Southeastern Naturalist 309 A.V. Fernando, K.B. Hecke, and M.A. Eggleton 22001188 SOUTHEASTERN NATURALIST 1V7o(2l.) :1370,9 N–3o2. 62 Length, Body Depth, and Gape Relationships and Inference on Piscivory Among Common North American Centrarchids Anthony V. Fernando1,2,*, Kyler B. Hecke1,3, and Michael A. Eggleton1 Abstract - Species of Centrarchidae are major components of inland fisheries in much of North America. Thus, information gained from the assessment of interspecies interactions and/or quantifying predator–prey relationships is a useful tool for fisheries managers. Using preserved fish specimens (n = 717) from 20 species of centrarchids, we made measurements of total length (TL), standard length (SL), horizontal gape, and body depth for each individual. We fitted mathematical models that included horizontal gape and body depth as functions of TL and SL, and TL as a function of SL. Linear-regression–model fits were generally good (r2 = 0.764–0.998) for all 20 species, with 61 of 78 possible models having r2 values exceeding 0.90. Horizontal gape–SL (F3,702 = 77.18, P < 0.001) and body depth– SL (F3,702 = 91.79, P < 0.001) ratios differed significantly along a gradient that reflected the species’ likelihood of piscivory. Slopes of TL–SL regressions did not vary by species, which enabled development of a generalized TL–SL model for centrarchids. Supplemental analyses supported that morphometric measurements had not been influenced significantly by preservation. Results of this study are useful to fisheries managers involved with understanding species interactions within centrarchid-dominated food webs, which are of high priority in most fisheries-management plans. Introduction The family Centrarchidae includes a number of North America’s most important freshwater sportfishes, including Micropterus spp. (black basses), Pomoxis spp. (crappies), and Lepomis spp. (sunfishes). Centrarchids dominate fish assemblages in many aquatic systems of North America, particularly within the Mississippi River basin (e.g., Alfermann and Miranda 2013, Lubinski et al. 2008, Olive et al. 2005). Centrarchids also fill multiple ecological niches in north-temperate lakes that tend to be structured more by species of Percidae (e.g., Sander vitreus [Mitchill] [Walleye]) and Esocidae (i.e., pikes and pickerels) (Tonn et al. 1990). Piscivorous centrarchids have long been recognized as being gape-limited predators at most sizes (Hambright 1991, Lawrence 1958, Nowlin et al. 2006) that often prefer smaller-sized prey that require less handling time (Hoyle and Keast 1987, Werner 1974). Collar et al. (2009) examined degree of piscivory in centrarchids as a causative mechanism that limited diversification in their feeding morphologies. They concluded that feeding-morphology diversification had been most limited in the centrarchid lineages that exhibited the greatest degrees of piscivory (e.g., Micropterus). 1Department of Aquaculture and Fisheries, University of Arkansas at Pine Bluff, Pine Bluff, AR 71601. 2Arkansas Game and Fish Commission, Russellville, AR 72802. 3Department of Forestry, Wildlife and Fisheries, University of Tennessee, Knoxville, TN 37996. *Corresponding author - anthonyvfernando@gmail.com. Manuscript Editor: Benjamin Keck Southeastern Naturalist A.V. Fernando, K.B. Hecke, and M.A. Eggleton 2018 Vol. 17, No. 2 310 Within the centrarchids, sunfishes (tribe Lepomini) are usually not considered to be key piscivores in many aquatic food webs, even though many lepomines will eat fishes opportunistically (Collar et al. 2009). Several studies have reported relationships between body length and horizontal gape (measured as mouth width) for various centrarchid species (e.g., Hale 1996, Hambright 1991, Hill et al. 2004, Keast and Webb 1966, Lawrence 1958, Werner 1974). Most fishes consume piscine prey held horizontally in the mouth; thus, it is more likely that horizontal gape is the more limiting dimension than vertical gape (Mihalitsis and Bellwood 2017). Wainwright and Richard (1995) demonstrated that the trophic level of centrarchids in food webs was often dependent on their size-dependent gape size. For example, invertebrate-feeding but smaller-bodied sunfishes such as Lepomis megalotis (Longear Sunfish) are routinely considered non-piscivorous (e.g., Collar et al. 2009); however, fishes can comprise up to 29% of their diet (Robison and Buchannan 1988). Although sunfishes are unlikely to be primary predators in most impoundments, sunfish predation on juveniles and fry of desirable sportfishes may be an unconsidered albeit significant source of first-year mortal ity. Accurate modeling of interspecies interactions among centrarchids requires an understanding of both the gape–body-length relationships and the body depth– body-length relationships for many species; these data are not readily available in the primary literature. Quantification of the physical limits to foraging behavior is essential to modern predator–prey modeling efforts (Audzijonyte et al. 2014, Persson and Diehl 1990). Understanding how horizontal gape and body depth relate to length are essential to identifying the size window when certain species become available as prey to sportfish, and at what size sportfish juveniles cease to be prey for the fishes that will ultimately comprise their forage base. Knowledge of these relationships is especially applicable to the centrarchids, considering their broad distributions, high abundances, and fisheries-management significance in many types of aquatic systems throughout North America. Although inland-fisheries management has been increasingly moving towards standardized methods (e.g., Bonar et al 2009), many large datasets contain only total length or standard length, but rarely both. Thus, conversion between the 2 length-measures may be useful and necessary for state