Behavioral Ecology Advance Access originally published online on January 4, 2007
Behavioral Ecology 2007 18(2):410-419; doi:10.1093/beheco/arl098
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Parasitic spawning in sand gobies: an experimental assessment of nest-opening size, sneaker male cues, paternity, and filial cannibalism
Department of Zoology, Stockholm University, SE-106 91 Stockholm, Sweden
Address correspondence to O. Svensson, who is now at the Department of Biological Sciences, University of Hull, Hull, HU6 7RX, UK. E-mail: o.svensson{at}hull.ac.uk. C. Kvarnemo is now at the Department of Zoology, Göteborg University, Box 463, SE-40530 Gothenburg, Sweden.
Received 21 January 2005; revised 27 November 2006; accepted 28 November 2006.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Sneaking is common in nest-building fish with paternal care, but the role of nest-opening size in protecting against entry by sneaker males has never been tested before. Using the sand goby (Pomatoschistus minutus), a fish with exclusive paternal care, experimental manipulations of nest openings provided no support for the hypothesis that nest openings serve as physical or visual defense or that sneaker males prefer to parasitize nests with wide openings. Female mating preference was also not influenced by nest-opening size. However, female courtship behavior and visibility were important cues for sneaker males. Most sneak entries occurred when the nest holder was occupied with courtship, chasing another sneaker male or nest building. In half the cases of observed sneak entry, the sneaker male fertilized eggs, also when sneaking only occurred before spawning. Sneak entry and its duration were good estimates of stolen paternity, but neither sneak entries nor degree of fertilizations were correlated with filial cannibalistic behavior. Testes size did not explain parasitic spawning success in replicates with genetically determined sneaking. However, all sneaker males without breeding coloration had huge testes and small sperm duct glands, whereas nest-holding males had small testes and large sperm duct glands, and sneaker males with breeding coloration were intermediate.
Key words: alternative reproductive tactics, concealment, filial cannibalism, paternity assurance, sperm competition, strategies.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Nest building has traditionally been regarded as a naturally selected behavior, and the literature on choice of safe nest sites as well as the role of nests in protecting against adverse environmental conditions and predation is extensive (Sargent and Gebler 1980
; Hostache and Mol 1998; Hansell 2000; Östlund-Nilsson 2000). However, in both birds and fish, nest-building behavior is frequently used in courtship, and nest quality often correlates with both female choice and paternal care (e.g., Hoi et al. 1994; Moreno et al. 1994; Kvarnemo et al. 1998; Soler et al. 1998; Jones and Reynolds 1999a; Östlund-Nilsson and Holmlund 2003). Male condition has also been shown to influence nest building, such that males in good condition build more ornamented or well-concealed nests than males in poor condition (Kvarnemo et al. 1998; Barber et al. 2001).
Male–male competition may also influence nest structure. In fish, smaller males that are less likely to defend a nest successfully and attract a female often steal fertilizations from other males by "sneaking." Male aggression toward individuals that are recognized as competing fertile males is common in most taxa (Taborsky 1994; Birkhead and Møller 1998). A cryptic nest may be difficult to discover by males seeking parasitic spawnings, as suggested for the three-spined stickleback, Gasterosteus aculeatus (Sargent and Gebler 1980). Furthermore, some nests may be easier to protect against nest takeover or sneaker male intrusions (Oliveira et al. 2002), and nest-building ability might thus be sexually selected through male–male competition (Svensson and Kvarnemo 2003). For example, defense against nest takeovers may actually be the primary function of nest sealing in the hornbill, Bycanistes subcylindricus (Kalina 1989). In a premating experiment on the sand goby, Pomatoschistus minutus, males built the smallest nest openings when sneaker males were present, smaller than those built by males held alone or together with males of the common goby, Pomatoschistus microps, which may prey upon eggs and compete for nest sites but not for fertilizations. Males with visual access to other nest-building males also tended to build smaller entrances than males held alone or with P. microps, whereas males held with egg predators, Carcinas maenas (shore crab) and Hinia reticulata (netted dog whelk), built nests with openings not differing significantly from any other treatment (Svensson and Kvarnemo 2003). These results support the hypothesis that the small nest openings made by males in the presence of sneaker males may have been favored as a protection against sneaking, or possibly through female choice.
The sand goby, P. minutus (Pallas), is a small, short-lived marine teleost fish. Males build nests by covering empty mussel shells with sand or by using cavities amid rocks (Lehtonen and Lindström 2004). Male display includes erect fins and tail beats. Females show their spawning interest by blackened eyes and a characteristic bobbing movement. Males exhibit breeding coloration by developing a melanized, black-colored edge along the anal fin and the tail fin. Some small males do not develop this melanization, or only rudimentarily so, whereas large males in full breeding coloration show intense melanization and an iridescent blue band inside the black edge of the anal fin and a blue and black spot on the first dorsal fin.
In the study area, this species usually has only one breeding season, during which both sexes breed repeatedly in multiple brood cycles. The male alone tends the eggs until hatching (Forsgren 1999). During each brood cycle, a male typically receives eggs from several females (Jones et al. 2001). In the field, half of all nests contain eggs fertilized by a male other than the nest-holding male, and one-tenth of all eggs are fertilized parasitically (Jones et al. 2001). All types of males spawn parasitically; subdominants and dominants, nest holders with and without eggs, and nonnesting males with and without breeding coloration (Singer et al. 2006 and the present study). Dissections of P. minutus males in catches taken early and late in the season revealed no males with immature testes. Rather, small males with no or only rudimentary melanization were found to have huge testes (in absolute size much larger than those found in any of the males in full breeding coloration) but only small sperm duct glands (involved in mucus secretion) (Kvarnemo C, Svensson O, unpublished data). The same pattern is found in the grass goby, Zosterisessor ophiocephalus (Scaggiante et al. 1999), and the black goby, Gobius niger (Rasotto and Mazzoldi 2002). In a subset of males that were dissected immediately after being sacrificed, all males with and without melanization had mature sperm in their testes, which were activated when subjected to seawater (Kvarnemo C, Svensson O, unpublished data). We will hereafter refer to those males lacking breeding coloration as "sneaker morph" males. In both P. minutus and P. microps, subordinate males that enter the nest of a dominant male have been referred to as sneaker males, regardless of whether they are a sneaker morph male lacking melanization or not (Magnhagen 1999). The sneaker males in the first experiment of the present study were a combination of subordinate males (with melanization) and sneaker morph males (without melanization), whereas those in the second experiment were only sneaker morph males.
