Egg Deposition by Atlantic Silverside, Menidia menidia: Substrate Utilization and Comparison of Natural and Altered Shoreline Type
- First Online:
- Cite this article as:
- Balouskus, R.G. & Targett, T.E. Estuaries and Coasts (2012) 35: 1100. doi:10.1007/s12237-012-9495-x
- 138 Views
Egg deposition by the intertidal spawning fish Atlantic silverside (Menidia menidia) was compared among six shoreline types (Spartina alterniflora, Phragmites australis, sandy beach, riprap, riprap-sill, and bulkhead) and various substrates. In spring 2010, M. menidia egg density was measured daily near Roosevelt Inlet, Delaware Bay, USA. Over 3,000,000 eggs were collected during 50 sampling days. Eggs were deposited at all six shoreline types, with >93 % of eggs collected from S. alterniflora shorelines. Choice of substrate for egg attachment was similar across shoreline types with >91 % of eggs collected from filaments of the green alga Enteromorpha spp., a disproportionately high utilization rate in comparison with Enteromorpha spp.'s relative coverage. This study demonstrates that S. alterniflora shoreline, in association with Enteromorpha spp., is the preferred spawning habitat for M. menidia and that hardened shorelines and shorelines inhabited by P. australis support substantially reduced egg densities.
The intertidal zone of estuaries serves as important spawning habitat for many species of fish (see review by DeMartini 1999). The physical structure of estuarine habitats, particularly the shore-zone, continues to be modified by human activity. During the past few decades, the area of intertidal fringing salt marsh, comprised of Spartina alterniflora (smooth cordgrass), has decreased in US Mid-Atlantic estuaries, while invasive Phragmites australis (common reed) (King et al. 2007) and several types of shoreline hardening including bulkhead, riprap, and riprap-sills have become more common (USGS 2011). Shoreline hardening has greatest effects on the zone of transition, including the intertidal and shallow subtidal zones, the most ecologically productive regions in estuaries (Toft et al. 2007; Bilkovic and Roggero 2008). Riprap and bulkhead structures have been shown to negatively impact, through reduced diversity, growth and abundance, a range of populations of fauna (primarily fishes and benthic invertebrates) living adjacent to them (Able et al. 1998; Seitz et al. 2006; Bilkovic and Roggero 2008; Pister 2009). Bulkhead and riprap structures alter the natural shape of the shoreline, remove undercut crevice habitat, change shore-zone wave dynamics, reduce shallow water habitat, and reduce or eliminate intertidal plant communities.
Disturbance of natural shorelines by shoreline hardening is one of the primary contributors to the expansion of a non-native genotype of P. australis along the Mid-Atlantic coast of the US (King et al. 2007). P. australis has led to a displacement of native macrophyte communities (such as S. alterniflora), degradation of wildlife habitat and alteration of ecosystem processes (Weinstein and Balletto 1999; Minchinton et al. 2006). P. australis-dominated marshes are utilized by estuary resident fishes for feeding and spawning (Able and Hagan 2003). However, due to increased elevation of established P. australis marshes, the length of time the intertidal marsh interior is submerged is reduced, resulting in a decrease in usage by estuarine fishes (Weinstein and Balletto 1999; Able and Hagan 2003; Able et al. 2003). Reduced access to intertidal area reduces viable locations for intertidal spawning.
Understanding how physical and biological structure of the shoreline and intertidal zone affect utilization by shore-zone biota for spawning habitat is important to further our understanding of the impacts of shoreline modification on estuarine systems. Hardened shorelines and invasive marsh plants represent an important spatial feature which influences spawning in the intertidal zone. Despite the magnitude of this issue, there is little information on the impact of shoreline habitat alteration effects on spawning by intertidally spawning fishes.
Menidia menidia is among the most abundant forage fish species in US Mid-Atlantic estuaries (De Sylva et al. 1962; Richards and Castagna 1970; Able and Fahay 2010). The importance of M. menidia as a food source for piscivores such as striped bass (Morone saxatilis), Atlantic mackerel (Scomber scombrus), bluefish (Pomatomus salatrix) and other fishes (Merriman 1941; Bayliff 1950; Bigelow and Schroeder 1953; Schaefer 1970) is well documented. Spawning of M. menidia occurs between April and July in the Mid-Atlantic (Middaugh et al. 1981). Spawning occurs in the intertidal zone, where fish lay demersal, filamentous, adhesive eggs at high tide (Kuntz and Radcliffe 1917; Hildebrand 1922; Wang 1974; Middaugh 1981). Just before high tide, schools of M. menidia swim along the shoreline until a suitable spawning substrate is encountered (Middaugh et al. 1981). The adhesive eggs are laid approximately 1.2 to 2.4 m above mean low water to reduce exposure to aquatic predators (Middaugh et al. 1981; Tewksbury and Conover 1987) and birds (Middaugh 1981). Numerous substrates for egg attachment have been noted in earlier studies, including: submerged vegetation (Bayliff 1950), eelgrass (Middaugh 1981), S. alterniflora (Middaugh et al. 1981), filamentous algae (Conover and Kynard 1984), sand (Wang 1974), and beach trash (Nichols 1908). Middaugh et al. (1981) reported S. alterniflora, detrital mats, and abandoned crab burrows to be the most common sites of M. menidia egg attachment in an estuary in South Carolina. In a Massachusetts estuary, M. menidia eggs were found to be attached only on filamentous algae (Pilayella littoralis and Enteromorpha spp.), despite the presence of many other substrates (Conover and Kynard 1984). M. menidia eggs have not been quantitatively collected for assessment of density across available egg attachment substrata.
