Referene: Biol. Bull. 218: 189 199. (April 1) 1 Marine Biologial Laboratory Understanding the Effets of Low Salinity on Fertilization Suess and Early Development in the Sand Dollar Ehinarahnius parma JONATHAN D. ALLEN 1 * AND JAN A. PECHENIK 2 1 Department of Biology, Bowdoin College, Brunswik, Maine 411; and 2 Department of Biology, Tufts University, Medford, Massahusetts 2155 Abstrat. Free-spawning marine invertebrates that live near shore or in estuaries may experiene redued fertilization suess during low-salinity events. Although several studies have doumented reprodutive failure at redued salinity in estuarine animals, few have looked at whether developmental failure is due to a failure of fertilization or to a failure of fertilized eggs to leave. In this study, we examined the effets of salinities ranging from 18 to 32 psu on fertilization suess and early development in the sand dollar Ehinarahnius parma. In addition to deoupling the effets of low salinity on fertilization from its effets on early leavage, we also assessed whether eggs or sperm were the weak link in aounting for reprodutive failure. We found that both fertilization and leavage failed at salinities below about 22 psu but that development ould be partially resued by returning zygotes to full-strength seawater. We also found that sperm remained ative and apable of fertilizing eggs even after being exposed to low salinities for 3 min.. Taken together, these results suggest that reprodutive failure at low salinities in E. parma is due more to an inability of the fertilized eggs to leave than to an inability of sperm to fertilize eggs. Introdution Shallow-water or estuarine speies exhibiting external fertilization may be espeially suseptible to low-salinity stress. Fertilization and early development of the polyhaete Nereis virens, for example, is rarely suessful at salinities below about 22 psu (Ushakova and Saranthova, 4). Reeived August 9; aepted 4 February 1. * To whom orrespondene should be addressed, at Department of Biology, College of William and Mary, Williamsburg, VA 23185. E-mail: jdallen@wm.edu Although some attention has been paid to the role of salinity variation in determining reprodutive suess, surprisingly little attention has been paid to how those effets are mediated, despite onsiderable interest in many other aspets of fertilization eology over the past 25 years (e.g., Pennington, 1985; Levitan and Young, 1995; Podolsky and Strathmann, 1996; Yund, ; Podolsky, 2; Johnson and Yund, 8). There is a substantial literature on the effets of salinity on gamete longevity and fertilization suess for marine algae (e.g., Brawley, 1992; Serrao et al., 1996; Steen, 4), but surprisingly little omparable information for marine animals (but for examples of renewed interest in this topi, see Bekova et al., 4; Kashenko, 7; and Pehenik et al., 7). While several studies have onsidered the effets of low salinity on larval development and metamorphosis (e.g., Anger et al., 1998; Metaxas, 1998; Qiu et al., 2), few have looked at the effets of low salinity on fertilization and leavage. In one reent study, Pehenik et al. (7) showed that in the free-spawning polyhaete Hydroides elegans, poor development below salinities of about psu was due more to an inability of eggs to leave than to an inability of sperm to suessfully fertilize the eggs. However, beause polyhaete eggs do not form a onspiuous envelope upon fertilization, there was no diret way to assess fertilization suess; fertilization suess had to be inferred from leavage. Ehinoderms seemed a promising alternative for suh studies beause their eggs typially form fertilization envelopes within 3 s of fertilization (Pearse and Cameron, 1991), allowing the effets of low salinity on fertilization to be distinguished from those on leavage. One of the few researhers to speifially investigate the effet of low salinity on both fertilization and leavage was the remarkable E. E. Just (Byrnes and Ekberg, 6). 189
19 J. D. ALLEN AND J. A. PECHENIK Working with the ehinoid Ehinarahnius parma, Just (1923, p. 19) onluded that the failure to leave is due to the ation of dilute sea-water in interfering with the leavage mehanism not with fertilization, sine eggs inseminated in dilutions of sea-water may separate membranes [i.e., form a fertilization envelope] though they do not leave. However, Just never speified the salinities he worked with. Instead his data were presented as a funtion of perent dilution of sea-water, with no report of the salinity of undiluted seawater or the atual data on perent leavage. Moreover, he used tap water to make his dilutions, so some of the effets observed ould have been at least partly aused by ontaminants. In this study we set out to repeat and extend key elements of Just s work on E. parma. In partiular, we define the salinities at whih fertilization envelopes form and at whih leavage ours, and by transferring well-rinsed eggs to different salinities after their mixture with sperm, we more learly differentiate the effets of low salinity on fertilization and leavage. We also doument the effet of redued salinity on the timing of fertilization and leavage, and examine for the first time the tolerane of sperm of this speies to low salinities, assessing the relative sensitivity of the two types of gametes to low-salinity stress. Materials and Methods General Adult speimens of the sand dollar Ehinarahnius parma (Lamark, 1816) were olleted from the low intertidal zone at Cedar Beah, Harpswell, Maine (43 44 37 N; 69 59 11 W) and transported to the Bowdoin College marine laboratory (Orr s Island, ME), where they were housed in flow-through seawater aquaria at ambient temperature. The natural spawning season