The aim of this study was to test experimentally whether the size of the nest opening functions as a physical or visual defense against parasitically spawning males as suggested in Svensson and Kvarnemo (2003). By observing all fish, we also noted which cues in addition to nest-opening size sneaker males may use to recognize a sneaking opportunity and whether a sneaker responds preferentially to female courtship. A second important aim of the study was to test whether observed sneaker male entry (which may reduce paternity assurance) and genetically assigned paternity correlate with rates of filial cannibalism, as has been observed in some species (Neff 2003; Manica 2004), but not in others (Svensson et al. 1998; Östlund-Nilsson 2002). Because there are reasons to believe that both the male and the nest are of high quality if the entrance is small (Kvarnemo et al. 1998; Jones and Reynolds 1999a), we also tested whether females prefer small nest openings over large ones and whether they court males with small nest openings more intensely than those with large openings. Finally, we measured the variation of nest openings in the field and looked at the relationship between breeding coloration, gonad size, and fertilization success of sneaker males.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
General experimental procedure
Experiment 1 (physical defense) was conducted at Tjärnö Marine Biological Laboratory (May–June 2000; 58°52'N, 11°10'E) and, due to logistic reasons, Experiment 2 (visual defense experiment) at Klubban Biological Station (May 2002; 58°15'N, 11°28'E). Both stations are on the West Coast of Sweden. Microsatellite DNA analyses were conducted in the DNA laboratory at the Department of Zoology, Stockholm University. Fish were caught with a hand trawl in bays nearby the stations and separated by sex kept in 130-l storage aquaria with a 2-cm layer of sand for at least 36 h before trials. All aquaria were, except where noted, provided with running natural seawater with a temperature of 10–13 °C. The fish were fed chopped mussel meat (Mytilus edulis), shrimp (Crangon crangon), and Alaska pollock (Theragra chalcogramma) daily during storage and every third day during the experiments.
Experiment 1: nest opening as a physical defense
In the field, we measured width and height of nest openings of natural nests using a ruler. We estimated the developmental status of the eggs in the nests and scored them as 1) newly laid (no eyes), 2) containing embryos with visible eyes, or 3) close to hatching (often starting to do so while we handled them). This was done relatively early in the breeding season (15–21 May) and was repeated later in the breeding season (9–14 June).
In the subsequent lab experiment, artificial nests were made out of PVC rings with a diameter of 7 cm, a height of 25 mm, and a 7 x 7–cm tile as roof. Each ring had either a tight opening (14 mm) or a wide opening (25 mm). The height of all nest openings was 14 mm (Figure 1A). The sizes of the openings were determined from the mean nest-opening size measured in a previous study with and without the presence of sneaker males, respectively (Svensson and Kvarnemo 2003), and fell within the natural distribution of nest sizes measured in the field (see Results).
|
The experiment was carried out in 20-l aquaria. Each aquarium contained 0.15 l of sand, one artificial nest, one large male (mean ± standard deviation [SD], 63 ± 4 mm total length), and a female in a net cage to encourage nest building. Due to the design of the artificial nest sites, the males were not able to change the shape of the nests or the width of the nest opening. The female and the cage were removed as soon as nest building started. At the start of the experiment, 2 smaller males (48 ± 2 mm) were added. They were designated as sneaker males only by being smaller than the nest-holding male. Two males were used to increase the amount of sneaking activity. One hour later, a female (62 ± 6 mm) was released and observations began. Continuous behavioral observations were made until spawning was complete. If the fish did not start to spawn within 6 h, observations were stopped and the fish were left to spawn unobserved. However, if the fish had started to spawn in the aquarium, or in an adjoining aquarium, observation continued beyond 6 h. Total observation time varied between 97 and 700 min (mean 341 ± 133 min). Replicates in which no spawning was observed were checked for eggs the next day and the day after. If there were no eggs after 2 days, the replicate was removed from the study and the observational data were not used. This generated 49 useful replicates of behavioral data (out of 96 undertaken). Behavioral data were recorded before and during spawning, and all frequencies were divided by total observation time, measured in minutes.
Behaviors included in statistical analyses were as follows: 1) Sneak entries: a sneaker male passed completely through the nest opening; 2) Sneak entry success: sneak entries divided by sneak attempts; 3) Sneak approaches: a sneaker male clearly swam toward the nest opening; 4) Female courted nest-holding male; 5) Female courted sneaker male; 6) Nest-holding male showed aggression toward sneaker males including attacks, chases, and occasional aggressive displays; 7) Nest-holding male courted female.
After spawning, all fish except the nest-holding male were removed, sacrificed (for future studies), and stored at –20 °C for later microsatellite DNA analysis. Egg areas were traced onto transparent plastic sheets, and egg-guarding males were left to tend their eggs for 9 days, whereupon egg areas were again traced onto transparent plastic sheets to determine egg losses. The absolute value of the area of egg loss was used as a measurement of filial cannibalism (although using the proportion of egg losses gave qualitatively similar results). Thereafter, eggs were sampled by removing strips from the left, middle, and right side of the clutch and were stored in dimethyl sulfoxide (DMSO) buffer (20% DMSO, 250 mM ethylenediaminetetraacetic acid, pH 7.5, saturated NaCl). The males were sacrificed and stored in –20 °C for later microsatellite DNA analysis.
