A key concept in evolutionary biology is that divergent selective regime will often generate and maintain some type of phenotypic diversity (Langerhans et al. 2003). This divergent selection can lead to differences in phenotypic expression either by a genetic differentiation or phenotypic plasticity (Levins, 1968; West-Eberhard, 1989: Robinson and Wilson, 1994; Orr and Smith, 1998; Schluter, 2000; cited in Langerhans, 2003). Such divergence is significant as it can influence microevolutionary changes and result in speciation (West-Eberhard, 1989).
Meristic morphological characteristics, such as fin rays, gill rakers, and scale rows have historically served as an important method for identifying fish. Count data can be used in statistical analysis, allowing for comparison of populations and sexes (Barlow, 1961 cited in Begg and Waldman, 1999). Meristic characteristics vary within and among species, making them useful for describing or identifying fishes (Strauss and Bond, 1990). Morphometrics is the analysis of body shape and or shape of other morphological features (Begg and Waldman, 1999).
Landmark methods in morphometric analysis uses data derived from discrete morphometric points, linear distances between points and the geometric relationships amongst points (Cadrin, 2005). Traditional systems of measurement including calipers and measuring boards are commonly used in morphometric studies. However, digital imaging with accurate calibration provides superior data format and allows data to be stored as coordinates which allows for geometric methods of analysis. Also, digital images are easily archived and saved allowing for one to reprocess and confirm results (Cadrin, 2005).
The objective of this study is to utilize meristic counts and morphological measurements to provide a comparison of a disjunct Elassoma zonatum population found above the fall line in Cypress Creek, located in the Green River system, with those found below the fall line in Bayou de Chien in the Coastal Plain region. Such analysis could help answer if populations of Banded Pygmy Sunfish in Kentucky, found above the fall line are morphologically similar to other populations of E. onatum found in Kentucky or if there is a distinct difference between populations supporting need for further studies to investigate potential causes.
The genus Elassoma encompasses seven recognized species: E. zonatum Jordan, E. evergladei Jordan, E okefenokee Bohlke, E. bohlke Rohde and Arndt, E. okatie Rohde and Arndt, E. alabamae Mayden, and E. gilberti Snelson, Krabbenhoft and Quattro. Four of these seven species (E. alabamae, E. boehlkei, E. gilberti, and E. okatie) have only been described in the past 30 years (Rohde and Arndt 1987, Mayden 1993, Snelson et al. 009). Five of the seven Elassoma species have restricted geographic distributions. Of these seven species, only the banded pygmy sunfish, Elassoma zonatum, occurs in Kentucky and was originally thought restricted to the coastal plain region (Clay, 1975). It is now known that E. zonatum populations in Kentucky inhabit three different drainages.
The Coastal Plain population drains into the Mississippi and Tennessee rivers, while the Green River population drains into the Ohio. Elassoma zonatum was originally assigned a conservation status of special concern (Branson et al. 981 cited in Burr and Warren, 1986). However, biological surveys conducted by the Kentucky State Nature Preserves Commission (KSNPC) in 1979–1980 discovered populations above the inner boundary of the Coastal Plain, often referred to as the fall line (Robison, 1986), in the Tradewater and Green River systems (Warren and Cicerello, 1982). Both populations represented new records for the state and extended the known Kentucky Elassoma zonatum range 160km east (Warren, 1980).
With this range expansion and new populations, Burr and Warren (1986) suggested E. onatum was more common than prior studies showed and should not be listed as a species of special concern. Prior studies on the life history and habitat preference of E. zonatum in the Coastal Plain region of Kentucky indicated association in the summer with swampy areas among dense beds of aquatic vegetation supporting bald cypress (Taxodium distichum), tupelo (Nyssa spp. ), oaks (Quercus spp. ) and willows (Salix spp. ) (Walsh and Burr, 1984). Barney and Anson (1920) noted typical habitat for E. zonatum in Cypress Bayou, Louisiana consisting of surface mats of vegetation with Coontail (Ceratophylum spp. growing underneath.
With the distribution of E. zonatum in Kentucky paralleling cypress swamps, it has been hypothesized that suitable habitat was once found over the western third of the state and that isolated populations found in the Tradewater and Green river systems may be a result of historical wetland loss (Warren, 1980). Methods Study Areas Two populations of Elassoma zonatum were identified for this study. One population was collected from below the fall line in the Mississippi system in a drainage ditch near Bayou de Chien in Hickman County.
