2007 SOUTHEASTERN NATURALIST 6(4):693–704 Observations on the Reproductive Biology of Two Notropis Shiner Species Bruce Stallsmith1,*, Kevin Butler1,2, Amy Woodall1,2, and Bob Muller3 Abstract - Important details of the reproductive ecology of many freshwater fishes of the species-rich southeastern USA are still poorly known. Two such species are Notropis asperifrons (Burrhead Shiner), and Notropis stilbius (Silverstripe Shiner). Both are endemic to the Mobile Basin, AL. To determine timing and patterns of reproductive effort, collections were made of as many as 20 individuals of each species at roughly four-week intervals from March through September. Female gonadal somatic index (GSI) data averages for each collection indicate that both species peak in reproductive activity in April. Both species still had reproductively competent female GSI values as late as July 31. Examination of ovarian tissue indicates that Burrhead Shiner oocytes are usually larger at equivalent developmental stages. Burrhead Shiners exhibit sexual size dimorphism with larger females, while Silverstripe Shiners do not. Introduction Important details of the reproductive ecology of many freshwater fishes of the species-rich southeastern USA are still poorly known. Many of these species are under growing pressure from anthropogenic activities such as urban expansion, agricultural uses, and waterway modification (Butler 2002, Powell 2003, Scott 2006). Proper management and protection of existing aquatic ecosystems requires the fullest possible knowledge of the life history of resident species. Two such species are the cyprinids Notropis asperifrons Suttkus and Raney (Burrhead Shiner) and N. stilbius Jordan (Silverstripe Shiner). Both are endemic to the Mobile Basin of Alabama, Mississippi, and Georgia, and are typically found in the riffles and pools of upland streams. Both species are often locally abundant, but little is known of the details of their reproduction. Boschung and Mayden (2004) state that the Silverstripe Shiner spawns from mid-April to mid-June. Burrhead Shiner, based on observations of tuberculate males, spawns from April to June. No information exists on whether these two generally similar species, often found in the same creeks, have different reproductive strategies such as producing fewer, larger eggs, or temporal spawning differences. A species’ schedule of reproduction (sensu Heins and Rabito 1986) is a key aspect of its life history and ecological relationships. Females in the genus Notropis are known to be multiple 1Department of Biological Sciences, University of Alabama in Huntsville, Huntsville, AL 35899. 2Current address - Open Biosystems, Huntsville, AL 35806. 3Royal Oak Nature Society, Royal Oak, MI 48067. *Corresponding author - fundulus@ hotmail.com. 694 Southeastern Naturalist Vol. 6, No. 4 spawners, spawning more eggs in a season than are found in the ovary at any one time (Dahle 2001, Heins and Clemmer 1976). The objective of this study is to assess seasonal variation in the reproductive competence of gonadal tissues for sympatric populations of Burrhead Shiners and Silverstripe Shiners. Materials and Methods Collection site and field collection Burrhead Shiners and Silverstripe Shiners were collected monthly from March 2004 to October 2004, and March 2005. Two collections were made in July, early and late, because the planned collection in June was cancelled due to heavy rains. A collection was made in March 2005 because few fish were collected in March 2004 due to high water. The collection site was at the Borden Creek Trailhead (34º18.569'N, 87°23.670'W; elevation 200 m) inside the Sipsey Wilderness Unit of the Bankhead National Forest in Lawrence County, AL. Borden Creek is a second-order tributary to Sipsey Fork of the Warrior River. Twenty individuals of each species were taken in each collection using seine nets (3.0-m L x 1.3-m D; 3.0-mm mesh). No effort was made to collect individuals of specific sex or size. Collected fishes were euthanized in a solution of MS-222 (tricaine methanesulfonate) and later transferred to 5% phosphate buffered formalin for transport and storage prior to histological examination. Water temperature (ºC), total dissolved solids (TDS, parts per million), and pH were recorded on each visit to Borden Creek. Temperature was measured by placing an alcohol thermometer on the streambed in a shady area of steady flow. TDS was measured using a Hanna Instruments TDS 1 meter. Water pH was measured using a Hanna Instruments