Diet and Habitat Comparison of Two Closely Related Darters (Percina bimaculata and Percina caprodes)

Abstract

The Chesapeake Logperch (Percina bimaculata) is a medium-sized darter that has had a limited distribution in Pennsylvania and Maryland. It is a threatened species native to the Susquehanna River and historically occurred in the Potomac River. It is currently being reviewed to determine if it should be listed as federally endangered. There are major efforts to reintroduce the Chesapeake Logperch to where it was historically native. Toward this end, we documented the selected habitats and diets, to aid in the selection of new habitats for reintroduction. Because historical data on habitat selection and diet of the Chesapeake Logperch were not available, we compared these habitats selected to those of the closely related Northern Logperch (P. caprodes semifasciata), and diet to the Ohio Logperch (P. caprodes caprodes). The habitat occupied by the Chesapeake Logperch was only a portion of that occupied by the Northern Logperch. Selection of streams for reintroduction can use data presented herein for the Chesapeake Logperch and historical data for the Northern Logperch. The diet of both the Chesapeake Logperch and the Ohio Logperch indicated that their diets differ, thus the prey items reported herein must be present when selecting putative reintroduction sites. These results will help to create more informed choices for streams considered for the reintroduction of the Chesapeake Logperch.

1. Introduction

The Chesapeake Logperch (Percina bimaculata) is a darter species in the subgenus Percina. Elevated to species status in 2008, it was previously thought to have been a subspecies of the Common Logperch (Percina caprodes) [1]. The current range of the Chesapeake Logperch is the Susquehanna River in both Pennsylvania and Maryland. While previously found in the Potomac River, it has since been extirpated (Figure 1) [2]. The Common Logperch (Percina caprodes) is one of the largest darters in Pennsylvania and found throughout the western portion of the Commonwealth (Figure 1). The range of the Ohio Logperch (P. caprodes caprodes) in Pennsylvania includes the Ohio River Basin and the Northern Logperch (P. caprodes semifasciata) is restricted to the Lake Erie Basin. (Figure 1) [2]. The Northern Logperch was commonly found in the Lake Erie drainage but has since seen a decline, thought to have been brought on by the introduction of the invasive Round Goby (Neogobius melanostomus) [3]. The Ohio Logperch is still commonly found in the Ohio drainage in both Pennsylvania and extending into Ohio [2]. Page and Burr [4] show that the Ohio Logperch extends throughout the Mississippi Drainage and the Northern Logperch is found throughout the northern states and Canada.

Chesapeake Logperch and both subspecies of P. caprodes commonly live in lotic systems, ranging from smaller streams to rivers, such as the Susquehanna and Ohio rivers. There are exceptions, as historically Northern Logperch were abundant in Lake Erie but spawned in its tributaries. Additionally, the Ohio Logperch has been observed by Stauffer in Conneaut Lake, PA. Welsh and Perry [5] described habitat use of the Ohio Logperch and we have observed habitat selection of the Northern Logperch. Both seem to prefer bodies of water with cobble (6–25 cm) and pebbles (0.5–6 cm) but are usually seen in slightly different locations in streams. Chesapeake Logperch have often been found in shallower riffles and sometimes runs at the edge of riffles. Percina caprodes are commonly found in slower moving runs and sometimes pools. Most observations have been made during the time they spend in streams. Limited observations and collections in larger water, such as rivers, have noted them spending time near vegetation, when pebbles and cobble were not present, as well as near larger boulders (>25 cm). One of the focuses of this research is to determine if there are differences in their microhabitats, such as preferred depth or flow rate.

Conservation of species has long been a topic concerning fishery biologists, with arguments ranging from conserving species to more general functional groups [6]. If species that occupy a particular drainage disappear, then the components of species-diversity change. Functional diversity can be seen as one of the more important approaches to conservation; however, diversity of species should still be considered as valuable [7,8,9]. Leitão et al. [10] found that within freshwater ecosystems, some of the rare species, the top 25% rarest fishes in their study, showed functional and habitat conditions that were not occupied by syntopic species. Another approach that is taken is looking at the monetary value of a species. Although attempts have been made, it usually requires a lot of information on the fish and the system it is in. Stauffer and Morgan [11] reviewed studies and hypotheses on valuing fish.