for E. parma along the southern Maine oast is June, July, and August, and all experiments were arried out during these months. All experiments were onduted at room temperature, about 22 C, and for eah experiment gametes were obtained from 3 4 males and 2 3 females by intraoelomi injetion of about 1 ml of.5 mol l 1 KCl. Eggs from eah female were ombined and well mixed prior to being subsampled for use in eah experiment. All seawater was filtered to.45 m and oxygenated by vigorous agitation before use. All dilutions were made by adding reverse-osmosis filtered water. Exept where otherwise indiated, all fertilizations were arried out using sperm onentrations of approximately 1 1 6 sperm ml 1, a hoie based on results presented below for fertilization suess at different sperm onentrations; and all experiments were onduted using 4-ml seawater suspensions in glass bowls, with three repliates per treatment unless otherwise noted. Eggs were examined for fertilization envelopes and leavage, using ompound mirosopes at a magnifiation of 4. The elevation of a fertilization envelope was taken to indiate fertilization suess, following standard protool (e.g., Greenwood and Bennett, 1981; Ringwood, 1992; Levitan, 2). We did not measure the ph of the seawater used in this study; however, the seawater at Bowdoin s marine laboratory typially ranges from 7.8 to 8.3, and a previous study (Pehenik et al., 7) found that seawater ph was not affeted by dilution with distilled water over a range of 1 to 35 psu. Salinity was diretly measured at the olletion site on two oasions in the summer of 9 by using a YSI 3 handheld ondutivity meter. Salinity measurements of nearby surfae waters were also olleted from the GOMOOS (Gulf of Maine Oean Observing System) buoy in lower Harpswell sound, about 2 km to the northwest of our olletion site (43 45 46 N; 69 59 16 W). Fertilization suess at different sperm onentrations Nearly dry sperm were obtained from three males and mixed together. A - l sample of the ombined sperm suspension was diluted in 1 ml of filtered seawater. A 1-ml subsample of this suspension (1%) was withdrawn, and one drop of Lugol s iodine was added to stain the sperm. Sperm onentration was then determined with a hemaytometer at a magnifiation of 1. The onentration of the 1% suspension was 1.4 1 6 sperm ml 1. Additional onentrations of 5%, 25%, 12.5%, 6.25%, and 3.12% were then made by serial dilution of the 1% suspension, and 4 ml of eah suspension was poured into eah of three repliate glass bowls. Eggs were obtained from two females and held at approximately 8 eggs ml 1 before use. A.5-ml subsample of the egg suspension was added to eah bowl of sperm suspension. The samples were examined for fertilization envelopes after about 3 1 min, and examined for leavage after about 2 2.5 h. Fertilization and development at onstant salinities This experiment was onduted to determine the range of salinities over whih eggs ould be fertilized and ould leave. In a pilot study, the following salinities were tested: 32, 3, 25,, 15, and 1 psu. Based on the results of that study, the following salinities were tested in a subsequent experiment: 32, 26, 24, 22,, and 18 psu. A.5-ml subsample of egg suspension was first added to eah dish (4 ml of seawater in eah dish), the appropriate (small) volume of sperm suspension was added within 1 min, and the suspension was then mixed gently by swirling eah dish. The presene of fertilization envelopes was assessed about 3 min later, and the extent of leavage was determined about 2 h after that.
SALINITY EFFECTS ON FERTILIZATION 191 Time ourse of leavage at onstant salinities Eggs from the above experiment at 32 psu were subsampled at about 3-min intervals to reord hanges in the frequeny of 1-ell, 2-ell, and 4-ell stages. A separate experiment was later onduted to determine the effets of salinity on the timing of fertilization and early leavage; in that experiment, eggs were examined for fertilization envelopes and leavage at intervals for up to 1 h. The following salinities were tested: 32, 26, 24, 22,, 18, and 16 psu. Effets of low salinity on leavage Eggs were fertilized at full salinity (32 psu) and allowed to sit for about min to allow fertilization envelopes to harden. The eggs were then arefully retained on a 5- m mesh filter that was then gently agitated in three suessive beakers of full-strength filtered seawater to remove exess sperm. The rinsed eggs were transferred to dishes of lowsalinity seawater and subsampled after about 3 6 min to assess fertilization and leavage. A suspension of unfertilized eggs was added to three samples of the final rinse water, and these eggs were heked for the presene of fertilization envelopes about 3 min later to test for the effetiveness of sperm removal. Additional eggs (with sperm added) remained at the initial full-strength salinity for the entire proedure (three repliates) to serve as a ontrol, and eggs to whih no sperm were added (three repliates) served as an additional ontrol, to be ertain that no eggs were inadvertently fertilized during olletion. Resue of eggs fertilized at low salinity The goal of this part of the study was to determine whether eggs mixed with sperm at low salinity ould be resued by transferring them to high-salinity seawater well before the time that first leavage would normally our. Conduting this experiment required that we be able to remove all non-adhering sperm before the transfer, so that there would be no free sperm available to fertilize eggs in full-strength seawater after the transfer. Gametes were mixed at eah of the salinities tested previously (18 32 psu) in about 1 liter of solution. After 5 min, during whih all eggs settled to the bottom of the