Are echinoderm egg size distributions bimodal?
In early mathematical modeling Vance (2, 3) viewed planktotrophy and lecithotrophy as extreme forms of larval development, and his models suggested that only the extreme egg sizes would confer maximum reproductive efficiency (number of newly metamorphosed adults per unit of energy devoted to reproduction). Subsequently, the prediction of bimodality in egg sizes within taxa of marine invertebrates was tested by a number of authors using previously published or new data sets. Most of these studies showed unimodal egg size distributions (see also fig. 2 in ref. 22). Examples include prosobranch gastropods (9, 10), opisthobranchs (11), muricid gastropods without nurse eggs (12), Indo-Pacific Conus (13), bivalves (8), chitons (23), thoracican barnacles (24), and stomatopods (25). Bimodal distributions in egg size were described in American species of the genus Crepidula (4), in prosobranch gastropods from Danish waters (4), and in echinoid and asteroid echinoderms (14).
Many of the early studies used data sets that were not adequate for a valid test of the distribution patterns of egg size. We defined three criteria for appropriate data sets: (a) the data should include both planktotrophic (feeding) and lecithotrophic (nonfeeding) forms of development; (b) the egg sizes should be from species that span a broad range of habitats and a wide geographic range; and (c) the data should exclude species that provide additional maternal investment (e.g., nurse eggs, capsules). Criterion (b) is necessary because egg size may be correlated with habitat - for example, more planktotrophs in the tropics (1) - and use of species from a limited geographic area might under- or overemphasize a particular developmental type. Criterion (c) is necessary because Vance's models (2, 3) assume that reproductive energy is invested solely in the egg. Additionally, because the models (2, 3) were developed in terms of energetic content rather than egg size per se, for all data sets it is assumed that there is a correlation between egg size and energetic content.
The primary criterion, that planktotrophic and lecithotrophic forms be present, is violated by three of the taxa: thoracican barnacles (24) and stomatopods (25), neither of which groups include species with lecithotrophic development, and chitons (23), which do not include species with planktotrophic development. Criterion (b) may be violated by Ockelmann's bivalve data set (8) because planktotrophs are rare in East Greenland (26), in compilations of egg sizes in American species of Crepidula (4) and Danish prosobranchs (4), and in studies with low sample sizes (4, 23, 24, 25). Violation of criterion (c) has been avoided by removal of species using nurse eggs (12, fig. 2 in ref. 22). With the removal of studies that clearly violate criterion (a) there are unimodal egg size distributions in six taxa (8-13) and bimodal distributions in four taxa (4, 14).
The best known, and most widely cited, example of the bimodality in egg sizes within a taxon is the extensive data set compiled by Emlet et al. (14) for asteroid and echinoid echinoderms. This data set meets criteria (a)(c) above, incorporating egg sizes from 89 asteroid species and 109 echinoid species from a variety of habitats. Moreover, in this phylum there is a strong correlation between egg size and energetic content (27). Of the four remaining classes of echinoderms, our criteria are not met in the crinoids, where there are no known planktotrophic species (28), or in the concentricycloids, where there are only two described species (29). In this paper we test the prediction of bimodality in egg sizes in the two remaining classes, the holothuroids and ophiuroids, which meet the criteria defined above, and we incorporate new data into the published egg size distributions of echinoids and asteroids (14).
The egg size data sets used in this analysis were compiled from the published literature and our own unpublished data, with the implicit assumption that all echinoderm eggs are basically alike. As noted by Turner and Lawrence (30), although egg diameter is an easy reproductive character to measure, there are shortcomings in its use for theoretical discussions of life-history patterns. Firstly, many species have oblately or prolately spheroidal eggs, so that eggs with the same diameter may have different egg volumes (30). We did not attempt to convert published egg diameters to volumes because in most cases only one of the two linear dimensions required for such a calculation was presented. Secondly, egg measurements made from histological sections will underestimate egg diameter because of tissue shrinkage during processing and the hydration of the egg during spawning (30). We used the diameters of both intraovarian and freshly spawned eggs in this paper to increase the sample sizes for statistical analysis (n [greater than] 100 species for all classes). Because the size classes in the raw (100 [[micro]meter]) and ln-transformed analyses (ln 0.25) are relatively large, we assumed that intraovarian egg diameters would have a low probability of moving into an adjacent size class should a freshly spawned egg diameter be available. Thirdly, eggs differ in their organic components - lipid, protein, etc - (27, 30) and our assumption that egg diameter is an effective measure of maternal investment may prove to be untenable.
