Predator Presence Influences Survival and Behavior of Translocated Stream Fish in Ponds

In many areas of the world, especially ranching or agricultural areas, freshwater ecosystems have been altered by the construction of small impoundments or farm ponds. Past studies have linked farm ponds to declines in native species, primarily through the stocking and subsequent escape and spread of predatory sport fish (Schrank et al. 2001; Perkin et al. 2016; Hedden et al. 2018). Slowing these declines will require multifaceted conservation strategies. One such approach in species conservation in regions with high densities of farm ponds might be the translocation or repatriation of threatened species to these environments coupled with control of sportfish communities, thereby establishing self-sustaining populations that buffer losses due to anthropogenic alterations (Copp et al. 2007; Thomson and Berry 2009; Schumann et al. 2020; Pfaff et al. 2023). Translocation of wild individuals from stable populations can be a useful tool to increase the redundancy of populations and prevent local extirpation while maintaining genetic diversity (Minckley 1995; Fischer and Lindenmayer 2000; Seddon et al. 2007).

The concept of our study arose from our efforts to conserve the federally endangered Topeka Shiner Notropis topeka in the Flint Hills region of Kansas. Topeka Shiner inhabit small headwater streams; however, they also use in-channel and near-channel natural (e.g., oxbows) or anthropogenically created (e.g., excavated ponds also known as dugouts and impoundments) habitats in the Flint Hills (Hedden et al. 2021) and other parts of their range (Thomson and Berry 2009; Pierce et al. 2019). Topeka Shiner often nest-associate with sunfishes, particularly Orangespotted Sunfish Lepomis humilis, which might improve their nesting success in ponds (Campbell et al. 2016). Further, when pond environments are periodically connected to streams, Topeka Shiners might escape farm ponds and supplement stream populations (Hedden et al. 2021).

One hurdle to establishing native fish populations in ponds is negative interactions with stocked predators. In addition to other factors, Topeka Shiner rarely coexist with predators such as Largemouth Bass Micropterus salmoides (Schrank et al. 2001; Mammoliti 2002; Campbell et al. 2016) and avoid them in an experimental setting (Knight and Gido 2005). However, the impact of predator–prey relationships varies across environmental conditions and spatial and temporal scales. Largemouth Bass and Smallmouth Bass Micropterus dolomieu presence in lakes negatively affects the abundance of some minnow species (MacRae and Jackson 2001; Jackson 2002) and can cause local extirpations (Kimberg et al. 2014; Van Der Walt et al. 2016; Pereira and Vitule 2019). In streams, Largemouth Bass presence can also decrease overall species richness (Van Der Walt et al. 2016), though not in all cases (Bruckerhoff et al. 2021a, 2021b). These impacts result from mortality from direct consumption (Schlosser 1988; Harvey 1991; Steinmetz et al. 2008; Marsh‐Matthews et al. 2013) or nonconsumptive effects such as alterations in behavior, which influences survival (Skalski and Gilliam 2002; Asaeda et al. 2007) or reproduction (Fraser and Gilliam 1992; Sheriff et al. 2020) and may impart a stronger interaction than direct consumption (Holomuzki et al. 2010). Translocation of native fishes to farm ponds may be of great conservation benefit for some species, but limitations on translocation success have not been thoroughly tested and are based on mostly correlative data (Lamothe and Drake 2019).

Experiments that track the success of translocated native fish into ponds are necessary to better understand factors limiting their success in these habitats. Although we were interested in testing limitations in translocation success of federally endangered Topeka Shiner, we conducted experiments with Bluntnose Minnow Pimephales notatus as a surrogate because its populations in the region are more stable. Both species are known to inhabit ponds (Hedden et al. 2021; Pfaff et al. 2023) and share similar morphology, feeding, and reproductive life histories, including nest association (Stark et al. 2002; Kansas Fishes Committee 2014). Additionally, Bluntnose Minnow show similar responses to predator presence as Topeka Shiner (Knight and Gido 2005) and other native minnow species (Bruckerhoff et al. 2021a, 2021b) in controlled mesocosm experiments.

The specific goal of this study was to quantify the impact predator presence has on translocation success of Bluntnose Minnow. Although we suspected mortality associated with predation would be high, we wanted to first quantify apparent survival of translocated Bluntnose Minnows between ponds with and without bass. Second, we tested if the probability of detections of translocated Bluntnose Minnows differed between ponds containing or not containing Largemouth Bass by means of probability of detection, which could indicate differences in behavior. We modeled apparent survival and probability of detection as differing among ponds, differing by presence of Largemouth Bass or not, and not differing among ponds. We predicted that Largemouth Bass presence would affect both apparent survival and probability of detection of translocated Bluntnose Minnow. Our experiments allowed us to test each factor independently.

