Developmental mode and species geographic range in regular sea urchins (Echinodermata: Echinoidea).
1. Is there a correlation between developmental mode and species geographic range, and does depth affect this relationship?
2. Egg size is a good indicator of developmental mode (Emlet et al. 1987). Is egg size also a good indicator of time to development through metamorphosis (e.g., length of the planktonic interval)?
3. Is time to develop through metamorphosis (i.e., length of the planktonic interval) correlated with species range?
Determinants of geographic range act at levels of the individual, population, and species and include physiological and ecological traits, genetic variation, ability to disperse, and historical changes in distributions of habitats, continents, and oceans (Jackson 1974; Jackson et al. 1985; Gaston 1990; Crame 1993). A broad geographic range is correlated with increased geological longevities in marine mollusks (Jackson 1974; Scheltema 1977, 1978; Hansen 1978, 1980; Jablonski 1982, 1986, 1987; but see also Russell and Lindberg 1988). Presumably, a broad geographic range buffers species against extinction caused by local disturbance events.
For marine benthic organisms, the ability to disperse is often considered a prime determinant of the extent of species geographic range, though other features such as environmental tolerance, adult body size, and depth have been correlated with extent of geographic range (bivalves, Jackson 1974; stomatopods, Reaka 1980). Because larval stages of marine organisms are the most common means of dispersal, variation in the length of larval period has the potential to influence geographic range. Closely related species of marine animals often differ in their mode of larval development, having pelagic, feeding larvae, pelagic, nonfeeding larvae, or nonpelagic (brooded or encapsulated) developmental stages. These developmental modes can differ dramatically in the length of the developmental period or time in the plankton. Feeding larvae typically spend weeks to months in the plankton; nonfeeding, pelagic larvae spend hours to days in the plankton; nonpelagic stages spend little or no time in the water column (Strathmann 1985; Emlet et al. 1987). Exceptions to this generalization are known (e.g., the nonfeeding pelagic larvae of the starfish Mediaster aequalis can remain pelagic for more than 1 yr, Birkeland et al. 1971), but these are rare (see Emlet et al.  for summary of the larval periods of echinoderms).
Evidence for a relationship between the larval period and extent of geographic range varies with the kind of organism under consideration. Long planktonic larval periods are widely thought to produce large geographic ranges among free-living invertebrates (Thorson 1950; Shuto 1974; Scheltema 1977, 1978; Jablonski and Lutz 1983; Perron and Kohn 1985) but not among sessile, aclonal or clonal, invertebrates (Jackson 1986). For fishes, quantitative studies on species within several families show no relationship between pelagic larval duration and extent of geographic range (Pomacanthidae, Thresher and Brothers 1985; Pomacentridae, Thresher et al. 1989; Pomacentridae, Wellington and Victor 1989). However, species of damselfishes (Pomacentridae) with above average pelagic durations were more likely to be distributed across major barriers to dispersal than those with shorter durations (Thresher et al. 1989).
Quantitative evaluations of the predicted relationship between developmental mode and species range are rare among invertebrate taxa, and all are based on molluscan data sets. Scheltema (1989) surveyed 87 species of Recent prosobranch gastropods from the western Atlantic and found that species with planktonic larvae had larger geographic ranges than those with nonplanktonic developmental stages. Scheltema (1989) also examined the length of pelagic interval and found no relationship to geographic range along American continental coastlines, but species with longer planktonic intervals (2-6 mo) were far more likely to have geographic ranges including both sides of the Atlantic ocean than those with shorter (2-6 wk) planktonic intervals. Perron and Kohn (1985) found that Conus species, which are widely distributed among oceanic islands, have longer planktonic periods than those occurring along continental areas. Kohn and Perron (1994) reported a significant, positive relationship between the extent of geographic range and minimum planktonic period for Conus species throughout the Indo-Pacific. In a study of lower Tertiary gastropods in the family Volutidae from the Gulf coast of the United States, Hansen (1978) found broader geographic ranges for species inferred to have pelagic feeding larvae than those with nonfeeding (planktonic or nonplanktonic) development. Jablonski (1982, 1986) reported similar results for a larger survey of late Cretaceous gastropods from Gulf and Atlantic coasts of the United States.
