Investigation of the relationship between the steroid hormone 11-ketotestosterone and reproductive status in the fish Lythrurus fasciolaris.
Abstract: In teleost fishes, 11-ketotestosterone (11KT) is a critical androgen regulating primary and secondary sex characteristics. In the sexually dimorphic Lythrurus fasciolaris (Cyprinidae), dominant nuptial males display heavy tuberculation on the head and nape, dark dorsolateral vertical bars, and dramatic red coloration in the fins, venter, and operculum area. This study aimed to quantify 11KT circulating levels in males and females, and determine any correlation with key male reproductive status indicators such as nuptial coloration, size, and gonadosomatic index (GSI). Thirty-one wild-caught L. fasciolaris were divided into three groups according to reproductive status: dominant males (D), nondominant males (ND), and females (F). Physical measurements, digital imaging, and blood samples were used to quantify body size, GSI, nuptial coloration, and 11KT circulating levels. Dominant males had higher 11KT levels and nuptial coloration traits compared to ND males and females (red area, hue, saturation), and a higher GSI than ND males. Non-dominant males had more 11KT and coloration than females. Increased 11KT levels corresponded to increased coloration, size and GSI.
Article Type: Report
Subject: Steroid hormones (Physiological aspects)
Steroid hormones (Health aspects)
Cyprinidae (Physiological aspects)
Cyprinidae (Health aspects)
Authors: Schade, Jennifer
Stallsmith, Bruce
Pub Date: 07/01/2012
Publication: Name: The American Midland Naturalist Publisher: University of Notre Dame, Department of Biological Sciences Audience: Academic Format: Magazine/Journal Subject: Biological sciences; Earth sciences Copyright: COPYRIGHT 2012 University of Notre Dame, Department of Biological Sciences ISSN: 0003-0031
Issue: Date: July, 2012 Source Volume: 168 Source Issue: 1
Product: Product Code: 2833220 Steroids NAICS Code: 325411 Medicinal and Botanical Manufacturing SIC Code: 2833 Medicinals and botanicals
Geographic: Geographic Scope: United States Geographic Code: 1USA United States
Accession Number: 296160030
Full Text: INTRODUCTION

Ornamentation plays an important role in animals in mate selection and social status by means of courtship (intersexual selection) and same-sex competition (intrasexual selection) for access to mates or resources (Noble, 1938; Kodric-Brown, 1998; Sargent et al., 1998; Andersson and Simmons, 2006). In teleost fishes, ornamentation may include seasonal physical features such as tubercles (e.g., cyprinids), or colorful pigmentation on the body and/or fins (e.g., red fins in breeding males of the scarlet shiner, Lythrurus fasciolaris (Gilbert)). In many groups of fishes that express dichromatism (color difference between sexes), males are typically the more ornately colored sex and, in many freshwater species, color expression is only during the breeding season (referred to as seasonal dichromatism).

Within species having seasonally dichromatic males, most females are attracted to and select mates with only the most colorful features. The colorful body patterns help to facilitate the assessment of social status and motivations of the displaying male during courtship and male-male competitions by enhancing the visibility of body posturing and behavior. Nuptial coloration has further significance in courtship in that temporal variations in coloration may act as an "honest signal" that indicates a male's physical condition and reproductive status or success (Kodric-Brown, 1998; Sargent et al., 1998). This visual cue may be used by discriminating females in selecting mates that are in the best physical or reproductive condition. More specifically, it may signal reproductive benefits such as sperm fertility traits, physical benefits related to health and parasite load, or even social conditions such as dominance and ability to possess nesting territory and other resources essential for female reproduction (Kodric-Brown, 1998; Sargent et al., 1998; Pitcher and Evans, 2001; Evans et al., 2002; Andersson and Simmons, 2006; Parikh et al., 2006a; Pitcher et al., 2007).

In teleost fishes, the brain-pituitary-gonad (BPG) axis regulates sexual development, reproductive activity, and expression of secondary sex characteristics, such as nuptial ornamentation. It is homologous to the hypothalamus-pituitary-gonad (HPG) axis in other vertebrates; however, unlike other vertebrates in which a blood portal system regulates gonadotropin-releasing hormone (GnRH) to the pituitary, in teleosts GnRH neurons of the preoptic area of the hypothalamus directly innervate the pituitary (Blazquez et al., 1998; Grober and Bass, 2002; Godwin et al., 2003). This direct connection between the hypothalamus and the pituitary gives teleost fishes a more rapid and direct response to environmental stimulants. Secretion of GnRH regulates the release of gonadotropin hormones (GtHs) from the pituitary, which in turn stimulates reproductive activity and secretion of gonadal steroid hormones such as androgens and estrogens. In male teleosts, androgens are an important output of the BPG axis because they stimulate gonadal development, initiate spermatogenesis, provide biofeedback regulation of the BPG axis and regulate the expression of secondary sexual characteristics such as nuptial color changes, behavior, and intromittent organs (Blazquez et al., 1998; Kodric-Brown, 1998; Grober and Bass, 2002; Yamaguchi et al., 2005; Parikh et al., 2006a; Yamaguchi et al., 2006).