Aside from gaining more information to attempt to assign a value to fishes, it is also key to successful reintroductions. Malone et al. [12] completed a study to develop methods for reintroductions and found that one of the most important factors to a successful reintroduction is matching the selected habitat to the location. Cochran-Biederman et al. [13] had similar considerations that the habitat was one of the most important factors, but that the reason a fish was extirpated in the first place should be the top priority. In addition, Cochran-Biederman et al. [13] also included that the prey availability should be included when considering a reintroduction. To help define the habitat and preferred food of the Chesapeake Logperch, microhabitat and diet studies were completed. In addition, two subspecies of P. caprodes were used as a comparison species to see if there was overlap between selected habitat and diet. If there was possible overlap between selected habitat or diet, it would allow for streams that the two subspecies of P. caprodes inhabit to be considered favorable habitats for potential streams to introduce Chesapeake Logperch. Streams considered for reintroduction could then be evaluated for similarities to streams where either the Northern Logperch or Ohio Logperch are found, leading to potentially higher success rates for reintroductions. Both of these are essential to the ongoing efforts to reintroduce the Chesapeake Logperch to historical locations [14].

2. Materials and Methods

2.1. Study Sites

The West Branch Octoraro and Chiques creeks were sampled for microhabitat data for the Chesapeake Logperch, while the same data were collected for the Northern Logperch in 12-Mile Creek, 20-Mile Creek, and Elk Creek. Fish collected for stomach analysis were collected from West Branch Octoraro Creek, Peter’s Creek, and Fishing Creek for the Chesapeake Logperch and the Shenango River for the Ohio Logperch (Figure 2). All sites were selected based on prior collections of high numbers of Ohio Logperch, or just the presence of Chesapeake Logperch. The presence of high numbers, >50 per 100 m, of Ohio Logperch at a site showed a possible preference for the habitat and diet present. With lower numbers of Chesapeake Logperch, sites were selected based on just the presence of fish where snorkel surveys were completed. Multiple streams were sampled to allow for the possibility of different streams having inherently different habitat. Data from all sites along a unique stream were combined to form the preferred microhabitat or diet for that stream.

The West Branch Octoraro and Chiques creeks are both tributaries of the Susquehanna River in Pennsylvania. They contain mainly gravel with stretches of silt and boulders spread throughout. Their benthic fishes’ communities are similar, consisting of mainly Tessellated Darter (Etheostoma olmstedi), Greenside Darter (Etheostoma blennioides), and various sucker species. Microhabitat data for West Branch Octoraro Creek were collected upstream of the White Rock Forge Covered Bridge (39.8258027; −76.090526). Microhabitat for Chiques Creek were collected near the East Donegal Township Chiques Creek Day Use Area (40.0554753; −76.5184977) (Figure 2).

Tributaries of Lake Erie included 12-Mile, 20-Mile, and Elk Creek in Pennsylvania. All three creeks have similar substrate consisting mainly of bedrock with patches of silt and gravel throughout. They all have similar fish compositions, including Banded Darter (Etheostoma zonale), Greenside Darter, Fantail Darter (Etheostoma flabellare), Rainbow Darter (Etheostoma caeruleum), various sucker species, Mottled Sculpin (Cottus bairdii), and the highly invasive Round Goby. Microhabitats for the Northern Logperch were taken above the Route 5 bridge for 12-Mile Creek (42.20716; −79.91493) and 20-Mile Creek (42.21154; −79.91508). Close to 12-Mile Creek’s mouth (42.21154; −79.91508) and near the Route 5 bridge were taken for 20-Mile Creek (45.26074; −79.78024). Elk Creek was sampled 100 m above the Route 5 bridge (42.007161; −80.354423) and at the first boat launch (42.018760; −80.370763) (Figure 2).