Previous studies of egg size distributions in marine invertebrates have determined the number of modes in the distribution by visual appraisal. In this paper we instead made use of statistical techniques designed to differentiate length-frequency classes of fish populations (31). Histograms of egg diameter frequencies were generated for each class using the raw and In-transformed diameters as shown in Figures 1-4. A modal analysis using the method of Schnute and Fournier (31) was then used to identify the Gaussian modes presented in these figures. Because the analysis is relatively insensitive to the dimensions of the size classes (31; B. Smith, pers. comm.), this method provides an objective way to examine the egg size distribution in a particular taxon and directly test the hypothesis of bimodality of egg sizes in marine invertebrates (2, 3). The complete raw data sets used for these analyses are available to interested readers from either of the authors.
Holothuroids show an extreme range in egg diameter, from a minimum of 50 [[micro]meter] in the apodid Synaptula reciprocans to a maximum of 4400 [[micro]meter] in the elasipod Psychropotes longicauda [ILLUSTRATION FOR FIGURE 1A OMITTED]. About 80% of the species have eggs smaller than 1000 [[micro]meter] in diameter [ILLUSTRATION FOR FIGURE 1A OMITTED], with four modes in the size-frequency distribution (Table I). The first mode is at an egg diameter of 194 [[micro]meter] ([ILLUSTRATION FOR FIGURE 1A OMITTED], Table I), and comprises species with eggs from 50 to about 300 [[micro]meter] with planktotrophic or unknown development [ILLUSTRATION FOR FIGURE 1A OMITTED]. A second, but overlapping, mode is seen at 421 [[micro]meter], with species with lecithotrophic, brooded, or unknown development (lecithotrophs: 150-950 [[micro]meter]; [ILLUSTRATION FOR FIGURE 1A OMITTED]). A large number of species in the unknown category in this mode are dendrochirote holothurians with egg sizes in the 200-800 [[micro]meter] range. As there are no known planktotrophs in this order (32), this mode may additionally include 20 lecithotrophic species. The last two modes, at 1024 and 2603 [[micro]meter], are primarily composed of species that are lecithotrophic, brood, or have unknown development type [ILLUSTRATION FOR FIGURE 1A OMITTED] - the majority of the latter species are elasipod holothurians and may have direct development (33). Thus, in the holothurians the first mode contains mainly species with planktotrophic development; the following three modes are species with nonfeeding [TABULAR DATA FOR TABLE I OMITTED] (planktonic or brooded lecithotrophs) or unknown development.
The natural log-transformed distribution appears visually to be unimodal with a slight right skew [ILLUSTRATION FOR FIGURE 1B OMITTED]. Statistical analysis, however, reveals two modes with means of In 5.62 and 7.08 respectively (Table I). There is no clear demarcation between planktotrophic, lecithotrophic, and brooding species, with representatives of all three developmental types in the first mode [ILLUSTRATION FOR FIGURE 1B OMITTED]. The second mode, however, is almost exclusively composed of species with lecithotrophic or brooded development.
