Effect of colony size, polyp size, and budding mode on egg production in colonial coral.
Recent studies on reproduction of colonial, scleractinian corals have revealed that polyps in small colonies produce no eggs or few eggs, and that in large, fecund colonies polyps at or near the margin of the colony produce fewer eggs than those in central positions (Harriott, 1983; Chornesky and Peters, 1987; Harrison and Wallace, 1990; Szmant, 1991; Soong and Lang, 1992; Van Veghel and Kahmann, 1994; Hall and Hughes, 1996). A coral colony is usually derived from a single polyp by budding, so that the polyps in the colony are usually genetically identical (Jackson and Coates, 1986) and likely to have the same structure and the same potential for any physiological function (Meester and Bak, 1995). Thus, the recent work has demonstrated that coral polyps may differ in reproductive activity even though they share the same structure and genetic identity.
In colonial corals, colony size is considered to be important for the maturation of polyps (reviewed in Harrison and Wallace, 1990), but the size of a coral has another aspect, i.e., polyp size. Although the polyp is the basic unit for physiological activities, its size has not often been studied. Two exceptions are the studies by Harriott (1983) and Van Vegbel and Kahmann (1994). Harriott (1983) investigated sexual reproduction in four scleractinian coral species, and suggested that polyp size rather than colony size limits gonad production in the mussid coral Lobophyllia corymbosa; she presented no data on polyp size. Van Veghel and Kahmann (1994) studied reproduction in three morphotypes of the faviid coral Montastrea annularis, measuring polyp size two-dimensionally at the base. They found that the fertility (measured as the percentage of polyps with eggs) of polyps from 27 colonies was less than 40% in small polyps ([greater than]1.5 mm) and increased with polyp diameter.
To date, however, no study of coral reproduction has quantitatively examined the relationship between the size of a polyp and its egg production. Furthermore, polyp size has previously been measured in two dimensions, even though it is the space within a polyp that limits gamete volume. Thus, I measured the three-dimensional size of polyps in the faviid coral Goniastrea aspera.
I examined the effect of colony size, polyp size, and location of polyps on egg production. I show that not only colony size but also polyp size is related to egg production, and further suggest that fertility in G. aspera is also affected by polyp age, determined as the time since formation by extratentacular budding.
Materials and Methods
Collection, colony size, and egg count
I collected specimens of G. aspera at the lowest low tide level on the fringing reef of Sesoko island, Okinawa, southern Japan (26[degrees] 38[feet]N, 127[degrees] 53[feet]E). This coral is a simultaneous hermaphrodite; in Okinawa it has one or two spawning peaks per year, in June and July (Heyward et al., 1987; Hayashibara et al., 1993; Sakai, 1997). To assess egg production by polyps, 23 G. aspera colonies of various sizes were collected on 21 May 1991, 10 days before the predicted date of the first spawning of the year. Before collection, the length, width, and height of the colony were measured to the nearest 1 min. Since egg production may be affected by disease or injury and by contact with large benthic organisms such as neighboring corals, I chose colonies without such features.
The collected colonies were fixed in 10% formalin in seawater for at least 24 h, and decalcified in a solution of 5% formalin + 5% acetic acid. I counted the number of polyps in the colony after decalcification and used this value as a measure of colony size, because the polyp number is highly correlated with colony somatic biomass (Sakai, 1998). Then, randomly selected polyps (n ranged from 1 to 20 according to colony size) were cut from the decalcified colony after the maximum and minimum diameters of the oral side were measured to the nearest 0.1 mm. Polyp height was measured after the polyps were excised from the colony. I estimated the volume of each polyp from its diameter and height. While non-marginal polyps (in polyp row 2 or higher, away from the colony margin; Fig. 1) were always cylindrical, marginal polyps (in the row at the colony edge) were frequently more conical. Hence, I estimated the polyp volume as k x [Pi] X [[(maximum diameter + minimum diameter)/4].sup.2] x height, where k ranged from 1/3 (conical) to 1 (cylindrical). The value of k was subjectively determined for each polyp according to its shape. Each polyp was dissected under a stereoscopic microscope, and all the eggs within a polyp were counted. I measured the diameter of five eggs, randomly selected from each of four polyps in each of ten fecund colonies. The overall mean of the egg diameter in the polyps was 310 [+ or -] 30 [[micro]meter] (Mean [+ or -] SD, n = 200). There was no significant colony effect on the egg diameter (Model III nested ANOVA, df = 9, SS = 9663.98, F = 0.99, P = 0.5; when polyps were nested in the colonies). Thus, egg production can be compared in terms of number of eggs per polyp volume (NE/PV, see below).