Filial cannibalism data were calculated for a subset of the initial replicates (N = 31). Replicates not included consisted of 4 cases where we failed to mark eggs on the ring, 3 where males became sick or died during the experiment, 1 where the male was accidentally removed instead of the female, and 1 where the female laid only a few eggs. Finally, in 9 replicates, males ate their entire clutches. Three of the 4 males with incomplete egg area markings did not eat all their eggs, and thus these were included in the behavioral data (N = 43), but not in the initial clutch size comparisons with the full clutch cannibalism data (N = 40).
Paternity analyses using microsatellite DNA
If the fish were observed until spawning finished and no sneaking occurred, 24 randomly chosen embryos from each clutch were analyzed. Otherwise, 36 randomly chosen embryos from each clutch were analyzed. DNA was extracted with the Gloor et al. (1993) protocol (the incubation step for fin clips was doubled to 60 min, and samples were vortexed a few times during incubation). The samples were stored in –20 °C before amplification by polymerase chain reaction (PCR). Each 10 µl PCR contained 2 mM MgCl2, 1x Qiagen PCR buffer (MgCl2 excluded), 0.2 µl each primer, 0.8 mM each dNTP, 0.25 units Qiagen HotStarTaq, and 1 µl DNA. One primer from each locus was dye labeled in the commercial synthesis (Sigma-Proligo, St. Louis, MO). Cycling using a PTC-100 programmable thermal controller (MJ Research, Inc., Ramsy, MN) consisted of an initial denaturation of 2 min at 95 °C followed by 32 cycles of 1 min of 95 °C, 1 min of 60 °C, and 2 min of 72 °C. Cycling was followed by 15 min extension at 72 °C. Fragment length was measured with CEQ 8000 Genetic Analysis System. Each well contained 0.25 µl PCR product, 20 µl CEQ Sample Loading Solution, and 0.25 µl CEQ DNA Size Standard-400. The CEQ Fragment Analyzing Software was used to analyze fragment length. The highly variable locus Pmin05 (Jones et al. 2001) was examined for all individuals, and Pmin10 (Jones et al. 2001) was added to assign paternity in 6 clutches. We had successful PCR reactions in 35 of the 37 egg samples (49 minus 9 full clutch cannibalism and 3 missing). All males in replicates in which successful parasitic fertilization had occurred were checked for melanization and dissected (6 replicates, 18 males). Testes, sperm duct glands, and gutted bodies were dried in 60 °C for more than 24 h and weighed on a Cahn electrobalance to the closest 0.001 mg (testes and glands) or 0.01 mg (bodies).
Experiment 2: nest opening as a visual defense
In contrast to the first experiment, the nests in this experiment were simulated, natural nests. Males were not given access to the nests and were therefore not able to change the shape of the nests, nor were any fish allowed to spawn. Thirty-six aquaria (75 x 21 x 25 cm) were divided into 3 compartments by placing transparent screens 18 cm from each end (Figure 1B). All aquaria were provided with a 3-cm layer of fine sand. Each end compartment was provided with a flowerpot cut in half and covered with sand like a natural nest. One of the nest openings was randomly assigned to be 1.5 cm and the other 4 cm in width. By manipulating the nest in this way, only the nest openings differed, not the sand cover.
In the evening, a clear plastic vial (0.5 l) provided with sand and containing a large male in breeding coloration (mean ± SD, 57 ± 3 mm) and a mature female (52 ± 6 mm) was placed beside each nest (Figure 1B). The 2 males and the 2 females in each aquarium were matched for size, and the 2 females were matched for extension of the abdomen, which is a good indicator of relative fecundity. A net over the top of the vial and holes on the sides enabled water circulation. Fish were enclosed in vials to prevent males from trying to alter their nests. The vials also prevented the enclosed fish from interfering with the sneaker morph males. A sneaker morph male (43 ± 3 mm, each with only rudimentary melanization) was placed in the middle section at 20 h, and the light was turned off.
The next morning at 8 h, the light was turned on, the water turned off to avoid effects of nonrandom water currents (the level of dissolved oxygen, measured at the end of each replicate, was always close to saturation), and the fish were left to acclimatize for 1 h. The fish were then observed for 3 min every half hour for 9 observation bouts (1620 s). The location of the sneaker morph male in the middle compartment (left third, middle third, or right third; Figure 1B) and the time spent there was measured. If the male stayed in the middle section, he was considered not to show a preference for any side. In P. minutus, a similar observation protocol was used in assigning female preferences (e.g., Forsgren 1997). Sneaker males showing aberrant behavior, such as continuously swimming back and forth in the water column or digging down in the sand in the same place throughout the trial, were not included in the analysis. To measure the degree of concealment of the nest opening, we noted the location, color, and behavior of the sneaker morph male during the first observation, provided he visited either of the 2 side sections containing nests. In addition, we measured preferences by calculating the time spent in front of fish that exhibited more and less of each of the following traits: male display, female blackened eyes, and male or female visibility (as reflected by presence or absence of digging down behavior, in which individuals "hide" in the sand). Only replicates in which there was a difference between the fish on the left and right side of the aquarium were used. We excluded replicates where males did not display or both females were continuously visible. Thus, sample sizes varied in these comparisons between 8 and 28.
After the sneaker trial, the females in the vials were removed, the sneaker morph male was exchanged for a fresh, fecund female (58 ± 5 mm), the water was turned on, and the light was turned off. The same observation protocol was then used for measuring female preference the following day. All fish were reused in other experiments.