Chesapeake Logperch were collected for a diet study from West Branch Octoraro Creek, Peters Creek, and Fishing Creek. Peters and Fishing Creek are both tributaries of Conowingo Pond in Pennsylvania, and both have similar substrate to West Branch Octoraro Creek. The fish composition is also similar, with sunfishes and some bass also present. Fish collected for diet in West Branch Octoraro Creek were collected at the same time microhabitat surveys were completed. Peter’s Creek was sampled at Stubbs Mill Road (39.761534; −76.227332) and Fishing Creek was sampled near the mouth of the creek (39.792642; −76.261579) (Figure 2).

Ohio Logperch were collected from the Shenango River in Pennsylvania. The river is a tributary of the Beaver River, that leads to the Ohio River. The streambed consists of gravel, silt, and larger cobble. The benthic fishes seen are similar to the other streams in this study, consisting of Banded Dater, Greenside Darter, Johnny Darter (Etheostoma nigrum), as well as Yellow Bullhead (Ameiurus natalis) and Flathead catfish (Pylodictis olivaris). Fish were collected near the Halfway Road Bridge (41.488190; −80.425205) and the Riverside Park (41.409570; −80.393932) in Mercer County Pennsylvania (Figure 2).

2.2. Habitat

All habitat surveys were collected through snorkel surveys following methods slightly altered from those assessing microhabitat partitioning [15,16]. West Branch Octoraro Creek and Chiques Creek were surveyed for Chesapeake Logperch during from June to August from 2019–2021, with September also being surveyed in 2020, while 12-Mile Creek, 20-Mile Creek, and Elk Creek were surveyed from July to August in 2010 and May to July in 2011 for the Northern Logperch. Habitat surveys for the Ohio Logperch were planned to be conducted the same time as the diet collections, but due to higher water and turbid conditions at the Shenango River, surveys could not be completed.

Collection efforts were completed between four transect lines made perpendicular to the stream, set 10 m apart, for a total of 40 m (Figure 3). All sampling events were planned based on the presence of the respective fish and that they included riffles, runs, and pools. Some sites for the Chesapeake Logperch contained little to no pools near where individuals were found. Late spring to summer months were chosen due to allow water temperatures to be high enough for snorkelers to spend more time in the water, and to sample when fish are most active. Snorkelers had a PVC wrist slate, pencil, and marker flags labeled with unique numbers. Starting downstream, the snorkelers worked between each transect in three zones. When a snorkeler observed any benthic fish, they recorded the species, how many were present, if it was above or below an object (position), and the direction they were facing (upstream, downstream, left or right bank) (orientation). Positions were collected to distinguish if fishes preferred hiding under structure. Orientations were collected to see if there was a preference for how they orient themselves, possibly supporting any feeding patterns found through diet, or showing other behavioral patterns. In addition, a marker flag was placed at the location of the fish and written on the slate to be revisited.

After all transect zones were surveyed, each flag was revisited with measuring tapes, a flowmeter, and a substrate board. Measurements for distance from the first transect line (m) and distance from the right bank while facing downstream (m) were taken using the measuring tapes. A Marsh-McBirney, Frederick, MD, USA, Flo-Mate 2000 flowmeters was used to collect depth (cm), bottom flow (m/s), and average flow (m/s). The average flow was estimated by taking the flow rate at 0.6 × depth. A substrate board, composed of a 25 cm × 25 cm clear piece of Plexiglas labeled with 5 cm × 5 cm cells, was used to collect a record of the substrate present at each flag (Figure 4). Once placed on the surface of the water, each object larger than one cell was noted by how many cells it filled. If a rock filled four cells, it would be recorded as a 1 × 4 rock (one rock that filled four cells). All remaining cells would be marked as a size one rock (Figure 4). Following the previous example, the flag’s substrate would be marked as 1 × 4 and 21 × 1.