Specimens above the fall line were collected from Cypress Creek State Nature Preserve in the Green River system in Muhlenberg County (Fig 1). Cypress Creek, a second order stream located in the Interior Low Plateaus Province, is bordered by cypress swamps and flows into Pond River, which drains into the Green River, which eventually drains into the Ohio River. Bayou de Chien a second order stream located in the Coastal plain province and is surrounded by heavy agriculture activity resulting in many agriculture ditches draining into the stream.
Both streams have been heavily impacted over the past century due to agricultural practices resulting in straightened stream channels and increased siltation. For site comparison species list of encountered fish species were collected and compared using Jaccard’s Index and Sorensen’s Index. Morphological Comparison Elassoma zonatum voucher specimens were procured from both streams using ? ” (6mm) dipnets. Collections from Cypress Creek yielded 24 individuals (12 males and 12 females), 25 individuals were collected from ditches around Bayou de Chien (9 males and 16 females).
Photographs were taken of specimens with a ruler for scale to allow visual comparison of body and fin coloration and assist with morphometric analysis and comparisons using tpsDIG2 and tpsUTIL for landmark digitalization both available from http://life. bio. sunysb. edu/morph/. IMP 8 package including Coordgen8, Twogroup8, PCA Gen8 and CVA Gen8 (available from http://www3. canisius. edu/~sheets/IMP%208. htm) was used to analyze landmark relationships and differences between means in Cypress Creek Populations and Bayou de Chien.
Prior to photographing specimens for morphometric analysis, they were anesthetized with Tricane (MS222), and then preserved in 90% ethanol for meristic counts and potential use in future genetic studies. Landmarks for morphological characters followed Armbruster (2012) guidelines. Meristic counts and measurements were conducted adhering to Hubbs and Lagler (1974) and Rohde and Arndt (1987) guidelines to determine if populations followed usual identification methods for E. zonatum.
Measurements of standard length, body depth and body width were taken to the nearest 0. mm using dial calipers under a dissecting microscope. Body depth was measured vertically from the origin of the dorsal fin. Dorsal and anal fin length measurements were taken from the base of the first spine to the tip of the longest ray. Fourteen other proportional body measurements were taken using Imagej software measurement function (https://imagej. nih. gov/ij/). In addition, all vertical trunk bars and preopercular pores were counted and compared between sexes and populations.
Breeding coloration of males was assigned a value by using GNU Image Manipulation Program (GIMP) (available from https://www. gimp. org/). Each male specimen had three hue values taken using the color selector function with sample average collected and radius setting of 5. One measurement was taken from the first third of the body around the lower jaw, a second measurement from the middle of the body around the trunk region below the origin of the dorsal fin. The final measurement was collected from the caudal peduncle after termination of the dorsal fin.
The three measurements were then averaged together for an overall score for each male. Students t-test was then used to compare the averages from each drainage to each other to detect if any difference in coloration could be observed. Results Body Measurement Comparison Comparisons of body measurements between Bayou de Chien and Cypress Creek using students T-test (table____) indicated significant differences in standard length (<0. 0001. Comparisons of males ( table_) resulted in a F-test F-score of 2. 22, p=0. 0100. Goodalls F-test resulted in F=2/22, df 20, p=0. 02 with 100 bootstraps.
PCA comparisons of 12 Procrustes fitted landmarks for both drainages indicated that PC1 and PC2 (table___) explained 54. 35% of the variation with PC3 explaining an additional 17. 6%. PCA plot (figure___) did not show any distinct separation of drainages. Comparing males across both drainages resulted in 57. 5% of the variation being explained by PC1 and PC2 (table___). The plotted PCA (figure___) showed little separation between groups. In females 60. 26% of variation was explained by PC1 and PC2 (table___).
The PCA plot (figure___) showed some separation between drainages with Cypress Creek falling more towards the left of the graph. Thin-plate spline transformations using color coded comparison between drainages (figure___) indicated that between drainages there was an some enlarging of landmarks located at the occiput and around the eyes. In male comparisons (figure__) there was only minor variations among landmarks. Females (figure___) appear more variable with variations more evident in the head region and caudal peduncle landmarks.