Statistical methods
In parametric tests when the data were not normally distributed, counts and areas were square root transformed, proportions were arcsine square root transformed, and weights (except testes) were log transformed. However, untransformed mean ± SD values are given in the text. Due to heteroscedasticity, in many cases, it was not possible to do partial correlation or analysis of covariance with observation time as covariate. Before statistical analyses were carried out, we decided which tests to do and how to divide the data set (Rice 1989) for the sequential Bonferroni test by the Dunn–Sidák method (Sokal and Rohlf 1995) at 
-levels 0.001, 0.01, 0.05, and 0.1. STATISTICA 5.5 software was used for the statistical calculations, and all tests were 2 tailed.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Experiment 1: nest opening as a physical defense
Nest-opening size in the field
Because there was no difference in the nest-opening size between May and June (width F1,48 = 0.92, P = 0.34, height F1,48 = 0.18, P = 0.67), data from nests measured in different months were pooled. Nest-opening width (mm) tended to differ among nests containing newly laid eggs (22.7 ± 9.1), those with visible eyes (20.3 ± 8.1), or those close to hatching (14.3 ± 3.4) (1-way analysis of variance [ANOVA]: F2,47 = 2.88, P =0.066). However, nest-opening height (mm) between nests in these categories did differ: eggs newly laid (12.3 ± 3.5), those with visible eyes (10.6 ± 2.4), or those close to hatching (9.3 ± 1.9) (ANOVA: F2,47 = 3.49, P = 0.039). Duncan's test showed a significant difference (P = 0.021) between nests with newly laid eggs and those with eggs close to hatching.
Effect of nest-opening size
Sneaker males successfully entered 7 of 21 nests with tight openings and 5 of 28 nests with wide openings (
2 = 1.55, P = 0.21). In several nests, sneakers entered more than once and sneak entry occurred a total of 26 times, with 15 entries in tight and 11 entries in wide openings. Of the 12 replicates in which sneaking was observed, 7 were observed until mating ended. In these, sneaking occurred during mating in 3 replicates, both before and during mating in 1 replicate and before mating in 3 replicates (Table 1). Nest-opening width had no effect on genetically assigned paternity or any measured behavior (Table 2). Similarly, nest-opening width did not affect the sneaker male's success rate scored as number of successful sneak entries divided by number of attempts (wide 0.46 ± 0.41, N = 7, tight 0.64 ± 0.37, N = 8; ANOVA: F1,13 = 0.78, P = 0.39). Moreover, neither nest-opening width had any effect on partial clutch cannibalism (egg area eaten, centimeter square, wide 5.6 ± 3.5, N = 18, tight 5.6 ± 6.6, N = 13; ANOVA: F1,29 = 0.41, P = 0.53) nor did it influence the incidence of full clutch cannibalism (6/20 tight, 4/23 wide, 2 = 0.59, P = 0.44). Consequently, because the 2 nest-opening treatments showed no significant difference in any of the variables measured, these data were pooled in further analyses.
|
|
Sneaker male behavior and success
Sneaker males were most often pale and slipped into the nest, usually while the nest-holding male was occupied with other activities, such as courting the female or chasing the other sneaker male (Table 1). The nest-holding male's aggression toward the sneaker males was positively correlated with the number of sneak approaches per minute, even when immediate responses to sneak attempts were excluded. Sneak approaches were significantly correlated with female courtship displays toward the nest-holding male (Table 3). Females showed no aggression toward the sneaker males, and nest-holding males were never seen courting them.
|
Paternity
In 6 out of 35 replicates, some eggs were fertilized by a sneaker male. All of them belonged to the group of 10 males in which successful sneak entry was observed and eggs survived (6/10, 0/25,
2 = 10.1, P < 0.001, sequential Bonferroni correction at = 0.01, P < 0.01). In 5 of the 6 replicates with genetically determined sneaking, one of the sneaker males was a sneaker morph male (i.e., rudimentary melanization and huge testes), whereas in the sixth, both the sneaker males showed breeding coloration. In each replicate with successful parasitic fertilization, only one of the 2 sneaker males fertilized eggs: 3 with and 3 without breeding coloration. All sneaker morph males had large testes and small sperm duct glands, whereas the opposite was true for the nest-holding males, and sneaker males with breeding coloration were intermediate (Table 4). Nevertheless, sneaker morph males were not more successful at fertilizing eggs parasitically, as the 3 successful sneaker morph males fertilized 12.9 ± 8.5% of eggs and the 3 successful sneaker males with breeding coloration fertilized 12.0 ± 6.4% of eggs (ANOVA, F1,4 < 0.01, P = 0.93).
|
Of the 4 replicates in which the sneaker male was present in the nest during spawning, 3 sneakers fertilized eggs, whereas of the 3 that entered before spawning but were absent during spawning itself, only one male still fertilized eggs (Table 1). The total time the sneaker male stayed in the nest was positively correlated with fertilization success (Spearman rank correlation, N = 7, Rs = 0.79, P = 0.039, sequential Bonferroni correction at
= 0.05, P < 0.05).
Filial cannibalism
Overall, 9 of 43 clutches (21%) were fully eaten, and 27 of 31 clutches (87%) were partially eaten. Males that experienced sneaking and were observed until the end of mating did not eat a significantly larger fraction of their eggs compared with males that did not experience sneaking. Neither were males that had in fact lost paternity to sneaker males more likely to eat their eggs (Table 5). Similarly, males that experienced sneaking did not eat all their eggs more often than those that did not experience sneaking (sneaked 0/7, not sneaked 2/17, 2 = 0.65, P = 0.42, all males observed until spawning finished). This was also true when including males that were known to have experienced sneakers but were not observed until spawning finished (partial clutch cannibalism: Table 5, full clutch cannibalism: sneaked 2/12, not sneaked 2/17, 2 = 0.14, P = 0.71).
|
Experiment 2: nest opening as a visual defense
The fish in the vials did not appear to react to the sneaker morph males in the choice compartment, whereas the sneaker morph males clearly observed the fish in the vials. The sneaker morph males had cryptic coloration rather than breeding coloration, and when they moved, they often, although not always, moved in a typical "sneaky" way, swimming close to the substrate but remaining vigilant. Of the males in the vials, 44% displayed toward the female and 94% were clearly visible and did not dig down into the sand of their vials. Of the females, 99% were visible. Across the trials and with all vials considered as independent units, some of the behaviors of the male and female within the same vial were correlated, whereas others were not (Spearman rank correlation, male visible vs. female visible N = 72, Rs = 0.11, P = 0.38; male visible vs. female eye color N = 72, Rs = –0.24, P = 0.039; male displaying vs. female visible N = 72, Rs = 0.01, P = 0.91; male display vs. female eye color N = 72, Rs = 0.23, P = 0.047).