pHep 2 meter. Length and mass data collection Standard length (0.01 mm) was recorded for each fish with digital calipers. Individuals were then weighed (0.01 g) after the removal of excess surface fluid by wrapping in a paper towel. Gonads (if present) and the digestive tracts were removed, and gonad mass was recorded to the nearest 0.01 g. The gonadal somatic index (GSI) was calculated as: (gonad mass/somatic mass) x 100. A Mettler H18 balance was used for the collection of all mass data. Sexual size dimorphism was evaluated by comparing mean standard length between all collected males and females for each species. Equality of variances was evaluated with a F-test and mean difference between samples was assessed with a 2-tailed t-test (Sokal and Rohlf 1973). Histological preparation Ovarian tissue from up to four females from each monthly collection, March through late July, were prepared histologically. Samples for 2007 B. Stallsmith, K. Butler, A. Woodall, and B. Muller 695 histological examination were fixed in Bouin’s fixative for 12 hours, then transferred through a series of ethanol dehydrations. Following dehydration, the samples were cleared in toluene and xylene. Samples were then embedded in paraffin by soaking in a paraffin bath overnight and allowed to cool. The embedded samples were faced off and then soaked in a solution of Downy® fabric softener and water (≈200 ml deionized H2O : ≈2 drops of Downy®) for 15 minutes before sectioning. The samples were sectioned at 4 μm, mounted on glass microscope slides, and stained with Gill-Modified Hematoxylin and Eosin Y. Examination and measurement of oocytes and ova To visually determine the status of oocyte development, stained slides of ovarian tissue were examined with a Wolfe Digivu TM CVM digital compound microscope using the software package Motic Images 2000 version 1.2. Four random digital images at 40X magnification were captured for each fish from among slides of an individual’s fixed tissue (Maddock and Burton 1999). Each developing ooctye fully visible in each of these images was characterized to stage following the methods of Rinchard and Kestemont (1996) and Maddock and Burton (1999). These four stages are: perinucleolar (visible nucleus but no vacuoles in the cytoplasm), cortical alveolar (appearance of yolk vesicles forming rings in the cytoplasm periphery), early exogenous vitellogenesis (oocyte full of yolk vesicles and the differentiation of cellular and follicular layers), and the nearly mature stage of late exogenous vitellogenesis (accumulation of yolk globules at the periphery of the cytoplasm). Mature oocytes were characterized by the appearance of the micropyle and the migration of the germinal vesicle to the micropyle. The area (A) of each fully visible oocyte in a captured image was calculated in square micrometers by tracing the outer edge of the on-screen image using the computer mouse. Diameter of each oocyte was then calculated as: Diameter = 2(√A/π). Fully mature ova released from the ovarian lumen by the visceral dissection of randomly selected females of both species from April, May, and July 2004 were also examined with the above microscope and software combination and measured for diameter. Aquarium spawning of Burrhead Shiner A group of four Burrhead Shiners, sex ratio unknown, was kept in an allglass 37-L aquarium. To simulate natural conditions, the aquarium was kept at about 17 ºC starting in January 2004, and allowed to rise to about 22 ºC by May 2004. The aquarium was kept on an elongated rack to allow the bottom of the aquarium to be viewed from underneath. The bottom of the aquarium was covered by a piece of plastic egg-crate light diffuser from a fluorescent light fixture and covered with netting. Stones (15–30 mm) were placed on 696 Southeastern Naturalist Vol. 6, No. 4 top of the diffuser to serve as spawning medium. This allowed the tank to be examined from the underside to see if any eggs had been deposited in the gravel and fallen through the egg crate diffuser. The bottom of the aquarium was examined for the presence of eggs on a near daily basis. A pipette was used to remove any eggs from the aquarium. Six eggs had their diameter measured microscopically, and all were removed to another aquarium for observation to determine incubation time. Results Water quality of Borden Creek Water temperature varied from a low of 12 ºC in March to a high of 23 ºC in late July. Total dissolved solids ranged from 76 ppm in March during a period of high