| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 School of Marine Sciences, Darling Marine Center, University of Maine, Walpole, Maine 04573
2 Biology Department, Augustana College, Sioux Falls, South Dakota 57197
* To whom correspondence should be addressed. E-mail: aimeep{at}maine.edu
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
Fertilization will not be successful if the distance between spawning individuals is too great. Experimental studies have documented decreasing fertilization with distance from a sperm source in diverse taxa of marine free-spawners (Pennington, 1985; Yund, 1990, 1995; Levitan, 1991; Levitan et al., 1991, 1992; Brazeau and Lasker, 1992; Oliver and Babcock, 1992; Babcock et al., 1994; Benzie and Dixon, 1994; Benzie et al., 1994; Yund and McCartney, 1994; Levitan and Young, 1995; Lasker et al., 1996; Coma and Lasker, 1997a, b; Meidel and Yund, 2001; Metaxas et al., 2002). As a logical consequence of these distance effects, reproduction in low-density populations may be limited by fertilization when individuals are separated by distances greater than the dispersal distance of fertilizing sperm. Fertilization failure at low population density (a form of negative density-dependent population dynamics known as an Allee effect; Levitan et al., 1992) has been investigated in commercially harvested species and considered in the development of management programs (Jamieson, 1993; Quinn et al., 1993; Myers et al., 1995; Pfister and Bradbury, 1996; Liermann and Hilborn, 1997; Levitan and Sewell, 1998; Shelton and Healey, 1999; Frank and Brickman, 2000). Models, coupled with laboratory experiments on the effects of variation in sperm concentration on fertilization, have predicted declining fertilization levels with decreasing population size or density for free-spawners that broadcast their eggs (Levitan, 1991; Levitan and Young, 1995; Claereboudt, 1999; Metaxas et al., 2002). Field experiments with three egg-broadcasting species of sea urchins have supported the predicted trend (Levitan, 1991; Levitan et al., 1992; Wahle and Peckham, 1999). Density has also been predicted to have a greater effect on fertilization when populations are small (Levitan and Young, 1995), suggesting the need to simultaneously evaluate the effects of population size and density (Levitan et al., 1992).
To date, research on how the density and size of a population affects fertilization has focused on organisms that broadcast eggs, but variation in density may have very different consequences for free-spawners that fertilize eggs internally (brooders). In external fertilizers, both gametes are subject to dilution; in brooders, only sperm are diluted. If brooders are able to capture or concentrate dilute sperm, fertilization levels may be high even when individuals are far apart and few in number (Yund, 2000).
We examined the effects of population density on fertilization in Botryllus schlosseri (Pallas, 1766), a sessile, hermaphroditic, free-spawning colonial ascidian with internal fertilization. Sperm limitation has been demonstrated under experimental conditions in the field (Yund and McCartney, 1994; Yund, 1995), and populations are characterized by substantial spatial and temporal variation in population density (Yund and Stires, 2002). However, fertilization levels in natural spawns have not previously been reported, nor have density effects on fertilization been rigorously evaluated. Past manipulations (designed to test for sperm competition) have manipulated male density but not total population density (Yund and McCartney, 1994; Yund, 1995, 1998). We conducted a mix of manipulative field experiments and surveys of spawning in natural populations in the Damariscotta River estuary (Maine) during the summers of 1999 and 2000. Population size and density were manipulated in experimental populations to independently test the effects of these two factors. Density effects were also assessed in natural spawns by deploying laboratory-cultured colonies in populations that varied in density in space and time.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
Colonies of B. schlosseri on rocks and shells were collected from the Damariscotta River estuary, Maine, by divers and transported to the laboratory. Fragments of each colony were explanted onto microscope slides and maintained in a flowing seawater system at the University of Maine’s Darling Marine Center (Walpole, Maine). Animals cultured on glass can be examined under a microscope to assess egg production and stage in the reproductive cycle. Colonies for the manipulative experiment and the field survey were selected on the basis of sexual stage. Male-phase colonies (those releasing sperm) were deployed at about stage three (by the criteria of Milkman, 1967). Female-phase colonies (those with eggs ready to be fertilized) were deployed at late stage four or early stage five (Milkman, 1967), just prior to takeover by the new asexual generation of zooids. Females were recovered between stages 3 and 4 in the subsequent cycle (total cycle duration ranged from 7–10 days), prior to the release of brooded larvae. All embryos were then surgically removed from the zooids for enumeration. The percent of eggs fertilized was calculated as the number of larvae brooded upon return from the field divided by the number of eggs brooded prior to deployment, multiplied by 100. Data from colonies that were unhealthy, damaged, or dead upon return from the field were excluded.