In the first observation of each replicate, neither the sneaker morph males were found more often at nests with wider openings (Binomial test, Ntight = 15, Nwide = 11, P > 0.50) nor did they spend more time in front of these nests throughout the experiment (Table 6, sneaker morph male preference). Similarly, they showed no preference for the side of the aquarium where the male in the vial was visible or displayed most frequently. Rather, they showed a significant preference for the side of the aquarium where the female in the vial was visible most frequently (Table 6, sneaker morph male preference).
|
In the setup in which the females were allowed to choose between wide and tight nest openings, 50% of the males were observed displaying toward the females. The females spent most of their time in front of the wide nest opening, although not significantly so. In those cases in which one of the males in the aquarium was visible more often or courted more frequently than the other, females showed no preference (Table 6, female preference).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Nest-opening size
In the physical defense experiment, we found no genetic or behavioral support for our main hypothesis that nest-opening size is a physical defense against sneaker males. Because we previously observed a strong effect of sneaker males on nest-opening width constructed by nest-holding males (Svensson and Kvarnemo 2003
), this is quite intriguing. We expected sneaking into nests with wide nest openings to be easy and cheap. The wide nest openings were almost twice as wide as the tight nest openings, and the heads of P. minutus males blocked a large fraction of the tight openings, which was not the case for those with wide openings. Furthermore, sneaking males of a congenor, P. microps, are known to take nest-holding male size (Magnhagen 1994) and predation risk (Magnhagen 1995) into consideration in their decision to parasitize a nest. However, in our study, the sneaker males often entered when the male was occupied outside the nest, engaged in behaviors such as courtship, nest building, or chasing the other sneaker male, and this is a possible explanation for the lack of treatment effect. Similarly, in several species without paternal care, it has been reported that subdominant males spawn in a foreign territory while the territory owner is occupied with other tasks (Taborsky 1994). However, because only one sneaker male managed to spawn parasitically in each nest, sneaker males apparently sneaked independently of each other.
A nest may protect the offspring as well as the parents from being discovered and preyed upon (Lindström and Ranta 1992; Jones and Reynolds 1999a; Hansell 2000), and there is no doubt that the P. minutus nests' main function is protection against egg predators. An exposed egg mass would be devoured within minutes. Furthermore, demand for oxygenated water leads to larger nest openings (Lissåker et al. 2003). The increased demand for oxygen during egg development should lead to an increased size of the nest opening over time, which is the case in P. microps (Jones and Reynolds 1999a, 1999b). Our current field data show the opposite pattern, suggesting defense against egg predation to be more important during late development, which could be the product of increased reproductive value of the clutch and/or increased predation risk (detectability). Lissåker and Kvarnemo (2006) found that egg-guarding P. minutus decrease their nest openings when C. maenas are introduced, although they do not show this behavior before mating when no eggs are present (Svensson and Kvarnemo 2003). Furthermore, in P. microps, a small nest opening has been shown to function as concealment against C. maenas (Jones and Reynolds 1999a), although the presence of a C. maenas had no effect on the nest opening during care (Jones and Reynolds 1999b).
The P. minutus nest opening may have a concealing function against sneaker males, although in the present study there was no support for this hypothesis. The sneaker morph males in the visual defense experiment did not discriminate between nests with wide openings and nests with tight openings. Likewise, in the physical defense experiment, they did not perform more sneak approaches or attempts to sneak toward nests with wide openings. Similarly, in the Azorean rock-pool blenny, Parablennius sanguinolentus parvicornis, nests in the field with associated satellite males did not have larger openings than nests without satellites. On the other hand, all nests with 2 openings had an associate satellite male, whereas this was true only in one-third of the nests with one opening, arguably due to nests with 2 openings being easier to parasitize (Oliveira et al. 2002).
Could the nest still provide concealment, even though we did not observe a treatment effect? It is possible that the sneaker morph males in our visual defense study were equally aware of the existence of both nests because the enclosed males were not able to block the tight nest openings with their heads as they do in nature, thereby making them cryptic, at least to the human eye. Also, the concealing effect may be important only over longer distances. Nevertheless, males frequently leave their nests to chase away other males or predators, as well as to court females (Table 1 and Svensson et al. 2004). Thus, the setup of the visual defense experiment simulates a natural situation fairly well. Our results support the hypothesis that the behavior of the potentially spawning female may be much more important to a sneaker male than the nest appearance, at least at close distance. Even though it is unlikely that the sneaker morph males in the visual defense study posed a threat as nest competitors, the discussion above would also apply to situations involving potential nest takeovers, sensu Sargent and Gebler (1980).
Nest choice and sexually selected nest building
Hoelzer (1989) proposed that epigamic traits could evolve as a means to advertise nonheritable components of parental quality, and female choice for good fathers has been demonstrated in different taxa (Møller and Jennions 2001) including P. minutus (Forsgren 1997). Nest building in Pomatoschistus is affected by condition (Kvarnemo et al. 1998; Jackson et al. 2002) and dissolved oxygen (Lissåker et al. 2003), and by being costly, it fulfills part of the essential assumptions for sexually selected ornaments (Kotiaho 2001). Direct benefits by choosing well-covered nests include reduced adult predation by birds (Lindström and Ranta 1992) and reduced filial cannibalism by egg-tending males (Kvarnemo et al. 1998). Indeed, P. minutus females prefer to spawn in the nests with relatively more sand on top, although nest-opening width has no effect on choice (Svensson and Kvarnemo 2005). In the physical defense experiment in the present study, females did not court males with tight nest openings more often than males with wide nest openings, and in the visual defense experiment, females showed no preference for tight nest openings. Thus, nest cover seems to be a more reliable estimate of male and nest quality (Svensson and Kvarnemo 2005). Female choice for well-covered nests, which strongly covaries with nest-opening width (Svensson and Kvarnemo 2005), is an alternative explanation for why males build small nest openings in the presence of sneaking males (Svensson and Kvarnemo 2003).