water to 133 ppm in September during low flow. The pH varied from 8.2 to 8.5. Oocyte development and size Measurements of the different stages of oocyte development for both species are summarized in Table 1. Mature oocytes at the stage of late exogenous vitellogenesis were found in females of both species from April through late July, as well as oocytes at all three of the earlier stages of development. In March, females of both species carried developing oocytes, but none were beyond the stage of early exogenous vitellogenesis. The monthly proportions of oocyte developmental stages for Burrhead Shiner and the Silverstripe Shiner are shown in Fig. 1. In both species, over 80% of March oocytes were in early stages of development, more than 50% of May oocytes were in the advanced maturation stages of vitellogenesis, and roughly 20% of the late July oocytes were in vitellogenesis. The major difference between the species was that Burrhead Shiner ooctyes were more mature in April with over 50% in vitellogenesis compared to roughly 25% for Silverstripe Shiner. Gonadal tissues from females collected in August was sufficient to weigh, but too small for successful histological preparation. The examination of fully mature but unhydrated ova found during dissection of Silverstripe Shiners from May and early July yielded a mean diameter of 1.4 mm, with a range of 1.2–1.6 mm. Fully mature ova of Burrhead Shiner from the same time periods had a mean diameter of 1.8 mm, with a range of 1.7–2.0 mm. Female GSI and length Reproductively active female Burrhead Shiners were present in Borden Creek from April through late July, as indicated by GSI values and the presence of oocytes in the advanced developmental stage of late exogenous vitellogenesis (Fig. 2, Table 1). Peak reproductive condition was in April 2007 B. Stallsmith, K. Butler, A. Woodall, and B. Muller 697 with a mean GSI of 16.3. At least some females showed evidence of being reproductively active in late July, with a mean GSI value of 5.6 and the continued presence of oocytes in the developmental stage of late exogenous Table 1. Oocyte size at different developmental stages during the spawning season. Mean diameter in μm is reported monthly for each developmental stage for each species. Mean n diameter Standard Range (fish/ Oocyte stage Month Species (μm) error (μm) oocytes) Perinucleolar March Burrhead Shiner 69.3 0.81 39–99 1/289 Silverstripe Shiner 68.3 1.25 37–99 1/121 April Burrhead Shiner 66.2 1.26 33–105 3/161 Silverstripe Shiner 70.1 1.04 30–97 2/173 May Burrhead Shiner 65.1 1.11 34–129 3/185 Silverstripe Shiner 75.2 1.57 42–131 3/119 Early July Burrhead Shiner 64.3 1.59 34–94 1/77 Silverstripe Shiner 68.0 0.99 33–100 2/231 Late July Burrhead Shiner 61.7 1.26 32–91 1/133 Silverstripe Shiner 62.6 0.72 32–90 2/388 Cortical alveolar March Burrhead Shiner 144.7 2.04 90–237 1/247 Silverstripe Shiner 123.3 1.67 90–183 1/142 April Burrhead Shiner 142.5 2.80 90–204 3/85 Silverstripe Shiner 153.9 2.80 88–240 2/149 May Burrhead Shiner 138.0 4.21 86–213 3/58 Silverstripe Shiner 121.2 1.96 92–219 3/83 Early July Burrhead Shiner 167.5 8.02 121–221 1/17 Silverstripe Shiner 147.8 3.45 89–227 2/102 Late July Burrhead Shiner 120.9 4.02 90–211 1/43 Silverstripe Shiner 119.5 1.71 82–201 2/201 Early exogenous vitellogenesis March Burrhead Shiner 256.3 5.66 159–341 1/44 Silverstripe Shiner 224.5 5.28 137–317 1/62 April Burrhead Shiner 265.8 4.24 153–410 3/180 Silverstripe Shiner 285.8 5.08 206–382 2/59 May Burrhead Shiner 275.5 4.47 133–439 3/188 Silverstripe Shiner 252.9 3.70 154–355 3/142 Early July Burrhead Shiner 322.2 14.74 216–450 1/24 Silverstripe Shiner 284.2 7.76 198–401 2/51 Late July Burrhead Shiner 270.0 11.94 181–390 1/24 Silverstripe Shiner 251.9 5.54 124–348 2/78 Late exogenous vitellogenesis March Burrhead Shiner none n/a n/a n/a Silverstripe Shiner none n/a n/a n/a April Burrhead Shiner 454.4 7.88 298–779 3/116 Silverstripe Shiner 466.5 8.84 340–636 2/60 May Burrhead Shiner 537.9 10.54 340–762 3/109 Silverstripe Shiner 423.3 7.98 305–564 3/83 Early July Burrhead Shiner 566.1 16.60 316–813 1/46 Silverstripe Shiner 459.8 6.41 283–596 2/97 Late July Burrhead Shiner 522.7 17.41 366–677 1/24 Silverstripe Shiner 436.9 7.50 249–565 2/67 698 Southeastern Naturalist Vol. 6, No. 4 vitellogenesis. By August, GSI values were lower than in March. Sexually mature females varied in size between 37.8–59.3 mm. Female Silverstripe Shiners show evidence of reproductive activity from April to late July, with a May peak