The two primary androgens regulating reproductive activity in male fish are testosterone (T) and 11-ketotestostosterone (11KT). Testosterone is important in sex determination and development, whereas 11KT is crucial to testes maturation and the expression of secondary sex characteristics (Godwin et al., 2003). There is great reproductive plasticity (gonochoristic, protogynous, protandry, and hermaphroditic) among fishes and thus a large variation in steroidal influences on sex characteristics. Testosterone levels between sexes vary significantly between species and seasonality. In some species, testosterone levels are higher in dominant males, whereas in others testosterone levels do not differ between sexes and even in other species testosterone levels are higher in females than males (Grober and Bass, 2002; Godwin et al., 2003). This latter difference where females have higher testosterone levels than males may not be unusual since testosterone is an estrogen precursor which is readily converted to estradiol-17[beta] by cytochrome P-450 aromatase, 11KT, however, is not convertible and thus is thought to be more potent and play a more important role in male secondary sexual characteristics and behavior (Godwin et al., 2003; Remage-Healey and Bass, 2006).

Lythrurus fasciolaris is a member of the Cyprinidae which includes freshwater minnows and carps. The species inhabits small to medium-sized clear-water steams with rock, pebble, and gravel substrates and medium water flow. It is found from the Ohio basin in Ohio and southeastern Indiana throughout the Cumberland River drainage in Kentucky and Tennessee to the Tennessee drainage in north Alabama (Boschung and Mayden, 2004).

Like many cyprinids, Lythrurus fasciolaris is a "tournament" species: males compete intraspecifically for mate-alluring resources or access to mates, with some males developing a larger, more colorful phenotype and aggressive behaviors to assert their dominance over other males. During the breeding season, males without the more colorful phenotype are not easily distinguishable from drab females, but nuptial males attempting to assert their physical dominance develop a series of blue-gray saddle bands across the back that extends ventrally to the lateral line. The fins (dorsal, caudal, anal, and leading ray of the pectoral fin), preopercle area (anterior margin of gill covering) and venter display bright red to orange coloration. Dominant males vigorously fight for access to females and defend territories while waiting to attract females to their spawning site. Once a male has attracted a female they spawn over the territory, with no further brood care by either parent (Dimmick et al., 1996; Boschung and Mayden, 2004).

This study looks at the relationship of the androgen 11KT to nuptial coloration and gonad size in males. The following hypotheses are tested: (1) Dominant males are larger in size (length and weight) and have greater gonad size, nuptial coloration (as measured by area of red pigmentation, degree of hue and percent saturation and brightness) and plasma 11KT levels than non-dominant males and females; and (2) plasma 11KT is positively correlated with all properties of nuptial coloration and gonadal development.

MATERIALS AND METHODS

Thirty-one Lythrurus fasciolaris were used to examine nuptial coloration characteristics and 11KT concentrations. All physical manipulations (blood collection, photographs, length, and weight) of the subjects were done sequentially, one subject at a time, due to the complexity of the procedures being performed on anesthetized subjects. Immediately upon complete anesthetization, first a blood sample was collected; second, digital photographs were taken; third, body length and weight were measured; and finally the subject was placed in phosphate-buffered 10% formaldehyde for later gonadal inspection. Gonad weight, 11KT plasma concentrations, and color analyses were measured on subsequent days.

Subjects were placed into three classes: dominant male (D), non-dominant male (ND), and female (F), based on specific ornamental features. Dominant status of male specimens was determined by two most identifiable nuptial secondary sex characteristics of reproductive Lythrurus fasciolaris males: (1) heavy tuberculation (large, horny protuberances covering most of head/nape epidermis) and (2) presence of the saddle (dark, vertical dorsolateral bars) pattern dorsally (Dimmick et al., 1996). Males not meeting both criteria were classified as non-dominant. Studies on social status and behavior of other ornamented male fish species showed positive correlations between social dominance and two characteristics--tuberculation and nuptial coloration--in males (Kodric-Brown, 1998; Grober and Bass, 2002; Kortet et al., 2004; Parikh et al., 2006a). Personal observations of captured L. fasciolaris males suggest that this species demonstrates the same correlations between ornamentation and social status and supports the assumption of social classification for the purpose of this study. Sex was confirmed by gonad inspection. Among the 15 males, eight were classified as dominant and seven as non-dominant.

FISH COLLECTION AND CARE

Subjects were captured using a seine on two dates in May and Jun., from Limestone Creek located at U.S. Highway 53 in Madison County, Alabama, a type locality for this species. Limestone Creek is a clear, spring fed stream in the Tennessee River watershed. As its name suggests, the substrate is primarily limestone bedrock with small areas of pebble substrate in pools and beds of vegetation near shallow riffles. Captured fish were transported alive to the facility and were kept in aquaria for no more than seven days before the experiment.