We conducted this study in 2020 and 2021 at the Tallgrass Prairie National Preserve, located within the Upper Cottonwood River basin in central Kansas and contains 24 farm ponds. Ten of the 24 ponds contain populations of Largemouth Bass. In 2020, we used four ponds that contained Largemouth Bass (ponds 3, 10, 15, and 19) and four non-Largemouth Bass ponds that did not contain bass (ponds 5, 8, 9, and 22; Figure 1) to test if Largemouth Bass affected survival and probability of detection of Bluntnose Minnow. Two of the non-Largemouth Bass ponds completely dried (ponds 5 and 8) and one pond nearly went dry (Pond 9) during summer 2020. As such, we excluded 2020 data from mark–recapture analysis. We replicated the experiment the following year using the same bass ponds while substituting two nonbass ponds more likely to maintain water throughout the experiment (ponds 4 and 11 were included, whereas 5 and 8 were no longer used). Unfortunately, Pond 9 experienced total drying, resulting in loss of all individuals in 2021 and was excluded from 2021 analysis. The ponds included in this study were of similar size and habitat structure. Fish tagged in 2020 were only used in the analyses for 2020 and not those in 2021. Non-Largemouth Bass ponds averaged 0.7 ha (standard deviation 0.2), whereas bass ponds averaged 0.64 ha (0.3). All ponds were vegetated around the margin, with the level of vegetation increasing though the summer. No ponds contained wood or had overhanging trees. Additionally, all ponds had primarily silt substrate with pockets of courser substrate near the dam. All ponds except for Pond 11 contained Green Sunfish Lepomis cyanellus. Pond 11 contained Yellow Bullhead Ameiurus natalis, whereas ponds 3 and 22 also contained Bluegill Lepomis macrochirus. All ponds except Pond 10 were in grazed areas and accessed by cattle.

We stocked each pond with 100 Bluntnose Minnows captured from a robust and naturally occurring population (Pond 11) and implanted with 8-mm passive integrated transponder (PIT) tags (Mini HPT8 Pit Tag, Biomark, Boise, ID). We implanted individuals ≥46 mm total length with PIT tags following standard procedures that yield high survivorship and tag retention (Pennock et al. 2016, Schumann et al. 2020). We monitored implanted individuals for 1 h and replaced individuals that did not survive. Overall, 98.6% of individuals survived the 1-h monitoring period. The average total length of Bluntnose Minnow was 53 mm and average total lengths released in each pond were between 49 and 55 mm. We conducted all tagging on 2 consecutive days each year that shared similar weather conditions. We released all Bluntnose Minnow on a left or right corner of the dam.

Two weeks after the release of minnows, we began to monitor populations with submersible PIT antennas to quantify detections over time. The first year we used a single circular (1-m-diameter) submersible PIT tag antenna (RM310, Biomark) with a read range of approximately 20 cm with 8-mm PIT tags. We placed antennas at approximately 1 m depth near the corner of the dam in each pond. We deployed antennas in four of the eight ponds for 1 wk and rotated to the next four ponds for 1 wk because of a limit on the number of antennas available. We repeated this process after 4 wk for a total of three sampling periods. Sampling occurred between May and August 2020. Because of low detections in bass ponds, we extended the duration of submersible antenna deployment for all ponds in 2021 to 2 wk and included four sampling periods such that each antenna was deployed every 6 wk from May to September. Additionally, we built a kayak-mounted PIT antenna that was designed to actively detect fish in ponds. We did this to explore the possibility of fish being present but not active. Rather than using a floating antenna, as is common in other studies (Zentner et al. 2021), our antenna swivels from the front of the kayak, allowing it to dip down through the water column and sample a depth of up to 2 m (Figure S1). We constructed the antenna using stranded wire enclosed in a frame of 2.54-cm-thick polyvinyl chloride tubing. The antenna was composed of two separate inductance loops, each measuring 2 m by 0.66 m and held together with a 0.66-m-long polyvinyl chloride tube segment. This gave us an effective size of a 2 × 2 m antenna though the use of two inductance loops, which made it much easier to detect PIT tags moving through the middle of the antenna. The antenna was powered using a Biomark IS1001 PIT tag reader and 24-V lithium-ion battery fitted into a custom waterproof enclosure. The read range of the antenna using 8-mm PIT tags was approximately 25 cm depending on tag orientation. We sampled each pond using the kayak antenna before submersible deployment in the first three sampling periods.

