams. Five metrics (GLIMPSS score, density 0–3 mm, overall and EPT richness, number of intolerant taxa) were significantly greater in unimpacted streams and % tolerant taxa were significantly greater in impacted streams (Table 2). Although few of the density and biomass metrics differed significantly between impacted and unimpacted streams, the general trend was for unimpacted streams to have much larger density and biomass values than impacted streams. Waterthrush territory density had significant, strong, and positive correlations with six of the benthic metrics (GLIMPSS, two biomass metrics, and three density metrics [Table 3; Figure 3]). All of the R2 values were .0.56; .0.5 is considered a strong relationship (Cohen 1988). The strongest correlation was with EPT density (Table 3; Figure 3). Waterthrush average clutch size (for 35 nests with complete clutches) had a significant, positive correlation with the GLIMPSS score (Table 3; Figure 3). Average number of fledglings per nest attempt (for 43 nests monitored) was significantly and positively correlated with biomass 0–3 mm and approached significance (P ¼ 0.11) for density 0–3 mm (Table 3). Overall daily survival rate for the 43 nests monitored in 2011 was 0.968 (SE ¼ 0.007). This yielded a breedingseason survival rate of 38.9% based on a 29-d nesting period. In the temporal model set (Table 4), nest stage was the only supported model (DAICc , 2) and accounted for most of the variation in DSR (wi ¼ 0.82). In the habitat model set, the HSI index was the most plausible and had strong support (wi ¼ 0.70). The benthic model set had 12 supported models but all had relatively low weights (wi 0.12) indicating a weak influence of individual benthic metrics on DSR. All supported models (those with DAICc , 2) from these three model sets were included in the final combined model set. In the combined model set (Table 4), the nest stage model received the most support (wi ¼ 0.45) followed by HSI (wi ¼ 0.21). The remaining models had DAICc . 4 and beta coefficient confidence intervals overlapped zero, indicating that these variables had little influence on nest survival, particularly when compared with nest stage and HSI. December 2016 | Volume 7 | Issue 2 | 428


P.B. Wood et al.

Table 2. Mean, standard error (SE), and range (Min, Max) of benthic variables and t-test results comparing streams impacted by shale gas well development vs. unimpacted streams during 2011 on the Lewis Wetzel Wildlife Management Area, West Virginia. We considered differences significant at P 0.10; they are emphasized with bold text. See Table 1 for variable definitions.

Discussion Our study adds to evidence that benthic macroinvertebrate communities change on streams impacted by Marcellus shale gas development, similar to evidence reported by Grant et al. (2015) and similar to effects on streams in Arkansas influenced by Fayetteville shale gas development reported by Johnson et al. (2015). We found lower GLIMPSS scores on impacted streams, suggesting a general decline in aquatic ecosystem health. Lower richness of all taxa—specifically of EPT Table 3. Pearson correlation coefficients (R2) and P-values relating Louisiana waterthrush Parkesia motacilla metrics to benthic biomass and density metrics during 2011 on the Lewis Wetzel Wildlife Management Area, West Virginia. We considered correlations significant at P 0.10; they are emphasized with bold text. See Table 1 for variable definitions.


taxa—and fewer intolerant taxa suggest that water quality of the impacted streams was worse. Grant et al. (2015) observed differences in stream water quality and aquatic biodiversity between fracked and nonfracked streams in northwestern Pennsylvania. They also found that macroinvertebrate taxa richness was negatively correlated with the number of well pads within a watershed. Our impacted streams also had greater percentages of tolerant taxa than unimpacted streams. Similarly, Johnson et al. (2015) detected an increase in tolerant taxa in streams where more well pads were closer to stream channels. The changes we detected in benthic macroinvertebrate communities were correlated with waterthrush territory density. Specifically, more waterthrush established territories along higher quality streams (those with higher GLIMPSS scores) and where our macroinvertebrate biomass and density metrics were greater. Mattsson and Cooper (2006) found that waterthrush occupancy was greater along streams with higher % EPT and greater macroinvertebrate biomass. Waterthrush territories typically are established in early April to early May, while young are hatched and raised in late May and June on our study area (Prosser and Brooks 1998). Our benthic sampling occurred in early May, which more closely coincided with territory establishment. The more similar timing may have contributed to our finding that territory density was strongly correlated with benthic metrics while nest survival was not. Waterthrush DSR was most strongly influenced by nest stage, which is common in songbirds (Dinsmore et al. 2002; Grant et al. 2005; Burhans et al. 2010). In future studies with a larger sample of nests, an informative next step would be to examine associations between DSR and December 2016 | Volume 7 | Issue 2 | 429


P.B. Wood et al.