Population size and density manipulation
We examined the potentially independent effects of population size and density on fertilization during the summer of 2000 by manipulating population size and colony spacing in experimental populations. Our experimental design employed four combinations of two population size treatments (4 and 16 colonies; Fig. 1) and two density treatments (equivalent to 13 and 157 colonies per square meter, calculated on the basis of the separation distances between the centers of the colonies). Population sizes in this experiment were smaller than many natural populations, but comparable to colony numbers on isolated pieces of hard substratum. Density treatments spanned most of the reported range for natural variation in this estuary (about 5 to 170 colonies per square meter; Yund and Stires, 2002), though for logistical reasons we were unable to explore the very lowest densities observed in nature. Experimental populations were assembled by mounting laboratory-cultured colonies growing on 2.5 x 7.6 cm glass microscope slides on 1.21-m–2 sheets of 0.6-cm-thick marine plywood that were in turn mounted on top of flat concrete patio blocks weighing 40 kg. Colonies were arranged so that male and female phases alternated as nearest neighbors (Fig. 1). Female colonies brooded an average of 59 eggs (±2.5 SE).
|
Four replicate trials were conducted of all four treatment combinations, with one replicate of each treatment performed simultaneously. Because each 16-colony (large population) array potentially yielded fertilization levels for four times as many females as the 4-colony (small population) arrays (8 vs. 2 females per trial), we conducted two extra trials of the small population size treatments to produce more equitable replication across treatments. Owing to the loss of data on some females that did not survive through their deployment, we obtained an average of 5.5 fertilization values per replicate trial for the large population size treatments, and 1.7 for the small population size treatments. The percent of eggs fertilized was arcsin-transformed to meet ANOVA assumptions, and then this dependent variable was analyzed by an ANOVA package that accommodates an unbalanced design (Statistica ’99, Statsoft, Inc.) with population size, population density, and trial as the three main effects. The data were initially analyzed with interaction effects included in the model. No interactions between any of the main effects were significant, so the analysis was repeated excluding interaction effects to maximize power for the main effects. We also performed a power analysis to assess our ability to detect differences among treatments.
Natural spawning populations
To assess fertilization levels in natural spawns, we deployed laboratory-cultured female-phase colonies in two field populations in the Damariscotta River during the summer of 1999. The two populations were located adjacent to the shores of small islands in the river. The seaward site was located at Carlisle Island (hereafter CI), while the landward site (DM) was at Glidden Ledge (see Yund and Stires, 2002, for a map and additional site information). The DM site experiences warmer temperatures during the summer, which decreases the duration of each reproductive cycle (Yund and Stires, 2002). Egg production is much higher at DM, apparently because of greater food availability (Stewart-Savage et al., 2001b). Both sites exhibit seasonal variation in population density, with peak densities occurring in August, but peak density is typically about seven times greater at DM than at CI (Yund and Stires, 2002).
To assess fertilization levels in the field, female-phase colonies growing on 2.5 x 7.6 cm microscope slides were placed individually in open-sided plastic slide boxes that were attached to a cinderblock-weighted rope supported by a surface buoy. Colonies were oriented upside down to reduce the effects of sedimentation, and were located approximately 0.5 m above the substratum. Three to eight (mean = 5.5 ± 0.3 SE) colonies were deployed at each site approximately weekly from early June to mid-August, and at site CI through the end of September. Colonies were positioned along a transect oriented parallel to the shore, and spaced about 7 m from their nearest neighbor.
Egg counts prior to deployment averaged 105 per colony (± 12 SE). All colonies remained in the field through the beginning of the next asexual generation (hereafter termed takeover) and the full period of egg viability, so that eggs were fertilized by naturally available sperm. Fertilization levels were determined as described above. Throughout most of the season, all embryos within a colony were at the same developmental stage. The presence of a mixture of developmental stages indicated that embryos lagging in development were fertilized later than the others (Meidel and Yund, 2001). Typically, all embryos surgically removed from a colony were at about stage four, and possessed a tail that wrapped completely around the embryo (Milkman, 1967). Embryos at a tail-bud stage or earlier (by the criteria of Milkman, 1967) were at least 24 h behind in development and therefore were reported as unlikely to complete development Stewart-Savage et al. 2001a).