Examination of the egg size distribution in the ophiuroids reveals a similar distribution in the untransformed egg sizes [ILLUSTRATION FOR FIGURE 2A OMITTED]. Ophiuroid eggs are smaller than holothuroid eggs, ranging in diameter from 50 [[micro]meter] in Ophiactis kroeyeri to 1500 [[micro]meter] in Astrospartus mediterraneus [ILLUSTRATION FOR FIGURE 2A OMITTED]. Ninety-five percent of the species produce eggs smaller than 800 [[micro]meter], and only two species have egg sizes greater than 1000 [[micro]meter] [ILLUSTRATION FOR FIGURE 2A OMITTED]. There are three statistical modes in the egg size distribution: the first, at 165 [[micro]meter], of species with primarily planktotrophic (80-185 [[micro]meter]) development; the second, at 450 [[micro]meter], of species with abbreviated (150-350 [[micro]meter]), brooded (100-1000 [[micro]meter]), or unknown development; and the third, at 1528 [[micro]meter], which results from a single species (Astrospartus mediterraneus) with unknown development ([ILLUSTRATION FOR FIGURE 2A OMITTED], Table I).
The In-transformed egg size distribution in the ophiuroids appears visually to be unimodal [ILLUSTRATION FOR FIGURE 2B OMITTED] and similar in shape to that observed in the holothuroids [ILLUSTRATION FOR FIGURE 1B OMITTED]. Although there are lower numbers of species with eggs in the ln 4.75-5.0 size class (115-148 [[micro]meter]), there is only one statistical mode (ln 5.46, Table I). In this class, therefore, all developmental types are included within the single mode of the distribution.
In literature published since the review of Emlet et al. (14) and in our own unpublished data, we added egg sizes for 61 asteroid species. Asteroids have four modes in the untransformed egg size distributions ([ILLUSTRATION FOR FIGURE 3A OMITTED], Table I). The first mode is of species with primarily planktotrophic development; the following three modes comprise species with lecithotrophic, brooded, or unknown development [ILLUSTRATION FOR FIGURE 3A OMITTED].
The addition of the new asteroid data to the natural log-transformed egg size distribution does not alter the bimodality described by Emlet et al. (14). Statistically there are two separate modes, at ln 5.03 and 6.75 ([ILLUSTRATION FOR FIGURE 3B OMITTED], Table I) - the first [ILLUSTRATION FOR FIGURE 3B OMITTED] contains almost exclusively planktotrophic species (79%), with one lecithotrophic species (Astropecten latespinosus) and three brooders, two of which are intraovarian brooders, Patiriella vivipara and P. parvivipara (34). The second mode [ILLUSTRATION FOR FIGURE 3B OMITTED] is predominantly lecithotrophic (62%) and brooding (12%) species, with the exception of the planktotrophic Porania antarctica (35). The asteroid species with the exceptional developmental type for that mode are often species with rare developmental types. For example, Astropecten latespinosus (300 [[micro]meter]) has an abbreviated lecithotrophic development, lacks both a bipinnaria and a brachiolaria stage, and is in the plankton for only about 4 days (36). Porania antarctica similarly has a large egg (550 [[micro]meter]) and planktotrophic development, but has the shortest period of development (65 days) of any of the Antarctic asteroids studied by Bosch and Pearse (35). For these species, particularly, it would be of interest to examine whether the energetic content or biochemical composition of their eggs differs from those of other asteroids of the same developmental type.
In the four echinoderm classes analyzed, it is only the asteroids that show two separate, distinct, and statistically differentiated modes [ILLUSTRATION FOR FIGURE 3B OMITTED]. There are few species with egg sizes in the size class ln 5.5 to 5.75, equivalent to egg diameters of 245 to 314 [[micro]meter] [ILLUSTRATION FOR FIGURE 3B OMITTED].
Twenty-two egg sizes were added to the echinoid data set of Emlet et al. (14). The untransformed distribution comprises four modes ([ILLUSTRATION FOR FIGURE 4A OMITTED], Table I). The first two modes, at 102 and 317 [[micro]meter], overlap considerably and are composed of planktotrophic, facultative planktotrophic, and some lecithotrophic species [ILLUSTRATION FOR FIGURE 4A OMITTED]. The last two modes are of lecithotrophic and brooding species [ILLUSTRATION FOR FIGURE 4A OMITTED].