fisheries-management agencies. Readily available conversion formulas, such as those presented by FishBase, are often unsatisfactory because they are based on photographic measurements from single individuals. The primary objective of the present study was to develop mathematical models that depicted relationships between horizontal-gape size and body length (both as total length [TL] and standard length [SL]), and body depth and body length (as TL and SL) for 20 centrarchid species common to North America. Data were largely measured from fish specimens contained in museum collections. Our second objective was to examine whether these mathematical relationships were consistent with the predicted likelihood of piscivory for individual centrarchids as reported by Collar et al. (2009). A third objective was to compare mathematical relationships between TL and SL and seek to develop a generalized TL–SL equation for Southeastern Naturalist 311 A.V. Fernando, K.B. Hecke, and M.A. Eggleton 2018 Vol. 17, No. 2 centrarchids. We also assessed whether preservation effects had affected the ratios of interest. Results of this study will support fisheries scientists and managers involved with understanding species interactions within food webs dominated by centrarchids, which represent the entire spectrum of piscivory in fish assemblages. Methods Sources of fishes We obtained the fish used in this study from 3 sources: (1) specimens preserved in 99% ethanol, following initial 10% formalin fixation (n = 301) from the Ichthyology Teaching Collection at the University of Arkansas at Pine Bluff, AR (UAPB); (2) specimens preserved in 100% isopropanol, following 10% formalin fixation (n = 260) from the Zoology Teaching Collection at nearby Henderson State University, Henderson, AR (HSU); and (3) juvenile Micropterus salmoides (Lacepede) (Largemouth Bass) freshly preserved in 99% ethanol with no fixation (n = 156). This latter group of Largemouth Bass was collected and measured in another study that had been recently completed (Fernando 2015). Specimens from the UAPB and HSU teaching collections were collected from the wild and had been preserved for more than 1 year. The juvenile Largemouth Bass from Fernando (2015) had been preserved for less than 1 year and were destroyed following otolith removal prior to commencement of the present study. Measurements of SL were not taken from these specimens. Assessment of preservation effects To assess whether long-term preservation of UAPB and HSU specimens had altered the dimensions being examined, we compared morphometric ratios of longterm preserved L. macrochirus Rafinesque (Bluegill) (n = 41) to those determined from live Bluegills (n = 150). We randomly selected live Bluegill specimens from 3 segregated groups maintained at UAPB for other experiments. These groups included specimens collected from a livestock watering pond, state hatchery, and commercial farm. For each specimen (live and preserved), we measured TL and SL to the nearest whole millimeter using rulers or measuring boards. We defined body depth as the maximum distance from the dorsal surface of the fish to the ventral surface of the fish along a line perpendicular to the specimen’s longest axis. We used a ruler to measure body depths to the nearest whole millimeter. We defined horizontal gape as the maximum distance between left and right maxilla of the fish with the mouth closed, and measured this character to the nearest 0.01 mm using digital calipers. We conducted an analysis of covariance (ANCOVA) to test relationships between horizontal gape and body depth, and TL and SL between preserved Bluegill from the UAPB collection and live Bluegill taken from the 3 sources listed above. We also compared mean horizontal gape–SL and body depth– SL ratios between preserved and live Bluegill. Objective 1 We measured TL, SL, body depth, and horizontal gape for each specimen of the 20 species examined, using the same methods as described for the assessment Southeastern Naturalist A.V. Fernando, K.B. Hecke, and M.A. Eggleton 2018 Vol. 17, No. 2 312 of preservation effects. For each species, we assessed the influence of TL and SL on horizontal gape and body depth using ordinary least-squares linear regression. Models were fitted using horizontal gape and body depth as the dependent variables, with TL and SL serving as the independent variables. Objective 2 We classified the likelihood of piscivory for all 20 centrarchid species into 4 categories following Collar et al. (2009). These categories were intended to reflect degrees of piscivory for each species, which we rated as high (3 species), moderate (8 species), low (2 species), and not piscivorous (7 species) (Table 1). Two species, Centrarchus macropterus (Lacepede) (Flier) and Lepomis miniatus (Redspotted Sunfish), classified by Collar et al. (2009) as “not/moderate” piscivores were classified here as having a “low” likelihood of piscivory based on local life-history information that suggested these species occasionally consumed fishes (Robison and Buchanan 1988). We determined mean horizontal gape–SL and body depth–SL ratios from specimens of each centrarchid species. We employed one-factor analysis of variance (ANOVA) to assess differences in mean horizontal gape–SL and body depth–SL ratios among the likelihood of piscivory categories. We performed Tukey’s honest significant difference (HSD) to test pairwise differences among likelihoods of piscivory with individual fish serving as replicates. Objective 3 For each species, we assessed the relationship between TL and SL using ordinary least-squares linear regression. Models were fitted using TL as the dependent variable and SL as the independent variable. We conducted ANCOVA to test among-species differences in TL–SL relationships, and to generate a generalized TL–SL equation for centrarchids. We set