ontainer, most of the water was drained from eah beaker by reverse filtration: a siphon tube was plaed inside a mesh-bottomed beaker so that only water passing through the mesh was removed from the ontainer, leaving the eggs behind. An additional liter of seawater at the same test salinity was then added to eah beaker and the filtration repeated. Another liter of water at the test salinity was added and filtered; then, finally, a liter of full-strength seawater was added to eah beaker and was also mostly drained off. Unfertilized eggs were then added to three subsamples of that rinse water to assess the effetiveness of sperm removal. The experimental eggs were examined after about 3 min to assess fertilization and again 2 3 h later to assess leavage suess. Additional samples (three repliates) remained the entire time at the original test salinities (inluding full-strength seawater) with sperm, as ontrols. Additional samples (three repliates) of eggs were never deliberately exposed to sperm and were subsequently monitored for fertilization and leavage, to ontrol for the possibility that eggs might have been inadvertently exposed to sperm during olletion. Assessing the salinity tolerane of sperm Two sperm-tolerane experiments were onduted. In the first, 5 l of sperm suspension was added to dishes of seawater at either 18 or psu (three repliates at eah salinity). Egg suspension (3 l) was then added to eah dish after, 5, 1, 15, or 3 min. The eggs were examined about 1 min later to look for bound sperm and sperm swimming ativity. In the seond experiment, 5 l of onentrated sperm suspension was added to 1 ml of seawater at 18 or psu; sperm remained in that low-salinity solution for, 5, 1, 15, or 3 min. Forty ml of full-strength (32 psu) seawater was then added to eah bowl, and.5 ml of egg suspension was added immediately afterward. Three repliates were onduted for eah treatment. Eggs were examined for fertilization envelopes about 15 min after salinity was restored to about 32 psu. Salinity stress ontrol experiment One ontrol experiment was onduted to test the possibility that sudden exposure to low salinity would itself initiate development of a fertilization envelope or ell division. A.5-ml volume of egg suspension (approximately 35 eggs) was added to eah of three dishes ontaining 4 ml of seawater at either 18 or psu. Three additional dishes of egg suspension at 32 psu served as ontrols. The eggs were examined at 1 after 1 and after 3.5 h for the presene of fertilization envelopes and multi-ellularity. Statistial analysis Where appropriate, one-way ANOVA was used to ompare mean differenes aross treatments. Perentage data were arsin transformed before analysis. When there was zero variane within a treatment (i.e., % fertilization aross all repliates) those data were exluded from the analysis to meet assumptions of the ANOVA. When signifiant treatment effets were found, post ho tests were onduted for all pairwise omparisons using Student s t tests with Bonferroni s orretion. All tests were onduted using SPSS ver. 17..
192 J. D. ALLEN AND J. A. PECHENIK 1 Mean perent fertilized 8 6 4 * 2 4 6 8 1 12 14 Sperm ml -1 (x 1 5 ) Figure 1. The influene of sperm onentration on average fertilization suess in Ehinarahnius parma. Between 15 and 3 eggs were generally subsampled per repliate. Error bars indiate one SD about the mean. The * indiates the one mean that is signifiantly different from the others (post ho test, P.5). Results Sea surfae (2-m depth) salinity at the lower Harpswell buoy ranged between 22 and 3 psu during the summer of 8, and between 15 and 3 psu in the summer of 9. In August 9, diret measurements of salinity over the sand dollar bed ranged between 21 and 32 psu. No eggs showed fertilization envelopes or leavage without the deliberate addition of sperm, demonstrating that eggs were not inadvertently ontaminated with sperm during olletion and handling and that exposure to low salinity did not in itself initiate development. Moreover, fewer than 5% of eggs were ever fertilized when added to the final rinse water after transfer of eggs between salinities, showing that the washing tehniques were effetive in removing most sperm. At normal salinities of 32 psu, sand dollar eggs formed onspiuous fertilization envelopes over a wide range of sperm onentrations, from about 4.2 1 4 sperm ml 1 to the highest onentration tested, 1.4 1 6 sperm ml 1 (Fig. 1). More than 6% of eggs were fertilized at the lowest onentration tested (about 4.2 1 4 sperm ml 1 ). As expeted, there was a signifiant effet of sperm onentration on fertilization suess (ANOVA; F 5,12 7.5; P.2), with the lowest sperm onentration resulting in signifiantly lower fertilization ompared with that obtained at all higher sperm onentrations (post ho test, P.5). All further studies were onduted using sperm onentrations of about 1 1 6 sperm ml 1. In a preliminary study, salinity had a signifiant effet on fertilization suess (ANOVA; F 4,1 1421.4; P.1) and leavage suess (ANOVA; F 3,8 124.6; P.1). Fertilization and leavage rates were signifiantly higher at salinities of 25 psu and above than they were at psu and below (post ho test, P.5; Fig. 2). At 1 psu, the lowest salinity tested, no eggs leaved or formed fertilization envelopes. In more detailed studies onduted subsequently (Fig. 2), redued salinity again affeted the suess of both fertilization (ANOVA; F 4,1 15.3; P.1) and leavage (ANOVA; F 5,12 97.2; P.1). Both fertilization and leavage were signifiantly greater at salinities of 26 psu and above than at salinities of 24 psu and below (post ho test, P.5). In partiular, fewer than % of eggs showed fertilization envelopes or leavage at salinities of 22 psu or less. Fertilization envelopes rarely formed (generally 1% of the time) at salinities between 15 and 18 psu. When gametes were mixed at normal (32 psu) salinity, 9% of the eggs onsistently beame fertilized (Fig. 3a); of those eggs that beame fertilized, 8% underwent at least one leavage within 1.5 h of fertilization and nearly 1% exhibited at least one leavage over the next 1.5 h (Fig. 3b). Many eggs took longer to produe fertilization envelopes (Fig. 3a) and to leave (Fig. 3b) at intermediate salinities ( 26 psu and 22 26 psu, respetively). Even 8 1 h after insemination, redued salinity signifiantly lowered perent fertilization (ANOVA; F 6,14 19.1;