The ln-transformed histogram shows statistically two modes ([ILLUSTRATION FOR FIGURE 4B OMITTED], Table I). The first (ln 4.60) comprises 68% of the species and is composed almost entirely of planktotrophic species, with the facultative planktotroph Encope michelini (37) at the point of overlap between modes 1 and 2 [ILLUSTRATION FOR FIGURE 4B OMITTED]. The second mode, at ln 6.22, has a high standard deviation (Table I) and includes two species of facultative planktotrophs (Clypeaster rosaceus, Brisaster latifrons, 14) as well as all the lecithotrophic and brooding species [ILLUSTRATION FOR FIGURE 4B OMITTED].
Two conclusions can be drawn from the statistical analyses presented here. Firstly, in all classes the untransformed egg size distributions show three to four modes, the first being primarily of planktotrophic species, and the remaining two to three containing mainly species with nonfeeding development (i.e., lecithotrophic, abbreviated, brooded, or direct development). Secondly, the ln-transformed distributions are bimodal in the holothurians, asteroids, and echinoids, but unimodal in the ophiuroids. In those classes with a bimodal distribution, it is only the asteroids and echinoids that show a demarcation between modes with feeding and nonfeeding development. In the holothuroids, the modes are less visually distinct, and lecithotrophic and brooding developmental types are found in both modes.
The egg size data examined here [ILLUSTRATION FOR FIGURES 1-4 OMITTED] illustrate the need to view marine invertebrate development as a reproductive continuum (5, 17, 22). Histograms of the egg size distribution show species with planktotrophic development forming a distinctive group at smaller egg sizes (max. egg size [less than]300 [[micro]meter]; [ILLUSTRATION FOR FIGURES 1-4 OMITTED]). However, in all classes, there is some overlap in egg size between these planktotrophic species and species with lecithotrophic or brooded development [ILLUSTRATION FOR FIGURES 1-4 OMITTED]. Egg sizes of species with "intermediate" forms of development, such as facultative planktotrophy (e.g., 17) or modified benthic development (e.g., Patiriella exigua, egg 400 [[micro]meter]; 34), generally fit within the mode for other lecithotrophic and brooding species [ILLUSTRATION FOR FIGURES 3 AND 4 OMITTED].
An important direction for future research on echinoderm eggs is examination of the assumption made here, and in other comparative studies (e.g., 14), that all eggs are basically equal (27). This question should be addressed at two levels. Firstly, within an echinoderm class we need information on the energetic content and biochemical composition of species with a particular developmental type. For example, in the Holothuroidea it would be of interest to determine whether the eggs of known lecithotrophs (egg diameter 150-950 [[micro]meter]) differ in any way from those of the planktotrophic species or from the large (150-4400 [[micro]meter]) eggs of the deep-sea elasipods that are suspected of having direct development (33). Secondly, when sufficient data have been compiled, it would be of interest to compare egg characters (size, energetic content, and composition) between echinoderm classes. In a recent review of the relationship between egg size and energetic content in echinoderms, Jaeckle (27) showed that there were average differences in egg size, energetic content, and biochemical composition between developmental types, but the low number of available studies made it necessary to combine egg parameters from different echinoderm classes. With increasing study on the energy content and biochemical composition of eggs of the less well known echinoderm classes (holothuroids, ophiuroids, crinoids) we may be better able to interpret patterns of egg size distribution [ILLUSTRATION FOR FIGURES 1, 2 OMITTED] and formulate or test new hypotheses on factors important for the selection of egg size.
Research support was provided by a post-doctoral fellowship to MAS from the Division of Marine Science, Harbor Branch Oceanographic Institution. We especially thank Dr. B. Smith, Environment Canada, for performing the statistical analyses. We also thank Drs. R. Mooi, D.S.M. Billett, and D. Pawson for help with egg sizes and taxonomy, and Drs. A. Metaxas, M. Temkin, R. Emson, and the anonymous reviewers for their assistance in clarifying the manuscript. Supported by NSF grants OCE-8717922, 8916264, and 9116560. Harbor Branch Contribution No. 1217.