Figure 3. Scatterplots showing relationship of Louisiana waterthrush Parkesia motacilla territory density to Ephemeroptera, Plecoptera, Trichoptera (EPT) density (top panel) and the relationship of territory density and clutch size to genus-level index of most probable stream status (GLIMPSS) scores (bottom two panels) during 2011 on the Lewis Wetzel Wildlife Management Area, West Virginia.




Table 4. Model selection results for daily survival rate (DSR) of Louisiana waterthrush Parkesia motacilla nests during 2011 on the Lewis Wetzel Wildlife Management Area, West Virginia, using Program MARK and Akaike’s Information Criterion adjusted for small sample sizes (AICc). K is the number of parameters in the model; DAICc is the difference in AICc values between individual models and the top model; and wi is the model weight, which indicates how much of the variation is explained by a given variable. Variables with DAICc 2 were considered to have an influence on DSR and are in bold text. See Table 1 for model notations.

habitat and benthic covariates during each nest stage separately. For example, parental investment theory (Montgomerie and Weatherhead 1988) suggests that impacts such as traffic and noise from well pads and infrastructure may be important correlates of nest failure early in the nesting cycle when females more readily abandon nests. Conversely, benthic covariates may be associated with survival during the nestling stage because nestling waterthrush may rely heavily on heptageniid mayfly taxa for nourishment (Trevelline et Journal of Fish and Wildlife Management | www.fwspubs.org

P.B. Wood et al.

al. 2016). The great influence of the HSI index on DSR was expected because this index is effective at characterizing the overall foraging and nesting habitat of waterthrush (Prosser and Brooks 1998). The strong support for these two covariates in the combined model set effectively swamped out the weaker effect of benthic covariates on DSR observed in the benthic model set. The individual covariates in the benthic model set also may have been weaker because the density and biomass categories were correlated. However, we retained these categories to better evaluate whether certain benthic taxa or size classes related to nest survival because some studies have suggested that waterthrush prefer specific taxa such as EPT (Mattsson et al. 2009) or commonly feed heptageniid mayfly taxa to their nestlings (Trevelline et al. 2016). Additional research comparing benthic samples to nest survival within individual territories is needed. In our study, we sampled near the base of streams, which characterizes the benthic community of the overall stream but appears to be less directly linked to individual waterthrush territories and nests. At the end of our study in 2011, the study area still had ~93% mature forest cover and little forest fragmentation had occurred. Each stream had ,50% of its length impacted in some way by gas development activities (mean ¼ 22%). Consequently, unimpacted stream segments and side tributaries occurred on every stream, providing the opportunity for waterthrush to forage and place nests in unimpacted sections of their territories. Waterthrush also are known to compensate for the loss of food resources by increasing their territory sizes, foraging in nearby unimpacted areas, and foraging off-stream (Mulvihill et al. 2008; Frantz and Wood unpublished data). These factors also likely contributed to the low importance of benthic variables on waterthrush nest survival. In summary, our study found that benthic macroinvertebrate communities changed on streams impacted by Marcellus shale gas development and that waterthrush demography responded to this change even though relatively low levels of canopy disturbance and gas development activities had occurred on our study area by 2011. These are some of the first analyses of the response of aquatic biota to shale gas extraction in the Appalachian Region and reveal how territory density and clutch size of an important avian habitat specialist of conservation concern are linked to changes in its aquatic resource base. We are not aware of any other studies that have simultaneously examined waterthrush demography and benthic macroinvertebrate communities in response to shale gas development. Thus, although our study spanned only one breeding season, the results provide important baseline information and suggest that further in-depth research is needed to better assess response and to develop mitigation measures. Proactive development that locates well pads and infrastructure farther from stream channels and slows development in stream catchments that support species of concern may reduce the potential for negative impacts to streams and their associated biota (Davis and Robinson 2012; Johnson et al. 2015). December 2016 | Volume 7 | Issue 2 | 431