significance for all statistical tests for all 3 objectives at an alpha level of 0.05. In all analyses, individual fish served as the experimental unit. We performed all statistical comparisons in the program R (R 3.2.0, The R Foundation for Statistical Computing, Vienna, Austria). Results Preservation effects There was a significant difference in horizontal gape–SL ratios among sources of live Bluegill (F3,187 = 21.76, P < 0.001; Fig. 1). However, Tukey’s HSD indicated no differences in horizontal gape–SL ratio between the preserved fish and those originating from the commercial hatchery (P = 0.420) or the wild population (P = 0.055). All other possible pairwise comparisons of horizontal gape–SL ratio were different (P < 0.001). With respect to body depth–SL ratios, sources of live Bluegill significantly differed (F3,187 = 35.77, P < 0.001; Fig. 2). In particular, pairwise comparisons detected no significant differences among the 2 hatcheries and the preserved fish (P = 0.870–0.989), though body depth–SL ratios of wild Bluegill were significantly less than all 3 other sources ( P < 0.001; Fig. 2). Southeastern Naturalist 313 A.V. Fernando, K.B. Hecke, and M.A. Eggleton 2018 Vol. 17, No. 2 Objective 1 Regression statistics depicting the relationships of horizontal gape and body depth to TL and SL are presented for 20 centrarchid species in Tables 1–4. Sample sizes were variable across species (mean = 43 [SE = 11]), although n exceeded 20 for 13 of 20 species modeled. With respect to horizontal gape–TL models, fits were generally good for most species; r2 values exceeded 0.90 for 16 of 20 models, with Figure 1. Horizontal gape–SL ratio of live and preserved Bluegill; error bars represent 1 SE. Figure 2. Body depth–SL ratio of live and preserved Bluegill; error bars represent 1 S E. Southeastern Naturalist A.V. Fernando, K.B. Hecke, and M.A. Eggleton 2018 Vol. 17, No. 2 314 Table 1. Model parameters and associated descriptive statistics for mathematical relationships between horizontal gape (mouth width) and total length (TL) for 20 species of Centrarchidae. Likelihood of piscivory (modified from Collar et al. 2009) was categorized as H = high, M = moderate, L = low, and N = not piscivorous. [Table continued on following page.] Horizontal TL (mm) gape (mm) Common name Scientific name n Min Max Model r2 P 95% CI of b Source Shadow Bass (M) Ambloplites ariommus Viosca 19 45 236 -2.559 + 0.146 TL 0.961 less than 0.001 0.132–0.161 Present study Ozark Bass (M) A. constellatus Cashner & Suttkus 7 44 200 -2.315 + 0.131 TL 0.988 less than 0.001 0.116–0.146 Present study Rock Bass (M) A. rupestris Rafinesque 7 24 172 -0.599 + 0.112 TL 0.995 less than 0.001 0.104–0.120 Present study Flier (L) Centrarchus macropterus Lacepède 34 27 165 0.235 + 0.081 TL 0.955 less than 0.001 0.075–0.088 Present study Green Sunfish (M) Lepomis cyanellus Rafinesque 39 51 181 -3.579 + 0.140 TL 0.954 less than 0.001 0.130–0.150 Present study Pumpkinseed (M) L. gibbosus L. 6 47 120 -1.249 + 0.083 TL 0.983 less than 0.001 0.069–0.096 Present study Warmouth (M) L. gulosus (Cuvier) 43 37 219 -2.358 + 0.150 TL 0.971 less than 0.001 0.142–0.158 Present study Orangespotted Sunfish (N) L. humilis (Girard) 25 45 83 -1.156 + 0.105 TL 0.764 less than 0.001 0.080–0.129 Present study Bluegill (N) L. macrochirus Rafinesque 191 37 191 -1.205 + 0.082 TL 0.916 less than 0.001 0.079–0.086 Present study Dollar Sunfish (N) L. marginatus (Holbrook) 21 49 110 -2.319 + 0.123 TL 0.933 less than 0.001 0.108–0.138 Present study Longear Sunfish (N) L. megalotis (Rafinesque) 92 62 160 -2.095 + 0.104 TL 0.856 less than 0.001 0.095–0.112 Present study Redear Sunfish (N) L. microlophus (Günther) 19 42 190 -0.201 + 0.083 TL 0.982 less than 0.001 0.077–0.088 Present study Redspotted Sunfish (L) L. miniatus Jordan 5 46 116 -0.561 + 0.099 TL 0.926 0.006 0.055–0.143 Present study Spotted Sunfish (N) L. punctatus (Valenciennes) 36 34 130 -1.539 + 0.112 TL 0.832 less than 0.001 0.094–0.129 Present study Bantam Sunfish (N) L. symmetricus Forbes 10 28 75 -0.462 + 0.109 TL 0.920 less than 0.001 0.085–0.134 Present study Smallmouth Bass (H) Micropterus dolomieu Lacepède 33 55 316 -1.828 + 0.120 TL 0.958 less than 0.001 0.111–0.129 Present study Spotted Bass (H) M. punctulatus Rafinesque 33 30 314 -1.574 + 0.117 TL 0.967 less than 0.001 0.109–0.124 Present study Largemouth Bass (H) M. salmoides Lacepède 156 39 150 -0.888 + 0.101 TL 0.843 less than 0.001 0.094–0.108 Present study A 31 99 1.88 + 0.078 TL Lawrence 1958 A 100 199 -1.88 + 0.111 TL Lawrence 1958 A 200 299 -5.16 + 0.129 TL Lawrence 1958 A 300 399 -7.96 + 0.137 TL Lawrence 1958 A 400 499 -29.41 + 0.196 TL Lawrence 1958 A 500 595 -56.36 + 0.248 TL Lawrence 1958 B 55 200 0.000 + 0.0968 TL 0.890 Shireman et al. 1978 B 200 603 -11.10 + 0.157 TL 0.950 Shireman et al. 1978 Southeastern Naturalist 315 A.V. Fernando, K.B. Hecke, and M.A. Eggleton 2018 Vol. 17, No. 2 Table 1, continued. Horizontal TL (mm) gape (mm) Common name Scientific name n Min Max Model r2 P 95% CI of b Source Largemouth Bass (H) M. salmoides C 6 15 -0.363 + 0.133 TL 0.269 Timmerman et al. 2000 C 16 24 0.114 + 0.134 TL 0.272 Timmerman et al. 2000 C 25 50 -0.152 + 0.116 TL 0.904 Timmerman et al. 2000 121 50 425 -5.590 + 0.140 TL 0.960 Hill et al. 2004 220 7 100 0.0507 TL1.149 0.994 Johnson and Post 1996 White Crappie (M) Pomoxis annularis Rafinesque 48 32 311 -1.919 + 0.091 TL 0.938 less than 0.001 0.084–0.098 Present study 4401 -2.155 + 0.141 TL 0.999 Hale 1996 Black Crappie (M) P. nigromaculatus Lesueur 38 67 306 -1.194 + 0.087 TL 0.946 less than 0.001 0.080–0.094 Present study ALawrence (1958) measured 1699 Largemouth Bass but did not indicate how many fell into each size group. BShireman et al. (1978) measured 310 Lar gemouth Bass but did not indicate how many fell into each size group. CTimmerman et al. (2000) measured 267 fish