1. Thorson, G. 1950. Reproductive and larval ecology of marine bottom invertebrates. Biol. Rev. 25: 1-45.
2. Vance, R. R. 1973a. On reproductive strategies in marine benthic invertebrates. Am. Nat. 107: 339-352.
3. Vance, R. R. 1973b. More on reproductive strategies in marine benthic invertebrates. Am. Nat. 107: 353-361.
4. Christiansen, F. B., and T. M. Fenchel. 1979. Evolution of marine invertebrate reproductive patterns. Theor. Popul. Biol. 16: 267-282.
5. Strathmann, R. R. 1985. Feeding and nonfeeding larval development and life-history evolution in marine invertebrates. Annu. Rev. Ecol. Syst. 16: 339-361.
6. Roughgarden, J. 1989. The evolution of marine life cycles. Pp. 270-300 in Mathematical Evolutionary. Theory, M.W. Feldman, ed. Princeton University Press, Princeton, NJ.
7. Havenhand, J. N. 1995. Evolutionary ecology of larval types. Pp. 79-122 in Ecology of Marine Invertebrate Larvae, L. McEdward, ed. CRC Press, New York.
8. Ockelmann, K. W. 1958. The zoology of East Greenland marine Lamellibranchia. Medd. Groenl. 122: 1-256.
9. Kohn, A. J. 1961. Studies on spawning behavior, egg masses, and larval development in the gastropod genus Conus. II. Observations in the Indian Ocean during the Yale Seychelles expedition. Bull. Bingham Oceanogr. Coll. 17: 3-51.
10. Amio, M. 1963. A comparative embryology of marine gastropods, with ecological emphasis. J. Shimonoseki Coll. Fish. 12: 229-253.
11. Thompson, T. E. 1967. Direct development in the nudibranch Cadlina laevis, with a discussion of developmental processes in Opisthobranchia. J. Mar. Biol. Ass. U.K. 47: 1-22.
12. Spight, T. M. 1976. Ecology of hatching size for marine snails. Oecologia 24: 283-294.
13. Kohn, A. J., and F. E. Perron. 1986. Life history and biogeography patterns in Conus. Clarendon Press, Oxford.
14. Emlet, R. B., L. R. McEdward, and R. R. Strathmann. 1987. Echinoderm larval ecology viewed from the egg. Pp. 55-136 in Echinoderm Studies, Vol. 2, M. Jangoux and J. M. Lawrence, eds. A. A. Balkema, Rotterdam.
15. Lessios, H. A. 1990. Adaptation and phylogeny as determinants of egg size in echinoderms from the two sides of the Isthmus of Panama. Am. Nat. 135: 1-13.
16. Levitan, D. R. 1993. The importance of sperm limitation to the evolution of egg size in marine invertebrates. Am. Nat. 141: 517-536.
17. Herrera, J. C., S. K. McWeeney, and L. R. McEdward. 1996. Diversity of energetic strategies among echinoid larvae and the transition from feeding to nonfeeding development. Oceanol. Acta 19: 313-321.
18. Podolsky, R. D., and R. R. Strathmann. 1996. Evolution of egg size in free-spawners: consequences of the fertilization-fecundity trade-off. Am. Nat. 148: 160-173.
19. Levitan, D. R. 1996. Predicting optimal and unique egg sizes in free-spawning marine invertebrates. Am. Nat. 148: 174-188.
20. Dickie, L., M. Hart, and R. Helling. 1989. Prefeeding larval development time is not correlated with egg size in regular echinoids. Invertebr. Reprod. Develop. 15: 229-232.
21. Strathmann, R. R. 1977. Egg size, larval development, and juvenile size in benthic marine invertebrates. Am. Nat. 111: 373-376.
22. Perron, F. E., and R. H. Carrier. 1981. Egg size distributions among closely related marine invertebrate species: are they bimodal or unimodal? Am. Nat. 118: 749-755.
23. Pearse, J. S. 1979. Polyplacophora. Pp. 27-85 in Reproduction of Marine Invertebrates, Vol. 5, A. C. Giese and J. S. Pearse, eds. Academic Press, New York.
24. Barnes, H., and M. Barnes. 1965. Egg size, nauplius size, and their variation with local, geographical, and specific factors in some common cirripedes. J. Anim. Ecol. 34: 391-402.