Preliminary examination of the effect of polyp position on egg production in four fecund colonies showed that marginal polyps had significantly lower NE/PV than nonmarginal polyps (Table I). NE/PV of polyps in row 2 or higher did not differ significantly (P [greater than] 0.05). In subsequent analyses I therefore grouped polyps in a colony into marginal and non-marginal sets. Polyp-group fertility, as percentage of polyps with eggs to total number of sampled polyps, was calculated separately for marginal and nonmarginal polyps in each colony.
Budding mode and polyp biomass
To examine the relationship between estimated polyp volume and polyp biomass, I collected 11 colonies of G. aspera of various sizes (from 40 to 420 polyps) in October 1993, when reproductive tissue was negligible (Sakai, 1997). These colonies were fixed and decalcified as described for the egg-count specimens. When I counted the number of polyps in the decalcified colonies, I also recorded [TABULAR DATA FOR TABLE I OMITTED] the mode of budding (i.e., intratentacular or extratentacular) for marginal and non-marginal polyps separately. Formation of a new bud outside the polyp wall was recorded as extratentacular budding, and formation of a new polyp wall dividing the original polyp was recorded as intratentacular budding.
Four polyps, two marginal and two non-marginal, were randomly selected from each colony after decalcification. These 44 polyps were measured and excised from the colonies as described for egg-count polyps, and then rinsed in distilled water and dried at 60C to constant weight. Each polyp was weighed to the nearest 0.1 mg. The estimated polyp volume showed a strong linear correlation with the somatic biomass, both in marginal ([r.sup.2] = 0.87, P [less than] 0.0001, n = 22) and non-marginal ([r.sup.2] = 0.84, P [less than] 0.0001, n - 22) polyps, with no significant difference between the two regression lines (ANCOVA, df = 1, SS = 0.29, F = 0.17, P - 0.7). Therefore, assuming that the polyp volume represents the somatic biomass, I calculated the average number of eggs per polyp volume (NE/PV, [mm.sup.-3]) for marginal and non-marginal polyps in each colony.
Egg production by a whole colony (colony fecundity) was estimated as the sum of egg production by marginal and non-marginal polyps. The total egg production for each polyp type was calculated as the mean number of eggs per polyp multiplied by the number of polyps of that type in the colony.
I examined the relationship between colony size (polyp number) and fertility and mean NE/PV, fitting a curve using the logistic equation, since egg production by coral polyps over colony size is expected to be sigmoidal (Babcock, 1991). For the remaining regression analyses, I examined linear, exponential, power, and logarithmic functions, and used the one with the highest correlation coefficients after examining the residuals (Zar, 1996). I used Model III ANOVA, nesting random-effects such as individual colonies within fixed-effects such as colony size classes (Zar, 1996). The assumptions for ANOVA (i.e., normality and homogeneity of variance) were examined with the Shapiro-Wilk W test and O'Brien's tests. Logarithmic transformation was made if necessary; when the assumptions were severely violated even after the transformation, a nonparametric analysis such as the Wilcoxon paired-sample test was performed using means from colonies. Statistical analyses were performed with computer software packages (Data Desk, Data Description Inc., Ithaca, NY; JMP, SAS Institute, Cary, NC; Kaleida Graph, Synergy Software, Reading, PA).
Marginal and non-marginal polyps
Six of twenty-three colonies smaller than 2 cm in diameter (9 polyps) had only marginal polyps. Hence, the comparisons between marginal and non-marginal polyps were conducted for the 17 colonies that had both polyp types. Values for mean polyp volume, fertility, and mean NE/PV of non-marginal polyps were significantly greater than those of marginal polyps from the same colony (Wilcoxon paired-sample test, all P [less than] 0.01; Fig. 2A-C). The mode of budding was also significantly different between the two groups (Fisher exact test, P [less than] 0.0001). All the budding (total = 58) in non-marginal polyps was intratentacular, and 96% of the budding (total = 53) in marginal polyps was extratentacular.