MATERIALS AND METHODS
Determination of Developmental Mode
Three developmental modes were recognized: species with pelagic, feeding larvae (planktotrophy); species with pelagic, nonfeeding larvae (pelagic lecithotrophy), and species with nonpelagic, brooded development (brooding). The developmental mode of a species was determined in one of the following ways: (1) information on larval development from published literature or personal observation (104 species, 48% of data); (2) inference from egg size given in the literature or measured by the author (37 species, 17% of data); (3) inference from crystallographic patterns of adult apical plates (74 species, 34% of data). In a review of published literature, Emlet et al. (1987) reported that echinoids with eggs smaller than 220 [[micro]meter] diameter invariably produce planktotrophic larvae and that echinoids with eggs larger than 350 [[micro]meter] produce pelagic, nonfeeding larvae or brooded embryos. Inference of developmental mode from egg diameter was made for species with eggs within these size ranges but not for species with intermediate egg diameters. The crystallographic patterns of apical plates have been shown to be reliable means for distinguishing between species with and without feeding larvae (Emlet 1985, 1988, 1989). This method of inferring development is based on the presence in feeding larvae of a calcitic skeleton, which determines crystallographic patterns in certain adult apical plates. Because most species with nonfeeding larvae lack a larval skeleton, the crystallographic patterns in the apical plates of these species are not determined by a larval skeleton and are distinct from those with feeding larvae (Emlet 1985, 1989). Neither method for inferring development distinguishes between species with pelagic, nonfeeding larvae or brooded development. All brooding species are recognized because they were collected with developing young among the adult spines or because they possess morphological specializations for brooding (Mortensen 1928-1951). Although there is probably some underestimate of brooding, it is unlikely that brooding is grossly underestimated as many of the species that were inferred to have pelagic, nonfeeding larval development are represented by numerous specimens in museums, and they do not show evidence of brooding (Emlet pers. obsv.) In addition, there may also be a small underestimate of non-feeding larval development as the method of inference based on crystallographic patterns depends on the absence of a larval skeleton, and a very few species with nonfeeding development may retain the larval skeleton. All of these limitations of inference make detection of differences in geographic range between modes of development less likely.
Determination of Species Ranges
Species' distributions and depth ranges of regular echinoids were determined by a literature review, including published monographs, biogeographical studies, and distribution records. Primary sources included: Mortensen's (1928-1951) Monograph of the Echinoidea (the authoritative, worldwide compilation on taxonomy and distributions of all echinoids, prior to the 1950s); Clark and Rowe 1971; Clark and Courtman-Stock 1976; Fell 1976; Serafy 1979; Serafy and Fell 1985; Shigei 1986. These were supplemented by a survey of the pertinent echinoid literature cited in the Zoological Record from 1970 to 1992. In addition, distributional information for some species was supplemented by surveys of collections in several museums (National Museum of Natural History, Washington, D.C.; California Academy of Sciences, San Francisco; Los Angeles County Museum of Natural History; National Museum of New Zealand, Wellington; Australian Museum, Sydney).
The size or extent of a species geographic range was estimated by measuring the greatest linear distance in kilometers across its global distribution. The length of a straight line across water (and land) between chosen end points was measured on maps with different scales and projections in the National Geographic Atlas of the World, 5th edition (1981). Compared with great circle distances, these estimated distances were up to 10% greater for the longer range lengths. Because the data were grouped into bins of 1000 km and all species with ranges greater than 12,000 were lumped together (see next section), the effects of biases in distances caused by map projections should be negligible.
This linear estimate of range differs from other published means of measuring marine geographic ranges, for example, number of biogeographical provinces or subprovinces in which a species occurs, number of (approximately) equal-area regions in which a species occurs, or estimates of areal extent of a species. For a global survey covering low and high latitudes, the more usual means of estimating geographical areas presented the following difficulties: (1) dividing the entire globe into equal-area regions bounded by latitudinal and longitudinal lines is done easily only in the tropics where meridians are approximately parallel and mercator projections offer ease of measuring areal extent. (2) Provinces or subprovinces are of quite different sizes and the locations of their boundaries depend on the taxon being examined. Ideally, one would like to have an areal measure of habitats in which a species is found, but again a global survey of a marine taxon makes such a task formidable. Use of greatest linear distance is intended only as a proxy of relative size of the geographic range and seems to be as reasonable as any other of the above mentioned estimates. See Appendix 1 for a complete listing of species, their known or inferred development, their range end-points, and estimated ranges.
Analysis of Data
Species range distributions were analyzed with Kolmogorov-Smirnov two-sample test (Sokal and Rohlf 1981) by making three pairwise comparisons between the range distributions of the three groups of species with differing modes of development. Initial distributional analyses used bin sizes of 2000, 1000, and 500 km, and all species with ranges greater than 12,000 km were lumped together in a single bin. Analyses with bin sizes of 1000 km resolved more differences than analyses based on bin sizes of 2000 km; but reduction of bin sizes to 500 km did not improve distinction between groups. The range data were also subjected to a Kruskal-Wallis test using developmental mode as a grouping factor and followed by nonparametric multiple comparisons tests for groups of unequal sample sizes (Zar 1984, p. 200).
To examine more closely the relationships between developmental mode and geographic range for species with pelagic larvae, information on egg diameter, time to develop through metamorphosis, and species range size was examined for approximately 28 species with pelagic, feeding larvae and 5 species with pelagic, nonfeeding larvae. These data were obtained from the published literature or from unpublished observations of the author or others (see Appendix 2). Egg diameter was used to estimate egg volume assuming eggs were approximately spherical. Data on times to metamorphosis were based on laboratory studies and serve only as approximations for times in the plankton.