The objectives of this study were to (1) compare egg deposition by M. menidia along six shoreline types (bulkhead, riprap, riprap-sill, sandy beach, P. australis and S. alterniflora), (2) compare substrate utilization for egg attachment, and (3) determine if a relationship exists among shoreline type and substrate type for egg deposition/attachment and therefore determine preferential spawning habitat for M. menidia in a US Mid-Atlantic estuary.
Materials and Methods
Bulkhead shorelines (constructed in 1974) consisted of the vertical bulkhead structure, constructed of wood or metal, depending on station. Riprap shorelines (constructed in 1974) were comprised of large rocky structure directly positioned along the shoreline extending from the low to high intertidal zone. The riprap-sill shoreline (constructed in 2009) was comprised of a riprap structure (composed of large rock) positioned low in the intertidal zone with S. alterniflora planted in the mid intertidal zone, landward of the riprap structure and Spartina patens (saltmeadow cordgrass) planted in the high intertidal zone. Plantings occurred at the time of construction; S. alterniflora had grown to 60 cm and occurred at about 75 % of average natural density, S. patens had grown to 30 cm and occurred in approximately natural density at the time of the present study. P. australis has been expanding its range within the sampling area and surrounding estuary for several decades (Bruce Campbell, personal communication). The subtidal zones at all shoreline types are characterized by generally featureless, slowly sloping, sandy to muddy substrate.
Substrate for egg attachment varied, in both variety and density, among shoreline types. All bulkhead stations were covered with large quantities of the green filamentous alga Enteromorpha spp. (most likely Enteromorpha intestinalis). Barnacle colonies and small amounts of the brown alga Fucus spp. occurred in the lower intertidal zone of most bulkhead stations. Most rock surfaces on riprap shorelines had extensive coverage by Enteromorpha spp., with Fucus spp. found slightly lower in the intertidal zone. The green alga Ulva lacuta was present in low densities in both the mid and lower intertidal zones of riprap stations. Small tufts of S. alterniflora were present at many riprap stations. Blue mussels (Mytilus edulus) were present in the lower intertidal area of many riprap stations. Detritus (primarily dead P. australis, loose Enteromorpha spp., and beach trash) was found at the high tide wrack lines most sampling days at all riprap stations. Riprap-sill station's substrates included Enteromorpha spp. on the rocks of the riprap structure found in the mid to lower intertidal zone as well as on the sparse, recently planted S. alterniflora in the mid to upper intertidal zone. Large quantities of detritus were present in the S. alterniflora dominated mid to upper intertidal. Slightly higher in the intertidal zone, recently planted S. patens was present at all riprap-sill stations. Sandy beach stations offered the least variety of substrata for M. menidia spawning; all were primarily featureless, with a few stations having sparse shoots of P. australis in the upper intertidal zone. Loose Enteromorpha spp., dead shoots of P. australis, and beach trash were commonly present on the sand in the high tide wrack line. P. australis stations included dense stands of P. australis with occasional shoots of S. alterniflora both of which had Enteromorpha spp. growing densely near the base of stems. Detritus was found at the high tide wrack line of all P. australis stations. Blue mussels occurred sparsely in the lower intertidal zone. Many tunnels from Atlantic marsh fiddler crab (Uca pugnax) were found in the peat below P. australis stands. S. alterniflora stations had nearly all substrate types found at other shoreline types. Enteromorpha spp. was abundant between S. alterniflora stems, and U. lacuta and Fucus spp. were dispersed throughout stations in low abundance. S. patens was found in the upper intertidal zone and dense blue mussel beds occurred in the lower intertidal zone of some S. alterniflora stations. Tunnels from U. pugnax were present in the intertidal zone of several S. alterniflora stations. Detritus was found at the high tide wrack line of all S. alterniflora stations. Although substrate availability differed greatly among shorelines and stations, Enteromorpha spp. was found, in some density, at all sampling stations.