The substrate index used by van Snik and Stauffer [16] was used to quantify the size of substrate present:

$${{\displaystyle \sum }}_1^{25}(N_x\times x^2)$$

N_x represents the number of rocks found at the size recorded and x represents the number of cells occupied by the rock. Continuing the previous example of a 1 × 4 and 21 × 1 would lead to a score of 37; (21 × 1²) + (1 × 3²). Indexing the substrate allowed for statistical tests to quantify the types of substrates each fish was spending time at.

Data used to characterize available stream habitat were collected for each site following previously detailed methods for sampling at flags. All transect lines were sampled along five equidistant points to characterize bottom flow (m/s), average flow (m/s), depth (cm), distance from first transect (m), distance from left ascending bank (m), and substrate.

2.3. Diet

A total of 40 Common Logperch and 33 Chesapeake Logperch was collected for diet using backpack electrofishing and euthanized following protocols accepted by the Pennsylvania State University Institutional Animal Care and Use Committee (IACUC) (Protocols 42210 and 201800659) and Pennsylvania State University’s Protocol, Review, Approval, and Management System (PRAMS 200345764). The backpacks were powered on a 12 V battery and wattage was decided in the stream based on the water conditions. Wattage was chosen to keep the total amperage 7–10 amp, allowing for low to no mortality of fishes that were released. Sampling for fishes was conducted at sites where at least each species had been observed within a year of sampling. Passes along streams with the electrofishing backpack were completed until the desired number of fish were collected. The entire width of the stream was sampled in a zig-zag pattern to allow for collection of fish in all possible habitats and stream flows. Once the desired number of fish were collected, only Chesapeake Logperch or Ohio Logperch were anesthetized with clove oil of at least 80% purity and euthanized. After euthanasia, they were placed in a 10% formaldehyde solution for at least two weeks, allowing the muscles to stiffen and digestion to cease. After sitting in formaldehyde, the fish were washed with water for three days before being transferred to 70% ethanol for preservation and cataloging into the Pennsylvania State Fish Museum. Each collection event at a location was treated as a collection and preserved only with fishes collected at the same event. Collections from the same stream were combined and considered the sample of that stream.

After being cataloged, dissections of stomachs were completed for the Chesapeake Logperch and the Ohio Logperch. After being preserved in 10% formalin, washed and transferred to 70% ethanol, each fish was given a tag around its caudal peduncle and given a unique number. Small scissors were used to make an incision along the ventral side, starting from the anus and continuing toward the gills. Once opened, tweezers were used to remove the stomach, esophagus, and intestines. All organs were stored in 70% ethanol in separate vials for each fish. Contents were then removed from the organs by using small scissors to open and tweezers to remove the contents. All contents were removed and put in separate vials filled with 70% ethanol for later identification.

At each site where fish were collected for diet, kick net samples were taken using a D-frame kick net. Nine total 20-s kicks were completed at riffles, runs, and pools at each site, split evenly between the three. This modified procedure of Frost et al. [17] has been found to be effective at sampling macroinvertebrate diversity that darters would encounter [18,19,20]. After collection, all debris were saved and put in 70% ethanol. If the debris was too large, it was carefully searched for any macroinvertebrates before being discarded. In the lab, the debris was poured into white trays and examined for all macroinvertebrates. Macroinvertebrates for each site were sorted from debris into jars for later identification. All data from each site of a unique stream were combined for that stream’s collective macroinvertebrate assemblage.

Macroinvertebrates were then identified using a dissection scope with magnification of 10.5 to 45× (Leica Model Z45 L, Leica Inc., Buffalo, NY, USA). All macroinvertebrates from the stomach dissections were identified down to the family level to allow for little error with partially digested macroinvertebrates. Macroinvertebrates collected with kick nets were identified to genus level using a dichotomous key to allow for macroinvertebrate biodiversity to be calculated [21]. Kick samples were additionally used for the Strauss electivity index discussed below for preference and avoidance in diet. While macroinvertebrates in the stomach were only identified to family, it is assumed that macroinvertebrates in a family have similar functional properties, and therefore a similar location within the stream, even if species are different [7].