Colony size, polyp size, and egg production by polyps
To examine the relationships among colony size, polyp size, and egg production, I grouped the colonies into three size classes: small (S; [less than or equal to]60 polyps, or [less than or equal to]4.5 cm in mean colony diameter), medium (M; 60-150 polyps, or 4.59 cm), and large (L; [greater than]150 polyps, or [greater than]9 cm). The classification was made because (1) below 60 polyps, NE/PV was extremely small; and (2) after 150 polyps, the fitted curve using the logistic equation between the colony size and NE/PV leveled off (Table II; Fig. 2C).
No eggs were found in colonies with marginal polyps only or with marginal polyps and one non-marginal polyp [TABULAR DATA FOR TABLE II OMITTED] ([less than] 13 polyps, or [less than]3 cm in colony diameter; n = 8). The smallest colony that produced eggs had 13 polyps ([approximately equal to] 3 cm colony diameter), two of which were non-marginal and the only polyps with eggs.
Table III ANOVA results for mean polyp volume and mean NE/PV (number of eggs per polyp volume) among the colony size classes (S = small, M =medium, L = large) Sum of Source df squares F P A. Mean volume of marginal polyps Colony size class (S, M, & L) 2 2.81 5.05 0.02 Colony [Colony size class] 19 5.28 2.00 0.01 B. Mean volume of non-marginal polyps Colony size class (S, M, & L) 2 1.62 7.14 0.007 Colony [Colony size class] 14 1.59 1.62 0.08 C. Mean NE/PV of marginal polyps Colony size class (M, & L) 1 0.01 0.50 0.5 Colony [Colony size class] 8 0.24 3.92 0.001 D. Mean NE/PV of non-marginal polyps Colony size class (M, & L) 1 0.34 6.37 0.04 Colony [Colony size class] 8 0.43 3.44 0.002 Note: Analyses for all dependent variables (Model III ANOVA) contain colonies nested within the colony size class. The S size class was excluded from the NE/PV analyses because of a severe violation of ANOVA assumptions. Logarithmic transformation was used in all the cases, and the assumptions for ANOVA were satisfied, excepting normality in D.
Model III nested ANOVAs showed significant among class differences in mean polyp volume for both marginal and non-marginal polyps (Table IIIA, B). Post-hoc tests (Scheffe test) showed that the S class colonies had smaller mean polyp volume than the M class colonies for both marginal and non-marginal polyps (Table IV; [ILLUSTRATION FOR FIGURE 2A OMITTED]). The S class colonies had smaller non-marginal polyp volume than the L size class colonies, but this was not true [TABULAR DATA FOR TABLE IV OMITTED] for marginal polyps. The M and the L class colonies did not differ in mean polyp volume for either marginal or non-marginal polyps.
Fertility [ILLUSTRATION FOR FIGURE 2B OMITTED] was significantly lower in the S size class than in the M and L size classes for both marginal and non-marginal polyps (Table IV; P [less than] 0.001; Kruskal-Wallis test). As with mean polyp volume, fertility did not differ significantly between the M and the L size classes (Table Iv). Fertility of non-marginal polyps was 100% in all the colonies that were larger than 84 polyps. The complete fertility of the non-marginal polyps was not due to large polyp volume. Although the volume was significantly larger in non-marginal polyps than in marginal polyps within a colony, small non-marginal polyps were fertile, whereas some of marginal polyps within the same volume range were not [ILLUSTRATION FOR FIGURE 3 OMITTED]. The overall fertility of these marginal polyps was 66.7% (n = 57), and polyp volume was positively associated with fertility.
ANOVA could not be performed for mean NE/PV among colonies in all the size classes, because the many 0 scores in mean NE/PV in the S size class (Fig. 2C) was a severe violation of ANOVA assumptions. However, mean NE/PV of both marginal and non-marginal polyps in the S size class was an order of magnitude smaller than those in the M and the L size classes (Table IV). I performed Model III nested ANOVAs excluding colonies in the S size class. Mean NE/PV of marginal polyps was not significantly different between the L and the M size classes (Table IIIC); in contrast, the difference was significant for non-marginal polyps (Table IIID).
