Egg Diameter Variation in Eastern North American Minnows (Pisces: Cyprinidae): Correlation with Vertebral Number, Habitat and Spawning Behavior
Miles M. Coburn
1986-06-01
Abstract:Data on mean diameters of mature, unspawned ova were obtained for 12 genera and 71 species of eastern North American cyprinids. Sizes ranged from approximately 0.7 mm for several species of Notropis, Hybognathus hankinsoni, and Hybopsis aestivalis to 2.0 mm for Campostoma anomalum (Reed 1958). Mean diameter is strongly conserved in several taxa including different subgenera of Notropis. In Notropis (Luxilus), mean diameter ranged from 1.24-1.41 mm, in Notropis {Hydrophlox) from 1.13-1.18 mm, and in Notropis (Alburnops) from 0.71-0.93 mm. Notropis (ss) and Cyprinella were exceptions, with both having a range of mean egg diameters that varied more than 0.50 mm. Mean vertebral counts were tabulated for 61 species. Regression and correlation analyses were performed on 20 non-Notropis species, 41 species of Notropis, and both combined. In all cases the slope of the regression line differed from zero at P < 0.001. Correlation coefficients (r) for the relationships of vertebral number to egg size were 0.930, 0.715, and 0.810, respectively. In addition, upland and montane species have larger eggs than lowland species; territorial and/or nesting species have larger eggs than those lacking these behaviors. OHIO J. SCI. 86 (3): 110-120, 1986 INTRODUCTION Generally, it is well known that fish species with larger eggs tend to have more vertebrae (Lindsey and Ali 1971), spawn in cooler water (Ware 1977), and are more likely to build nests (Shine 1978) than related species with smaller eggs. Mean diameters of mature ova are reported regularly in life history studies of nearctic cyprinid fishes. They range from roughly 0.7 to 2.0 mm, or an increase in volume of 23X. Whereas egg size variation within a population is usually addressed, this obvious variation among many related species and its correlation with vertebral number, habitat, and spawning behavior have not been discussed adequately. Vertebral number and egg diameter have never been compared directly in a study of a nearctic minnow, but several investigators have compared female size with egg diameter. Mathur and Ramsey (1974), Heins and Bresnick (1975), Heins and Clemmer (1976), and Heins et al. (1980) reported no correlation between the two, whereas Hoyt (1971) found a positive correlation for Ericymba buccata. Although cyprinids exhibit pleomerism, or a positive correlation between maximum body size and vertebral number (Lindsey 1975), it is obvious that in life history studies of single populations size will be more a function of age than differences in vertebral number. Comparisons among isolated populations of a species, or among several related species are necessary to discern a relationship. Other than with female size, correlations are most often sought between egg diameter and environmental parameters, and are most common for commercially im1 Manuscript received 16 October 1985 and in revised form 25 March 1986 (#85-49). portant marine species. Bagenal (1971) summarized available literature on both marine fishes with pelagic and demersal eggs as well as freshwater fishes. He concluded that seasonal change in egg size within a species is a biological rather than physical phenomenon, and is not dependent on seasonal changes in temperature or salinity. The most important factor affecting egg size in marine fishes appears to be seasonal availability of food (Bagenal 1971, Ware 1975, 1977). Among freshwater species, for which data are sparse, Bagenal (1971) found some geographic (e.g., Oncorhynchus nerka, Cyprinus carpio) and seasonal (e.g., Salmo salar, Clupea harengus) variation, which he attributed to food availability. However, the importance of detritus and other foods of terrestrial origin obscured any connection between primary production in streams and egg size. In a recent study, Heins and Baker (1985) found a positive correlation (r = 0.825) between egg diameter and mean annual runoff in an analysis of 16 Gulf coast populations of Notropis venustus, ranging from eastern Texas to northwestern Florida. Drift and detritus are probably correlated with runoff. This may account for one environmental factor in egg diameter variation. The influence of environmental and genetic factors on egg size is unknown. The present study relies heavily on published records. It summarizes literature for 52 species (including 30 species of