mean GSI of 12.2 and the presence of oocytes in the developmental stage of late exogenous vitellogenesis (Fig.2, Table 1). Some individuals seemed to be reproductively active into late July with a mean GSI of 4.8. As with Burrhead Shiners, the mean August GSI value was lower than that for March. Sexually mature females varied in size from 41.7 to 70.1 mm. Individuals of both species collected in September had such reduced gonads that we were unable to determine gonadal mass to report GSI values for them. Male GSI and length Male mean GSI values for both species started relatively high in March at 0.8 (Fig. 3). Males of both species had variably high mean GSI values from April through early July, with a sharp drop in late July and very low Figure 1. Mean proportion of maturing oocyte stages observed in Notropis asperifrons (Burrhead Shiner) and N. stilbius (Silverstripe Shiner) over a breeding season. Different oocyte developmental stages are represented in the graphs as follows: PN = perinucleolar, CA = cortical alveolar, EEV = early exogenous vitellogenesis, LEV= late exogenous vitellogenesis. 2007 B. Stallsmith, K. Butler, A. Woodall, and B. Muller 699 mean GSI values in August. Reproductively active male Burrhead Shiners ranged in size between 28.1–49.7 mm, and male Silverstripe Shiners ranged between 38.2–64.5 mm. As with females, males collected in September had such reduced gonadal tissue that it was not possible to determine gonadal mass for the reporting of GSI values. Sexual size dimorphism For each species, average standard length of all collected males was compared to average standard length of all collected females. In both species, females were longer than males. For Burrhead Shiners, mean standard length of males was 40.8 mm (n = 29, standard error = 0.96) and mean standard length of females was 46.2 mm (n = 42, standard error = 0.81). For Silverstripe Shiners, mean standard length of males was 55.2 mm (n = 70, standard error = 0.76) and mean standard length of females was 56.7 mm (n = 63, standard error = 0.82). A two-tailed t-test comparing standard length of males to females for each species showed a significant difference in Burrhead Shiners (p < 0.001) but not in Silverstripe Shiners (p = 0.19). Figure 2. Mean monthly GSI values for female Notropis asperifrons (Burrhead Shiner) and N. stilbius (Silverstripe Shiner) from Borden Creek, Bankhead National Forest, Lawrence County, AL. Error bar indicates one standard error. Sample size is indicated by a number above each bar. 700 Southeastern Naturalist Vol. 6, No. 4 Aquarium observations of spawning by Burrhead Shiners All four adults in the spawning group were observed to spend much of their time buried in the gravel on the bottom of the aquarium. No color differences were discernible between the individuals, so the sex ratio was not determined. Individuals were not observed in the act of spawning. Freshly spawned eggs were clear, 2 mm in diameter, and non-adhesive. Eggs removed to another aquarium hatched in 82 hours at 22 ºC. Fry were observed to remain in the bottom 50 mm of the aquarium. At 60 days, the fry were about 15 mm long. Discussion Both Burrhead Shiners and Silverstripe Shiners show evidence of being reproductively active from April into July based on GSI values for both sexes and microscopic examination of ovarian tissue. Reproductive competence into July is later than what has been reported for these two species. Other Notropis species from both warmer and colder climates have been observed to have a spawning season of three or more months. Notropis Figure 3. Mean monthly GSI values for male Notropis asperifrons (Burrhead Shiner) and N. stilbius (Silverstripe Shiner) from Borden Creek, Bankhead National Forest, Lawrence County, AL. Error bar indicates one standard error. Sample size is indicated by a number above each bar. 2007 B. Stallsmith, K. Butler, A. Woodall, and B. Muller 701 longirostris Hay (Longnose Shiner) from warmer south Mississippi were in breeding condition from late March until October as indicated by GSI of females and males, and male tuberculation and coloration (Heins and Clemmer 1976). In colder Minnesota, Notropis topeka Gilbert (Topeka Shiner) individuals were observed to be in reproductive condition from mid-May to early August based on gonadal development, GSI, and field observations (Dahle 2001). These differences in breeding-season length seem to be correlated with variation in local climates, especially the onset of milder weather in the