but did not indicate ho w many fell into each size group. Table 2. Model parameters and descriptive statistics for mathematical relationships between body depth and total length (TL) for 20 species of Centrarchidae. [Table continued on following page.] TL (mm) Body depth (mm) Common name Scientific name n Min Max Model r2 P 95% CI of b Source Shadow Bass Ambloplites ariommus 19 45 236 -3.554 + 0.377 TL 0.976 less than 0.001 0.347–0.406 Present study Ozark Bass A. constellatus 7 44 200 -0.483 + 0.324 TL 0.998 < 0.001 0.307–0.341 Present study Rock Bass A. rupestris 7 24 172 21.991 + 0.184 TL 0.818 0.003 0.094–0.273 Present study Flier Centrarchus macropterus 34 27 165 -1.169 + 0.395 TL 0.976 less than 0.001 0.372–0.417 Present study Green Sunfish Lepomis cyanellus 39 51 181 -2.615 + 0.375 TL 0.947 less than 0.001 0.346–0.404 Present study -4.356 + 0.372 TL Lawrence 1958 Pumpkinseed L. gibbosus 6 47 120 -5.357 + 0.421 TL 0.971 less than 0.001 0.330–0.511 Present study Southeastern Naturalist A.V. Fernando, K.B. Hecke, and M.A. Eggleton 2018 Vol. 17, No. 2 316 Table 2, continued. TL (mm) Body depth (mm) Common name Scientific name n Min Max Model r2 P 95% CI of b Source Warmouth L. gulosus 43 37 219 -4.979 + 0.386 TL 0.984 less than 0.001 0.371–0.402 Present study 117 18 190 -5.850 + 0.380 TL 0.990 Hill et al. 2004 Orangespotted Sunfish L. humilis 25 45 83 -2.502 + 0.367 TL 0.825 less than 0.001 0.296–0.438 Present study Bluegill L. macrochirus 191 37 191 -11.334 + 0.455 TL 0.941 less than 0.001 0.438–0.471 Present study -7.983 + 0.418 TL Lawrence 1958 29 41 206 -8.880 + 0.460 TL 0.990 Hill et al. 2004 92 -2.070 + 0.372 TL 0.980 less than 0.001 Dewey et al. 1997 143 20 80 -3.662 + 0.378 TL 0.990 Schramm and Zale 1985 Dollar Sunfish L. marginatus 21 49 110 -3.898 + 0.421 TL 0.962 less than 0.001 0.382–0.461 Present study Longear Sunfish L. megalotis 92 62 160 -3.958 + 0.427 TL 0.863 less than 0.001 0.392–0.462 Present study Redear Sunfish L. microlophus 19 42 190 -3.140 + 0.379 TL 0.998 less than 0.001 0.369–0.388 Present study -2.096 + 0.346 TL Lawrence 1958 28 52 223 -6.230 + 0.400 TL 0.999 Hill et al. 2004 Redspotted Sunfish L. miniatus 5 46 116 -0.571 + 0.387 TL 0.877 0.012 0.160–0.613 Present study Spotted Sunfish L. punctatus 36 34 130 -5.248 + 0.449 TL 0.932 less than 0.001 0.408–0.491 Present study 81 24 158 -3.840 + 0.420 TL 0.990 Hill et al. 2004 Bantam Sunfish L. symmetricus 10 28 75 -2.19 + 0.411 TL 0.969 less than 0.001 0.356–0.468 Present study Smallmouth Bass Micropterus dolomieu 33 55 316 -3.086 + 0.245 TL 0.966 less than 0.001 0.228–0.261 Present study Spotted Bass M. punctulatus 33 30 314 -1.022 + 0.257 TL 0.855 less than 0.001 0.219–0.294 Present study Largemouth Bass M. salmoides 156 39 150 -3.392 + 0.306 TL 0.923 less than 0.001 0.283–0.328 Present study A 31 99 0.008 + 0.200 TL Lawrence 1958B A 100 199 -8.120 + 0.258 TL Lawrence 1958B A 200 299 -32.798 + 0.344 TL Lawrence 1958B A 300 399 -82.696 + 0.472 TL Lawrence 1958B A 400 499 -136.572 + 0.574 TL Lawrence 1958B A 500 595 -431.400 + 1.059 TL Lawrence 1958B 140 17 423 -3.380 + 0.250 TL 0.990 Hill et al. 2004 220 7 100 0.173 TL1.046 0.995 Johnson and Post 1996 White Crappie Pomoxis annularis 48 32 311 -3.515 + 0.332 TL 0.802 less than 0.001 0.284–0.381 Present study Black Crappie P. nigromaculatus 38 67 306 -5.200 + 0.353 TL 0.959 less than 0.001 0.329–0.378 Present study ALawrence (1958) measured 1699 Largemouth Bass but did not indicate how many fell into each size group. BAlgebraically re-arranged from source. Southeastern Naturalist 317 A.V. Fernando, K.B. Hecke, and M.A. Eggleton 2018 Vol. 17, No. 2 Table 3. Model parameters and descriptive statistics for mathematical relationships between horizontal gape (mouth width) and standard length (SL) for 20 species of Centrarchidae. SL (mm) Horizontal gape (mm) Common name Scientific name n Min Max Model r2 P 95% CI of b Source Shadow Bass Ambloplites ariommus 19 37 200 -1.979 + 0.172 SL 0.963 less than 0.001 0.155–0.188 Present study Ozark Bass A. constellatus 7 36 165 -2.458 + 0.161 SL 0.984 less than 0.001 0.139–0.183 Present study Rock Bass A. rupestris 7 18 138 -0.530 + 0.140 SL 0.995 less than 0.001 0.130–0.151 Present study 15 130 150 0.120 SLA Keast and Webb 1966 Flier Centrarchus macropterus 34 21 133 0.282 + 0.101 SL 0.952 less than 0.001 0.093–0.109 Present study Green Sunfish Lepomis cyanellus 39 40 155 -3.018 + 0.164 SL 0.946 less than 0.001 0.151–0.177 Present study -2.635 + 0.178 SL Werner 1974 Pumpkinseed L. gibbosus 6 42 95 -1.779 + 0.112 SL 0.983 less than 0.001 0.093–0.130 Present study Warmouth L. gulosus 43 30 184 -1.867 + 0.173 SL 0.976 less than 0.001 0.165–0.182 Present study Orangespotted Sunfish L. humilis 25 36 69 -1.937 + 0.141 SL 0.920 less than 0.001 0.124–0.159 Present study Bluegill L. macrochirus 191 32 153 -0.820 + 0.098 SL 0.907 less than 0.001 0.093–0.102 Present study 0.217 + 0.093 SL 0.980 Werner 1974 15 125 150 0.120 SLA Keast and Webb 1966 Dollar Sunfish L. marginatus 21 39 92 -2.167 + 0.147 SL 0.939 less than 0.001 0.129–0.164 Present study Longear Sunfish L. megalotis 92 50 130 -1.965 + 0.127 SL 0.856 less than 0.001 0.116–0.138 Present study Redear Sunfish L. microlophus 19 33 150 -0.143 + 0.103 SL 0.977 less than 0.001 0.096–0.111 Present study Redspotted Sunfish L. miniatus 5 36 98 -0.029 + 0.111 SL 0.920 0.006 0.060–0.163 Present study Spotted Sunfish L. punctatus 36 27 110 -1.098 + 0.129 SL 0.833 less than 0.001 0.109–0.149 Present study Bantam Sunfish L. symmetricus 10 23 62 -0.520 + 0.134 SL 0.919 less than 0.001 0.103–0.164 Present study Smallmouth Bass Micropterus dolomieu 33 47 274 -1.768 + 0.144 SL 0.958 less than 0.001 0.134–0.155 Present study Spotted Bass M. punctulatus 33 22 261 -1.667 + 0.141 SL 0.962 less than 0.001 0.131–0.151 Present study Largemouth Bass M. salmoides 15 350 500 0.16 SLA Keast and