BODY MEASUREMENTS

Subjects was placed in a freshly prepared MS-222 (Tricaine methanesulfonate) water bath (30 mg MS-222 per 200 ml water; for larger fish 40 mg of MS-222 was used). Complete sedation (<2 min) was determined by pinching the caudal peduncle of the fish and the fish was removed only when no response (no movement) was observed. To ensure optimal blood flow for blood sampling, sedation of the fish was closely monitored to prevent over sedation. MS-222 is also used to euthanize fish, and over exposure to MS-222 will decrease the heart rate and make blood sampling more difficult in small fish.

Standard length and weight were measured immediately after the blood sample and digital photographs were taken. Standard length (SL), defined as the length of the body from the tip of the mouth to the base of the caudal peduncle, was measured to the nearest 0.01 mm using a calibrated electronic digital caliper (Fisher Scientific). Body weight ([W.sub.B]) was measured to the nearest mg using a calibrated digital balance (Ohaus Explorer). After length and weight was recorded, specimens were placed in phosphate-buffered 10% formaldehyde for 2-3 d to fix gonads for easier excision. After the gonads were excised, gonad weight ([W.sub.G]) was measured to the nearest mg using the same digital balance used to measure body weight. Gonadosomatic index (GSI), which is the ratio of gonad to body weight, was calculated using the equation: ([W.sub.G]/([W.sub.B]-[W.sub.G])) * 100.

PLASMA COLLECTION AND 11KT ASSAY

Blood samples (between 6.5 and 86 [micro]l) were taken from anesthetized subjects using a heparinized capillary tube. Since it was unknown exactly how much plasma was required to effectively test 11KT levels in this species, as much blood as the subject would yield was collected, effectively euthanizing the subject by the end of the collection process. This process is commonly done to laboratory animals and is a humane method of euthanasia when the maximum quantity of blood is needed from a subject. Blood was collected within 5 min of removal from the tank by means of a puncture to the caudal sinus. The caudal sinus was located just ventral to the lateral line and vertically aligned with the caudal end of the dorsal fin. It was punctured from the subject's right side so that subsequent pictures taken of the subject's left side were not compromised. A sharp, pointed scalpel was used to puncture the sinus causing blood to quickly pool at the skin's surface. As blood pooled at this site, it was quickly collected using a heparinized capillary tube (80 [micro]l capacity), and then transferred to a chilled 1.5-ml centrifuge tube containing 20 [micro]l of sample buffer with protease inhibitor (1%) and kept on ice. All blood samples were then centrifuged for 3 min at 12,000 rpm to separate plasma from red blood cells. The plasma was isolated and stored at -80 C until assayed. Plasma volume after centrifugation was assumed to be 55% of blood volume for all samples, based on known blood composition figures.

Plasma 11KT concentrations of seven D males, seven ND males and six females were measured using an enzyme-linked immunosorbent assay, or ELISA (Cayman Chemical, 2003). To balance the sample size among the groups, only six of the 16 females sampled were used for measuring 11KT levels. In addition, one dominant male was omitted from 11KT measures due to the very low blood volume that was collected (6.65 [micro]l) relative to its size (70.88 mm), suggesting the sample would be unreliable. Preliminary assays showed purification of the plasma samples was required. Cold-spike extraction using diethyl ether was used to purify the samples. The purpose of this method was the determination of the percent recovery of the samples after extraction using a known concentration of 11KT standard added to a second aliquot of sample before extraction. Plasma samples were transferred to glass tubes containing 200 [micro]l of EIA buffer. Two randomly chosen samples were split into two separate tubes each, one tube containing only buffer and the other containing buffer spiked with 11KT standard (5 ng/ml). The spiked samples were used to determine the average percent recovery of the samples after purification. Samples were extracted three times using diethyl ether. The diethyl ether was evaporated by heating to 30 C under a gentle stream of nitrogen, and then samples were stored overnight under desiccation at -20 C. Average recovery from this extraction method was 50%. Purified plasma samples were thawed and reconstituted in a 1:10 solution with assay buffer. Samples were then assayed at two dilutions (1:50 and 1:100) in triplicate. The ELISA kit protocol was then followed, which had a specificity of 100% for 11KT and [less than or equal to] 0.01% for other similar analytes. Circulating 11KT levels between females, ND males, and D males varied widely. Five of the D male samples were outside of the linear range of the standard curve and were assayed again at higher dilutions (1:1000, 1:10,000) in duplicate. Only duplicates were used in the second assay because low inter-well variability within replicates showed that the pipetting technique was very consistent. Two samples, one ND male and one female, had to be omitted from the data set. Although these two samples were on the linear graph of the standard curve, there was a large disparity (>20%) between the dilutions of these samples (disparity between sample dilutions: ND = 40%, F = 26%) suggesting unreliability of the measured results for the two. Plates were read (Bio-Tek PowerWave HT) at 405 nm. Final sample size used for 11KT measurements was 18, composed of seven D males, six ND males and five females. Final dilution values used to calculate the concentrations of 11KT in the plasma samples were as follows: D males = 1:10,000 or 1:1000; ND males = 1:100; and females = 1:50 dilution factor. Results of the ELISA were analyzed using an Excel spreadsheet tool provided by the manufacturer of the ELISA kit (Cayman Chemical, 2003).