SALINITY EFFECTS ON FERTILIZATION 193 Mean perent fertilized 1 a) Fertilization envelopes ' ' 8 6 b' 4 b a' a a' a a' 5 1 15 25 3 35 Salinity (psu) Mean perent leaved 1 8 6 4 b) Cleavage a a a 5 1 15 25 3 35 a' a' Salinity (psu) a' b b' Figure 2. Influene of salinity on prodution of fertilization envelopes (a) and leavage (b) in Ehinarahnius parma. Results from 2 separate experiments are shown, with 3 repliates per treatment. In one experiment (open symbols), 44 115 eggs were examined per repliate. In the other experiment (losed symbols), 3 75 eggs were examined per repliate. Fertilization was assessed after 5 1 min; leavage was assessed about 2 h later. Error bars indiate one SD about the mean. Different letters above eah data point indiate signifiant differenes between means, with prime signs indiating the omparisons assoiated with the seond experiment (post ho test, P.5). Data for the 32 psu ontrol salinity are displaed slightly for both experiments to improve presentation larity. ' b b '
194 J. D. ALLEN AND J. A. PECHENIK Mean perent fertilized a) a 1 32 psu a 26 ab 24 8 b 22 6 18 16 4 dd 1 2 3 4 5 6 7 8 9 1 11 Hours after adding sperm Mean perent leaved 1 b) a 32 psu a 26 24 8 22 b 6 18 16 4 1 2 3 4 5 6 7 8 9 1 11 Hours after adding sperm Figure 3. Influene of salinity on the timing of prodution of fertilization envelopes (a) and leavage (b) in Ehinarahnius parma. Eah point is the mean of 3 repliates, with 42 116 eggs examined per repliate for (a) and 25 1 eggs examined per repliate for (b). Error bars indiate one SD about the mean. Letters to the right of eah line indiate signifiant differenes between means at the last time point measured (post ho test, P.5). P.1) for eggs held at 24 psu and below (post ho test, P.5). In addition, the perentage of eggs that had leaved 8 1 h post-insemination was signifiantly lowered (ANOVA; F 3,8 129.4; P.1) for eggs held at 24 psu and below (post ho test, P.5). We reorded no fertilization envelopes or leavage at the lowest salinities tested (18 and 16 psu) even 9 1 h after sperm were added. The perent of embryos leaving signifiantly inreased between the initial and final readings for salinities between 24 and 32 psu (t test, df 4, P.5); however, the
SALINITY EFFECTS ON FERTILIZATION 195 Mean perent leaved 1 8 6 4 Gametes mixed at 32 psu a ab b 5 1 15 25 3 35 Final salinity (psu) after transfer Figure 4. The influene of low salinity on leavage. Eggs and sperm of Ehinarahnius parma were mixed at 32 psu and then transferred to the salinities indiated. Cleavage was assessed about 9 min later. Eah point is the mean ( 1 SD) of 3 repliates, with 3 57 eggs examined per repliate. Letters above eah data point indiate signifiant differenes between means (post ho test, P.5). perent leaving did not signifiantly inrease between the 3-h time reading and the final reading for any of the salinities (t test, df 4, P.11). For example, at 24 psu the perentage of eggs seen leaving inreased substantially between 1.5 and 3 h after gametes were mixed together (t 6.67, df 4, P.27), but did not signifiantly inrease further over the next 5 h (Fig. 3b). When eggs were fertilized at high salinity and later transferred to low salinity, a surprisingly high perentage of eggs (about 3% 35% on average) leaved at the lowest salinities tested, 18 and psu (Fig. 4). However, post-transfer salinity still had a signifiant depressive effet on leavage suess (ANOVA; F 5,11 13.997; P.1). Similarly, transferring eggs from salinities of 18 and psu to fullstrength seawater (32 psu) after they had been mixed with sperm for 15 min at low salinity resued many of the eggs (Fig. 5). For example, whereas fewer than 2% of eggs leaved when the gametes were left at 18 psu, more than % of eggs, on average, leaved after being transferred from 18 psu to 32 psu. Again, this effet of post-fertilization salinity transfer was signifiant (ANOVA; F 5,14 212.1; P.1). Very few of the eggs (only 2.26%.95%) added to the rinse water for this treatment beame fertilized. Sperm were remarkably tolerant of low salinity. They retained nearly their full apaity to fertilize eggs even after spending 3 min at 18 psu, a salinity that supported little leavage (Figs. 2b, 3b). In our intial sperm-tolerane experiment, sperm were observed to swim and bind to eggs at 18 and psu even after 3 min of exposure prior to egg introdution. However, the low salinities of this initial experiment prevented us from assaying fertilization or leavage. In our seond sperm-tolerane experiment, sperm ativity was assayed at 32 psu. Exposures of sperm to 18 psu for 5, 1, 15, and 3 min resulted in high levels of fertilization (mean S.D.; 94.9%.22%; 94.6% 1.43%; 94.51% 2.69%; and 91.5%.71%, respetively). There were no signifiant effets of the low-salinity inubation on the ability of sperm to subsequently fertilize eggs at full salinity (ANOVA; F 3,8 3.47; P.92). Disussion The lower limit for suessful fertilization and leavage in Ehinarahnius parma was about 24 psu. Fertilization ourred, but was delayed, at salinities of and 22 psu; however, subsequent leavage ourred in less than 