We used accumulation curves to visualize differences in the number of fish detected between treatments and ponds. We also used Cormack–Jolly–Seber mark–recapture models (Pledger et al. 2003) using Program MARK v. 10.1 (White and Burnham 1999) to estimate apparent survival and probability of detection across redetection periods. We only used stationary antenna data in creating these models. The Cormack–Jolly–Seber model is an open-population live recapture model allowing for changes in population size through mortality and emigration. We selected this model because apparent survival, through mortality or emigration, and probability of detection were the estimates of interest in this study. We used model selection in Program Mark on the basis of Akaike information criterion corrected for small sample size (AICc). Before model selection, we tested goodness of fit on the full model using the median c-hat function in Program Mark to estimate overdispersion, which may indicate that model assumptions are violated (Cooch and White 2014). Values for c-hat between 1 and 3 indicate minor overdispersion, which requires the use of a quasi-likelihood adjusted AICc (Cooch and White 2014). Two nonbass ponds experienced total or partial drying in 2020, leading to no Bluntnose Minnow being detected in Pond 5 in periods 2 and 3 and no detection in Pond 7 in period 3. We were unable to model apparent survival and probability of detection in 2020 because of a lack of replication and overall detections. We only used data from Bluntnose Minnow that were PIT tagged and released in 2021 in modeling. Although our systems are open and periodically connected to the stream when pond level is high enough, no such overflow occurred during the second year of the study. Therefore, we assumed the estimate of apparent survival to be indicative of true survival. Additionally, we treated all detections as live individuals. Using stationary antennas, the only way to detect a “ghost tag” (a shed tag or a dead individual) would be if the tag was near the antenna and would be continually detected until the antenna was moved. We did not encounter such an event during this study. We modeled apparent survival and probability of detection as being constant or differing between treatment as per our hypotheses. We assumed that declines in apparent survival in bass ponds would reflect predation and lower probability of detection would reflect reduced movement in the presence of bass. This study took place in nature and not laboratory conditions and each pond had slightly different conditions including water depth, existing fish community, and densities of species within those communities. As such, we used pond identification as a covariate to test if differences in apparent survival or probability of detection may be attributed to confounding factors.

We used the amount of time we detected an individual Bluntnose Minnow at an antenna within a single day to further explore if behavior was different among treatments. We set antennas to detect unique individuals every 60 s, providing a fine-scale temporal resolution of fish presence around the antennas. Thus, a greater number of detections of a single individual within a day should reflect less activity because the individual stayed in the same location (i.e., close to the antenna) for a greater amount of time.

Of the 1,600 tagged Bluntnose Minnow (800 in Largemouth Bass ponds and 800 in non-Largemouth Bass ponds), the mean number of individuals detected on stationary receivers in non-Largemouth Bass ponds was 20 times greater than in Largemouth Bass ponds (Figure 2). We only detected two Bluntnose Minnow, both in the first time period, in Largemouth Bass ponds in 2020. In contrast, we detected 73 individuals in non-Largemouth Bass ponds across all time periods. In 2021, we detected 15 individual Bluntnose Minnow, 12 in the first period and three in the fourth period, in bass ponds. In contrast, we detected 252 individual Bluntnose Minnow in non-Largemouth Bass ponds across all time periods including 51 in the fourth period alone.

Median c-hat estimation of the full model (1.88) indicated minor overdispersion and we used quasi-likelihood AICc for model comparison. Model selection revealed competing top Cormack–Jolly–Seber models; apparent survival and probability of detection differed by treatment and apparent survival differed by treatment, whereas probability of detection differed by pond (Table 1). The model estimates indicate that apparent survival was higher in non-Largemouth Bass ponds than in Largemouth Bass ponds (Figure 3A). Apparent survival for non-Largemouth Bass ponds (76%) was approximately double that of Largemouth Bass ponds (44%). Probability of detection estimates for non-Largemouth Bass ponds (35.4 and 55.1%) were generally higher than estimates from Largemouth Bass ponds (2.6 to 11.4%; Figure 3B). The 95% confidence intervals were overlapping between one Largemouth Bass pond (3) and one non-Largemouth Bass pond (11).