Overall and Taxonomic Patterns
Analysis of geographic ranges includes 215 species of regular echinoids for which developmental mode was known or inferred. This number represents approximately 24% of all living echinoids and 45% of all living regular echinoids (n = 474, counted by Kier 1977). This data set grew out of an initial compilation of 161 echinoid species, including 95 regular echinoids, whose development was known or inferred from egg size (Emlet et al. 1987). Despite a more than a doubling of the original sample of regular echinoids and inclusion of data inferred from apical plate crystallography, the representation of different modes of development has not changed substantially. Percentages of planktotrophic species in the initial and current data sets were very similar, 72% and 70%, respectively. Species known or inferred to have pelagic nonfeeding development was 23% of the current sample compared with 18% in the smaller data set, and this increase came at the expense of brooding, which declined to 7% in the current sample from 11% in the initial sample. The general stability of the relative abundances of species with the different modes of development suggests the larger sample is representative of developmental modes of regular echinoids. For a systematic summary of developmental modes of echinoids, see Emlet (1990) and Appendix 1.
When all urchin species across all depths were considered, the distribution of range sizes of species with planktotrophic larvae was significantly different from that of species with pelagic nonfeeding development ([ILLUSTRATION FOR FIGURE 1 OMITTED], table 1). Somewhat surprisingly, there was not a significant difference in the distribution of range sizes between species with planktotrophic development and those with brooded development (P = 0.09, table 1). However, if the one brooding species with the greatest and outlying range size was omitted (Aporocidaris milleri, range = 21,800 km), then distributions of range sizes of planktotrophic and brooding species were significantly different (P = 0.03, table 1). Whether this brooder with the extremely broad species range was included or not, there is no significant difference in the distributions of range sizes of species with pelagic nonfeeding larvae and brooded development (table 1).
Species with feeding larvae had significantly larger range sizes than those with pelagic, nonfeeding development and those with brooded development, when the one brooding species with an extreme range was excluded (table 1, Kruskal-Wallis and multiple-comparisons tests).
The orders Cidaroida and Temnopleuroida (each distinct clades) contain the great majority of regular echinoid species known or inferred to have pelagic, nonfeeding larvae (40 of 51 species) and the Cidaroids contained 13 of the 15 known brooding species. Each of these orders also has considerable numbers of species with pelagic feeding larvae. The data sets available for these taxa were reasonably large (60 and 40 species for cidaroids and temnopleuroids, respectively), and therefore geographic range patterns were examined for these taxonomic groups [ILLUSTRATION FOR FIGURE 1 OMITTED]. The distributions of species ranges for each taxon parallel those for the complete data set described above (table 1). For both cidaroids and temnopleuroids, species with pelagic, feeding larvae had significantly longer ranges than those of species with pelagic, nonfeeding development (table 1: cidaroids, Kruskal-Wallis test followed by multiple comparisons; temnopleuroids, Mann-Whitney U-test). For cidaroids, distributions of range sizes were not significantly different for species with pelagic, feeding larvae and species with brooded development, even when the one brooder with the extreme range was excluded (table 1). Among cidaroids, there was no difference in distribution of range sizes for species with pelagic, nonfeeding larvae and those with brooded development. Because brooded development is known for only one living temnopleuroid species, it was not compared within this taxon.
The brooding cidaroid, Aporocidaris milleri, was both included and excluded in the above analyses because its unusually large geographic range (circum-Antarctic and southern and northern Pacific Ocean) cannot be easily interpreted. Based on studies of morphological variation, Fell (1976) synonomized A. antarctica with A. milleri, and this has resulted in the very large geographic range. It is possible that this taxon is a complex of cryptic species. Because of its deep water occurrence, very little is known of its biology and no molecular data on populations are available.
Analysis by Depth
Emlet et al. (1987) examined patterns of development as a function of latitude and depth for all echinoids. That data set showed 80% of species in shallow water had planktotrophic development and this percentage dropped to considerably less than 50% in deeper waters (e.g., 9% for species occurring only deeper than 100 m). The present, larger data set supported the previous one in showing that planktotrophic development dominated (ca. 80% of species) in shallow waters (table 2), but planktotrophic larvae were also found in approximately 55% of the regular echinoid species occurring in deeper waters (table 2). This increased occurrence of species with feeding larvae in deeper waters is due to the increased sampling permitted by crystallographic techniques for inferring development.
The abundant representation of planktotrophic development among echinoids in deeper waters and the increased number of shallow-water species with pelagic, nonfeeding larvae reported here make reasonable an analysis of species range size distributions as a function of depth. When all regular echinoid species occurring 100 m deep or less were examined (n = 130, 60% of the data), the distributions of species range sizes paralleled the patterns for those reported [TABULAR DATA FOR TABLE 1A OMITTED] at all depths ([ILLUSTRATION FOR FIGURE 1 OMITTED], shallow water). Only the range-size distributions for pelagic feeding and nonfeeding larvae were significantly different (P = 0.005, table 1B). When cidaroids occurring in shallow waters (n = 22) are considered alone, only range distributions of species with feeding larvae and brooded development are significantly different (P = 0.01, table 1B). For shallow-water temnopleuroids, distributions of ranges of species with pelagic feeding and nonfeeding larvae are significantly different (P = 0.025, table 1B).