1% of fertilized embryos. At salinities less than psu, no eggs produed fertilization envelopes or leaved, even by 1 h after insemination. Although few studies have expliitly examined the link between salinity and the ability to produe fertilization envelopes, a redution in fertilization suess between 18 and 24 psu does not appear to be unusual for ehinoderms. For example, in the asteroid Asterina petinifera, fertilization envelopes do not form at salinities below 18 psu (Kashenko, 6). Similarly, in the holothuroid Eupentata fraudatrix, psu marked the lower salinity limit for suessful fertilization (Kashenko, ). For the ehinoid Ehinoardium ordatum, the lowest salinity at whih fertilization envelopes were produed was 24 psu (Kashenko, 7). In the sand dollar Sapehinus mirabilis, fertilization envelopes were produed down to salinities of 22 psu, but embryos maintained at these low salinities did not survive to the blastula stage (Bekova et al., 4). Although ehinoderms are typially regarded as stenohaline, restrited to areas of undiluted seawater, some speies do our in estuarine habitats. The gametes of some of these estuarine speies an alimate to the altered salinity at whih the adult resides during gametogenesis (Hintz and Lawrene, 1994). However, the more typial response may be a lak of alimation to redued salinities, as has been shown for several ehinoid speies (Roller and Stikle, 1993, 1994). In our study we did not allow adults or gametes to alimate to hanges in salinity. Future studies may benefit from allowing adults to alimate to suh hanges prior to spawning or from induing animals to release gametes diretly into ontainers at the experimental salinity. The general inability of ehinoderm gametes to alimate or adapt to low salinities may partly explain why members of this phylum have yet to invade freshwater habitats. Adult ehinoderms also our in hypersaline environments up to 46 psu (Binyon, 1966), and we suggest that future work on the effets of elevated salinities on fertilization suess is warranted. Ehinoderms are not alone among marine organisms in displaying redued developmental suess at low salinities. In the polyhaetes Nereis virens and Hydroides elegans, normal fertilization and development do not usually our
196 J. D. ALLEN AND J. A. PECHENIK 1 Continuous exposure Transferred to 32 psu Rinse ontrol 32 psu a Mean perent leaved 8 6 4 b b b 5 1 15 25 3 35 Initial salinity (psu) Figure 5. Assessing the ability of fertilized eggs to reover when transferred to normal salinity after having been fertilized at low salinity. Gametes of Ehinarahnius parma were mixed at the salinities indiated and either transferred bak to seawater of the original low salinity (ontrols) or to full-strength seawater (32 psu). In the rinse, ontrol unfertilized eggs were plaed in water used to rinse exess sperm from eggs prior to transfer to full-strength seawater. Perent leavage was assessed about 2 h later. Eah point is the mean ( 1 SD) of 3 repliates, with about 3 1 eggs examined per repliate. Letters above eah data point indiate signifiant differenes between means (post ho test, P.5). below salinities of 18 psu and psu respetively (Ushakova and Saranthova, 4; Pehenik et al., 7). Developmental suess was also redued when gametes of marine fishes (e.g., the California grunion Matsumoto and Martin, 8) and estuarine fishes (e.g., Fundulus heterolitus Able and Parmer, 1988) were exposed to low-salinity seawater. While previous studies have shown that the eggs and embryos of E. parma (Just, 1922a, b) and some other invertebrates are sensitive to hanges in salinity, the reason for this sensitivity has remained unlear. One explanation is that a failure of development reflets a failure of fertilization, due to effets on sperm motility or binding ability. In ontrast, development may fail after the eggs have been fertilized but before leavage an our. In one of the few studies to attempt to tease out the auses of failed development at low salinities for marine animals, Pehenik et al. (7) found that at least the early stages of fertilization in the polyhaete Hydroides elegans ould be suessful at low salinities; as in the present study with E. parma, development was substantially resued when well-rinsed eggs were transferred from a sperm-egg mixture at low salinity to a substantially higher salinity. Pehenik et al. (7) did suggest, however, that sperm funtion was also redued at lower salinities, so that both gametes played a role in poor development under low-salinity onditions. In our study, the sperm of E. parma were very resilient to low-salinity stress and were able to fertilize eggs even after spending 3 min at 18 psu before being exposed to eggs at normal salinity. We also found that eggs ould be resued by transferring zygotes from low salinity to high salinity, indiating that at least the early steps of fertilization had ourred in some eggs at the lower salinity, and that many eggs fertilized at normal salinity were unable to leave after being transferred to water of redued salinity. These results suggest that, in E. parma, developmental failure ours either at the later steps of fertilization (e.g., pronulear fusion and egg ativation) or at the zygote stage (mitosis) rather than at the level of individual gametes. The resue effet ould be due to sperm that remained bound to the egg surfae and that then fertilized the egg upon the egg s return to full-strength seawater. However, this seems unlikely given that only 2.26% of eggs were fertilized in the rinse ontrol, but 15.23% ( psu) and 23.67% (18 psu) were fertilized in our two resue treatments. These differenes between the ontrol and the resue treatments mean that