Throughout the reproductive season, we surveyed the density of B. schlosseri colonies at both sites monthly. Density surveys, made with the aid of scuba, involved counting all of the colonies present in eight randomly deployed 0.64-m–2 quadrats. We did not measure the size distribution, reproductive status, or reproductive output of the colonies because past work (Stewart-Savage et al., 2001b; Yund and Stires, 2002) permits reasonable inference about variation in these variables between sites. A subset of the field fertilization data was used to assess the possible relationship with population density. Fertilization values obtained during three week-long intervals (including the weeks immediately before, during, and after the week of a population survey) were paired with the associated population density survey for analysis. Fertilization values obtained outside of these intervals could not be reliably associated with a density and so were excluded from this portion of the analysis. The possible relationship between fertilization and population density was assessed by regression. We also calculated the variance in fertilization at each density and used a variety of nonlinear regressions to explore possible relationships between variance in fertilization and population density.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
= 0.05.
|
|
170 colonies m–2; Yund and Stires, 2002). Nevertheless, peak density was more than an order of magnitude higher than density early in the season. Densities at the seaward site (CI) were generally lower than at DM and exhibited less variation during the season (Fig. 3).
|
Beginning at week 38, embryos removed from colonies that had been deployed at site CI varied greatly in developmental staging (in contrast to all earlier weeks when the embryos within a colony were all at the same developmental stage). Embryos at earlier developmental stages would have had insufficient time to complete development before the colony began a new asexual generation and the old zooids were resorbed (Stewart-Savage et al., 2001a). By week 40, although 66% of eggs were fertilized, we estimate that only 39% of eggs would have resulted in fully developed larvae (Fig. 3).
There was no relationship between fertilization level and natural population density (Fig. 4A); the regression of percent of eggs fertilized on the log of density was not significant (P = 0.250, r2 = 0.016). The power analysis indicated that a slope estimate of double the value obtained would have been necessary for significance at = 0.05. In contrast to the absence of a pattern in overall fertilization level, a nonlinear (exponentially declining) relationship did exist between the variance in fertilization and colony density (r2 = 0.7115, 0.010 < P < 0.025; Fig. 4B).
|
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
While overall fertilization levels did not change significantly with population density (Fig. 4A), the variance in fertilization in natural populations did increase at lower density (Fig. 4B), indicating that reduced population density may result in greater variation in fertilization among different individuals. The very high variance in fertilization at the lowest population density was heavily influenced by a single colony that exhibited exceptionally low fertilization (8.8%). However, such individual effects may be relevant to natural populations and should not be dismissed as outliers.
Because B. schlosseri is a colonial species, population density alone is an imperfect predictor of population-wide reproductive output. Juvenile colonies delay sexual reproduction until they achieve a minimum size (Boyd et al., 1986; Chadwick-Furman and Weissman, 1995). Reproductive output in sexually mature colonies is highly variable and subject to both genetic (Grosberg, 1988; Yund et al., 1997) and environmental (Newlon et al., 2003) influences. Nevertheless, temporal variation in density is a reasonable first approximation of average gamete production at the population level, because it indicates times of population growth and contraction. Unlike many other colonial invertebrates, B. schlosseri rarely exhibits fragmentation (Grosberg, 1982). Consequently, recruitment of sexually produced larvae is essential for population growth (Yund and Stires, 2002).
In surveys, sperm appeared to become somewhat more limiting at the very end of the sample period. At that time, the annual reproductive season was ending and populations were shrinking (Fig. 3; Yund and Stires, 2002), so any marginal fertilization effect is best viewed as a minor component of a larger trend in population dynamics. Overall fertilization values exhibited only a slight downward trend, but the timing of fertilization varied more substantially (Fig. 3). The presence of embryos at various developmental stages indicated that fertilization within each brood occurred over a wider time span than during the rest of the reproductive season. Embryos at an earlier developmental stage at the conclusion of our sample intervals would probably not have had sufficient time to complete development before the adult zooids were resorbed and the next asexual generation began (Stewart-Savage et al., 2001a). Therefore, delayed fertilizations are unlikely to have produced viable progeny, and so may represent a form of temporal sperm limitation. Sperm availability may have been reduced at this time of year from decreases in both population density and energy allocated to sperm production (Stewart-Savage et al., 2001a). A crash in the local phytoplankton population each autumn Incze et al., 1980) is consistently associated with smaller testes and lower egg production in B. schlosseri (unpubl. data).