In contrast to the above patterns, when all species occurring predominantly deeper than 100 m were examined (n = 85 species; 40% of the data), there were no significant differences in distributions of species range sizes for any of the pairwise comparisons between species grouped by mode of development ([ILLUSTRATION FOR FIGURE 1 OMITTED], deep water; table 1B). Kruskal-Wallis [TABULAR DATA FOR TABLE 1B OMITTED] tests on ranges of deep-water species grouped by developmental mode corroborates the lack of differences between the groups (table 1B). Similar nonsignificant results were obtained for cidaroids (table 1B). No analysis was carried out for the eight species of deep water temnopleuroids.
Relative to shallow-water species, the deep-water species show apparent reductions in range sizes for species with feeding larvae and increases in range sizes for species with brooded development (cf. means and medians in table 1A). Species with pelagic, feeding larvae have significantly greater ranges in shallow water ([less than]100 m) than those occurring in deeper water ([greater than]100 m) (Mann-Whitney test, U = 3051, P [less than] 0.01). Species with other developmental modes do not differ in extent of ranges as a function of depth (Mann-Whitney tests: pelagic, nonfeeding larvae, U = 328, P = 0.86; brooded development, including A. milleri, the species with extreme range, U = 17, P = 0.24; brooded development, excluding A. milleri, U = 17, P = 0.37). This reduction in ranges of planktotrophic species in deeper waters accounts for the absence in deeper waters of differences in range as a function of development.
TABLE 2. Patterns of development as a function of species depth range. Grouping is that used in Emlet et al. (1987) and includes the "shallow" and "deep" groups used in the present study. P, pelagic feeding larvae; L, pelagic nonfeeding larvae, B, brooded development. Species considered as "deep" are labeled with an asterisk after their depth range in the Appendix 1.
No. of spp. P (%) L (%) B (%)
Littoral (shallowest depth 0 m) 86 80 16 3.5
In [less than or equal to] 10 m 113 79 17.5 3.5
Shallow (down to ca. 100 m) 130 78.5 17 4.5
Deeper than 1000 m 34 53 32.5 14.5
Only deeper than 50 m 81 53 37 10
Deep (ca. 100 m and deeper) 85 55 34 10.5
Only deeper than 100 m 65 54 34 12
Analysis of Egg Volume, Developmental Time, and Species Range
Two relationships were examined. (1) Is egg volume (an estimate of egg nutritional content; Emlet et al. 1987; McEdward and Chia 1991) correlated with time to develop through metamorphosis? (2) Is time to develop through metamorphosis correlated with the extent of species range?
The data on time to metamorphosis were highly variable. Although some variability may be due to egg nutritional value (parental investment), it is also widely recognized that temperature and food ration (for feeding larvae) strongly affect developmental rates of sea urchin larvae (e.g., Strathmann 1985, 1987; Emlet et al. 1987; Pearse and Cameron 1991). These later two factors may reduce or obscure a relationship between egg nutritional value and time to develop. The effects of temperature on rate processes are commonly summarized by a [Q.sub.10] value (Schmidt-Nielsen 1990). A [Q.sub.10] of 2 reflects a doubling of a rate process over an interval of 10 [degrees] C and is a commonly measured value (Hochachka and Somero 1984). To compensate for temperature effects, all data on times to develop through metamorphosis were adjusted to times at 20 [degrees] C with a [Q.sub.10] value ranging from 2 to 3.6. Cameron et al. (1985) reported [Q.sub.10]S of 3.6 and 3.0, respectively, for developmental times to metamorphosis for the tropical, regular urchins Lytechinus variegatus and Echinometra lucunter. McEdward (1985) reports a [Q.sub.10] of 3.6 for part of feeding larval development (two-armed to six-armed stages) of the sand dollar, Dendraster excentricus. No adjustments to developmental time for variation in food ration were made because no quantitative rules have been developed.
Without adjusting developmental times to a common temperature, there was a significant difference in the developmental times for regular echinoid species with feeding larvae and those with nonfeeding larvae (Mann-Whitney test, U = 135, P = 0.001, n = 28 and five species with feeding and nonfeeding larvae, respectively). However, the correlation [TABULAR DATA FOR TABLE 3 OMITTED] between egg volume and developmental time is nonsignificant when species with feeding and nonfeeding larvae are considered together or when species with feeding larvae are considered alone (table 3, [ILLUSTRATION FOR FIGURE 2A OMITTED]). When data on time to metamorphosis for species with feeding and nonfeeding larvae were adjusted to a common 20 [degrees] C, there was a significant negative correlation between egg volume and time to metamorphosis for each [Q.sub.10] value tested from 2.0 to 3.6 (table 3). The [Q.sub.10] values which gave the strongest correlation (Spearman rank correlation value, -0.634) were 3.0 and 3.1. When only species with feeding larvae were considered, the negative correlation between egg volume and time to metamorphosis were significant for a range of [Q.sub.10]s from 3.0 through 3.6 [ILLUSTRATION FOR FIGURE 2B OMITTED], with the strongest correlation (Spearman rank correlation value, -0.418) occurring at [Q.sub.10] = 3.6.