Supplemental Material Table S1. Data used in correlation and t-test analyses of Louisiana waterthrush Parkesia motacilla and benthic macroinvertebrate data collected from 12 headwater streams in 2011 on the Lewis Wetzel Wildlife Management Area, West Virginia. Found at DOI: http://dx/doi.org/10.3996/092015JFWM-084.S1 (13 KB XLSX). Reference S1. Barbour MT, Gerritsen J, Snyder BD, Stribling JB. 1999. Rapid bioassessment protocols for use in streams and wadeable rivers: periphyton, benthic macroinvertebrates and fish. Second edition. EPA 841-B99-002. Washington, D.C.: U.S. Environmental Protection Agency, Office of Water. Found at: http://nepis.epa.gov/Exe/ZyPDF.cgi/ 20004OQK.PDF?Dockey¼20004OQK.PDF (10373 KB PDF). Also found at DOI: http://dx/doi.org/10.3996/092015JFWM-084.S2 (10376 KB PDF).

Acknowledgments West Virginia Division of Natural Resources provided access to the study area and Wheeling Jesuit University provided access to field housing. We thank Debbie Archer, Darin Blood, and Jim Sheehan for field assistance. Kyle Aldinger, Jeremy Mizel, and Jim Sheehan assisted with data summaries and analyses. We thank Greg Pond and Kelly Krock (USEPA, Region III) for processing and identifying benthic macroinvertebrates and Greg Pond for summaries and assistance with analyses of the benthic data. Kyle Aldinger, Greg Pond, Jim Sheehan, Brian Trevelline, the Associate Editor, and three anonymous reviewers provided helpful comments on this manuscript. Banding was conducted under U.S. Geological Survey banding permit no. 23412. This study was completed under the auspices of West Virginia University IACUC protocol 04-0302, 07-0303. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

References Adams MB, Ford MW, Schuler TM, Thomas-Van Gundy M. 2011. Effects of natural gas development on forest ecosystems. Proceedings of the 17th Central Hardwood Forest Conference GTR-NRS-P-78:219–226. Available: http://www.nrs.fs.fed.us/pubs/gtr/gtr-p78papers/23adamsp78.pdf (February 2016). Askins, RA, Lynch JF, Greenberg R. 1990. Population declines in migratory birds in eastern North America. Current Ornithology 7:1–57. Baker BA, Warner TA, Conley JF, McNeil BE. 2013. Does spatial resolution matter? A multi-scale comparison of object-based and pixel-based methods for detecting change associated with gas well drilling operations. Journal of Fish and Wildlife Management | www.fwspubs.org

P.B. Wood et al.