Sneaking and genetically assigned paternity
The P. minutus sneaker males have cryptic coloration resembling slim females. However, because the nest-holding males are highly aggressive toward the sneaker males and never court them, they are obviously not considering them to be fertile females. In the physical defense experiment, the immediate responses (attacks and bites) were highly correlated with sneak attempts. Similar results have been found in P. minutus (Malavasi et al. 2001) and P. microps (Magnhagen 1998).
In the physical defense experiment, females sometimes courted sneaker males and never showed aggression toward them, raising the hypothesis that females might encourage sneaking (for a discussion of possible benefits, see Griffith et al. 2002). However, this did not appear to be the case as female courtship toward the sneaker male was not correlated with the number of sneak approaches. Consistent with this, a previous study showed that females do not spawn more (or less) often in a nest that has a confined sneaker morph male close to its entrance compared with a nest that does not (Svensson and Kvarnemo 2005). Likewise, because fertilization rate is close to 100% in P. minutus whether or not sneaker males are present, greater fertilization is an unlikely benefit to the female. Because P. minutus have a pelagic larval stage (low philopatry) and huge population sizes (outbreeding populations), inbreeding avoidance and genetically variable offspring are also unlikely benefits. Instead, female behavior toward the nest-holding male was correlated with the number of sneak approaches, whereas the behavior of the nest-holding male toward the female had no effect. Thus, the most important stimulus for sneaker males appeared to be female behavior. The visual defense experiment similarly showed that male behavior had no effect on sneaker males, whereas the number of times the females were visible had a strong and positive effect on sneaker male preference. Visibility of any fish (male and female combined) had no effect on sneaker males. Therefore, alternative explanations of sneaker male preference, such as preference for a larger number of fish are less likely. Female courtship may thus result in an increased sperm competition risk. This has also been shown in the peacock blenny, Salaria pavo (Gonçalves et al. 2003) and appears likely in P. sanguinolentus parvicornis (Oliveira et al. 2002).
In half the cases in which a sneaker was observed to enter the nest, parasitic fertilization was successful, and in no cases were eggs fertilized by sneakers unless they had also been observed to enter the nest. Furthermore, the time the sneaker male stayed in the nest was positively correlated with fertilization success. Therefore, sneak entry in combination with the duration of sneaking was a good estimate of paternity.
Sneaker males attempted and succeeded in entering nests both before and during mating. Importantly, our microsatellite DNA analyses showed that sneaker males could fertilize eggs even when entering the nest 23 min before spawning and leaving it 4 min later. This is possible because P. minutus sperm are motile in seawater for at least 2 h (Svensson O, Elofsson H, unpublished data). As in other species of Gobiidae, nest-holding P. minutus males attach sperm-containing mucus by rubbing the anal–urogenital area toward the ceiling of the nest (Svensson and Kvarnemo 2005). Sperm is then released from the mucus for several hours (Marconato et al. 1996; Ota et al. 1996; Scaggiante et al. 1999; Rasotto and Mazzoldi 2002). In the present study, not only the nest-guarding males but also the sneaker males were observed upside down in the nests, rubbing their anal–urogenital toward the ceiling, suggesting that also sneaker males may attach sperm-containing mucus. However, this needs further investigation because all sneaker morph males dissected had large testes but small sperm duct glands (involved in mucus secretion), whereas the opposite was true for the nest-holding males, and sneaker males with breeding coloration were intermediate.
In the present study, sneaking behavior or lost paternity did not affect any type of cannibalism. Similarly, an experimental study of the congenor P. microps, in which some males experienced sneak intrusions whereas others did not, showed no effect of observed sneaking on filial cannibalism, fanning, or nest defense (Svensson et al. 1998) nor did studies of less closely related species, the fifteen-spined stickleback Spinachia spinachia (Östlund-Nilsson 2002) and in the peacock wrasse, Symphodus tinca (van den Berghe et al. 1989). In birds, numerous studies have shown decreased care as an outcome of decreased paternity assurance, whereas others have found no effect. Theoretical models predict different outcomes of paternity on male care, depending on a number of underlying assumptions (Neff and Sherman 2002; Sheldon 2002). In fish, there are only 3 known examples of an effect of parasitic spawning on male care: the bluegill sunfish, Lepomis macrochirus (e.g., Neff and Gross 2001), the pumpkinseed sunfish, Lepomis gibbosus (Rios-Cardenas and Webster 2005), and the scissortail sergeant, Abudefduf saxatilis (Manica 2004). In P. minutus, increased partial cannibalism may not be expected because there is no correlation between clutch size and partial clutch cannibalism, and the average success of the sneaker males is typically low (Malavasi et al. 2001 and the present study). In contrast, full clutch cannibalism in which small clutches are consumed is clearly a threshold response (Manica 2002), and results by Okuda and Yanagisawa (1996a, 1996b) suggest that this threshold is not universal but depends upon the individual circumstances. In theory, sneaking could push a male with a small clutch over the threshold (Svensson et al. 1998), but the present study did not support this hypothesis.