Eggs were collected from shoreline types with a minimum length of 40 m (mean = 88 m, SE = 11 m). Eight sampling stations were established parallel with the shoreline at each of the six shoreline types (Fig. 1) (n = 48 sampling stations). Stations were positioned >4 m apart from one another and >4 m from the edge of that shoreline type. Stations were 3 m wide and extended perpendicular to the waterline from the mean low water mark to the highest high tide elevation of the sampling day (Fig. 1c).
Stations were sampled from April 14th–June 10th, 2010 (n = 120 for each shoreline type). M. menidia eggs were identified using morphological features and egg deposition location (Hardy 1978; Able and Fahay 1998). Three stations at each shoreline type were randomly selected each day, for a total of 18 stations sampled per day. Incubation time for eggs at the water temperatures of this study range from 10 to 28 days (Martin and Drewry 1978); no sampling station was sampled at a timeframe greater than 7 days during the study, making it unlikely that eggs were deposited and hatched in this interval. Sampling began at least 2 h after the first daytime high tide. Density of M. menidia eggs at each shoreline type was assessed using a 3 × 1/4 m PVC quadrat. The sample quadrat was placed at the high tide mark (noted by wrack and/or direct observation), parallel lengthwise to shore, at the selected station, and the enclosed area was visually examined for eggs. If no eggs were present, the quadrat was moved down in 1/4 m steps, toward the waterline, thus covering the entire elevation of the station until eggs were encountered or the waterline was reached. If eggs were encountered (eggs were frequently deposited in distinct <1/4 m wide bands parallel to shore), the quadrat was positioned such that it enclosed the area of the station with the densest abundance of eggs. All shoreline types were sampled consistently by objectively selecting the most representative area of high egg density within the sampling station for collections. Attempts at randomizing collection area by elevation resulted in detrimentally reduced egg collections. All eggs within the quadrat were collected by hand removal and placed in a jar with seawater labeled by the substrate the eggs were collected from. Due to the filaments of M. menidia eggs, substrate was often removed along with the eggs during collections. Removal of substrate was minimized through careful, deliberate collection of eggs and the use of scalpels to carefully remove plant stems or algae when required. Prior to egg collection, the two-dimensional areal percent coverage of Enteromorpha spp. (to 10 % intervals) was visually estimated within each quadrat. When walking in P. australis or S. alterniflora, care was taken to minimally disturb the marsh.
In the laboratory, eggs were placed in 95 % ethanol for 15 min until they became opaque. Eggs were enumerated, and the substrate they were attached to was noted. When excessively large numbers of eggs were collected, eggs were estimated using egg volume in group sizes of 200. Analysis of 30 estimated groups of 200 eggs proved this method to be within 8 % of the actual value.
Maxim iButton® thermochrons were positioned (using the previously described elevations) at low, mid, and upper intertidal elevations at bulkhead, riprap, sandy beach, P. australis, and S. alterniflora shorelines. Thermochrons collected atmospheric/water temperatures every 15 min from April 17th–June 18th, 2010.
Non-parametric analysis of variance tests (Kruskal–Wallis) were used to test the effect of shoreline type and substrate type on egg density. Non-parametric analysis of variance tests (Kruskal–Wallis) were used to test the effect of tidal height, moon phase, daily water temperature, percentage Enteromorpha spp. coverage, and days since stations had been last sampled on egg density. Predicted tidal height and moon phase data were obtained from Tides and Currents Pro © (v. 3.3). Non-parametric tests were used due to the large variation in egg density from day to day, shoreline to shoreline, and station to station. Transformations were not successful in normalizing egg density data.
Friedman's two-way analysis of variance was used to test differences in intertidal atmospheric/water temperature among shoreline types over time. A stepwise step-down multiple comparison post-hoc test, adopted from a method from Campbell and Skillings (1985), was performed to determine homogeneous subsets in each non-parametric analysis. Statistical analyses were performed using SPSS (v. 18.0.2).
Egg Deposition by Shoreline Type
Menidia menidia eggs collected and mean density (eggs/m2) by shoreline type collected near Roosevelt Inlet, Delaware Bay, during spring 2010. Significant differences denoted by superscript letters (p < 0.05)
Mean eggs (eggs per m2 per day) with SE
Percentage of total eggs collected
32,468 ± 10,400
1,046 ± 1,003
553 ± 196
516 ± 238
28 ± 14
0.04 ± 0.04
Egg Attachment by Substrate Type
Menidia menidia eggs collected by substrate type for egg attachment: total and percentage at all shoreline types and percentage by respective shoreline type
Substrate utilization percentage by shoreline type
Percentage of total eggs
All eggs collected from bulkhead and beach shorelines were deposited on Enteromorpha spp. (Table 2). Utilization of Enteromorpha spp. for egg attachment was also very high at riprap (96 % of eggs), riprap-sill (98 %), P. australis (99 %) and S. alterniflora (91%) shorelines, where egg densities were significantly greater than on any other substrate type (p < 0.01). At riprap shorelines, Fucus spp. was used in addition to Enteromorpha spp. At the riprap-sill and P. australis shorelines, detritus was used in addition to Enteromorpha spp. At the riprap-sill shoreline, despite access to Enteromorpha spp. covered rock and S. alterniflora stems on the landward side of the riprap structure, eggs were only deposited on the open water side of the riprap structure. At S. alterniflora shorelines, a greater number of substrate types were used for egg deposition than at any other shoreline type. In addition to dominance of egg attachment on Enteromorpha spp., the base of S. alterniflora stems (p < 0.01) was utilized more than other substrates, including U. lacuta, detritus, and Fucus spp.