We found that Bluntnose Minnow were more active in non-Largemouth Bass ponds than in Largemouth Bass ponds, as inferred by the amount of time individuals spent within detection range of PIT antennas. Activity was strongly bimodal in Largemouth Bass ponds, with individuals averaging 165.5 min/d within range of the antenna on days they were detected (Figure 4A). In non-Largemouth Bass ponds, activity was more uniformly distributed, averaging only 5.9 min/d within detection range of the antenna. Excluding the first redetection period, when individuals may have still been acclimatizing to the ponds, the contrast between treatments increases to 436.0 min in Largemouth Bass ponds and decreases to 4.7 min in non-Largemouth Bass ponds (Figure 4B).

We detected only 16 individuals across all pond and time periods using the mobile antenna, 8 individuals in Largemouth Bass ponds and 8 individuals in non-Largemouth Bass ponds. We found that the patterns of detections from the mobile antenna, albeit much less successful than stationary antennas, were consistent and suggested less minnow activity in Largemouth Bass ponds. Of the eight individuals detected in Largemouth Bass ponds with the mobile antenna, we only detected one on a stationary antenna, whereas in non-Largemouth Bass ponds only one of the eight detected individuals was not detected on a stationary antenna. These data support the hypothesis that lower detections in Largemouth Bass ponds is at least partially due to reduced activity, leading to lower detection probability.

Our results indicate that the apparent survival of translocated Bluntnose Minnow was nearly twice as high in ponds that do not contain Largemouth Bass (76% in non-Largemouth Bass ponds and 44% in Largemouth Bass ponds). Although direct predation by Largemouth Bass can strongly influence minnow survival (Schlosser 1988; Harvey 1991; Van Der Walt et al. 2016), relatively few studies have directly assessed true survival or apparent survival rates of translocated individuals and none in treatments of habitats with or without predators. Apparent survival in non-Largemouth Bass ponds (76% over the 20-wk study period in 2021) was higher than apparent survival estimates reported in other studies with Razorback Suckers Xyrauchen texanus (between 0 and 12% over a 6-mo period; Webber and Haines 2014) and Humpback Chub Gila cypha (between 22 to 41% estimated annual apparent survival; Spurgeon et al. 2015). These studies took place in lotic systems and over varying lengths of time and did not control for presence or absence of predators, which could account for the lower apparent survival rates.

In this study, we assessed only the influence of predator presence on apparent survival and behavior; however, observed differences among non-Largemouth Bass ponds, including permanency, highlight the importance of considering abiotic factors when selecting ponds for translocation. For example, Pond 9 was the smallest among non-Largemouth Bass ponds and dried in 2021, eliminating the translocated population by redetection period four. In 2020, two of the non-Largemouth Bass ponds went dry, resulting in the loss of all translocated individuals. Stream fish tend to inhabit intermediate-sized ponds that have some degree of permanency (Pfaff et al. 2023), and great care should be taken to select ponds that are in this intermediate zone when considering conservation action. Other interpond factors that may influence survival or continued fitness include habitat and water quality. Yates et al. (2019) found that temperature and pH can limit fitness of small, translocated fish populations. Similarly, of 148 Gila Topminnow Poeciliopsis occidentalis translocations, habitat was one of the most important factors predicting persistence time of translocated populations (Sheller et al. 2006). Although habitat factors may still play an important role in translocation success, direct predation likely influenced minnow survival in this study and differences between treatments were greater than that within treatments when excluding ponds that dried. Differences also existed among ponds in existing communities and densities. Apparent survival estimates were similar for non-Largemouth Bass ponds that contained or did not contain Green Sunfish; however, this interaction was not the focus of this study and additional replication is needed to assess this interaction.