To conclude, we found no support for the hypothesis that the size of the opening of the P. minutus nest is important in the protection against sneaker males or that sneaker males prefer to parasitize nests with wide openings. Therefore, it is unlikely that this was the primary function of the "extratight" nest openings built by nest-holding males in the presence of sneaker males (Svensson and Kvarnemo 2003). Rather, sneaking typically occurred when the male was occupied with courtship, chasing sneaker males, or nest building. In addition, small nest openings did not appear to function as cryptic concealment against sneaker males. However, concealment may still play a role in other conditions and at larger distances. In half of the cases when a sneak entry was observed, but in no cases when sneak entry was not observed, did sneaker males succeed in fertilizing eggs, and the time the sneaker male stayed in the nest was correlated with his fertilization success. However, filial cannibalism was not correlated with either successful sneak entries or sneak fertilizations. Further research on the nest and its adaptive significance as a visual and physical defense against both pre- and postejaculate male–male competition is needed.
|
ACKNOWLEDGEMENTS |
|---|
We thank Wesley Manson and Daniel Simonsson, who helped us with massive amounts of behavioral observations and in the field, Maria Lissåker who helped in the field and measured the males' testes, sperm duct gland, and body weights, Kristina Axelson, Lotta Laurent, and Patrik Lindberg, who provided additional help in the field, Anna Singer, who carried out parts of the microsatellite analyses, Minna Miettinen, who gave us invaluable help in the laboratory, and Bertil Borg, Sigal Balshine, Topi Lehtonen, Naomi Pierce, and anonymous referees for valuable comments on the manuscript. We are especially grateful to the helpful staff at Tjärnö and Klubban. Financial support was given by the Smitts Foundation (to O.S.), the Crafoord Foundation, and the Swedish Research Council (to C.K.).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Barber I, Nairn D, Huntingford FA. (2001) Nests as ornaments: revealing construction by male sticklebacks. Behav Ecol 12:390–396.
Birkhead TR and Møller AP. (1998) Sperm competition and sexual selection. (Academic Press, London).
Forsgren E. (1997) Female sand gobies prefer good fathers over dominant males. Proc R Soc Lond B Biol Sci 264:1283–1286.
Forsgren E. (1999) Sexual selection and sex roles in the sand goby. In Almada VC, Oliveira RF, Gonçalves EJ (Eds.). Behaviour and conservation of littoral fishes(ISPA, Lisboa (Portugal)) pp. 249–274.
Gloor GB, Preston CR, Johnsonschlitz DM, Nassif NA, Phillis RW, Benz WK, Robertson HM, Engels WR. (1993) Type-I repressors of P-element mobility. Genetics 135:81–95.[Abstract]
Gonçalves D, Oliveira RF, Korner K, Schlupp I. (2003) Intersexual copying by sneaker males of the peacock blenny. Anim Behav 65:355–361.[CrossRef]
Griffith SC, Owens IPF, Thuman KA. (2002) Extra pair paternity in birds: a review of interspecific variation and adaptive function. Mol Ecol 11:2195–2212.[CrossRef][Medline]
Hansell M. (2000) Bird nests and construction behaviour. (Cambridge university press, Cambridge (UK)).
Hoelzer GA. (1989) The good parent process of sexual selection. Anim Behav 38:1067–1078.[CrossRef][Web of Science]
Hoi H, Schleicher B, Valera F. (1994) Female mate choice and nest desertion in penduline tits, Remiz pendulinus—the importance of nest quality. Anim Behav 48:743–746.[CrossRef]
Hostache G and Mol JH. (1998) Reproductive biology of the neotropical armoured catfish Hoplosternum littorale (Siluriformes—Callichthyidae): a synthesis stressing the role of the floating bubble nest. Aquat Living Resour 11:173–185.[CrossRef]
Jackson AC, Rundle SD, Attrill MJ. (2002) Fitness consequences of prey depletion for the common goby Pomatoschistus microps. Mar Ecol Prog Ser 242:229–235.[CrossRef]
Jones AG, Walker D, Kvarnemo C, Lindström K, Avise JC. (2001) How cuckoldry can decrease the opportunity for sexual selection: data and theory from a genetic parentage analysis of the sand goby, Pomatoschistus minutus. Proc Natl Acad Sci USA 98:9151–9156.
Jones JC and Reynolds JD. (1999a) The influence of oxygen stress on female choice for male nest structure in the common goby. Anim Behav 57:189–196.[CrossRef][Web of Science][Medline]
Jones JC and Reynolds JD. (1999b) Oxygen and the trade-off between egg ventilation and brood protection in the common goby. Behaviour 136:819–832.[CrossRef][Web of Science]
Kalina J. (1989) Nest intruders, nest defense and foraging behavior in the black-and-white casqued hornbill Bycanistes subcylindricus. Ibis 131:567–571.[CrossRef][Web of Science]
Kotiaho JS. (2001) Costs of sexual traits: a mismatch between theoretical considerations and empirical evidence. Biol Rev 76:365–376.[Medline]
Kvarnemo C, Svensson O, Forsgren E. (1998) Parental behaviour in relation to food availability in the common goby. Anim Behav 56:1285–1290.[CrossRef][Web of Science][Medline]
Lehtonen T and Lindström K. (2004) Changes in sexual selection resulting from novel habitat use in the sand goby. Oikos 104:327–335.[CrossRef][Web of Science]
Lindström K and Ranta E. (1992) Predation by birds affects population-structure in breeding sand goby, Pomatoschistus minutus, males. Oikos 64:527–532.[CrossRef][Web of Science]
Lissåker M and Kvarnemo C. (2006) Ventilation or nest defence—parental care trade-offs in a fish with male care. Behav Ecol Sociobiol 60:864–873.[CrossRef][Web of Science]
Lissåker M, Kvarnemo C, Svensson O. (2003) Effects of a low oxygen environment on parental effort and filial cannibalism in the male sand goby, Pomatoschistus minutus. Behav Ecol 14:374–381.
Magnhagen C. (1994) Sneak or challenge—alternative spawning tactics in nonterritorial male common gobies. Anim Behav 47:1212–1215.[CrossRef]
Magnhagen C. (1995) Sneaking behavior and nest defense are affected by predation risk in the common goby. Anim Behav 50:1123–1128.[CrossRef][Web of Science]
Magnhagen C. (1998) Alternative reproductive tactics and courtship in the common goby. J Fish Biol 53:130–137.[CrossRef]
Magnhagen C. (1999) Alternative mating tactics, nest defence and courtship in gobies—just a matter of size? In Almada VC, Oliveira RF, Gonçalves EJ (Eds.). Behaviour and conservation of littoral fishes(ISPA, Lisboa (Portugal)) pp. 275–300.