International Journal of Remote Sensing 34:1633– 1651. Barbour MT, Gerritsen J, Snyder BD, Stribling JB. 1999. Rapid bioassessment protocols for use in streams and wadeable rivers: periphyton, benthic macroinvertebrates and fish. Second edition. EPA 841-B-99-002. Washington, D.C.: U.S. Environmental Protection Agency, Office of Water (see Supplemental Material, Reference S1, http://nepis.epa.gov/Exe/ZyPDF.cgi/20004OQK. PDF?Dockey¼20004OQK.PDF (February 2016). Benke AC, Huryn AD, Smock LA, Wallace JB. 1999. Length–mass relationships for freshwater macroinvertebrates in North America with particular reference to the southeastern United States. Journal of the North American Benthological Society 18:308–343. Bibby CJ, Burgess ND, Hill DA, Mustoe SH. 2000. Bird census techniques. 2nd edition. London: Academic Press. Bond NR, Downes BJ. 2003. The independent and interactive effects of fine sediment and flow on benthic invertebrate communities characteristic of small upland streams. Freshwater Biology 48:455–465. Brittingham MC, Maloney KO, Farag AM, Harper DD, Bowen ZH. 2014. Ecological risks of shale oil and gas development to wildlife, aquatic resources and their habitats. Environmental Science and Technology 48:11034–11047. Brooks RP, O’Connell TJ, Wardrop DH, Jackson LE. 1998. Towards a regional index of biological integrity: the example of forested riparian ecosystems. Environmental Monitoring and Assessment 51:131–143. Burhans DE, Root BG, Shaffer TL, Dey DC. 2010. Songbird nest survival is invariant to early-successional restoration treatments in a large river floodplain. Wilson Journal of Ornithology 122:307–317. Burnham KP, Anderson DR. 2002. Model selection and multimodel inference: an information-theoretic approach. New York: Springer-Verlag. Cohen J. 1988. Statistical power analysis for the behavioral sciences. Second edition. Mahwah, New Jersey: Lawrence Erlbaum Associates. Cunningham MA, Johnson DH. 2006. Proximate and landscape factors influence grassland bird distributions. Ecological Applications 16:1062–1075. Davis JB, Robinson GR. 2012. A geographic model to assess and limit cumulative ecological degradation from Marcellus Shale exploitation in New York, USA. Ecology and Society 17:25. Available: http://dx.doi.org/ 10.5751/ES-04822-170225 (February 2016). Dinsmore SJ, White GC, Knopf FL. 2002. Advanced techniques for modeling avian nest survival. Ecology 83:3476–3488. Drohan PJ, Brittingham M, Bishop J, Yoder K. 2012. Early trends in landcover change and forest fragmentation due to shale-gas development in Pennsylvania: a potential outcome for the northcentral Appalachians. Environmental Management 49:1061–1075. Farwell LS, Wood PB, Sheehan J, George GA. 2016. Shale gas development effects on the songbird community in a central Appalachian forest. Biological Conservation 201:78–91. December 2016 | Volume 7 | Issue 2 | 432


Flaspohler DJ, Temple SA, Rosenfeld RN. 2001. Species specific edge effects on nest success and breeding bird density in a forested landscape. Ecological Applications 11:32–46. Grant CJ, Weimer AB, Marks NK, Perow ES, Oster JM, Brubaker KM, Trexler RV, Solomon CM, Lamendella R. 2015. Marcellus and mercury: assessing potential impacts of unconventional natural gas extraction on aquatic ecosystems in northwestern Pennsylvania. Journal of Environmental Science and Health, Part A: Toxic/Hazardous Substances and Environmental Engineering 50:482–500. Grant TA, Shaffer TL, Madden EM, Pietze PJ. 2005. Timespecific variation in passerine nest survival: new insights into old questions. Auk 122:661–672. Johnson E, Austin BJ, Inlander E, Gallipeau C, Evans-White MA, Entrekin S. 2015. Stream macroinvertebrate communities across a gradient of natural gas development in the Fayetteville Shale. Science of the Total Environment 530–531:323–332. Kargbo DM, Wilhelm RG, Campbell DJ. 2010. Natural gas plays in the Marcellus shale: challenges and potential opportunities. Environmental Science and Technology 44:5679–5684. Kiviat E. 2013. Risks to biodiversity from hydraulic fracturing for natural gas in the Marcellus and Utica shales. Annals of the New York Academy of Sciences 1286:1–14. Ligeiro R, Ferreira W, Hughes RM, Callisto M. 2013. The problem of using fixed-area subsampling methods to estimate macroinvertebrate richness: a case study with Neotropical stream data. Environmental Monitoring and Assessment 85:4077–4085. Marcellus Center for Outreach and Research. 2015. Tri-state unconventional shale wells (PA, WV, OH). Available: http://www.marcellus.psu.edu/resources/maps.php. Last updated March 31, 2015 (February 2016). Martin TE. 1992. Interaction of nest predation and food limitation in reproductive strategies. Current Ornithology 9:163–197. Martin TE, Geupel GR. 1993. Nest-monitoring plots: methods for locating nests and monitoring success. Journal of Field Ornithology 64:507–519. Martindale S. 1982. Nest defense and central place foraging: a model and experiment. Behavioral Ecology and Sociobiology 10:85–89. Mattsson BJ, Cooper RJ. 2006. Louisiana waterthrushes (Seiurus motacilla) and habitat assessments as costeffective indicators of instream biotic integrity. Freshwater Biology 51:1941–1958. Mattsson BJ, Cooper RJ. 2009. Multiscale analysis of the effects of rainfall extremes on reproduction by an obligate riparian bird in urban and rural landscapes. Auk 126:64–76. Mattsson BJ, Master TL, Mulvihill RS, Robinson WD. 2009. Louisiana waterthrush (Seiurus motacilla). Account 151 in Poole A, editor. The birds of North America online. Ithaca, New York: Cornell Lab of Ornithology; Retrieved from the Birds of North America Online. Available: http://bna.birds.cornell.edu.bnaproxy.birds. cornell.edu/bna/species/151 (February 2016).