Webb 1966 B 265 19 324 Log10 (Gape width) = 0.991 . . Huskey and Turingan 1.06 Log10 (SL) - 0.90 2001 272 20 324 Log10 (Gape width) = 0.991 Huskey and Turingan 1.11 Log10 (SL) - 0.98 2001 White Crappie Pomoxis annularis 48 25 251 -1.783 + 0.113 SL 0.953 less than 0.001 0.106–0.120 Present study Black Crappie P. nigromaculatus 38 50 251 -0.469 + 0.104 SL 0.946 less than 0.001 0.095–0.112 Present study 15 130 220 0.09 SLA Keast and Webb 1966 AReported as a percentage of SL. BPossibly M. floridanus (Lesueur) (Florida Bass). Southeastern Naturalist A.V. Fernando, K.B. Hecke, and M.A. Eggleton 2018 Vol. 17, No. 2 318 Table 4. Model parameters and descriptive statistics for mathematical relationships between body depth and standard length (SL) for 20 species of Centrarchidae. SL (mm) Body depth (mm) Common name Scientific name n Min Max Model r2 P 95% CI of b Source Shadow Bass Ambloplites ariommus 19 37 200 -2.265 + 0.444 SL 0.986 less than 0.001 0.418–0.470 Present study Ozark Bass A. constellatus 7 36 165 -0.886 + 0.399 SL 0.996 less than 0.001 0.372–0.424 Present study Rock Bass A. rupestris 7 18 138 21.805 + 0.235 SL 0.852 0.002 0.134–0.337 Present study 15 130 150 0.410 SLA Keast and Webb 1966 Flier Centrarchus macropterus 34 21 133 -0.847 + 0.487 SL 0.958 less than 0.001 0.451–0.523 Present study Green Sunfish Lepomis cyanellus 39 40 155 -1.243 + 0.441 SL 0.946 less than 0.001 0.406–0.475 Present study Pumpkinseed L. gibbosus 6 42 95 -8.283 + 0.570 SL 0.985 less than 0.001 0.482–0.659 Present study 72 -3.250 + 0.463 SL 0.990 Hambright 1991 Warmouth L. gulosus 43 30 184 -3.578 + 0.445 SL 0.982 less than 0.001 0.426–0.464 Present study Orangespotted Sunfish L. humilis 25 36 69 -4.426 + 0.481 SL 0.931 less than 0.001 0.426–0.536 Present study Bluegill L. macrochirus 191 32 153 -9.374 + 0.544 SL 0.939 less than 0.001 0.524–0.564 Present study 15 125 150 0.520 SLA Keast and Webb 1966 Dollar Sunfish L. marginatus 21 39 92 -3.189 + 0.500 SL 0.954 less than 0.001 0.449–0.552 Present study Longear Sunfish L. megalotis 92 50 130 -2.379 + 0.509 SL 0.820 less than 0.001 0.459–0.559 Present study Redear Sunfish L. microlophus 19 33 150 -2.920 + 0.473 SL 0.995 less than 0.001 0.456–0.490 Present study Redspotted Sunfish L. miniatus 5 36 98 1.583 + 0.435 SL 0.866 0.014 0.167–0.703 Present study Spotted Sunfish L. punctatus 36 27 110 -3.737 + 0.522 SL 0.947 less than 0.001 0.480–0.564 Present study Bantam Sunfish L. symmetricus 10 23 62 -2.432 + 0.504 SL 0.974 less than 0.001 0.441–0.568 Present study Smallmouth Bass Micropterus dolomieu 33 47 274 -2.869 + 0.294 SL 0.961 less than 0.001 0.273–0.316 Present study Spotted Bass M. punctulatus 33 22 261 -1.400 + 0.312 SL 0.858 less than 0.001 0.266–0.357 Present study Largemouth Bass M. salmoides 15 350 500 0.370 SLA Keast and Webb 1966 White Crappie Pomoxis annularis 48 25 251 -2.696 + 0.408 SL 0.799 less than 0.001 0.348–0.468 Present study Black Crappie P. nigromaculatus 38 50 251 -2.604 + 0.425 SL 0.973 less than 0.001 0.402–0.449 Present study 15 130 220 0.420 SLA Keast and Webb 1966 AReported as a percentage of SL. Southeastern Naturalist 319 A.V. Fernando, K.B. Hecke, and M.A. Eggleton 2018 Vol. 17, No. 2 values varying from 0.764–0.995 (average = 0.929; Table 1). For body depth–TL models, r2 values varied from 0.802 to 0.998 (average = 0.939 and exceeded 0.900 for 14 of 20 models (Table 2). For horizontal gape–SL models, r2 values varied from 0.833 to 0.995 (average = 0.944) and exceeded 0.900 for 18 of 20 models (Table 3). For body depth–SL models, r2 values varied from 0.799 to 0.996 (average = 0.936) and exceeded 0.900 for 15 of 20 models (Table 4). Objective 2 There were distinct differences in mean horizontal gape–SL ratios among piscivory likelihood categories (F3,702 = 70.66, P < 0.001); all piscivory likelihood categories differed significantly from each other (Fig. 3). We observed the smallest horizontal gape–SL ratio in non-piscivorous centrarchids (0.083 ± 0.005 for Lepomis gibbosus [Pumpkinseed], and 0.086 ± 0.001 for Bluegill), while the greatest horizontal gape–SL ratio was observed in highly piscivorous centrarchids (0.126 ± 0.002 for M. dolomieu [Smallmouth Bass], and 0.122 ± 0.003 for Micropterus punctulatus [Spotted Bass]). Mean body depth–SL ratios also differed among piscivory likelihood categories (F3,702 = 91.24, P < 0.001). Although the pairwise difference between non-piscivorous species and species with a low likelihood of piscivory was not significant (P = 0.052), the other pairwise comparisons were all significant (P < 0.001) (Fig. 4). We found the largest body depth–SL ratios from species with low likelihoods of piscivory (0.468 ± 0.008 for Flier, and 0.461 ± 0.020 for Redspotted Sunfish); the smallest body depth–SL ratio occurred in the species with the highest likelihoods of piscivory (0.265 ± 0.004 for Smallmouth Bass, and 0.290 ± 0.11 for Spotted Bass). Figure 3. Mean horizontal gape–SL classified by likelihood of piscivory; error bars represent 1 SE. Likelihood of piscivory categories modified from Coll ar et al. (2009). Southeastern Naturalist A.V. Fernando, K.B. Hecke, and M.A. Eggleton 2018 Vol. 17, No. 2 320 Objective 3 Linear-regression analysis indicated that the slope of the composite TL–SL relationship for all 20 centrarchids differed from zero (t = 243.3, df = 703, r2 = 0.988, P < 0.001; Fig. 5). ANCOVA results suggested that slope coefficients from the Figure 4. Mean body depth–SL classified by likelihood of piscivory; error bars represent 1 SE. Likelihood of piscivory categories modified from Collar et a l. (2009). Figure 5. Generalized relationship of SL to TL for the family Centrarchidae. Each data point represents an individual fish from 1 of the 20 species assessed. Trend line indicates significant positive relationship (t = 243.3, df = 703, r2 = 0.988, P < 0.001). Southeastern Naturalist 