COLORATION ASSESSMENT

In Lythrurus fasciolaris, reproductive males express intense red coloration in their fins and operculum during the breeding season. The degree of coloration in L. fasciolaris has two aspects: the ratio of color cover relative to body area and the intensity of the color. To estimate the degree of intensity of the red coloration, three properties of coloration--hue, saturation, and brightness--were measured. Hue is commonly thought of as the color (red, yellow, etc.) and is the property associated with the wavelength (spectral perspective) that is emitted by pigments in the integument. The carotenoid-based and pteridine pigments found in the pigmented skin of L. fasciolaris emit a hue range from yellow to orange to red. Saturation refers to the purity of the color, with fully saturated representing the truest version of that color. Brightness describes the amount of white in a color, or relative lightness, on a black to white scale.

To measure the variables of nuptial coloration previously discussed, subjects were photographed using a digital camera (Fujifilm FinePix A400) immediately following blood collection. In accord with the standard method for photographing fish, subjects were photographed from their left side with the dorsal, caudal, and anal fins spread open similar to a swimming position. A ruler, a color palette, and an identification tag were included in each photo. Subjects were placed on a non-reflective glass surface atop a white background. The non-reflective glass served two purposes. First, it helped reduce glare from the lighting while keeping the background clean. Second, and more importantly, it provided a slightly textured surface for manipulating and keeping the erect fins in position without the need for pins.

Subjects were photographed under standardized lighting conditions consisting of two full spectrum light sources positioned about fifteen cm above the subject's body surface, with ambient fluorescent room lighting in the background. All images were originally saved as JPG file formats and were later converted to TIFF as the working storage format.

To estimate the percentage of pigmented area in the fins and body, the total surface area of the body and specific areas of interest (AOI) displaying red coloration were measured from the digital images using NIH image analysis software (Image J: free software available at http://rsb.info.nih.gov/nih-image/). Presence of pigmentation was determined using a non-reproductive female expressing no coloration or other nuptial characteristics as the null model. Any area of visible coloration in the AOIs of all other subjects compared to the null model, regardless of degree of color (ranging from pale orange to vibrant red) was considered pigmented and thus outlined and quantified.

The areas of interest include the dorsal fin, caudal fin, anal fin, and operculum. For each image, before measuring the red color area of the AOIs, the ruler in the photograph was used to set the measurement scale (mm) of the polygon tool. Using the polygon tool, an outline was drawn around the perimeter of the body and around the pigmented (red to red-orange) color patches that appeared in the AOIs shown in Figure 1A. Other areas of pigmentation were noted as observations but were not measured due to the lack of clear perimeters around these areas. Only the most prominent areas of red coloration considered the most significant to dominance, and easily measurable, were chosen. Preliminary tests of the polygon tool's precision showed no significant differences in "area" results between multiple polygon replications measuring a given AOI. Therefore, a single polygon measure of each AOI was used to represent the value of the AOI.

[FIGURE 1A OMITTED]

To determine the ratio of red coloration on the body as discussed above, the sum of the colored AOIs in the dorsal, caudal, and anal fins and operculum was divided by the total body surface area of the subject. To determine the intensity of red coloration, numerical estimates of three color properties--hue, saturation, and brightness--were measured using Adobe Photoshop[R] (Skarstein and Folstad, 1996; Liljedal et al., 1999). In Adobe Photoshop, numerical values are produced for each property. Hue is measured as an angular location (between 0 and 360 degrees) on a standard color wheel with red ranging from 0 degrees (pure red) to about 45 degrees (yellow-orange). Here, the color of hue values is defined as: red = 0-25[degrees], red-orange to orange-yellow = 26-50[degrees] and no pigmentation present ->50[degrees]. Saturation is measured as a percentage of gray in proportion to hue from 0% (gray) to 100% (fully saturated), and brightness is measured as a percentage of lightness from 0 (black) to 100% (white).

Any slight lighting variance between digital images was standardized using Adobe Photoshop before measuring the color properties in order to provide the most accurate estimate of the differences between subjects. To standardize the images, the brightness of the white paint chips in each image was adjusted as necessary so that the average histogram (grayscale 0-255) value was approximately 217 (the average value among the images).