Strauss electivity index was used to analyze the diet preference of the fishes. The electivity index calculates a value (L) for each macroinvertebrate in the stomach. The index was used because it allows macroinvertebrates to have values even if they are only found in the habitat or stomach [22]. The index uses the prey found in stomachs and found in the habitat to calculate the Strauss’ Linear Index, following the equation below:

$$L=r_i-p_i$$

Prey macroinvertebrates were denoted by i, and the abundances of it found in the stomach and the habitat were r_i and p_i, respectively. If the value was positive (L > 0), it would signify that the macroinvertebrate taxon was found in higher abundance in either individuals or the sample of fish. A negative value (L < 0) would signify that a macroinvertebrate taxon was found in the stream but was in low abundance in either individuals or the sample of fish. Significance was identified by L > 0.3 and L < −0.3 for preference and avoidance, respectively. To allow for the index to be calculated with the lower resolution of the stomach dissections, only the family level was used for calculations for both the stomach dissection and kick net samples. All L values averaged for each family and individual macroinvertebrates can be found on figshare [23].

2.4. Data Analyses

All data analyses were made using RStudio version 4.2.1 (23 June 2022), utilizing the tidyverse (1.3.2) [24], ggplot2 (3.4.2) [25], dplyr (1.1.2) [26], factoextra (1.0.7) [27], rstatix (0.7.2) [28], and tabula (3.0.0) [29] packages. Principal Component Analyses (PCA) were run for habitat availability, selected habitat, and diet, using bottom flow, average flow, depth, orientation, position, and macroinvertebrate taxa counts. Multivariate Analysis of Variance (MANOVA) were only run for habitat availability and selected habitat, testing bottom flow, average flow, depth, orientation, and position.

All PCA were completed using base RStudio to prepare the data and factoextra to define dimensions, variable contributions, and graphing the results. Individual observations (sample size) were summarized with ellipses, and vectors were added for all stream conditions and for the top ten contributions for the macroinvertebrates, when relevant. Individual observations for the stream comparison were the five equidistant points along each transect line. Individual observations for microhabitat partitioning were individual fish observations. PCA’s were scaled but did not follow assumptions, as they were used only to distinguish between groupings of variables by stream or fish species and were not used as statistical tests.

Multivariate Analysis of Variance (MANOVA) were used to test statistical significance of each variable. All observations were assumed to be independent, as each observation was based on the presence of random individuals. Stream data were also assumed to be independent as the points were chosen at random along the transect lines. Assumptions were tested using rstatix in RStudio. Multivariate normality was violated for microhabitat and stream habitat, but since the degrees of freedom were greater than 20, the central limit theorem allows for more robustness (W = 0.906, W = 0.698, p < 0.05) [30]. Multicollinearity was not found in all MANOVA tested variables through pairwise Pearson correlation tests (0.2 > r < −0.2, p < 0.0001). Homogeneity of covariances were violated through Levene’s test of equality of variances (p < 0.05). Pillai’s Trace MANOVA test and Games–Howell post hoc test were completed to be more robust with assumptions violated. The Games–Howell post hoc test also allows for differences in sample size, as it is taken into account when making the calculation [31].

MANOVA tests were completed in base RStudio to test and summarize the habitat differences among streams and microhabitats among species. Brillouin diversity of kicks were calculated from tabula. Multivariate outliers were removed from microhabitat data using Mahalanobis distance for all rows using rstatix. Mahalanobis distance (d_M) were used to account for the covariance of the data. Significant outliers were identified at p > 0.001; however, only outliers with d_M > 100 were removed to allow for data with large Strauss’ Linear Index to be included. Due to the nature of the possibilities of extremes in the habitat results, such as a large stone but shallow depth, data that should not be considered outliers were statistically significant outliers. The minimum d_M was found by removing data points with the highest d_M slowly until the smallest d_M that did not significantly alter the results was found.