Some colonies in the S size class produced eggs, but their colony fecundity was less than 60 eggs. In contrast, the fecundity of colonies in the M and the L size classes was more than 4500 eggs, and it increased linearly with colony size [ILLUSTRATION FOR FIGURE 4 OMITTED]. The X intercept of colony size vs. colony fecundity in the regression line for the M and L size class colonies was 58.3 polyps. Hence I considered the colony size for active maturation to be 60 polyps. Marginal polyps contributed less than 1% to the fecundity of colonies in the M size class. The contribution by marginal polyps decreased with colony size, and it was less than 0.1% in the L size class ([greater than]150 polyps).
The results suggest that the sexual maturation of a polyp is less a function of its size than of the mode of budding that produced it. In large colonies of mature G. aspera ([greater than]84 polyps), the non-marginal polyps, which showed intratentacular budding, always contained eggs regardless of polyp size, but less than 70% of the marginal polyps were fertile (Fig. 3). New polyps produced by intratentacular budding of mature non-marginal polyps appear to share the developmental stage of their parent polyps, whereas new small polyps produced by extratentacular budding are rarely fertile and may be regarded as developmentally "young" polyps. In the Caribbean massive faviid coral Montastrea annularis, not all the polyps in the central area are fertile, and polyp fertility is positively correlated with polyp size (Van Veghel and Kahmann, 1994), as in marginal polyps from the large colonies of G. aspera [ILLUSTRATION FOR FIGURE 3 OMITTED]. One of the generic characters of Montastrea is that it exhibits only extratentacular budding (Veron, 1986). So for this species, polyp size likely reflects the polyp age after extratentacular budding; i.e., each polyp has its own developmental clock. In contrast, new polyps produced by intratentacular budding in G. aspera are mature from inception. Thus the age of these polyps should also be measured from the time of original formation by extratentacular budding, not from the time of intratentacular budding.
Egg production of G. aspera was affected by colony size, polyp size, and budding mode. The colonies became actively reproductive at a size of 60 polyps, and the polyp volume of the reproductive colonies was significantly larger than that of the pre-reproductive colonies (Tables IIIA, IIIB, and IV). An effect of polyp age on egg production has been suggested by Kojis and Quinn (1985). They divided mature colonies of Goniastrea favulus into fragments smaller than the minimum size at maturity observed in naturally growing colonies, and found that some of the fragments produced gametes. In G. aspera, non-marginal polyps were formed in colonies with more than eight polyps, which then began intratentacular budding. All non-marginal polyps in colonies larger than 84 polyps were fertile, presumably because all the polyps have attained maturity upon extratentacular budding.
This study demonstrates that polyps within a mature colony of G. aspera play different roles according to their position, even though all the polyps in a colony share the same structure and are likely to be genetically identical and physiologically connected to each other. Significant differences in polyp volume, NE/PV, and mode of budding were detected between marginal and non-marginal polyps within a colony. Moreover, NE/PV of marginal polyps was relatively constant in mature colonies, whereas NE/PV of non-marginal polyps was significantly higher in the larger mature colonies ([greater than]150 polyps) than in smaller mature colonies (60-150 polyps; Tables IIIC, IIID and IV). This implies that marginal polyps function as a buffer for the non-marginal polyps, which produce the bulk of the gametes (Fig. 4), against external perturbations such as competition and grazing, which are likely to be heavier at the colony edge (Jackson, 1979; Hughes and Jackson, 1985; Chornesky and Peters, 1987; Soong and Lang, 1992). In addition, marginal polyps may also contribute to the expansion of the attachment area of the colony (Chomesky and Peters, 1987). In a massive coral colony, polyps in the central area contribute to colony growth, increasing the surface area of the colony. However, large colony size cannot be attained without expansion of the attachment area as well. As a colony becomes larger, its total fecundity becomes greater with an increase in the number of fecund, non-marginal polyps. Thus, the differences between the marginal and the non-marginal polyps play a role in promoting high colony fecundity.
I thank R. A. Kinzie III, E. F. Cox, and J. S. Stimson for comments on the manuscript. I also thank K. Karino, T. Kuwamura, Y. Nakashima, Y. Loya, and H. Yamashiro for comments on the earlier manuscripts, and E. Hirose for discussion on polyp age. This study was partly supported by grants from the Japanese Ministry of Education, Science, Sports and Culture. This is a contribution from Sesoko Station, University of the Ryukyus.