spring. The fact that at least some individuals of both species have elevated GSI and well-developed gonadal tissues including maturing oocytes of all stages in July is also consistent with multiple spawnings over a reproductive season. This asynchronous ovarian development is a trait observed in many other cyprinids (Heins and Rabito 1986, Roberts and Grossman 2001). Many other North American cyprinids show evidence of producing 6 or more clutches in a continuous spawning season, including Longnose Shiner (Heins and Clemmer 1976), Cyprinella leedsi Fowler (Bannerfin Shiner; Heins and Rabito 1986), Rhinichthys cataractae Valenciennes (Longnose Dace; Roberts and Grossman 2001) and Topeka Shiner (Dahle 2001). The overall pattern of the timing of reproduction as indicated by mean GSI values is similar between the two species, but there appears to be differences of timing and investment in the quality and quantity of eggs between the species within the shared spawning season. Silverstripe Shiner is the larger of the two species, but produces smaller mature ova on average (1.4 mm) than Burrhead Shiner (mean of 1.8 mm). Our data on mean oocyte size at different development stages reveal two tendencies. First, Silverstripe Shiners produce on average slightly larger perinucleolar oocytes, the first stage in maturation. Secondly, Burrhead Shiner produces larger oocytes at later developmental stages, except in the peak GSI month of April. The mean size of Burrhead Shiner late-exogenous vitellogenesis oocytes increases by ≈100 μm for the three months after April, while those of Silverstripe Shiner slightly decrease in size. Figure 1 show that Burrhead Shiner ovaries contain a higher fraction of oocytes in advanced stages of development than those of Silverstripe Shiner. Our interpretation is that Burrhead Shiner females produce larger eggs more evenly over the reproductive season than Silverstripe Shiner females, which show evidence of peak egg production in May. Previous studies have found no inverse relationship between clutch size and egg size in North American cypriniformes (Winemiller and Rose 1992). These authors describe Notropis species as showing an opportunistic reproductive strategy as small fish with seasonal spawning, moderately large clutches, small eggs, and relatively few spawning bouts per year. This description fits both Burrhead Shiner and Silverstripe Shin702 Southeastern Naturalist Vol. 6, No. 4 er. Data presented in this paper illustrate potential differences between the two species’ reproductive biology that may reflect trade-offs with other life-history parameters. During seining to collect these fish in Borden Creek, we collected many more Silverstripe Shiners than Burrhead Shiners, often by a ratio of ten to one. The Silverstripe Shiner was by far the most abundant species at the research site. One possible explanation for this difference in abundance may be that Silverstripe Shiner females produce a larger number of smaller eggs than Burrhead Shiner females, and this larger number of larvae can better use available microhabitat(s). In addition, we observed habitat partitioning between adult Burrhead Shiners and Silverstripe Shiners in the creek. While snorkeling in Borden Creek, we saw Burrhead Shiners in small groups near the substrate, hiding under rocks or burrowing in sand, while large schools of Silverstripe Shiners occupied the upper water column. In our aquarium, we provided few options for spawning habitat and did not observe any spawning activity of Burrhad Shiner. However, spawning in Burrhead Shiner is likely more closely associated with the substrate rather than the water’s surface given that we noted a greater propensity for adults to be buried in the aquarium substrate, and that spawned eggs were retrieved from the bottom of the aquarium below the substrate. There are no published spawning observations of Silverstripe Shiners, but based on the lack of sexual size dimorphism, it is likely that the reproductive mode is egg scattering during group spawning as has been observed with another member of the subgenus Notropis, N. atherinoides Rafinesque (Emerald Shiner) (summarized in Boschung and Mayden 2004). From our GSI data, neither Burrhead Shiners nor Silverstripe Shiners have testes that are unusually enlarged, an expectation if a species’ reproductive mode involves sperm competition (Pyron 2000). This is consistent with observations of other North American minnows (Pyron 2000). Differences between reproductive mode in the