SALINITY EFFECTS ON FERTILIZATION 197 between 12.98% and 21.41% of the eggs were resued at and 18 psu respetively. The fats that many fertilized eggs did not leave after being transferred to low-salinity seawater and that eggs ould be resued by transferring them from a sperm suspension at low salinity to full-strength seawater in the absene of sperm suggest that leavage (e.g., mitosis) is diretly impated by redued salinity. Further studies will be needed to determine how this effet is mediated. Redued osmoti pressure might affet mirotubule-stabilizing proteins (Sharp, 2) or the operation of the protein motors involved in separating the spindles during mitosis (Sharp et al., a, b; Rogers et al., 4), or might affet ytokinesis diretly (Gilbert, 3). In an eologial ontext, the ability of eggs to be fertilized at low salinity may be of value in nearshore environments with short-term osillations in salinity. If natural populations of sand dollar eggs are fertilized at low salinity and subsequently return to undiluted seawater (or possibly just less dilute seawater) through mixing of the water olumn, then it appears likely that subsequent development an proeed normally. Future studies should investigate whether this resue effet is present in other marine organisms. A handful of other studies have also examined the relative abilities of sperm and eggs to tolerate salinity stress. Greenwood and Bennett (1981) exposed eggs and sperm of the sea urhin Parehinus angulosus to low salinities prior to fertilization and found that subsequent fertilization suess at 34 psu was strongly affeted by prior low-salinity exposure, dropping below 5% fertilization at about psu. Greenwood and Bennett (1981) also found that sperm were more resilient than eggs to lowered salinities, retaining the ability to fertilize eggs down to 3 psu, whereas eggs lysed at salinities below 15 psu. Similarly, in marine fish, sperm were able to retain their motility and tolerate salinities as low as 1 psu (Billard, 1978). Sperm may be robust to hanges in salinity, but they are not immune to its effets. In an estuarine polyhaete, for example, sperm retained funtionality down to psu, but at 1 psu and below the arosome was destroyed and organelles were lost (Hsieh, 1997). In other speies, sperm may indeed be the gamete that limits fertilization suess at redued salinity. In herring, sperm motility was redued at low salinities, and this redued motility was impliated as a main fator in poor fertilization suess (Griffin et al., 1998). Similarly, in the fifteen-spined stiklebak, the longevity of sperm was signifiantly redued at salinities below 1 psu (Elofsson et al., 3). Additionally, exposure of eggs to low salinities may ontribute to inreased polyspermy beause the fast blok to polyspermy is sodium-dependent in many marine invertebrates (Jaffe, 198). Sine sodium is also required for proper formation of the fertilization envelope (the slow blok to polyspermy) and for embryogenesis in ehinoids (Shuel et al., 1982), some of the failure to leave may be diretly attributable to low onentrations of sodium ions. In marine algae, redued salinity has been shown to inrease polyspermy (Brawley, 1992) and to delay fertilization and leavage (Serrao et al., 1999). This is similar to our result that at intermediate salinities ( 26 psu) the appearane of fertilization envelopes and early leavage were delayed; the mehanism for this delay should be explored in future studies. To properly address the interations between salinity, polyspermy, and fertilization suess in this and other speies, future studies should be arried out over a wide range of sperm onentrations, as has been suggested by Marshall (6). Our finding that fertilization envelopes do not always rise at low salinities (e.g., 18 psu), even in suessfully fertilized eggs, was unantiipated. In one experiment, fewer than 2% of eggs, on average, evidened fertilization envelopes, and yet about 25% of those eggs leaved after their transfer to full-strength seawater (32 psu). Clearly, at least the early events of fertilization (e.g., sperm-egg binding) were suessful at low salinity. So, while it is ommon to judge fertilization suess by the elevation of a fertilization envelope (e.g., Greenwood and Bennett, 1981; Ringwood, 1992; Levitan, 2), this method may underestimate fertilization suess at low salinity, and perhaps also under some other stressful onditions suh as extreme polyspermy (Levitan, 2). Further studies will be needed to determine why some fertilized eggs failed to form fertilization envelopes at low salinities. For example, binding of sperm with the egg envelope may have failed to release internal alium stores (Swann and Jones, 2; Gilbert, 3), or ortial granules may have failed to bind to the egg envelope or to release their ontents (Swann and Jones, 2; Gilbert, 3), or the osmoti gradient aross the envelope at redued salinity may simply have been insuffiient to allow for suffiient inward diffusion of water (Gilbert, 3). Finally, dereased salinities in the marine realm are often oupled with inreases in temperature. For example, in reent deades sea surfae temperatures in the Gulf of Maine have been inreasing at the same time that salinities have been dereasing (Drinkwater et al., 9). It is possible that in nature the gametes of marine invertebrates experiene salinity and temperature stresses that at synergistially or antagonistially on development, as has been shown to our for zooplankton experiening multiple stressors (Folt et al., 1999). Aknowledgments Thanks to Olivia Ambrogio for suggesting one of the sperm tolerane protools and to Rahel Diker and Ashton Bune for help soring fertilization in the lab. Thanks also to Margaret Pizer for helpful omments on a draft of this manusript.