321 A.V. Fernando, K.B. Hecke, and M.A. Eggleton 2018 Vol. 17, No. 2 TL–SL model were similar among species (F18,667 = 1.438, P = 0.107), which indicated that the slopes for individual species were statistically parallel. We detected significant among-species differences with respect to the intercept coefficients (F19,667 = 5.239, P < 0.001) of the TL–SL models. Intercept coefficients varied from –5.74 to 10.16 mm across species, and averaged 2.30 mm. There was no apparent pattern based on feeding mode with respect to the species-specific intercept coefficients. This model produced a generalized equation for conversion of SL to TL in centrarchids as: TL = 2.761 + 1.202 SL All measurements taken and the script to perform the analyses described above have been archived using Dryad and are available online at doi:10.50 61/dryad.17m23. Discussion Long-term preservation techniques are known to induce changes in fish morphometrics over long time-periods (Peterson and VanderKooy 1996), but we detected little evidence of these effects in this study. We found no differences between the horizontal gape–SL or body depth–SL ratios of live fish specimens compared to those preserved for >12 months. Ratios from preserved fish were intermediate in the range of measured values from all 3 sources of live fish. Although this finding was based only on Bluegill, it seemed unlikely that this characteristic would vary greatly across different centrarchid species. Thus, models developed in the present study should be generalized enough to be applicable for fisheries-management purposes. These models would be especially useful for studies attempting to assess interspecies interactions or quantify predator–prey interaction s. Differences in horizontal gape–SL ratios between the live and preserved fish were small (0.080–0.094; Fig. 1). The greater variation observed with statehatchery fish compared to commercial-hatchery fish or wild fish was unexpected. However, this finding could have reflected commercial breeding practices, whereby a specific Bluegill morphology was artificially selected for, perhaps even inadvertently. Similarly, low variation in horizontal gape–SL ratios in both wild and preserved fish (which were initially collected as wild fish) also might reflect natural selection forces occurring in the wild. Wild Bluegill demonstrated a substantially smaller body depth–SL ratio (0.36) than both the preserved fish and fish from the other 2 hatcheries (0.43–0.44; Fig. 2). The wild Bluegill used in this study were collected from a small, unmanaged pond on the UAPB campus with virtually no structure or large-bodied piscivorous fishes. It was possible that Bluegill inhabiting this pond have adapted a more streamlined, limnetic-zone body form (e.g., Layzer and Clady 1987). Our results have been influenced by the fact that the HSU and UAPB collections were substantially based on field collections made in Arkansas. Within Arkansas, Bluegill have been widely and indiscriminately stocked as forage by both private landowners and government agencies since the 1940s. Intermediate values in the morphometric ratios we measured possibly reflected a mix of hatchery and wild Southeastern Naturalist A.V. Fernando, K.B. Hecke, and M.A. Eggleton 2018 Vol. 17, No. 2 322 origins in the preserved specimens, even though all fish were originally collected in the wild. Although it may have been preferable to use live fish throughout this study, doing so would have required intensive state-wide fish collections that we deemed unfeasible. We feel it was improbable that such a study could have been conducted in the current funding environment. Use of preserved fishes may have underestimated the maximum body depth of prey for each species, but our horizontal-gape measurements were similar to the horizontal maxillary oral gape as defined by Mihalitsis and Bellwood (2017). The only exception was that those authors had manipulated the mouth to its maximum stretched position whereas we measured fish with the mouth closed. Length, depth, and gape relationships This study documents strong linear relationships between both horizontal gape and body depth, and TL and SL for all 20 centrarchid species examined. This statement was true even when thoroughly considering the influence of preservation. Overall, 61 of 78 models fitted exhibited r2 values exceeding 0.90. Determination of such sizes is necessary for modern mechanistic approaches to population modeling (Audzijonyte et al. 2014, Persson and Diehl 1990). In fisheries-management contexts, opportunistically piscivorous sunfishes may need to be considered as predators of juvenile sportfishes until they outgrow gape limitations. More research utilizing existing institutional collections of fishes to find morphological limits to behavior should be conducted on other regionally important freshwater fishes, including the North American families Moronidae and Percidae. Nearly all centrarchids will consume fish prey opportunistically, provided their gape is large enough to permit consumption (Wainwright and Richard 1995). Horizontal gape–SL ratios increased directly relative to likelihood of piscivory, which was consistent with model predictions. In particular, species with the greatest likelihoods of piscivory (e.g., Micropterus spp.) displayed significantly greater horizontal gape–SL ratios than species with lower likelihoods of piscivory (e.g., all Lepomis spp., Ambloplites spp., Pomoxis spp., and Centrarchus macropterus). Collar et al. (2009) suggested that the diversity of mouth shapes in centrarchids was constrained due to some level of piscivory being present throughout the family. Although our interpretation of these results may be partly confounded by effects of shared ancestry, centrarchids appear to have at least some morphological variety driven by feeding type; more-piscivorous