Although temperature may affect developmental time and thus time in the plankton, there was no difference in geographic range sizes for species with pelagic feeding and non-feeding larvae (Mann-Whitney test, U = 78, P = 0.69, n = 28 and 5) and no correlation between developmental time and species range size (table 3, [ILLUSTRATION FOR FIGURE 3 OMITTED]). Only developmental times, unadjusted for temperature, were used because most laboratory temperatures approximate the environmental temperatures, and these are the most realistic estimates available for time in the plankton.
Dispersal of larvae and a long larval period are widely cited reasons for large geographic ranges (e.g., Scheltema 1989; Thresher et al. 1989). This study on echinoids shows (1) that developmental mode and egg volume affect developmental time to metamorphosis, (2) that developmental mode affects the extent of geographic range among species occurring in water depths less than 100 m, but (3) the length of the planktonic interval (developmental time) is not related to geographic range.
Because feeding larvae on average take longer to develop in the water column than do nonfeeding larvae (table 3; Emlet et al. 1987), it is tempting to attribute the difference in geographic ranges between species with pelagic, feeding larvae and those with pelagic, nonfeeding larvae to the time larvae have to disperse. Measures of the planktonic interval, estimated from time to develop through metamorphosis in the laboratory, should indicate the ability of an individual to spread its offspring and might ultimately translate into a broader geographic range for a species as a whole. With a somewhat limited data set (n = 33 species, see Appendix 2), there is no correlation between time to develop through metamorphosis and geographic range size for species, even though species with nonfeeding larvae have shorter developmental times than those with feeding larvae.
The potential influence of a planktonic interval on species range can also be evaluated by comparing species that have pelagic larvae to those that brood their young. Although some benthic invertebrates (with or) without pelagic larvae may disperse by rafting of juvenile or adult stages (e.g., Highsmith 1985; Jokiel 1989), this has not been reported for echinoids. Neither is there reason to suspect that echinoids that brood are more likely than echinoids with pelagic larvae to disperse by rafting. Range sizes are significantly greater for species with pelagic feeding larvae and those with brooded development only when one brooder with an extreme range is removed from the full data set. Even when this species is ignored, five brooding species have substantial range sizes between 4000 and 7000 km. Differences in ranges for planktotrophic and brooding species are not found within the shallow-water subset of the data. There is no difference in the distributions of ranges or range sizes when brooding species are compared with those that have pelagic, nonfeeding larvae. Finally, all differences in range extent as a function of development disappear in species occurring at depths over 100 m. Though pelagic larval dispersal may partially affect species ranges, the above comparisons suggest such an effect is weak and not directly related to potential dispersal distances judged by the planktonic interval.
For echinoids, there is no evidence that time to develop through metamorphosis is related to the extent of a species' geographic range. This latter result stands in strong contrast to data for the tropical gastropod, Conus, which show a strong correlation between length of the precompetent (pelagic) period and areal extent of geographic range (Kohn and Perron 1994). The result for echinoids seems paradoxical since developmental mode and time to develop through metamorphosis are strongly correlated. One possible solution to this paradox is that species range may be influenced by the indirect effects of dispersal rather than its direct effects. Time to development through metamorphosis most directly reflects the ability of an adult to spread its sibling offspring (Strathmann 1974), and that is not correlated with the extent of species range. Dispersal has other consequences such as promoting gene flow and genetic similarity between populations and replacement of discrete populations eliminated by disturbance. Perhaps developmental mode influences the extent of species geographic range, indirectly, through one or several of these consequences of dispersal. Demonstration of the indirect effects of developmental time on geographic range will require combined temporal and spatial studies on genetic and populational variability between species with differing dispersal (larval developmental) times.
Regardless of developmental mode echinoids have very large species ranges compared to other benthic taxa, for example, mollusks (Vermeij 1987; Scheltema 1989; Kohn and Perron 1994). Many echinoid species in the data set have Indo-Pacific distributions ranging from the Red Sea or Africa to the South Pacific and sometimes Hawaii. Other broadly distributed taxa are amphi-Atlantic, and still others are circum-Antarctic. The tendency of echinoids to have such large ranges may contribute to their low diversity relative to other marine benthic taxa.