25. Reaka, M. L. 1979. The evolutionary ecology of life history patterns in stomatopod crustacea. Pp. 235-260 in Reproductive Ecology of Marine Invertebrates, S. E. Stancyk, ed. University of South Carolina Press, Columbia.
26. Grant, A. 1983. On the evolution of brood protection in marine benthic invertebrates. Am. Nat. 122: 549-555.
27. Jaeckle, W. B. 1995. Variation in the size, energy content, and biochemical composition of invertebrate eggs: correlates to the mode of larval development. Pp. 49-77 in Ecology of Marine Invertebrate Larvae, L. McEdward, ed. CRC Press, New York.
28. Holland, N. D. 1991. Echinodermata: Crinoidea. Pp. 247-299 in Reproduction of Marine Invertebrates, Vol. VI, Echinoderms and Lophophorates, A. C. Giese, J. S. Pearse, and V. B. Pearse, eds. Boxwood Press, Pacific Grove, CA.
29. Rowe, F. W. E., D. T. Anderson, and J. M. Healy. 1991. Echinodermata: Concentricycloidea. Pp. 751-760 in Reproduction of Marine Invertebrates, Vol. VI, Echinoderms and Lophophorates, A. C. Giese, J. S. Pearse, and V. B. Pearse, eds. Boxwood Press, Pacific Grove, CA.
30. Turner, R. L., and J. M. Lawrence. 1979. Volume and composition of echinoderm eggs: implications for the use of egg size in life-history models. Pp. 25-40 in Reproductive Ecology of Marine Invertebrates, S. E. Stancyk, ed. University of South Carolina Press, Columbia.
31. Schnute, J., and D. Fournier. 1980. A new approach to length-frequency analysis: growth structure. Can. J. Fish. Aquat. Sci. 37: 1337-1351.
32. Strathmann, R. R. 1978. The evolution and loss of feeding larval stages of marine invertebrates. Evolution 32: 894-906.
33. Billett, D. S. M. 1991. Deep-sea holothurians. Oceanogr. Mar. Biol. Annu. Rev. 29: 259-317.
34. Byrne, M., and A. Cerra. 1996. Evolution of intragonadal development in the diminutive asterinid sea stars Patiriella vivipara and P. parvivipara with an overview of development in the Asterinidae. Biol. Bull. 191: 17-26.
35. Bosch, I., and J. S. Pearse. 1990. Developmental types of shallow-water asteroids of McMurdo Sound, Antarctica. Mar. Biol. 104: 41-46.
36. Komatsu, M. 1975. On the development of the sea-star, Astropecten latespinosus Meissner. Biol. Bull. 148: 49-59.
37. Eckert, G. L. 1995. A novel larval feeding strategy of the tropical sand dollar, Encope michelini (Agassiz): adaptation to food limitation and an evolutionary link between planktotrophy and lecithotrophy. J. Exp. Mar. Biol. Ecol. 187: 103-128.
38. Hendler, G. 1991. Echinodermata: Ophiuroidea. Pp. 355-511 in Reproduction of Marine Invertebrates, Vol. VI, Echinoderms and Lophophorates, A. C. Giese, J. S. Pearse, and V. B. Pearse, eds. Boxwood Press, Pacific Grove, CA.
39. Rao, C. R. 1973. Linear Statistical Inference and Its Applications. 2nd ed. John Wiley, New York. 625 pp.
40. Akaike, H. 1974. A new look at the statistical model identification. IEEE Trans. Autom. Control AC-19: 716-723.
41. Zar, Z. H. 1984. Biostatistical Analysis. 2nd ed. Prentice-Hall, Englewood Cliffs, NJ. 718 pp.
|Printer friendly Cite/link Email Feedback|
|Author:||Sewell, Mary A.; Young, Craig M.|
|Publication:||The Biological Bulletin|
|Date:||Dec 1, 1997|
|Previous Article:||Report of the library director.|
|Next Article:||Spermiogenesis and sperm structure in relation to early events of fertilization in the limpet Tectura testudinalis (Muller, 1776).|