observed two species seem to be differences between female reproductive effort. Other cyprinid minnows present in Borden Creek have more elaborate spawning strategies, such as crevice spawning by Cyprinella species (Heins 1990, Rabito and Heins 1985) and the construction and guarding of stone nests by dominant male Campostoma oligolepis Hubbs and Greene (Largescale Stoneroller). Neither the Burrhead Shiner nor Silverstripe Shiner is in immediate danger of extinction. The key to management for species such as these two with an opportunistic reproductive strategy is to protect their habitat from largescale or chronic disturbances that would eliminate all refugia (Winemiller and Rose 1992). A better understanding of what aspects of these species’ life-history strategies have made them tolerant to human disturbance to date can help our efforts to understand why other, often closely related, species are in decline. 2007 B. Stallsmith, K. Butler, A. Woodall, and B. Muller 703 Acknowledgments We would like to thank Ruth Fledermaus, Bill Garstka, Dewey Mason, Amy Bishop, and members of NANFA for help with the field collection and laboratory preparation necessary to this project. Jim Daniels and two anonymous reviewers offered suggestions that much improved this manuscript. Fishes were collected under Special Use Permit BAN700114 from the US Department of Agriculture, US Forest Service. Literature Cited Boschung, H.T., Jr., and R.L. Mayden. 2004. Fishes of Alabama. Smithsonian Books, Washington, DC. 736 pp. Butler, R.S. 2002. Imperiled fishes of the lower Tennessee Cumberland ecosystem,with emphasis on the non-federally listed fauna. Prepared for the Lower Tennessee Cumberland Ecosystem Team, US Fish and Wildlife Service, Asheville, NC. 39 pp. Dahle, S.P. 2001. Studies of Topeka Shiner (Notropis topeka) life history and distribution in Minnesota. M.Sc. Thesis. University of Minnesota, Minneapolis, MN. 67 pp. Heins, D.C. 1990. Mating behaviors of the Blacktail Shiner, Cyprinella venusta, from southeastern Mississippi. Southeastern Fishes Council Proceedings 21:5–7. Heins, D.C., and G.H. Clemmer. 1976. The reproductive biology, age, and growth of the North American cyprinid, Notropis longirostris (Hay). Journal of Fish Biology 8:365–379. Heins, D.C., and F.G. Rabito, Jr. 1986. Spawning performance in North American minnows: Direct evidence of the occurrence of multiple clutches in the genus Notropis. Journal of Fish Biology 28:343–357. Maddock, D.M., and M.P.M. Burton. 1999. Gross and histological observation of ovarian development and related condition changes in American plaice. Journal of Fish Biology 53:928–944. Powell, J.R. 2003. Response of fish communities to cropland density and natural environmental setting in the Eastern Highland Rim Ecoregion of the Lower Tennessee River Basin, Alabama and Tennessee, 1999. Water-Resources Investigations Report 02-4268, National Water-Quality Assessment Program, US Geological Survey. 48 pp. Pyron, M. 2000. Testes mass and reproductive mode in minnows. Behavioral Ecology and Sociobiology 48:132–136. Rabito, F.G., Jr., and D.C. Heins. 1985. Spawning behaviour and sexual dimorphism in the North American cyprinid fish Notropis leedsi, the Bannerfin Shiner. Journal of Natural History 19:1155–1163. Rinchard, J., and P. Kestemont. 1996. Comparative study of reproductive biology in single- and multiple-spawner cyprinid fish. I. Morphological and histological features. Journal of Fish Biology 49:883–894. Roberts, J.H., and G.D. Grossman. 2001. Reproductive characteristics of female Longnose Dace in the Coweeta Creek drainage, North Carolina, USA. Ecology of Freshwater Fish 10:184–190. 704 Southeastern Naturalist Vol. 6, No. 4 Scott, M.C. 2006. Winners and losers among stream fishes in relation to land use legacies and urban development in the southeastern US. Biological Conservation 127:301–309. Sokal, R.R., and F.J. Rohlf. 1973. Introduction to Biostatistics. W.H. Freeman and Company, San Francisco, CA. 368 pp. Warren, M.L., Jr., B.M. Burr, S.J. Walsh, H.L. Bart, Jr., R.C. Cashner, D.A. Etnier, B.J. Freeman, B.R. Kuhajda, R.L. Mayden, H.W. Robison, S.T. Ross, and W.C. Starnes. 2000. Diversity, distribution, and conservation status of the native freshwater fishes of the Southeastern United States. Fisheries 25(10):7–29. Winemiller, K.O., and K.A. Rose. 1992. Patterns of life-history diversification in North American fishes: Implications for population regulation. Canadian Journal of Fisheries and Aquatic Sciences 49:2196–2218.