Limitations and Biases in the Data Set
Though extensive, this global data set is not comprehensive and may be subject to collecting biases that influence the patterns. Possible biases include: (1) systematic biases such as under or over representation of certain groups; (2) taxonomic biases caused by incorrect species identifications or incorrect alpha taxonomy; (3) biogeographic biases including under- or overrepresentation of certain areas; (4) biases related to depth (i.e., localities of shallower taxa are better known); (5) biases caused by differences between rare and common species.
There are no obvious systematic omissions. The sample includes species from all nine orders of regular echinoids recognized by Kier (1977). Based on his species counts, the data set contains at least 20% of the species of each order. For cidaroids and temnopleuroids (including toxopneustids in Kier's count), respectively, 42% of 144 species and 48% of 119 species were sampled. In two other large orders, Echinoida and Diadematoida, respectively, 69% of 65 species and 54% of 48 species were sampled.
Taxonomic biases are certainly possible. All species used in the study are recognized by morphological characters and thus cryptic species may exist. The great majority of the taxa in this sample were included in Mortensen's Monograph of the Echinoidea (1928-1951), recognized as the major treatise on the group. Mortensen's work has and continues to dominate the alpha taxonomy of the echinoids. Recently his species characters have been used in construction of cladistic phylogenies at generic and higher levels (e.g., Smith 1988; Smith and Wright 1989, 1990). Recent studies have suggested one nominal species, Echinometra mathaei, is a complex of four species (e.g., Uehara et al. 1991; Metz et al. 1991). Though this may occur for other echinoids, I know of no other attempts to split or lump regular echinoid species. Incorrect species identification are also possible sources of error, but most literature citations can be traced back to Mortensen, and where I collected data on museum specimens, I attempted to compare material with Mortensen's Monograph.
Biogeographical biases may arise from imprecisely known distributions of species, especially deep water and rare species. My measure of geographic range (greatest distance across a species distribution) is not exempt from this bias. An alternative approach that avoids this bias was used by Vermeij (1987) and Vermeij et al. (1990), who compared the occurrence of gastropod species across specific barriers as a function of different traits (adult shell characteristics and larval developmental mode, respectively). If the faunas of two areas separated by a barrier are well known, and the subset of species shared by the two areas is dominated by species with a particular trait, then that trait may have influenced their distribution. A similar analysis could be done on a subset of the data for echinoids, but the patchy distribution of species with nonfeeding development (see below) makes the choice of areas and barriers difficult. Which barriers should be used in a global survey are not obvious either.
The distribution of water and land masses, specific habitat, and earth history all influence species distributions (e.g., Mayr 1954; Smith 1984). Emlet (1990) reported a high incidence of echinoid species with pelagic nonfeeding larvae along the southern and western coasts of Australia. Another collection of species with this developmental mode occurs along the Pacific coast of Japan and into Indonesia. The high incidence of brooding among Antarctic echinoids has been long known (e.g., Mortensen 1928-1951), and Emlet (1990) reconfirmed a separate, small group of brooding species from southern Australia and New Zealand. Beyond these regional concentrations, a few species with nonfeeding development also occur in tropical waters (13 regular species in the data set: 5 in shallow and 8 in deeper water) and brooding is known for one North Atlantic regular echinoid. The paucity of regular echinoids with nonfeeding development in tropical seas does not allow a statistical evaluation of the influence of development on species range in a tropical subset of taxa.
I am grateful to A. Kohn, D. Lindberg, T. Mace, R. Mooi, and G. Vermeij for many helpful comments on the manuscript. This research was supported by setup funds from the University of Oregon and by National Science Foundation grant IBN 93-96004. This is Oregon Institute of Marine Biology Contribution number 94-02.
Birkeland, C., F.-S. Chia, and R. R. Strathmann. 1971. Development, substratum selection, delay of metamorphosis and growth in the seastar Mediaster aequalis Stimpson. Biological Bulletin 141:99-108.
Cameron, R. A., I. Boidron-Metairon, and O. Monterrosa. 1985. Does the embryonic response to temperature and salinity by four species of Caribbean sea urchins parallel the reproductive synchrony? Proceedings of the 5th International Coral Reef Congress, Tahiti. 5:273-278.
Clark, A. M., and J. Courtman-Stock. 1976. The Echinoderms of southern Africa. British Museum of Natural History, London.
Clark, A. M. and F. W. E. Rowe. 1971. Monograph of shallow-water Indo-West Pacific Echinoderms. Pitman Press, Bath, England.
Crame, J. A. 1993. Latitudinal range fluctuations in the marine realm through geological time. Trends in Ecology and Evolution 8:162-166.
Emlet, R. B. 1985. Crystal axes in recent and fossil adult echinoids indicate trophic mode in larval development. Science 230:937-940.
-----. 1988. Crystallographic axes of echinoid genital plates reflect larval form: some phylogenetic implications. Pp. 299-310 in C. R. C. Paul and A. B. Smith, eds. Echinoderm phylogeny and evolutionary biology. Clarendon Press, Oxford.