3. Results

3.1. Habitat

The available habitat for streams showed two general groups based on the ellipses. The streams with Northern Logperch (12-Mile Creek, 20-Mile Creek, and Elk Creek) all showed wider ranges of habitats when compared to the streams with Chesapeake Logperch (West Branch Octoraro and Chiques Creek) (Figure 5). Dimension 1 (Dim1) made the highest contributions by the average flow (47.2%) and bottom flow (46.1%). Dimension 2 (Dim2) was influenced mainly by depth (55.3%) and substrate index (43.8%) (

Table 1. Contributions of all variables, bottom flow, average flow, depth, and substrate index, to Dimension 1 and Dimension 2 of available habitat in studied streams PCA.

	Dimension 1	Dimension 2
Bottom Flow	46.1%	0.2%
Average Flow	47.1%	0.6%
Depth	2.4%	55.3%
Substrate Index	4.3%	43.8%

). Streams with Chesapeake Logperch showed more tendency for deeper water and smaller substrate, while the streams with Northern Logperch had wider ranges of both flow rates.

The MANOVA results did not show significant differences among the bottom flow, average flow, depth, and substrate index within each group of streams. The only significant differences were between streams that contained Chesapeake Logperch (Chiques Creek and West Branch Octoraro Creek) and streams that contained Northern Logperch (12-Mile Creek, 20-mile Creek, and Elk Creek). Chiques Creek differed from all streams with Northern Logperch in most variables (t* > 3.64, df ≥ 120.3, p < 0.05), with the only nonsignificant being average flow (t* < 2.44, df ≥ 245.3, p > 0.05). West Branch Octoraro Creek showed similar trends with primarily significant variables from streams with Northern Logperch (t* > 2.89, df ≥ 108.4, p < 0.05); however, both average flow and depth were not significant (t* < 0.68, t* < 1.94, df ≥ 60.4, df ≥ 52.9, p > 0.05). Only 12-Mile Creek was significantly different from all other streams with larger substrate indices (t* > 3.06, df ≥ 203.6, p < 0.05).

The Chesapeake Logperch and Northern Logperch overlapped in their selected conditions, but with different ranges. The Chesapeake Logperch revealed a smaller range of conditions but were commonly inside the ranges for the Northern Logperch (Figure 6). The greatest contributions to Dim1 were bottom flow (33.7%), average flow (32.2%), and orientation in the stream (12.4%). Dim2 was contributed primarily by the depth (28.2%), position (25.9%), distance from the start of the transects (17.3%), and orientation in the stream (13.8%) (

Table 2. Contributions of all variables, bottom flow, average flow, depth, orientation, position, distance from Bank, and substrate index, to Dimension 1 and Dimension 2 of selected habitat PCA.

	Dimension 1	Dimension 2
Bottom Flow	29.1%	13.7%
Average Flow	28.7%	5.8%
Depth	1.9%	32.9%
Orientation	16.8%	9.2%
Position	11.2%	26.5%
Distance from Bank	4.5%	11.1%
Substrate Index	4.3%	43.8%

). The only differences in ellipses between the Chesapeake Logperch and Northern Logperch were depth (Figure 6).

Results of MANOVA testing confirmed the elliptical difference, distinguishing depth, but added that there were significant differences for most of the other variables as well. Bottom flow (t* = 4.37, df = 238.4, p < 0.05) and average flow rate (t* = 2.38, df = 183.9, p < 0.05) for the Northern Logperch were significantly higher than the Chesapeake Logperch. The Chesapeake Logperch did show significantly higher depth (t* = 7.18, df = 241.7, p < 0.05) and distance from the bank (t* = 6.77, df = 260.6, p < 0.05). The only condition that was not significantly different was the substrate index (t* = 1.09, df = 214.3, p < 0.05) (Figure 7). The Northern Logperch showed wider ranges of values with the widest ranges seen in bottom flow (0.62 m/s), average flow (0.98 m/s), and depth (58 cm).