[FIGURE 1B OMITTED]

Next, the distal area of the dorsal fin was used to estimate the hue, saturation, and brightness of red coloration. The "point sample" tool in Adobe Photoshop was set to measure and display the HSB values. The properties were measured simultaneously using four point samples transecting the distal tip of the dorsal fin, specifically placed on the flesh between the fin rays to measure only the pigmented tissue, shown in Figure 1B. For each subject, the value of each property was taken as the average of the four point measures.

STATISTICS

All statistical tests were performed using SPSS statistical analysis software (version 14.0 student version). Not all data met normal assumptions of equal variance among groups and normally distributed populations. Initially, all proportional data (i.e., GSI, saturation, brightness, and area of red coloration) were arcsine square root transformed before parametric statistical analyses in order to make them more normally distributed, and hue, 11KT and weight were [log.sub.10] transformed to render them more normally distributed (Zar, 1999). However, some data (i.e., hue, saturation, and brightness) still did not meet complete normality after transformations; therefore, the nonparametric Mann-Whitney statistical test was used to compare variation among D male, ND male, and female (control) groups. Because 22 of these Mann-Whitney tests were performed, sequential Bonferroni correction was applied to the resulting P values to avoid Type 1 errors of interpretation (Rice, 1989). Spearman's rank correlation was used to determine the relationships among the variables. All means are reported [+ or -] SE.

RESULTS

Statistical analyses were used to test for any significant differences in coloration and size between the three classes: dominant male (D: n = 8), non-dominant male (ND: n = 7), and females (F: n = 16). Sample sizes of 11 KT were a subset of this population (D: n = 7, ND: n = 6, F: n = 5).

RELATIONSHIPS BETWEEN CLASSES

A comparison of the mean [+ or -] SE and range of all measured variables is presented in Table 1. Mann-Whitney nonparametric rank scores indicated highly significant differences between groups for all variables except brightness among males and standard length and weight between ND males and females (Table 2). Using a sequential Bonferroni correction on the 22 rank scores in Table 2, a statistically significant P value was one of 0.006 or less. As expected for a tournament species, D males (mean: 68.3 [+ or -] 0.8 mm) were significantly larger in standard length than ND males (mean: 57.0 [+ or -] 1.8 mm, P = 0.001; Table 2) and females (mean: 52.7 [+ or -] 1.3 mm, P < 0.001; Table 2), and had more mass (ND: P = 0.001; Female: P < 0.001; means: D = 4.9 [+ or -] 0.2, ND = 2.6 [+ or -] 0.3, F = 2.0 [+ or -] 0.1 g; Table 2). ND males were not larger in size than females (standard length: P = 0.256; mass: P = 0.181; Table 2). Gonadosomatic Index (GSI) of D males was larger than ND males but not significantly so (P = 0.008, Table 2).

11-ketotestosterone levels of D males were more than 20 times higher than ND males or females, but only the difference between D males and females was significant (means: D = 39.3 [+ or -] 14.6, ND = 1.9 [+ or -] 0.4, F = 0.5 [+ or -] 0.2 pg/ul; P = 0.02 between D and ND males, P = 0.004 between D males and females; Tables 1 and 2). Although ND males had considerably lower 11KT levels, they were still higher than the small amount found in females but not significantly so (P = 0.02). Among D males, 11KT levels varied considerably (Table 1).

The area of red coloration, hue, and saturation differed among all three classes (Tables 1 and 2). Brightness difference was significant between D males and females (P = 0.002) but not significant between ND males and females or between D and ND males.

RELATIONSHIP AMONG VARIABLES IN MALES

Using Spearman's rank correlation to analyze untransformed data of all 31 subjects (males and females combined), there was a significant correlation between 11KT and all color intensity properties except brightness ([r.sub.s] = 0.361, P = 0.141, n = 18). 11KT was positively correlated with area of red coloration ([r.sub.s] = 0.744, P = 0.001, n = 18) and saturation ([r.sub.s] = 0.720, P = 0.001, n = 18), and negatively correlated with hue ([r.sub.s] = -0.777, P = 0.001, n = 18). The degree of red hue is lower than orange on the color wheel, thus the negative correlation value represents a positive relationship in which higher plasma concentrations of 11KT correspond to a more red (i.e., red = 0-25 degrees) than orange (i.e., orange = 26-50 degrees) or absence of color (i.e., no pigmentation [greater than or equal to] 50 C) pigmentation in the fins. GSI of females was 20 times higher than males and, therefore, could not be correlated with 11KT using the collective data set. The relationship between GSI and 11KT was analyzed in male subjects only.

The three parts of Figure 2 graphically show the separation between D males, ND males and females in the relationships between 11KT level and hue (Fig. 2A), 11KT and saturation (Fig. 2B), and 11KT and red area (Fig. 2C).