centrarchids consistently had proportionately larger mouths than non-piscivorous or less-piscivorous species. Conversely, centrarchids exhibiting greater likelihoods of piscivory (e.g., Micropterus spp.) also had lower body depth–SL ratios, which was consistent with more streamlined body forms. Although this body form is more associated with pursuit predators (e.g., Morone spp. or Sander spp.), this trait would still be consistent with an ambush predator (e.g., Micropterus spp.) when only centrarchids are considered. Body depth–SL ratios increased monotonically as the likelihood of piscivory for species decreased, but then declined for non-piscivorous species (e.g., Longear Southeastern Naturalist 323 A.V. Fernando, K.B. Hecke, and M.A. Eggleton 2018 Vol. 17, No. 2 Sunfish and Spotted Sunfish) that feed mostly on small crustaceans, mollusks, and insect larvae. In effect, a greater body depth–SL ratio is advantageous when maneuvering through vegetation or other littoral-zone structure for mobile prey. There is little advantage to a greater body depth–SL ratio when feeding on less-mobile prey. This general relationship may be modified by the presence of predators. In prey species, rapidly increasing body depth in the presence of predators may also provide a size-refugium that permits access to additional diet resources (e.g., Persson et al. 1996). Presence of predator cues has been demonstrated to increase the body depth–fork length ratio with Pumpkinseed (Januszkiewicz and Robinson 2007). In addition, within-species variation in body morphology has been found to be associated with predation risk rather than heritability in Rutilus rutilus L. (Common Roach) (Scharnweber et al. 2013). The present study reports a strong linear relationship between TL and SL for the 20 centrarchid species examined, which produced a reliable generalized TL–SL equation. Carlander (1977) summarized many aspects of studies of fish morphometrics, including TL–SL ratio. Within the centrarchids, Carlander (1977) reported a minimum TL–SL ratio of 1.121 for Largemouth Bass, and a maximum TL–SL ratio of 1.388 for Pomoxis nigromaculatus (Black Crappie). Although these represent relatively extreme values, even for these species, the majority of TL–SL ratios suggested by Carlander (1977) fell within the 95% confidence limits of the TL–SL slope generated in the present study. Thus, the generalized equation reported here may be useful for converting SL to TL in centrarchids, though the researcher should be mindful that individual populations of any species can exhibit local variability unique to that population. The funnel-shaped plots of horizontal gape against TL (Fig. 6) and body depth against TL (Fig. 7) support using species-specific models rather than a generalized model for those characteristics. The models developed in this study should be sufficiently robust to be of wide use. The 20 species that we included represent the majority of the North American centrarchid fauna, and all species important to fisheries management. Centrarchid species excluded from this study were few, and none were species readily available for measurement. The only centrarchid genera excluded from this study were Acantharchus, Archoplites, and Enneacanthus. Acantharchus includes only 1 extant species, Acantharchus pomotis (Baird) (Mud Sunfish). This species is a smallbodied fish occurring only in Atlantic Slope rivers of the eastern US. Archoplites interruptus (Girard) (Sacramento Perch) is the only member of its genus, and the only centrarchid naturally found west of the Rocky Mountains. The 3 species of Enneacanthus are diminutive centrarchids that live in marginal habitats and would not likely be included in any fisheries management plan. In modern fisheries management, fishes are often released alive unless hard structures such as otoliths are needed for ageing. As a result, the managers often only have measurements of length (either SL or TL depending on the standards of the agency conducting the sampling) and weight. Sunfishes (Bluegill in particular) are often assumed to be prey for sportfishes such as Largemouth Bass. However, an appropriate subdivision of the sunfish population to identify what proportions Southeastern Naturalist A.V. Fernando, K.B. Hecke, and M.A. Eggleton 2018 Vol. 17, No. 2 324 are available as prey and predator requires an understanding of how body depth increases with length. It is likewise important to understand how sunfish gape increases with length because sunfish predation on fry and juveniles may be more prevalent than previously believed (e.g., Robinson and Buchanan 1988). Although modeling of this interaction has rarely been attempted, more detailed modeling of Figure 6. Scatterplot of all horizontal-gape measurements against TL. Funnel shaped plot suggests species-specific models are more appropriate than a gen eral regression. Figure 7. Scatterplot of all body-depth measurements against TL. Funnel shaped plot suggests species-specific models are more appropriate than a genera l regression. Southeastern Naturalist 325 A.V. Fernando, K.B. Hecke, and M.A. Eggleton 2018 Vol. 17, No. 2 interspecies interactions may permit more efficient allocation of hatchery resources in the future. This need has become more apparent as natural resource agencies deal with decreasing budgets for hatchery operations. Acknowledgments The authors acknowledge the assistance of S. Lochmann for access to the UAPB Ichthyology Teaching Collection and laboratory space, R. Tumilson for access to the HSU Zoology Collection and laboratory space, and A. Sanders, H. Dickey, and P. Fernando for assistance with measurements. Literature Cited Alfermann, T.J., and L.E. Miranda. 2013. Centrarchid assemblages in floodplain lakes of the Mississippi Alluvial Valley. Transactions of the American Fisheries Society 142:323–332. Audzijonyte, A., E.A. Fulton, and A. Kuparinen. 