3.2. Diet

The diet results showed similar results to habitat, with Chesapeake Logperch’s diet being a subset of Ohio Logperch’s diet. Dipteran pupae, Isonychiidae, Philopotamidae, and Perlidae were all more associated with the Chesapeake Logperch, while Ohio Logperch ate more Potomanthidae, Hydropsychidae, Leptoseridae, and Molannidae. Chironomidae was one family found commonly in both diets, while the portions were found to be higher for the Chesapeake Logperch (Figure 8). Neither dimension explained most of the variation, 12.5% and 8.3%, respectively.

The only biologically meaningful preference found by the Strauss Linear Index (L) for any macroinvertebrate families consumed by the Chesapeake Logperch was Chironomidae (L > 0.3), with no biologically meaningful avoidance was found (−0.3 < L < 0). No biologically meaningful preference or avoidance was found for the Ohio Logperch (−0.3 < L < 0.3). The highest preference for the Ohio Logperch was Daphniidae, as prey was found in stomachs but not in the habitat kick nest (

Table 3. Strauss Linear Index (L) averages for Chesapeake Logperch in Peter’s Creek (n = 7), West Branch Octoraro Creek (n = 22), and Fishing Creek (n = 4) and for Ohio Logperch in the Shenango River (n = 40), L > 0 shows preference, L < 0 shows avoidance. Bolded values represent biologically meaningful preference.

Macroinvertebrate Family	Peter’s Creek	West Branch Octoraro Creek	Fishing Creek	Shenango River
Tipulidae	−0.004	0.023	0.018	−0.028
Chironomidae	0.516	0.502	0.749	−0.020
Baetidae	0.132	0.021	0.025	0.005
Hydropsychidae	−0.149	0.040	0.247	0.036
Daphniidae	0	0	0	0.095
Gammaridae	0	0	0	0.068
Potomanthidae	0	0	0	−0.042
Elmidae	0	0	0	−0.070

). While not significant, more avoidance values were found for the Chesapeake Logperch than the Ohio Logperch (L < 0).

4. Discussion

The microhabitat results demonstrated a principal difference between the Chesapeake Logperch and Northern Logperch. With the overlapping selected habitats, observations were not able to find visually different ranges. The Chesapeake Logperch showed significant differences in the form of narrower ranges. With habitat degradation being a large factor in endangerment and extinction, a possible reason for their declining numbers is a loss of habitat [32]. Even if the habitat is not directly lost, more restricted animals have been found to be affected more by disturbances as well [33].

The deeper depths for the Chesapeake Logperch were inverse of previous observations made by Stauffer. Northern Logperch have been commonly observed in deeper waters by Stauffer, seen at the border between riffles and runs, while Chesapeake Logperch is found primarily in riffles [2]. Some of the deeper depths can possibly be explained by the specific streams for each species. The greatest difference between streams was the depth, where over half of Dim2 was explained by depth and variation was seen between Chesapeake Logperch and Northern Logperch streams (Figure 6). The deepest stream was Chiques Creek, which was a reintroduced population. This would explain some of the results for why the Chesapeake Logperch were found at deeper depths than Northern Logperch.

When considering streams for reintroductions, there should be abiotic factors that align with the findings. The first is the selection for riffles with on average, water velocity 0.18–0.33 m/s with a bottom flow 0.02–0.14 m/s. The substrate index should include substrate 31–99, translating to gravel, cobble, and some larger boulder mixed in. The substrate should be present in depths between 40 cm and 55 cm (Figure 7). While the current findings are limited, more information can be found by selecting for streams with conditions that suit Northern Logperch.