Comparison with results from other taxa
The effects of population size and density on fertilization have been investigated in other free-spawning marine invertebrates. Field experiments with three species of sea urchins used eggs retained in Nitex mesh bags and either induced males to spawn or used sperm-filled syringes to simulate males. Increasing the number or density of spawning individuals increased fertilization in all three species (Levitan, 1991; Levitan et al., 1992; Wahle and Peckham, 1999). Ultimately, the differing consequences of population size and density for fertilization in sea urchins and ascidians can probably be explained by the different spawning strategies of these taxa. Sea urchins broadcast both sperm and eggs, so rapid gamete dilution may have a greater impact on sperm-egg encounters than in brooders, where egg retention prevents the dilution of female gametes. However, all three of the sea urchin studies cited above manipulated gametes so that eggs were held at fixed concentrations in sperm-permeable containers. Consequently, though egg dilution may be important in natural spawns, it played no role in the experimental results reported in these papers. Hence the proximate explanation for the difference between our results and those of the sea urchin studies must involve a process other than egg dilution.
As part of a reproductive strategy that involves retaining eggs internally, brooders usually possess some mechanism for capturing sperm. The method by which sperm enter a female-phase B. schlosseri colony is not known, but all indications are that this organism is exceedingly efficient at acquiring dilute, long-lived sperm from the water (Johnson and Yund, 2004). Additionally, fertilization in B. schlosseri is a time-integrated process with eggs viable to be fertilized for 24 h (Stewart-Savage et al., 2001a). Consequently, it seems probable that sperm are filtered out of the water as a by-product of feeding activity, which is likely to involve the processing of a relatively large volume of water. If so, sperm would in effect be concentrated, limiting the impact of sperm dilution. Thus increased efficiency of sperm collection coupled with time-integrated fertilization, rather than reduced egg dilution, may explain the absence of density effects on fertilization in B. schlosseri.
Density effects on fertilization have also been explored in the internally fertilizing Queen conch, Strombus gigas, which does not free-spawn, but instead transfers sperm by copulation. At very low adult densities (<100/ha or <0.01/m2), reproductive success was found to be density dependent in S. gigas (Stoner and Ray-Culp, 2000). However, above a critical density, evidence for Allee effects in S. gigas dissipated as the frequency of observed reproductive behavior plateaued. By analogy, sperm limitation is to be expected in B. schlosseri at some very low population density. However, that density condition does not appear to occur in the Damariscotta River estuary.
Although previous studies of fertilization in egg-brooding free-spawners have not directly addressed the effects of population density, a comparison of reported efficiencies of sperm capture is nonetheless illuminating. Both the colonial ascidian Diplosoma listerianum and the bryozoan Celleporella hyalina achieve maximum fertilization at sperm concentrations on the order of 102 ml–1, in contrast to the 104–105 ml–1 concentrations required for fertilization in sea urchins (Pemberton et al., 2003). When sperm from a single male-phase D. listerianum colony were diluted in a 3840-1 tank, the male was nevertheless able to sire abundant progeny with 20 female-phase colonies (Bishop, 1998). Consequently, these two brooding species appear likely to have ecological fertilization dynamics similar to those of B. schlosseri. In marked contrast to these results, female colonies of the brooding octocoral Briareum asbestinum were severely sperm limited when placed only 5 m away from a male, and reproductive success was positively correlated with male density (Brazeau and Lasker, 1992). Though based on a very limited number of brooding species, the comparison between the octocoral and the bryozoans and asidians hints at another general principle. Bryozoans, like ascidians, are active suspension feeders who use feeding structures to filter sperm or sperm packets from the seawater (Temkin, 1994, 1996), whereas octocorals are passive suspension feeders. Although the mechanism by which sperm gain access to brooded eggs is unknown in octocorals, this taxon lacks a feeding mechanism that could be co-opted for sperm capture. So while active suspension-feeding brooders may be largely immune to sperm limitation in nature, passive suspension-feeding brooders may be among the most sperm-limited of marine invertebrates (Yund, 2000).