Babcock, R. C. 1991. Comparative demography of three species of scleractinian corals using age- and size-dependent classifications. Ecol. Monogr. 61: 225-244.
Chornesky, E. A., and E. C. Peters. 1987. Sexual reproduction and colony growth in the scleractinian coral Porites astreoides. Biol. Bull. 172: 161-177.
Hall, V. R., and T. P. Hughes. 1996. Reproductive strategies of modular organisms: comparative studies of reef-building corals. Ecology 77: 950-963.
Harriott, V. J. 1983. Reproductive ecology of four scleractinian species at Lizard island, Great Barrier Reef. Coral Reefs 2:9-18.
Harrison, P. L., and C. C. Wallace. 1990. Reproduction, dispersal and recruitment of scleractinian corals. Pp. 133-207 in Ecosystems of the World, Vol. 25. Coral Reefs, Z. Dubinsky, ed. Elsevier, Amsterdam.
Hayashibara, T., K. Shirnoike, T. Kimura, S. Hosaka, A. Heyward, P. Harrison, K. Kudo, and M. Omori. 1993. Patterns of coral spawning at Akajima Island, Okinawa Japan. Mar. Ecol. Prog. Ser. 101: 253-262.
Heyward, A., K. Yalnazato, T. Yeemin, and M. Minei. 1987. Sexual reproduction of corals in Okinawa. Galaxea 6:331-343.
Hughes, T. P., and J. B.C. Jackson. 1985. Population dynamics and life histories of foliaceous corals. Ecol. Monogr. 55:141-166.
Jackson, J. B.C. 1979. Morphological strategies of sessile animals. Pp. 499-555 in Biology and Systematics of Colonial Organisms, G. Larwood and B. R. Rosen, ed. Academic Press, New York.
Jackson, J. B.C., and A. G. Coates. 1986. Life cycles and evolution of clonal (modular) animals. Phil. Trans. R. Soc. Lond. B 313: 7-22.
Jackson, J. B. C., and T. P. Hughes. 1985. Adaptive strategies of coral-reef invertebrates. Am. Sci. 73: 265-274.
Kojis, B. L., and N.J. Quinn. 1985. Puberty in Goniastrea favulus: age or size limited? Proc. 5th Int. Coral Reef Congr., Tahiti 4: 289-293.
Meester, E. H., and R. P.M. Bak. 1995. Age-related deterioration of a physiological function in the branching coral Acropora palmata. Mar. Ecol. Prog. Ser. 121: 203-209.
Sakai, K. 1997. Gametogenesis, spawning, and planula brooding by the reef coral Goniastrea aspera (Scleractinia) in Okinawa, Japan. Mar. Ecol. Prog. Ser. 151: 67-72.
Sakai, K. 1998. Delayed maturation in the colonial coral Goniastrea aspera (Scleractinia): whole-colony mortality. colony growth and polyp egg production. Res. Popul. Ecol. (Kyoto) 40: (in press).
Soong, K., and J. L. Lang. 1992. Reproductive integration in reef corals. Biol. Bull. 183: 418-431.
Szmant, A. M. 1986. Reproductive ecology of Caribbean reef corals. Coral Reefs 4: 43-54.
Szmant, A. M. 1991. Sexual reproduction by the Caribbean reef corals Montastrea annularis and Montastrea cavernosa. Mar. Ecol. Prog. Ser. 74:13-25.
Van Veghel, M. L. J., and M. E. H. Kahmann. 1994. Reproductive characteristics of the polymorphic Caribbean reef building coral Montastrea annularis. II. Fecundity and colony structure. Mar. Ecol. Prog. Ser. 109: 221-227.
Veron, J. E. N. 1986. Corals of Australia and the Indo-Pacific. Angus & Robertson, North Ryde, Australia.
Zar, J. H. 1996. Biostatistical Analysis. 3rd ed. Prentice-Hall, Upper Saddle River, NJ.
|Printer friendly Cite/link Email Feedback|
|Publication:||The Biological Bulletin|
|Date:||Dec 1, 1998|
|Previous Article:||Reproduction-associated immunoreactive peptides in the nervous systems of prosobranch gastropods.|
|Next Article:||Late postembryonic development of the symbiotic light organ of Euprymna scolopes (cephalopoda: sepiolidae).|