Our results also suggest that predator presence influences the probability of detection and potential behavior of translocated minnows. Largemouth Bass presence led to lower probability of detection compared with non-Largemouth Bass ponds. Additionally, detected individuals in Largemouth Bass ponds spent 28 times longer within antenna range than those in non-Largemouth Bass ponds, indicating less total movement when Largemouth Bass were present. This pattern intensified over the course of the study. Excluding the first redetection period, individuals detected in Largemouth Bass ponds spend nearly 100 times longer within antenna range than those in non-Largemouth Bass ponds. Further, the redetection distribution in Largemouth Bass ponds was bimodal. No individuals redetected in period one for <10 min/d were redetected in subsequent periods. Moreover, fish detected on mobile antennas were much less likely to be captured on stationary antennas in Largemouth Bass ponds relative to non-Largemouth Bass ponds. Those individuals that survived past the first time period may have modified their behavior to remain sheltered and less active. This finding is consistent with past literature. Knight and Gido (2005) found significant differences in habitat use and response to Largemouth Bass by four prey species, including both Bluntnose Minnow and Topeka Shiner, in a mesocosm experiment. Further, they found that when cover was present, Bluntnose Minnow would become inactive and stay within cover exclusively. Similarly, Bruckerhoff et al. (2021a, 2021b) found significant reductions in movement of Red Shiner Cyprinella lutrensis and Bluntnose Minnow in mesocosms when Largemouth Bass were present. Changes in behavior when predators are present can influence overall movement and foraging and spawning success (Fraser and Gilliam 1992; Katano and Aonuma 2002; Divino and Tonn 2007; Peterson and Kitano 2021). Thus, mortality and altered behavior present a challenge to any minnow population that is translocated into a pond containing Largemouth Bass.

The mobile antenna was more difficult to glean useful information from because it had poor detection probability, and it might detect tags that were from dead fish (i.e., ghost tags). Regardless, most individuals detected in Largemouth Bass ponds were not detected on stationary antennas (seven of eight), whereas the same pattern was reversed in non-Largemouth Bass ponds, with most individuals also being detected on stationary antennas (seven of eight). This could be due to detecting ghost tags in Largemouth Bass ponds that were shed after the fish was consumed by a Largemouth Bass. Given the relatively narrow read range of our antenna and silty substrate in the ponds, we do not suspect this is the case, as those tags would likely settle into the sediment at the bottom of the pond and not be detected by our mobile antenna. Alternatively, it could be due to having a higher number of unique detections in non-Largemouth Bass ponds; thus more mobile detections would be represented in the stationary antenna data. This, coupled with lower minnow movement in Largemouth Bass ponds, would lead to the observed pattern. Lentic pond environments may not be conducive to mobile PIT antenna monitoring because fish are able to move away from the antenna in any direction. Mobile PIT antennas are commonly used in lotic systems that are by nature more constrained, though detection rates can be highly variable across studies (Musselman et al. 2017; Zentner et al. 2021). Studies using PIT tags in lentic estuary environments tend to have low probability of detection: 1.0 to 4.0% (Ledgerwood et al. 2006), 0.9 to 3.5% (Morris et al. 2018), 0.7 to 2.2% (Holcombe et al. 2019). Additionally, sampling efficiency in ponds is likely low because of difficulty in moving the mobile antenna over and through vegetation cover. Overall, this method was not effective in the lentic systems in this study.

Another limitation of the present study is that we only consider interaction with a single predator and do not consider other predatory species such as Green Sunfish, bullheads, or avian predators that might have been present in or around these ponds. Among the non-Largemouth Bass ponds, Pond 4 and Pond 22 contained Green Sunfish, whereas Pond 11 contained Yellow Bullhead. Whereas Green Sunfish may influence survival rates of some minnows (Marsh‐Matthews et al. 2013), ponds 4 and 22 had the highest apparent survival rates among all ponds, including fishless ponds (5, 7, and 9). In addition to predation from other fish species, avian predators act as structuring mechanism in shallow habitats (Power 1984; Power et al. 1989). The number of fish removed by avian predators accounted for more than those consumed by piscivorous fish in shallow eutrophic lakes (Winfield 1990). Bird exclusion experiments indicate that avian predators may even be the top predators in a variety of habitats including shallow lentic systems (Steinmetz et al. 2003). Further, altered behavior due to avian predation threat can lead to less foraging and slower growth in minnows (Allouche and Gaudin 2001). All ponds included in this study, however, were of similar depth and likely had similar rates of avian predation. The shallowest ponds included in this study (5, 7, and 9) would likely have the greatest risk of avian predation, though habitat loss through drying poses a greater risk to translocation failure, as we observed.

This study suggests that the presence of Largemouth Bass negatively affected translocated Bluntnose Minnow apparent survival and probability of detection and suggests that translocation efforts to ponds with Largemouth Bass are less likely to be successful. In addition to mortality through direct predation, altered behavior might change habitat use, feeding behavior, and spawning success, all of which can negatively influence the persistence of translocated populations. These indirect effects may be stronger and more impactful to translocated population survival than direct effects through predation. It is important to note that predation by Green Sunfish was likely occurring in two of the non-Largemouth Bass ponds and that the apparent survival estimates in this study may not represent the best-case scenario for minnows translocated to ponds. Although it is important to take predator presence into account when selecting translocation sites, other factors such as habitat stability and water permanency also need to be considered. Translocation is a powerful conservation tool and selecting optimum target locations is key in determining translocation and conservation success for fish species such as Topeka Shiner. Further, removing existing Largemouth Bass from small ponds may increase the habitat available to Topeka Shiner for unassisted colonization, thereby further helping in the conservation of this species.

Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding author for the article.

Figure S1. Picture of mobile kayak-mounted passive integrated transponder (PIT) antenna used to detect PIT tag implanted Bluntnose Minnow Pimephales notatus at the Tallgrass Prairie National Preserve from 2020 to 2021. Note the forward antenna attachment, which allows the rear portion of the antenna to dip down to a depth of 2 m.

Available: https://doi.org/10.3996/JFWM-22-069.S1 (1,620 KB DOC)

Data S1. Supplemental data of tagged Bluntnose Minnow Pimephales notatus, recapture histories, and minutes per day individuals were redetected at the Tallgrass Prairie National Preserve from 2020 to 2021. Record of tagged Bluntnose Minnow is contained in tab “tagged PIMNOT.” This tab contains the unique identifier of each passive integrated transponder (PIT)-tagged individual (Tag), if they were tagged or not (Tagged), year of tagging (Year), species tagged (Species), length at time of tagging (Length), original capture location (Origin), and translocation destination (Release). Recapture histories are contained in tab “Redetections by Period.” This tab includes the unique identifier of each PIT-tagged individual (Tag), year of tagging (Year), species tagged (Species), translocation destination (Release Pond), recapture year (Year), and recapture period within each year (Round). Minutes per day individuals were redetected is contained in tab “Redetection Minutes.” This tab contains the tagging year (Tag Date), species tagged (Species), translocation destination (Release Pond), pond treatment (Treatment), the unique identifier of each PIT-tagged individual (HEX Tag ID), recapture period (Round), date of recapture (Scan Date), and total minutes an individual was redetected on each day (Total Minutes).

Available: https://doi.org/10.3996/JFWM-22-069.S2 (97 KB XLXS)

Reference S1. Holcombe EF, Borsky AJ, Biron JM, Bentley PJ, Sanford BP. 2019. Detection of PIT-tagged juvenile salmonids migrating in the Columbia River Estuary, 2018. Report to the Bonneville Power Administration, Contract 46273 RL58, Portland, Oregon. 2.2 MB.

Available: https://doi.org/10.3996/JFWM-22-069.S3 (1,359 KB PDF)

Reference S2. Ledgerwood RD, Cameron AS, Sandford BP, Way LB, Matthews GM. 2006. Detection of PIT-tagged juvenile salmonids in the Columbia River Estuary using pair trawls, 2003 and 2004. National Oceanic and Atmospheric Administration, Northwest Fisheries Science Center, Seattle. Report to U.S. Army Corps of Engineers, Contract 56ABNF100030, Walla Walla, Washington. 9.6 MB.

Available: https://doi.org/10.3996/JFWM-22-069.S4 (8,114 KB PDF)

Reference S3. Morris MS, Holcombe EF, Borsky AJ, Bentley PJ, Sandford BP. 2018. Detection of PIT-tagged juvenile salmonids migrating in the Columbia River Estuary, 2017. National Oceanic and Atmospheric Administration, Northwest Fisheries Science Center, Seattle. Report to U.S. Department of Energy, Contract 46273 RL58, Walla Walla, Washington. 2.3 MB.

Available: https://doi.org/10.3996/JFWM-22-069.S5 (1,481 KB PDF)

We thank the National Park Service and the Nature Conservancy for access to study locations. We also thank Darin McCullough, Andrew Hagemann, Kelsey Porter, Trevor Jones, Maddy Siller, and Aiden Masek for field assistance. We thank Elizabeth Renner, Sophia Bonjour, Matt Bogaard, and John Cleveland for field assistance and discussion. We also thank the editors and several anonymous reviewers who provided valuable feedback that greatly improved the manuscript. All applicable institutional and national guidelines for the care and use of animals were followed under Kansas Department of Wildlife and Parks permit no. 842778789, U.S. Fish and Wildlife Service permit no. TE067729-4, and Kansas State University Institutional Animal Care and Use Committee permit no. 4494.

Any use of trade, product, website, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. Government.