2014. The impacts of fish-body size changes on stock recovery: A case study using an Australian marine ecosystem model. ICES Journal of Marine Science 72:782–792. Bonar, S.A., W.A. Hubert, and D.W. Willis. 2009. Standard methods for sampling North American freshwater fishes. American Fisheries Society, Bethesda, MD. 335 pp. Carlander, K.D. 1977. Handbook of Freshwater Fishery Biology. Iowa State University Press, Ames, IA. 432 pp. Collar, D.C., B.C. O’Meara, P.C. Wainwright, and T.J. Near. 2009. Piscivory limits diversification of feeding morphology in centrarchid fishes. Evolution 6 3:1557–1573. Dewey, M.R., W.B. Richardson, and S.J. Zigler. 1997. Patterns of foraging and distribution of Bluegill Sunfish in a Mississippi River backwater: Influence of macrophytes and predation. Ecology of Freshwater Fish 6:8–15. Fernando, A.V. 2015. Largemouth Bass recruitment in southeastern Arkansas Lakes. M.Sc. Thesis. University of Arkansas at Pine Bluff, Pine Bluff, AR. 139 pp. Hale, R.S. 1996. Threadfin Shad use as supplemental prey in reservoir White Crappie fisheries in Kentucky. North American Journal of Fisheries Management 16:619–632. Hambright, K.D. 1991. Experimental analysis of prey selection by Largemouth Bass: Role of predator mouth-width and prey body-depth. Transactions of the American Fisheries Society 120:500–508. Hill, J.E., L.G. Nico, C.E. Cichra, and C.R. Gilbert. 2004. Prey vulnerability to Peacock Cichlids and Largemouth Bass based on predator gape and prey body-depth. Proceedings of the Southeastern Association of Fish and Wildlife Agencies 58:47–56. Huskey, S.H., and R.G. Turingan. 2001. Variation in prey-resource utilization and oral-jaw gape between two populations of Largemouth Bass, Micropterus salmoides. Environmental Biology of Fishes 61:185–194. Hoyle, J.A., and A. Keast. 1987. The effect of prey morphology and size on handling time in a piscivore, the Largemouth Bass (Micropterus salmoides). Canadian Journal of Zoology 65:1972–1977. Januszkiewicz, A.J., and B.W. Robinson. 2007. Divergent Walleye (Sander vitreus)-mediated inducible defenses in the centrarchid Pumpkinseed Sunfish (Lepomis gibbosus). Biological Journal of the Linnean Society 90:25–36. Johnson, J.M., and D.M. Post. 1996. Morphological constraints on intracohort cannibalism in age-0 Largemouth Bass. Transactions of the American Fisheries Society 125:809–812. Southeastern Naturalist A.V. Fernando, K.B. Hecke, and M.A. Eggleton 2018 Vol. 17, No. 2 326 Keast, A., and D. Webb. 1966. Mouth and body form relative to feeding ecology in the fish fauna of a small lake, Lake Opinicon, Ontario. Journal of the Fisheries Research Board of Canada 23:1845–1874. Lawrence, J.M. 1958. Estimated sizes of various forage fishes Largemouth Bass can swallow. Proceedings of the Southeastern Association of Fish and Wildlife Agencies 11:220–225. Layzer, J.B., and M.D. Clady. 1987. Phenotypic variation of young-of-year Bluegills (Lepomis macrochirus) among microhabitats. Copeia 1987:702–707. Lubinski, B.J., J.R. Jackson, and M.A. Eggleton. 2008. Relationships between floodplainlake fish communities and environmental gradients in a large river-floodplain ecosystem. Transactions of the American Fisheries Society 137:895–908. Mihalitsis, M., and D.R. Bellwood. 2017. A morphological and functional basis for maximum prey size in piscivorous fishes. PLoS One 12:e0184679. Nowlin, W.H., R.W. Drenner, K.R. Guckenberger, M.A. Lauden, G.T. Alsonso, J.E. Fennell, and J.L. Smith. 2006. Gape limitation, prey-size refuges, and the top-down impacts of piscivorous Largemouth Bass in shallow pond ecosystems. Hydrobiologia 563:357–369. Olive, J.A., L.E. Miranda, W.D. Hubbard. 2005. Centrarchid assemblages in Mississippi state-operated fishing lakes. North American Journal of Fisheries Management 25:7–15. Persson, L., and S. Diehl. 1990. Mechanistic individual-based approaches in the population/ community ecology of fish. Annales Zoologici Fennici 27:165–182. Persson, L., J. Anderson, E. Wahlstrom, and P. Eklov. 1996. Size-specific interactions in lake systems: Predator-gape limitation and prey-growth rate and mortality. Ecology 77:900–911. Peterson, M.S., and S.J. VanderKooy. 1996. Preservation-induced changes in morphometrics of fishes: Influence on prey-size choice in juvenile Walleye, Stizostedion vitreum (Mitchill). Ecology of Freshwater Fish 5:133–139. Robison, H.W., and T.M. Buchanan. 1988. Fishes of Arkansas. University of Arkansas Press, Fayetteville, AR. 536 pp. Scharnweber, K., K. Watanabe, J. Syvaranta, T. Wanke, M.T . Monaghan, and T. Mehner. 2013. Effects of predation pressure and resource use on morphological divergence in omnivorous prey fish. BMC Evolutionary Biology 13:132. Schramm, H.L., and A.V. Zale. 1985. Effects of cover and prey size on preferences of juvenile Largemouth Bass for Blue Tilapias and Bluegills in tanks. Transactions of the American Fisheries Society 114:725–731. Shireman, J.V., D.E. Colle, and R.W. Rottman. 1978. Size limits to predation on Grass Carp by Largemouth Bass. Transactions of the American Fisheries Society 107:213–215. Timmerman, C.M., C.A. Annett, and C.F. Bailey. 2000. Determination of factors limiting prey-size swallowed by larval and small juvenile Largemouth Bass. Transactions of the American Fisheries Society 129:618–622. Tonn, W.M., J.J. Magnuson, M. Rask, and J. Toivonen. 1990. Intercontinental comparison of small-lake fish assemblages: The balance between local and regional processes. The American Naturalist 136:345–375. Wainwright, P.C., and B.A. Richard. 1995. Predicting patterns of prey use from morphology of fishes. Environmental Biology of Fishes 44:97–113. Werner, E.E. 1974. The fish size, prey size, handling time in relation in several sunfishes, and some implications. Journal of the Fisheries Research Board of C anada 31:1531–1536.