DISCUSSION

There were significant correlations between the secondary sex androgen 11KT, various nuptial coloration attributes and GSI. Only one color intensity property, brightness, showed no significant difference between social groups or correlations in this study. A probable cause for this was the transparency of the colorless fins in females that revealed the white background color through their fins. In Adobe Photoshop, brightness is the measure of white in a color, or relative lightness, on a black to white scale, and so the dramatic contrast between the red pigmented fins and transparent fins is not surprising or very informative.

Excluding the brightness variable, D males were significantly different than both ND males and females in most characteristics, and near significant difference in the others. The mean standard length of the D male was about 20% longer and the mean mass was about twice that of ND males and females (Table 1). It is not uncommon for dominant males of tournament species to be larger in size than the other members of the population. For example, in the African cichlid Astatotilapia burtoni (Gunther), dominant (territorial) males are larger and more aggressive at defending their territory (Parikh et al., 2006a, b). In the current study ND males did not significantly differ from females in size (Tables 1 and 2).

In D males, the variances (expressed as the SE) of the body measurements and coloration values (except brightness) were very small (Table 1), suggesting that individuals in this class are completely committed to the full expression of nuptial coloration. In contrast, ND males displayed a wide range of ornamentation from a silvery body color and a trace of pigmentation in only the dorsal, caudal, and anal fins to nearly D male colorations with individuals at various stages in between. Furthermore, the variances (SE) of the traits for the ND class were large compared to those seen in dominant males.

One possible cause for this may be that individuals in the ND class may be at different ages and stages of reproductive development. Those on the lower range of ornamentation may be in their first year of reproductive maturity or may still be immature, whereas the more ornamented individuals may be second or third year and more mature. Since this species has a life span of about 3 y and dominant males are also the largest in standard length and weight, it is possible, and more than likely probable, that dominance is related to age and sexual maturity. Although literature suggests that they are reproductively capable after 1 y, there may be an increasing scale of gonadal maturity that peaks in their last year if they survive their full life span.

Gonadosomatic index values and 11KT concentrations also support the idea that Lythrurus fasciolaris D males are at their peak reproductive condition. Mean GSI of D males was more than twice that of ND males (D = 0.5 [+ or -] 0.1, ND = 0.2 [+ or -] 0.1) and the mean 11KT plasma concentrations of D males was 20 times higher than ND males (D = 39.3 [+ or -] 14.6, ND = 1.9 [+ or -] 0.5). In Astatotilapia burtoni, territorial (dominant) males which are brightly colored, large, and reproductively capable had significantly higher levels (~8.5 fold) of circulating 11KT compared with non-territorial males (Parikh et al., 2006a). The D males were more reproductively mature as a group than ND males; however, there were two atypical individuals within each group.

[FIGURE 2 OMITTED]

In D males of this study there was a large variance (SE = 14.6, see Table 1) in 11KT levels. Values ranged from 1.61 pg/[micro] to 104.36 pg/[micro]l. One might expect to see consistent and nearly equal concentrations if individuals were expressing nearly the same degree of nuptial coloration and GSI as in this group of D males, but that is not the case. The variances in GSI and color related traits in D males were not large. Even with this wide range of levels D males have much more 11KT than ND males. This suggests that rather than a gradual gradient of phenotypic effects of 11KT, there is some sort of threshold mechanism in 11KT level which acts more like an activator producing dramatic phenotypic changes. Therefore, 11KT and gonad growth may not be regulated solely by the BPG pathway. For species adapted to seasonal breeding conditions (i.e., Lythrurus fasciolaris), utilizing rapid physiological changes and alternative regulatory mechanisms controlling reproductive traits and behavior in response to varying environmental cues may be necessary for reproductive success. The disparity between GSI and 11KT of some males could have been due to influences by a broader regulatory mechanism such as the arginine vasotocin (AVT) pathway. Fluctuating AVT levels have been shown to control reproductive social behavior and sex characteristics in many teleost species. In Cyprinodon pupfishes native to Death Valley in California, the AVT system was found to have high plasticity in subspecies being studied; it also influenced social behavior and regulated physiological changes and social behavior from environmental cues (Lema, 2008).

Females also had measurable concentrations of 11KT although they were slightly lower concentrations than ND males, with a mean of 0.5 [+ or -] 0.2 pg/[micro]l. Females lack a well-developed 11[beta]-hydroxylase enzyme system for converting testosterone to 11KT. The female that had the highest concentration of 11KT (1.09 pg/[micro]l) was one of two gravid females that expressed a faint orange tint in the dorsal fin.

In this study, when looking at the relationship between 11KT and ornamentation traits with respect to the entire sample size (males and females), 11KT was highly significantly correlated with the intensity (hue and saturation) and area of red coloration in scarlet shiners. This suggests a strong relationship between 11KT and the seasonal nuptial coloration observed in reproductively active males. However, when looking at the correlation between 11KT and the other traits in males, only a weak correlation between 11KT and saturation existed and the correlation with hue was reduced, suggesting that the percent of red coloration and color hue are the traits most affected by levels of circulating 11KT in males.