-----. 1989. Apical skeletons of sea urchins (Echinodermata: Echinoidea): two methods for inferring mode of larval development. Paleobiology 15:223-254.
-----. 1990. World patterns of developmental mode in echinoid echinoderms. Pp. 329-310 in M. Hoshi and O. Yamashita, eds. Advances in invertebrate reproduction, Vol 5. Elsevier, Tokyo.
Emlet, R. B., L. R. McEdward, and R. R. Strathmann. 1987. Echinoderm larval ecology viewed from the egg. Pp. 55-136 in M. Jangoux and J. M. Lawrence, eds. Echinoderm studies, Vol. 2. Balkema Press, Rotterdam.
Fell, F. J. 1976. The Cidaroida (Echinodermata: Echinoidea) of Antarctica and the southern oceans. Ph.D. diss. University of Maine, Orono.
Gaston, K. J. 1990. Patterns in the geographical ranges of species. Biological Reviews 65:105-129.
Hansen, T. A. 1978. Larval dispersal and species longevity in lower Tertiary gastropods. Science 199:885-887.
-----. 1980. Influence of larval dispersal and geographic distribution on species longevity in neogastropods. Paleobiology 6: 193-207.
Highsmith, R. C. 1985. Floating and algal rafting as potential dispersal mechanisms in brooding invertebrates. Marine Ecology Progress Series 25:169-176.
Hochachka, P. W., and G. N. Somero. 1984. Biochemical adaptation. Princeton University Press, Princeton, N.J.
Jablonski, D. 1982. Evolutionary rates and modes in late Cretaceous gastropods: role of larval ecology. Third North American Paleontological Convention, Proceedings 1:257-262.
-----. 1986. Larval ecology and macroevolution in marine invertebrates. Bulletin of Marine Science 39:565-587.
-----. 1987. Heritability at the species level: analysis of geographic ranges of Cretaceous mollusks. Science 238:360-363.
Jablonski, D., and R. A. Lutz. 1983. Larval ecology of marine benthic invertebrates: Paleobiological implications. Biological Reviews 58:21-89.
Jackson, J. B. C. 1974. Biogeographic consequences of eurytopy and stenotopy among marine bivalves and their evolutionary significance. American Naturalist 108:541-560.
-----. 1986. Modes of dispersal of clonal benthic invertebrates: consequences for species' distributions and genetic structure of local populations. Bulletin of Marine Science 39:588-606.
Jackson, J. B.C., J. E. Winston, and A. G. Coates. 1985. Niche breadth, geographic range, and extinction of Caribbean reef-associated cheilostome Bryozoa and Scleractinia. Proceedings of the 5th International Coral Reef Congress, Tahiti. 4:151-158.
Jokiel, P. L. 1989. Rafting of reef corals an other organisms at Kwajalein Atoll. Marine Biology 101:483-493.
Kier, P. M. 1977. The poor fossil record of the regular echinoid. Paleobiology 3:168-174.
Kohn, A. J., and F. E. Perron. 1994. Chapter 5 in Life history and biogeography: patterns in Conus. Clarendon Press, Oxford.
McEdward, L. R. 1985. Effects of temperature on the body form, growth, electron transport system activity, and development rate of an echinopluteus. Journal of Experimental Marine Biology and Ecology 93:169-181.
McEdward, L. R., and F.-S. Chia. 1991. Size and energy content of eggs from echinoderms with pelagic lecithotrophic development. Journal of Experimental Marine Biology and Ecology 147:95-102.
Mayr, E. 1954. Geographic speciation in tropical echinoids. Evolution 8:1-18.
Metz, E. C., H. Yanagimaschi, and S. R. Palumbi. 1991. Gamete compatibility and reproductive isolation of closely related Indo-Pacific sea urchins, genus Echinometra. Pp. 131-137 in T. Yanagisawa et al., eds. Biology of Echinodermata. Proceedings of the Seventh International Echinoderm Conference. A. A. Balkema, Rotterdam.
Mortensen, T. 1928-1951. A monograph of the Echinoidea I-V. C. A. Reitzel, Copenhagen.
National Geographic atlas of the world, 5th ed. 1981. National Geographic Society, Washington, D. C.
Pearse, J. S., and R. A. Cameron. 1991. Echinodermata: Echinoidea. Chap. 7, Pp. 513-662 in A. C. Giese et al., eds. Reproduction of marine invertebrates, Vol. VI. Echinoderms and Lophophorates. Boxwood Press, Pacific Grove, Calif..
Perron, F. E., and A. J. Kohn. 1985. Larval dispersal and geographic distribution in coral reef gastropods of the genus Conus. Proceedings of the 5th International Coral Reef Congress, Tahiti 4: 95-100.
Reaka, M. L. 1980. Geographic range, life-history patterns, and body size in a guild of coral-dwelling mantis shrimps. Evolution 34:1019-1030.
Russell, M. P., and D. R. Lindberg. 1988. Real and random patterns associated with molluscan spatial and temporal distributions. Paleobiology 14:322-330.