Comparison with previous results for B. schlosseri
It is useful to view our results within the context of past fertilization studies on B. schlosseri. A series of papers on sperm competition focused on relative male reproductive success, but also incidentally quantified levels of egg fertilization. While Atkinson and Yund (1996) found no significant difference in the proportion of eggs fertilized in combinations of high and low population density and size, three other studies (Yund and McCartney, 1994; Yund 1995, 1998) did report increased fertilizations with increased male density. If male density (past studies) has more of an effect than total population density (this study) on fertilization, fertilization in B. schlosseri may be more sensitive to the male:female ratio than to absolute density. At the level of individual gametes (i.e., the cellular level), this pattern in turn suggests that fertilization levels may be dictated more by sperm:egg ratios than by the absolute sperm concentration. Yund (1998) also used a rare biochemical marker for paternity determination and showed that nearby males monopolize fertilizations as long as the quantity of sperm they produce is greater than some threshold level. When local sperm production is below that threshold, eggs are fertilized nevertheless, but the sperm come from more distant sources. This result helps explain how eggs can be fertilized under a broad range of density conditions.
Maximum field fertilization levels
Fertilization levels for B. schlosseri were generally quite high, but even in the most dense conditions still averaged less than 100%. Fertilization levels in colonies placed in natural populations averaged 85.6% throughout the entire sampling season (Fig. 3); levels in the experimental populations were also very close to 85% (Fig. 1). Although it is tempting to interpret these data as evidence of a low level of sperm limitation, another explanation is more likely. Reproductive success below 100% may have been the result not of unsuccessful fertilization, but of unsuccessful development of embryos because of outbreeding depression (Grosberg, 1987). Due to the philopatric dispersal of larvae (that is, dispersal that keeps the larvae near their site of origin), B. schlosseri colonies typically live in kin groups (Grosberg, 1987, 1991; Yund and O’Neil, 2000), and consequently mate with relatives (Grosberg, 1987, 1991). Because the normal mating system involves inbreeding, this species may be subject to outbreeding depression (Grosberg, 1987). Our experimental manipulation assembled populations that lacked genetic structure, and our survey introduced randomly selected genotypes into natural populations. Thus, all fertilizations were the product of out-crossed matings. Grosberg (1987) has previously reported that outcrossing reduces the success of both fertilization and subsequent embryo development. The cumulative effect of outbreeding depression (through larval hatching) that he predicted is consistent with the 15% failure that we observed. Because we assayed fertilization by successful development, our approach would not have distinguished early developmental failures from fertilizations that did not occur.
Alternatively, recent ecological work on the occurrence of polyspermy in marine invertebrates (e.g., Franke et al., 2002) suggests a different explanation for our observed 85% fertilization maximum. All of the eggs in our colonies may have been fertilized, but some may have been fertilized by more than one sperm. Because embryos resulting from polyspermic fertilization would have failed to develop, and we assayed fertilization in terms of successful development, our results could also have incorporated the effects of polyspermy. Although we are convinced that polyspermy plays an important role in the fertilization dynamics of many marine free-spawners (Yund, 2000), we are skeptical of this explanation for our own results. All evidence suggests that successful fertilization in B. schlosseri tends to be a time-integrated process in which dilute sperm are slowly acquired from the water. If this scenario is valid, then eggs are unlikely to be subject to the high short-term sperm concentrations necessary to cause polyspermy. Furthermore, effective polyspermy blocks are present in ascidians (e.g., Lambert et al., 1997). Even when eggs of B. schlosseri are subjected in laboratory experiments to sperm concentrations much higher than those found in nature, we see no evidence of polyspermy (unpubl. data). In addition, if polyspermy had been an important factor in our experiments, fertilization levels (as assayed by development) should have declined at higher population densities.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Current Address: Marine Science Center, University of New England, Biddeford, ME 04005. 
|
Literature Cited |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. A. Pechenik, J. S. Pearse, and P.-Y. Qian Effects of Salinity on Spawning and Early Development of the Tube-Building Polychaete Hydroides elegans in Hong Kong: Not Just the Sperm's Fault? Biol. Bull., April 1, 2007; 212(2): 151 - 160. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. D. Bishop and A. J. Pemberton The third way: spermcast mating in sessile marine invertebrates Integr. Comp. Biol., August 1, 2006; 46(4): 398 - 406. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. R. Lasker High fertilization success in a surface-brooding Caribbean gorgonian. Biol. Bull., February 1, 2006; 210(1): 10 - 17. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. L. Johnson and P. O. Yund Remarkable Longevity of Dilute Sperm in a Free-Spawning Colonial Ascidian Biol. Bull., June 1, 2004; 206(3): 144 - 151. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