In addition, among males, there was a positive correlation between 11KT and GSI ([r.sub.s] = 0.632, P = 0.021), supporting other studies' findings that 11KT stimulates gonad development (Godwin et al., 2003). For example, in the bluehead wrasse, Thalassoma bifasciatum (a sequentially hermaphroditic, dimorphic reef fish), injection of 11KT into gonadally intact females not only induced the expression of male coloration and neural changes, but also induced sex change (Warner and Swearer, 1991). The direct relationship between gonad development and 11KT is not completely clear, however. In one species, 11KT may be crucial to the development of male gonads, whereas in another mature levels of circulating 11KT may be dependent on mature gonads. There is a high degree of plasticity among fish species regarding reproductive strategies and therefore mechanisms. The elucidation of this plasticity through such mechanisms as AVT influence on the preoptic region of the hypothalamus is important to understanding life history of many species. Such research is needed to better understand the relationship between 11KT and phenotype reported in this paper.

LITERATURE CITED

ANDERSSON, M. AND L. W. SIMMONS. 2006. Sexual selection and mate choice. Trends Ecol. Evol., 21:296-302.

BLAZQUEZ, M., P. T. BOSMA, E.J. FRASER, K.J. van LOOK, AND V. L. TRUDEAU. 1998. Fish as models for the neuroendocrine regulation of reproduction and growth. Comp. Biochem. Phys. C., 119:345-364.

BOSCHUNG, H. T. AND R. L. MAVEN. 2004. Fishes of Alabama. Smithsonian Books, Washington, D.C. 960 p.

CAYMAN CHEMICAL. 2003, Cold-spike extraction literature citation obtained from Cayman technical assistance center.

DIMMICK, W. W., K. L. FIORINO, AND B. M. BURR. 1996. Reevaluation of the Lythrurus ardens (Cypriniformes: Cyprinidae) complex with recognition of three evolutionary species. Copeia, 4:813-823.

EVANS, J. P., T. E. PITCHER, AND A. E. MAGURRAN. 2002. The ontogeny of courtship, colour, and sperm production in male guppies. J. Fish Biol., 60:495-498.

GODWIN, J., J. A. LUCKENBACH, AND R.J. Boom. 2003. Ecology meets endocrinology: environmental sex determination in fishes. Evol. Dev., 51:40-49.

GROBER, M. S. AND A. H. BASS. 2002. Life history, neuroendocrinology, and behavior in fish. Comp. Physiol., 17:331-47.

KODRIC-BROWN, A. 1998. Sexual dichromatism and temporary color changes in the reproduction of fishes. Am. Zool., 38:70-81.

KORTET, R., J. TASKINEN, A. VAINIKKA, AND H. YLONEN. 2004. Breeding tubercles, papillomatosis and dominance behaviour of male roach (Rutilus rutilus) during the spawning period. Ethology, 110:591-601.

LEMA, S. C. 2008. The phenotypic plasticity of Death Valley's pupfish. Am. Sci., 96:28-36.

LILJEDAL, S., I. FOLSTAD, AND F. SKARSTEIN. 1999. Secondary sex traits, parasites, immunity and ejaculate quality in the Arctic chair. P. R. Soc. B., 266:1893-1898.

NOBLE, G. K. 1938. Sexual selection among fishes. Biol. Rev., 13:133-158.

PARIKH, V. N., T. S. CLEMENT, AND R. D. FERNALD. 2006a. Androgen level and male social status in the African cichlid, Astatotilapia burtoni. Behav. Brain Res., 166:291-295.

--, --, AND R. D. FERNALD. 2006b. Physiological consequences of social descent: studies in Astatotilapia burtoni. J. Endocrinol., 190:183-90.

PITCHER, T. E. AND J. P. EVANS. 2001. Male phenotype and sperm number in the guppy (Poecilia reticulata). Can. J. Zoolog., 79:1891-1896.

--, F. H. ROOD, AND L. ROWE. 2007. Sexual colouration and sperm Waits in guppies. J. Fish Biol., 70:165-177.

REMAGE-HEALEY, L. AND A. H. BASS. 2006. A rapid neuromodulatory role for steroid hormones in the control of reproductive behavior. Brain Res., 1126:27-35.

RICE, W. R. 1989. Analyzing tables of statistical tests. Evolution, 43:223-225.

SARGENT, R. C., V. N. RUSH, B. D. WISENDEN, AND H. Y. YAN. 1998. Courtship and mate choice in fishes: integrating behavioral and sensory ecology. Am. Zool., 38:82-97.

SKARSTEIN, F. AND I. FOLSTAD. 1996. Sexual dichromatism and the immunocompetence handicap: an observational approach using Arctic charr. OIKOS, 76:359-367.