Malavasi S, Lindström K, Sundström L. (2001) Behaviour and success of sneaker males in the sand goby, Pomatoschistus minutus. Acta Ethol 4:3–9.[CrossRef]
Manica A. (2002) Filial cannibalism in teleost fish. Biol Rev 77:261–277.[Medline]
Manica A. (2004) Parental fish change their cannibalistic behaviour in response to the cost-to-benefit ratio of parental care. Anim Behav 67:1015–1021.[CrossRef]
Marconato A, Rasotto MB, Mazzoldi C. (1996) On the mechanism of sperm release in three gobiid fishes (Teleostei: Gobiidae). Environ Biol Fishes 46:321–327.[CrossRef]
Møller AP and Jennions MD. (2001) How important are direct fitness benefits of sexual selection? Naturwissenschaften 88:401–415.[CrossRef][Web of Science][Medline]
Moreno J, Soler M, Møller AP, Linden M. (1994) The function of stone carrying in the black wheatear, Oenanthe leucura. Anim Behav 47:1297–1309.[CrossRef]
Neff BD. (2003) Paternity and condition affect cannibalistic behavior in nest-tending bluegill sunfish. Behav Ecol Sociobiol 54:377–384.[CrossRef][Web of Science]
Neff BD and Gross MR. (2001) Dynamic adjustment of parental care in response to perceived paternity. Proc R Soc Lond B Biol Sci 268:1559–1565.[Medline]
Neff BD and Sherman PW. (2002) Decision making and recognition mechanisms. Proc R Soc Lond B Biol Sci 269:1435–1441.[Medline]
Okuda N and Yanagisawa Y. (1996a) Filial cannibalism by mouthbrooding males of the cardinal fish, Apogon doederleini, in relation to their physical condition. Environ Biol Fishes 45:397–404.[CrossRef]
Okuda N and Yanagisawa Y. (1996b) Filial cannibalism in a paternal mouthbrooding fish in relation to mate availability. Anim Behav 52:307–314.[CrossRef][Web of Science]
Oliveira RF, Carvalho N, Miranda J, Gonçalves EJ, Grober M, Santos RS. (2002) The relationship between the presence of satellite males and nest-holders' mating success in the Azorean rock-pool blenny Parablennius sanguinolentus parvicornis. Ethology 108:223–235.[CrossRef][Web of Science]
Östlund-Nilsson S. (2000) Are nest characters of importance when choosing a male in the fifteen-spined stickleback (Spinachia spinachia)? Behav Ecol Sociobiol 48:229–235.[CrossRef][Web of Science]
Östlund-Nilsson S. (2002) Does paternity or paternal investment determine the level of paternal care and does female choice explain egg stealing in the fifteen-spined stickleback? Behav Ecol 13:188–192.
Östlund-Nilsson S and Holmlund M. (2003) The artistic three-spined stickleback (Gasterosteous aculeatus). Behav Ecol Sociobiol 53:214–220.[CrossRef][Web of Science]
Ota D, Marchesan M, Ferrero EA. (1996) Sperm release behaviour and fertilization in the grass goby. J Fish Biol 49:246–256.[CrossRef]
Rasotto MB and Mazzoldi C. (2002) Male traits associated with alternative reproductive tactics in Gobius niger. J Fish Biol 61:173–184.[CrossRef]
Rice WR. (1989) Analyzing tables of statistical tests. Evolution 43:223–225.[CrossRef][Web of Science]
Rios-Cardenas O and Webster AS. (2005) Paternity and paternal effort in the pumpkinseed sunfish. Behav Ecol 16:914–921.
Sargent RC and Gebler JB. (1980) Effects of nest site concealment on hatching success, reproductive success, and paternal care in the threespined stickleback Gasterosteus aculeatus. Behav Ecol Sociobiol 7:137–142.[CrossRef][Web of Science]
Scaggiante M, Mazzoldi C, Petersen CW, Rasotto MB. (1999) Sperm competition and mode of fertilization in the grass goby Zosterisessor ophiocephalus (Teleostei: Gobiidae). J Exp Zool 283:81–90.[CrossRef][Web of Science]
Sheldon BC. (2002) Relating paternity to paternal care. Philos Trans R Soc Lond B Biol Sci 357:341–350.
Singer A, Kvarnemo C, Lindström K, Svensson O. (2006) Genetic mating patterns studied in pools with manipulated nest site availability in two populations of Pomatoschistus minutus. J Evol Biol 19:1641–1650.[CrossRef][Web of Science][Medline]
Sokal RR and Rohlf FJ. (1995) Biometry. 3rd ed (W. H. Freeman and company, New York).
Soler JJ, Cuervo JJ, Møller AP, De Lope F. (1998) Nest building is a sexually selected behaviour in the barn swallow. Anim Behav 56:1435–1442.[CrossRef][Web of Science][Medline]
Svensson O and Kvarnemo C. (2003) Sexually selected nest-building—Pomatoschistus minutus males build smaller nest-openings in the presence of sneaker males. J Evol Biol 16:896–902.[CrossRef][Web of Science][Medline]
Svensson O and Kvarnemo C. (2005) The importance of sperm competition risk and nest appearance for male behavior and female choice in the sand goby, Pomatoschistus minutus. Behav Ecol 16:1042–1048.
Svensson O, Magnhagen C, Forsgren E, Kvarnemo C. (1998) Parental behaviour in relation to the occurrence of sneaking in the common goby. Anim Behav 56:175–179.[CrossRef][Web of Science][Medline]
Svensson O, Nyman A, Kvarnemo C. (2004) Costly courtship or dishonest display? Intensely displaying sand goby males have lower lipid content. J Fish Biol 64:1425–1429.[CrossRef][Web of Science]
Taborsky M. (1994) Sneakers, satellites, and helpers: parasitic and cooperative behavior in fish reproduction. Adv Study Behav 23:1–100.
van den Berghe EP, Wernerus F, Warner RR. (1989) Female choice and the mating cost of peripheral males. Anim Behav 38:875–884.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||