Notropis) in 12 genera of eastern cyprinids, presents new data for 19 additional species of Notropis, and duplicates measurements for another 20 species. Its purpose is to demonstrate the correlation of egg diameter with three parameters: mean vertebral count, habitat, and spawning behavior. The interplay among these factors is not generally appreciated in the literature on cyprinid systematics and life histories, but may be of some Ohio J. Science EGG DIAMETER VARIATIONS IN MINNOWS 111 importance in formulating ideas on life history strategies and evolutionary trends. METHODS AND MATERIALS SOURCES OF SPECIMENS FOR ANALYSES. Specimens used for egg diameter measurements were obtained from the Ohio State Museum of Zoology (OSUM) collections, and are listed in Appendix I. Specimens used for vertebral counts were obtained predominantly from OSUM; some were also received from the University of Michigan Museum of Zoology (UMMZ) and Cornell University (CU). A list of these can be obtained on request. In cases where both were published, mean vertebral counts for the entire range of a species were used rather than counts for a local area. No attempt was made to match samples used for egg and vertebral studies. RATIONALE FOR ESTIMATES OF MATURE EGG DIAMETER. Ripe females were opened, and mature, deep yellow eggs were removed from the posterior ovary. Ten eggs per female were measured with an ocular micrometer mounted on a Wild dissecting scope. Measurements on eggs from at least three females per species were desired, but not always obtained. Total egg counts were not made. In most cases the standard deviations (SD) of egg diameter measurements were well under 0.10; the 95% confidence intervals (CI) were in the range of ± 0.02-0.03mm. The largest variation occurred in Notropis rubellm (SD = 0.130, 95% CI — 1.08-1.14 mm) and N. coccogenis (SD = 0 .132, 95% CI = 1.26-1.36 mm). A mean value for mature egg diameter was necessary for the regression and correlation analyses and for analysis of variance (ANOVA). This was obtained by averaging all literature values and direct measurements for each species, regardless of the number of females in each sample. For species in which ranges, but not means, were published, the average of the range was used. Six literature citations in Table I report a wide range of mature egg diameters for a single species. Each of these deserves comment. Mathur and Ramsey (1974) classified mature ova for Notropis baikyi in two ways. To estimate the number of clutches per spawning season, mature ova were given a range of 0.65 to 1.66 mm, with few larger than 1.40 mm. The second method involved the measurement of 10 to 20 of the largest ova in each ovary. This yielded a mean diameter that peaked in May at 1.15 mm. I used the latter for the average mean diameter for N. baikyi. Becker (1983) similarly estimated the diameters of mature ova from N. anogenus (0.7-1.3 mm for a single female) and N. heterodon (0.8-1.2 mm for three females). For this study, I took the midpoint of each range as an estimate of the mature mean diameter for N. anogenus and N. heterodon. There are other data available for both species, however. Starnes and Starnes (1981) reported an egg diameter range of 0.82 to 1.36 mm for Phoxinus cumberlandensis. They removed individual eggs from the entire ovary and estimated mature ova by color. Although this is a broad interpretation of what constitutes mature ova, I took the value of 1.09 mm as an estimate of the mean diameter of mature ova ready to be spawned. Jenkins and Burkhead (1984) reported a range of 0.8 to 1.4 mm for the eggs of Hybopsis monacha. Jenkins (pers. comm.) suggested that some females may have been in a rest interval between spawns, in accordance with their fractional spawning behavior, and perhaps lacked fully developed ova. The value of 1.10 mm is probably a low estimate for the mean diameter of mature ova for H. monacha, as is the average value for P. cumberlandensis. Finally, Brown and Hammer (1970) reported a range of 1.29 to 1.67 mm in the diameters of mature ova from Couesius plumbeus. Their data show a skewed distribution for May, 1967, with a mean value of 1.48 mm. This value was used in the present study. With three exceptions, values in Table I were taken from preserved specimens. For N. rubellus and Campostoma anomalum, Reed (1958) placed eggs from fresh specimens in water prior to measurement. The TABLE 1 Egg diameters, vertebral counts, habitat, and spawning behavior for some eastern North American cyprinids