Selected conditions recorded for both species show that streams that have similar habitats to the Northern Logperch would provide system resources that could support Chesapeake Logperch streams. The narrower range of preferences, however, should not be neglected. To better understand why their range is narrower, possible competitions should be addressed in the future. Merely mentioning the existence of different fishes would not suffice, as previous work has shown that only some benthic fishes, specifically darters, compete directly with each other [19]. One way of competition previously documented is through aggression. Despite the fact, Tessellated Darters, one of the most common darters found in the same streams, has shown to not be aggressive with other darters [34]. While benthic fishes may not show aggression, they may form their own habitat preferences within the stream to reduce the competition. This would cause them to have narrower realized niches, leading to narrower microhabitat conditions. The last detail for consideration when comparing the microhabitat preferences is that the ranges, while statistically significant, the real differences were not large. With median average and bottom flow rates < 0.1 m/s apart and a difference of only 5 cm in depth, the biological meaningfulness would be less (Figure 7).

The biologically meaningful preference of Chironomidae shows that Chesapeake Logperch have a somewhat restrictive diet, preferring to predate on non-biting midge larvae, Chironomidae. The Ohio Logperch diet exhibited a difference as many prey items were found, but there were no biologically meaningful preferences or avoidances found (Table 3). It should be noted that prey commonly found in slow to still moving water, such as Daphniidae, were found in the Ohio Logperch stomachs. This is the opposite of the intuitive thought that they only flip rocks to find prey. The addition of prey from slow to still moving water supports the previous observation of Ohio Logperch in and around pools. Since prey not only found under rocks were found, it is considered that the method of flipping rock may not be either species’ primary feeding strategy. While the diet of the Chesapeake Logperch still needs to be studied further, previous work on Ohio Logperch showed similar trends of being generalists [19]. Considerations should be made when interpreting the relevance of the results. The main consideration is that while the Chesapeake Logperch and Common Logperch were compared, different subspecies of the Common Logperch were used for the microhabitat and diet portions. Furthermore, the Northern Logperch has been suggested to be reviewed to be elevated to a species [35]. This would mean that the results for microhabitat should be considered as a comparison of the Chesapeake Logperch and the Northern Logperch, not the Common Logperch, and likewise for diet and the Ohio Logperch. Another consideration should be made for the recent diet work on the Ohio Logperch, as all individuals were taken from the Shenango River. More work would be needed to confirm that the Ohio Logperch still have generalist feeding patterns and it was not an effect of the Shenango River.

To be able to create better reintroduction and conservation plans, more information about the Chesapeake Logperch life history is needed. There is a consensus of how they reproduce, possibly going up tributaries of larger rivers to reproduce [2]. Observations of juveniles, however, have been recorded in the Susquehanna River Flats in the Chesapeake Bay, where there would be little tributary access for adults. Most work is also performed in smaller streams, and their presence in larger bodies of water is constricted to Chesapeake Bay. To better understand what all stages of their life cycle need to be accommodated, these studies could help to determine what is required for all life stages to be successful.

5. Conclusions

While the diet results may not prove as useful for planning reintroduction based on streams that can hold Ohio Logperch, the habitat results showed that streams that hold Northern Logperch have the possibility of being blueprints for streams planned for the reintroduction of Chesapeake Logperch. Although there was statistical significance between the habitat of Northern Logperch and Chesapeake Logperch, the results show the possibility of Chesapeake Logperch being able to live in streams viable for Northern Logperch. When significantly different, the selected habitats were a subset of what Northern Logperch preferred. If the habitat range for Northern Logperch is present, Chesapeake Logperch should be able to occupy part of the available habitat.

Unlike with habitat, the support for the generalist nature of Ohio Logperch leaves little reinforcement that Chesapeake Logperch would prey on a smaller portion of the diet. Ohio Logperch could survive without part of their known prey present, while if that portion is missing, Chesapeake Logperch might not have a preferred prey present in the stream. However, just the diet results of Chesapeake Logperch may help in consideration process of streams for reintroduction. With the preference for Chironomids, the stream would ideally hold an abundant population of Chironomids. While not significant, Chesapeake Logperch did also show a preference for Ephemeropterans and usually Trichopterans, both known to live in higher quality water.