198 J. D. ALLEN AND J. A. PECHENIK Literature Cited Able, K. W., and R. E. Parmer. 1988. Salinity effets on fertilization suess and larval mortality of Fundulus heterolitus. Copeia 2: 345 35. Anger, K., E. Spivak, and T. Luppi. 1998. Effets of redued salinities on development and bioenergetis of early larval shore rab, Carinus maenas. J. Exp. Mar. Biol. Eol. 2: 287 34. Bekova, N. V., E. V. Zhuravel, and N. K. Khristoforova. 4. Effets of desalination and the detergent sodium dodeylsulphate on the early development of the sand dollar Saphehinus mirabilis. Russ. J. Mar. Biol. 3: 175 182. Billard, R. 1978. Changes in struture and fertilizing ability of marine and freshwater fish spermatozoa diluted in media of various salinities. Aquaulture 14: 187 198. Binyon, J. 1966. Salinity tolerane and ioni regulation. Pp. 359 377 in Physiology of Ehinodermata, R. A. Boolotian, ed. Intersiene, New York. Brawley, S. H. 1992. Fertilization in natural populations of the dioeious brown alga Fuus eranoides and the importane of the polyspermy blok. Mar. Biol. 113: 145 157. Byrnes, W. M., and W. R. Ekberg. 6. Ernest Everett Just (1883 1941). An early eologial developmental biologist. Dev. Biol. 296: 1 11. Drinkwater, K. E., F. Mueter, K. D. Friedlan, M. Taylor, G. L. Hunt, J. Hare, and W. Melle. 9. Reent limate foring and physial oeanographi hanges in Northern Hemisphere regions: a review and omparison of four marine eosystems. Prog. Oeanogr. 81: 1 28. Elofsson, H., K. Van Look, B. Borg, and I. Mayer. 3. Influene of salinity and ovarian fluid on sperm motility in the fifteen-spined stiklebak. J. Fish Biol. 63: 1429 1438. Folt, C. L., C. Y. Chen, M. V. Moore, and J. Burnaford. 1999. The effets of multiple stressors on freshwater and marine eosystems. Limnol. Oeanogr. 44: 864 877. Gilbert, S. 3. Developmental Biology, 7th ed. Sinauer Assoiates, Sunderland, MA. Greenwood, P. J., and T. Bennett. 1981. Some effets of temperaturesalinity ombinations on the early development of the sea urhin Parehinus angulosus (Leske) fertilization. J. Exp. Mar. Biol. Eol. 51: 119 131. Griffin, F. J., M. C. Pillai, C. A. Vines, J. Kaaria, T. Hibbard-Robbins, R. Yanagimahi, and G. N. Cherr. 1998. Effets of salinity on sperm motility, fertilization, and development in the Paifi herring, Clupea pallasi. Biol. Bull. 194: 25 35. Hintz, J. L., and J. M. Lawrene. 1994. Alimation of gametes to redued salinity prior to spawning in Luidia lathrata (Ehinodermata: Asteroidea). Mar. Biol. 1: 443 446. Hsieh, H. L. 1997. Self-fertilization: a potential fertilization mode in an estuarine polyhaete. Mar. Eol. Prog. Ser. 147: 143 148. Jaffe, L. A. 198. Eletrial polyspermy blok in sea urhins: niotine and low sodium experiments. Dev. Growth Differ. 22: 53 57. Johnson, S. L., and P. O. Yund. 8. Multiple paternity and subsequent fusion-rejetion interations in a kin-strutured population. Mar. Eol. Prog. Ser. 364: 129 134. Just, E. E. 1922a. Studies of ell division. I. The effet of dliute sea-water on the fertilized egg of Ehinarahnius parma during the leavage yle. Am. J. Physiol. 61: 55 515. Just, E. E. 1922b. The fertilization-reation in Ehinarahnius parma. V. The existene in the inseminated egg of a period of speial suseptibility to hypotoni sea-water. Am. J. Physiol. 61: 516 527. Just, E. E. 1923. The fertilization-reation in Ehinarahnius parma. VIII. Fertilization in dilute sea-water. Biol. Bull. 44: 17 21. Kashenko, S. D.. Combined effet of temperature and salinity on the development of the holothurian Eupentata fraudatrix. Russ. J. Mar. Biol. 26: 188 193. Kashenko, S. D. 6. The ombined effet of temperature and salinity on development of the sea star Asterina petinifera. Russ. J. Mar. Biol. 32: 37 44. Kashenko, S. D. 7. Adaptive responses of embryos and larvae of the heart-shaped sea urhin Ehinoardium ordatum to temperature and salinity hanges. Russ. J. Mar. Biol. 33: 381 39. Levitan, D. R. 2. Density-dependent seletion on gamete traits in three ongeneri sea urhins. Eology 83: 464 479. Levitan, D. R., and C. M. Young. 