Scheltema, R. S. 1977. Dispersal of marine invertebrate organisms: paleobiogeographic and biostratigraphic implications. Pp. 73-107 in E.G. Kauffman and J. E. Hazel, eds. Concepts and methods of biostratigraphy. Dowden, Hutchinson and Ross, Stroudsburg, Pa.
-----. 1978. On the relationship between dispersal of pelagic veliger larvae and the evolution of marine prosobranch gastropods. Pp. 303-321 in B. Battaglia and J. A. Beardmore, eds. Marine organisms. Plenum Press, New York.
-----. 1989. Planktonic and non-planktonic development among prosobranch gastropods and its relationship to the geographic ranges of species. Pp. 183-188 in J. S. Ryland and P. A. Tyler, eds. Reproduction, genetics, and distributions of marine organisms. Olsen and Olsen, Fredensborg, Denmark.
Schmidt-Nielsen, K. 1990. Animal physiology: adaptation and environment, 4th ed. Cambridge University Press, Cambridge.
Serafy, D. K. 1979. Echinoids (Echinodermata: Echinoidea). Memoirs of the Hourglass Cruises 5(3):1-120.
Serafy, D. K., and F. J. Fell. 1985. Marine flora and fauna of the northeastern United States. Echinodermata: Echinoidea. National Oceanic and Atmospheric Administration Technical Report, National Marine Fisheries Service 33:1-27.
Shigei, M. 1986. The sea urchins of Sagami Bay. Maruzen, Tokyo.
Shuto, T. 1974. Larval ecology of prosobranch gastropods and its bearing on biogeography and paleontology. Lethaia 7:239-256.
Smith, A. B. 1984. Echinoid paleobiology. George Allen and Unwin, London.
-----. 1988. Phylogenetic relationship, divergence times, and rates of molecular evolution for Camarodont sea urchins. Molecular Biology and Evolution 5:345-365.
Smith, A. B., and C. W. Wright. 1989. British Cretaceous Echinoids. Part 1, General introduction and cidaroida. Monograph of the Palaeontographical Society, London 1-101.
-----. 1990. British Cretaceous Echinoids. Part 2, Echinothurioida, Diadematoida and Stirodonta (1, Calycina). Monograph of the Palaeontographical Society, London 101-198.
Sokal, R. R., and F. J. Rohlf. 1981. Biometry. W. H. Freeman, San Francisco.
Strathmann, R. R. 1974. The spread of sibling larvae of sedentary marine invertebrates. American Naturalist 108:29-44.
-----. 1985. Feeding and nonfeeding larval development and life-history evolution in marine invertebrates. Annual Review of Ecology and Systematics 16:339-361.
-----. 1987. Larval feeding. Pp. 465-550 in A. C. Giese et al., eds. Reproduction of marine invertebrates, Vol. IX. General aspects: seeking unity in diversity. Blackwell Scientific, Palo Alto, Calif.
Thorson, G. 1950. Reproductive and larval ecology of marine bottom invertebrates. Biological Reviews 25:1-45.
Thresher, R. E., and E. B. Brothers. 1985. Reproductive ecology and biogeography of Indo-west Pacific angelfishes (Pisces: Pomacanthidae). Evolution 39:878-887.
Thresher, R. E., P. L. Colin, and L. J. Bell. 1989. Planktonic duration, distribution and population structure of western and central Pacific damselfishes (Pomacentridae). Copeia 1989:420-434.
Uehara, T., M. Shingaki, K. Taira, Y. Arakaki, and H. Nakatomi. 1991. Chromosome studies in eleven Okinawan species of sea urchins, with special reference to four species of the Indo-Pacific Echinometra. Pp. 119-129 in T. Yanagisawa et al., eds. Biology of Echinodermata. Proceedings of the Seventh International Echinoderm Conference. A. A. Balkema, Rotterdam.
Vermeij, G. J. 1987. The dispersal barrier in the tropical Pacific: implications for molluscan speciation and extinction. Evolution 41:1046-1058.
Vermeij, G. J., A. R. Palmer, and D. R. Lindberg. 1990. Range limits and dispersal of mollusks in the Aleutian Islands, Alaska. Veliger 33:346-354.
Wellington, G. M., and B. C. Victor. 1989. Planktonic larval duration of one hundred species of Pacific and Atlantic damsel-fishes (Pomacentridae). Marine Biology 101:557-567.
Zar, J. H. 1984. Biostatistical analysis. Prentice-Hall, Englewood Cliffs, N.J.
[TABULAR DATA FOR APPENDIX 1 OMITTED]
[TABULAR DATA FOR APPENDIX 2 OMITTED]
|Printer friendly Cite/link Email Feedback|
|Author:||Emlet, Richard B.|
|Date:||Jun 1, 1995|
|Previous Article:||Sympatric convergence and environmental correlation between two land-snail species.|
|Next Article:||The selection barrier between populations subject to stabilizing selection.|