WARNER, R. R. AND S. E. SWEARER. 1991. Social Control of Sex Change in the Bluehead Wrasse, Thalassoma bifasciatum. Biol. Bull., 181:199-204.

YAMAGUCHI, S., K. GEN, K. OKUZAWA, N. KUMAKURA, M. MATSUYAMA, AND H. KAGAWA. 2005. Effect of 11 ketotestosterone and gonadotropin-releasing hormone on follicle-stimulating hormone and luteinizing hormone gene expression in castrated and sham-operated male red seabream Pagrus major. Fish. Sci., 71:1049-1058.

--, --, --, M. MATSUYAMA, and H. KAGAWA. 2006. Influence of estradiol-17[beta], testosterone, and 11-ketotestosterone on testicular development, serum steroid hormone, and gonadotropin in male red sea bream Pagrus major. Fish. Sci., 72:835-845.

ZAR, J. H. 1999. Biostatistical Analysis. 4th ed. Prentice-Hall, New Jersey. 929 p.

SUBMITTED 18 FEBRUARY 2011

ACCEPTED 22 NOVEMBER 2011

JENNIFER SCHADE AND BRUCE STALLSMITH (1)

Department of Biological Sciences, University of Alabama in Huntsville, Hun tsville 35899

(1) Corresponding author: e-mail: stallsb@uah.edu; Telephone: (256) 824-6992
TABLE 1.--Mean [+ or -] SE estimates and ranges of 11KT, body
measurements and color related variables. Mean [+ or -] SE of
untransformed data. For 11KT, D males, n = 7; ND males, n = 6;
Females, n = 5; for all other variables, D males, n = 8; ND
males, n = 7; Females, n = 16. SL, standard length (length
between anterior margin of mouth to posterior margin of caudal
peduncle)

Variable            Mean [+ or -] SE      Range

11KT (pg/ul)
  D males          39.3 [+ or -] 14.6   1.6-104.4
  ND males          1.9 [+ or -] 0.5     0.7-3.8
  Females           0.5 [+ or -] 0.2     0.2-1.1
GSI
  Dominant males    0.5 [+ or -] 0.1     0.3-0.8
  ND males          0.2 [+ or -] 0.1     0.1-0.5
  Females          12.4 [+ or -] 2.1     7.2-19.5
SL (mm)
  Dominant males   68.3 [+ or -] 0.8    64.8-70.9
  ND males         57.0 [+ or -] 2.0    49.7-62.3
  Females          55.4 [+ or -] 1.0    45.6-61.4
Weight (g)
  Dominant males    4.9 [+ or -] 0.2     4.2-5.9
  ND males          2.6 [+ or -] 0.3     1.4-3.6
  Females           2.3 [+ or -] 0.1     1.3-3.1
Red area (%)
  Dominant males   24.2 [+ or -] 0.9    20.4-28.1
  ND males         17.4 [+ or -] 0.7    15.0-19.5
  Females           9.2 [+ or -] 1.7     0.5-19.5
Hue (degrees)
  Dominant males   16.7 [+ or -] 0.6    13.8-18.5
  ND males         34.2 [+ or -] 3.2    25.3-46.5
  Females          64.3 [+ or -] 2.3    41.8-84.0
Saturation (%)
  Dominant males   91.8 [+ or -] 0.7    88.0-94.5
  ND males         61.2 [+ or -] 8.7    26.5-82.5
  Females          26.5 [+ or -] 1.8    15.8-42.0
Brightness (%)
  Dominant males   76.1 [+ or -] 1.2    69.3-80.3
  ND males         74.3 [+ or -] 1.3    70.0-77.3
  Females          70.7 [+ or -] 0.6    65.8-76.3

TABLE 2.--Mann-Whitney U statistics of body measurements and
color intensity properties between dominant males, non-dominant
males and females (df = 1). For 11KT, D males, n = 7; ND males, n
= 6; Females, n = 5; for all other variables, D males, n = 8; ND
males, n = 7; Females, n = 16. Asterisks (*) following P values
indicate a significant difference with [alpha] = 0.05 and using a
sequential Bonferroni correction to avoid Type 1 errors

                                      P-Values

Group tests          11KT         SL       Body weight     GSI

D--ND males         0.015      0.001 *       0.001 *     0.008
D males--Females    0.004 *    0.000         0.000 *        --
ND males--Females   0.018      0.256         0.181          --

                                      P-Values

Group Tests           Hue     Saturation   Brightness    Red Area

D--ND males         0.001 *    0.001 *       0.642       0.001 *
D males--Females    0.000 *    0.000 *       0.002 *     0.000 *
ND males--Females   0.000 *    0.001 *       0.030       0.009
Gale Copyright: Copyright 2012 Gale, Cengage Learning. All rights reserved.