1995. Reprodutive suess in large populations: empirial measures and theoretial preditions of fertilization in the sea bisuit Clypeaster rosaeus. J. Exp. Mar. Biol. Eol. 19: 221 241. Marshall, D. J. 6. Reliably estimating the effet of toxiants on fertilization suess in marine broadast spawners. Mar. Pollut. Bull. 52: 734 738. Matsumoto, J. K., and K. L. M. Martin. 8. Lethal and sublethal effets of altered sand salinity on embryos of beah-spawning California Grunion. Copeia 2: 484 491. Metaxas, A. 1998. The effet of salinity on larval survival and development in the sea urhin Ehinomentra luunter. Invertebr. Reprod. Dev. 34: 323 33. Pearse, J. S., and R. A. Cameron. 1991. Ehinodermata: Ehinoidea. Pp. 513 662 in Reprodution of Marine Invertebrates, Vol. 6, A. C. Giese, J. S. Pearse, and V. B. Pearse, eds. Boxwood Press, Paifi Grove, CA. Pehenik, J. A., J. S. Pearse, and P.-Y. Qian. 7. Effets of salinity on spawning and early development of the tube-building polyhaete Hydroides elegans in Hong Kong: not just the sperm s fault? Biol. Bull. 212: 151 16. Pennington, J. T. 1985. The eology of fertilization of ehinoid eggs: the onsequenes of sperm dilution, adult aggregation, and synhronous spawning. Biol. Bull. 169: 417 43. Podolsky, R. D. 2. Fertilization eology of egg oats: physial versus hemial ontributions to fertilization suess of free-spawned eggs. J. Exp. Biol. 5: 1657 1668. Podolsky, R. D., and R. R. Strathmann. 1996. Evolution of egg size in free-spawners: onsequenes of the fertilization-feundity trade-off. Am. Nat. 148: 16 173. Qiu, J. W., R. Tremblay, and E. Bourget. 2. Ontogeneti hanges in hyposaline tolerane in the mussels Mytilus edulis and M. trossulus: impliations for distribution. Mar. Eol. Prog. Ser. 228: 143 152. Ringwood, A. H. 1992. Comparative sensitivity of gametes and early developmental stages of a sea urhin speies (Ehinometra mathaei) and a bivalve speies (Isognomon alifornium) during metal exposures. Arh. Environ. Contam. Toxiol. 22: 288 295. Rogers, G. C., S. L. Rogers, T. A. Shwimmer, S. C. Ems-MClung, C. E. Walzak, R. D. Vale, J. M. Sholey, and D. J. Sharp. 4. Two mitoti kinesins ooperate to drive sister hromatid separation during anaphase. Nature 427: 364 37. Roller, R. A., and W. B. Stikle. 1993. Effets of temperature and salinity alimation of adults on larval survival, physiology, and early development of Lytehinus variegatus. Mar. Biol. 116: 583 591. Roller, R. A., and W. B. Stikle. 1994. Effets of adult salinity alimation on larval survival and early development of Strongyloentrotus droebahiensis and Strongyloentrotus pallidus (Ehinodermata: Ehinoidea). Can. J. Zool. 72: 1931 1939. Shuel, H., R. Shuel, P. Dandekar, J. Boldt, and R. G. Summers. 1982. Sodium requirements in hardening of the fertilization envelope and embryoni development in sea urhins. Biol. Bull. 162: 2 213. Serrao, E. A., L. Kautsky, and S. H. Brawley. 1996. Distributional suess of the marine seaweed Fuus vesiulosus L. in the brakish Balti sea orrelates with osmoti apabilities of Balti gametes. Oeologia 17: 1 12.
SALINITY EFFECTS ON FERTILIZATION 199 Serrao, E. A., S. H. Brawley, J. Hedman, L. Kautsky, and G. Samuelsson. 1999. Reprodutive suess of Fuus vesiulosus (Phaeophyeae) in the Balti sea. J. Phyol. 35: 254 269. Sharp, D. J. 2. Cell division: MAST sails through mitosis. Curr. Biol. 12: R585 R587. Sharp, D. J., H. M. Brown, M. Kwon, G. C. Rogers, G. Holland, and J. M. Sholey. a. Funtional oordination of three mitoti motors in Drosophila embryos. Mol. Biol. Cell 11: 241 253. Sharp, D. J., G. C. Rogers, and J. M. Sholey. b. Mirotubule motors in mitosis. Nature 47: 41 47. Steen, H. 4. Effets of redued salinity on reprodution and germling development in Sargassum mutium (Phaeophyeae, Fuales). Eur. J. Phyol. 39: 293 299. Swann, K., and K. T. Jones. 2. Membrane events of egg ativation. Pp. 319 346 in Fertilization, D. M. Hardy, ed., Aademi Press, San Diego. Ushakova, O. O., and O. L. Saranthova. 4. The influene of salinity on fertilization and larval development of Nereis virens (Polyhaeta, Nereidae) from the White Sea. J. Exp. Mar. Biol. Eol. 31: 129 139. Yund, P. O.. How severe is sperm limitation in natural populations of marine free-spawners? Trends Eol. Evol. 15: 1 13.