P.B. Wood et al.

Montgomerie RD, Weatherhead PJ. 1988. Risk and rewards of nest defence by parent birds. Quarterly Review of Biology 63:167–187. Mulvihill RS, Newell FL, Latta SC. 2008. Effects of acidification on the breeding ecology of a streamdependent songbird, the Louisiana waterthrush (Seiurus motacilla). Freshwater Biology 53:2158–2169. Paton PWC. 1994. The effect of edge on avian nest success: how strong is the evidence? Conservation Biology 8:17–26. Petit LJ, Petit DR. 1996. Factors governing habitat selection by prothonotary warblers: field tests of the Fretwell– Lucas models. Ecological Monographs 66:367–387. Pond GJ, Bailey JE, Lowman BM, Whitman MJ. 2013. Calibration and validation of a regionally and seasonally stratified macroinvertebrate index for West Virginia wadeable streams. Environmental Monitoring and Assessment 185:1515–1540. Price FE, Bock CE. 1983. Population ecology of the dipper (Cinclus mexicunus) in the front range of Colorado. Studies in Avian Biology 7:1–84. Prosser DJ, Brooks RP. 1998. A verified habitat suitability index for the Louisiana waterthrush. Journal of Field Ornithology 69:288–298. Rotella JJ, Dinsmore SJ, Shaffer JA. 2004. Modeling nestsurvival data: a comparison of recently developed methods that can be implemented in MARK and SAS. Animal Biodiversity and Conservation 27:187–205. Sabo JL, Bastow JL, Power ME. 2002. Length–mass relationships for adult aquatic and terrestrial invertebrates in a California watershed. Journal of the North American Benthological Society 21:336–343. Sauer JR, Hines JE, Fallon JE, Pardieck KL, Ziolkowski DJ Jr, Link WA. 2012. The North American breeding bird survey, results and analysis 1966–2011. Version 12.13.2011. Laurel, Maryland: U.S. Geological Survey Patuxent Wildlife Research Center. Available: http:// www.mbr-pwrc.usgs.gov/bbs/bbs2011.html (February 2016). Souther S, Tingley MW, Popescu VD, Hayman DTS, Ryan ME, Graves TA, Hartl B, Terrell K. 2014. Biotic impacts of energy development from shale: research priorities and knowledge gaps. Frontiers in Ecology and the Environment 12:330–338. Trevelline BK, Latta SC, Marshall LC, Nuttle TJ, Porter BA. 2016. Molecular analysis of nestling diet in a longdistance Neotropical migrant, the Louisiana waterthrush (Parkesia motacilla). Auk 133:415–428. [WVDEP] West Virginia Department of Environmental Protection. 2011. 2011 standard operating procedures. Charleston, West Virginia: WVDEP, Division of Water and Waste Management, Watershed Assessment Branch. Available: http://www.dep.wv.gov/WWE/ watershed/Pages/WBSOPs.aspx (February 2016). White GC, Burnham KP. 1999. Program MARK: survival estimation from populations of marked animals. Bird Study 46 Supplement:120–138. Available: http:// warnercnr.colostate.edu/~gwhite/mark/mark.htm (February 2016).