The percentage of Enteromorpha spp. coverage within sampling stations did not change over time. Although Kruskal–Wallis analysis did find significant differences in Enteromorpha spp. coverages among sample dates (p < 0.01), a Spearman's rho correlation did not reveal any statistical trends (ρ = 0.086, p = 0.135). Furthermore, visual inspection of plotted Enteromorpha spp. coverage percentages over time demonstrates that coverage was not reduced during sampling at any shoreline type.
Egg removal during sampling did not reduce egg numbers at a given station on subsequent sampling days as no relationship existed between the number of days since a station had been previously sampled and egg density (p = 0.822). Height of the high tide immediately prior to sampling was not significantly related to egg density across shorelines (p = 0.052). When analyzed within shoreline type, high tide height immediately prior to sampling was not significantly related to egg density at bulkhead (p = 0.995), beach (p = 0.153), or P. australis (p = 0.433) shorelines. However, tidal height was significantly related to egg density at riprap (p < 0.01), riprap-sill (p < 0.01), and S. alterniflora (p < 0.01) shorelines, with higher tidal heights having greater egg densities. Similar results were found for maximum height of high tide at 24, 48, and 72 h prior to sampling. Moon phase was not significantly related to egg density (p = 0.867). Atmospheric/water temperature differed among the five shoreline types (bulkhead, riprap, beach, P. australis, S. alterniflora) at which thermochrons were present during the sampling period. Significant differences in mean ranks of temperature were found (p < 0.01) when all 15 locations (low, mid, and upper intertidal elevations at five shoreline types) were analyzed. Upper intertidal locations had significantly greater atmospheric/water temperatures than those at the lower intertidal locations across all shoreline types.
Mean and maximum (in parentheses) water/air temperatures (°C) at upper, mid and lower intertidal elevations at shoreline types near Roosevelt Inlet, Delaware Bay on days when eggs were present (April 17th–May 27th)
Efforts were made to locate eggs for 4 weeks after the last egg was collected from the sampling area. These efforts were unsuccessful both in and around the sampling area; indicating that M. menidia egg deposition had ended within the sampling area for the spawning season.
Egg Deposition by Shoreline Type
The present results demonstrate that hardened and P. australis invaded shorelines support substantially reduced densities of M. menidia eggs compared with S. alterniflora shorelines. Approximately 94 % of eggs collected were found at S. alterniflora stations, suggesting depositional preference for the native, vegetated, unhardened shoreline type. According to the description of M. menidia spawning behavior reported by Middaugh et al. (1981), schools of adults actively choose the location of egg deposition. This concept is well supported by data from the present study.
High variability of egg density was found within S. alterniflora stations, including at adjacent stations, in this present study (Fig. 2). The degree of variability in egg density among S. alterniflora stations contrasts with that at other shoreline types. Physical differences in these S. alterniflora stations, particularly the elevation of the marsh plain, may have strongly contributed to this variability in egg density. M. menidia have been found to spawn at high tide on S. alterniflora marsh surface where water depth is only 0–30 cm (Middaugh et al. 1981). Tewksbury and Conover (1987) noted certain locations within their sampling area to consistently be areas of intense M. menidia spawning activity, surrounded by areas of much lower spawning activity. This likely reflects the searching behavior exhibited by M. menida schools (Middaugh et al. 1981) as S. alterniflora marshes exhibit variable elevation and substrate availability.
Hardened shorelines homogenize the intertidal zone and reduce variability in substrate and elevation, potentially removing areas suitable for high M. menidia spawning activity. Riprap and riprap-sill stations were comparable physically and biologically homogenous with one another (with respect to genera and coverage of macroalgae available in the mid intertidal zone). This similarity was reflected in a 4 % difference in egg density between these two shoreline types. The physical structure of bulkhead shorelines is the most unique of any of the sampled shorelines and had by far the fewest eggs deposited. M. menidia avoided the vertical structures, both wood and metal, of bulkheads for egg deposition. This result is in contrast with earlier studies in which M. menidia were reported to utilize the algae-covered vertical structure of a floating dock as well as the vertical structure provided by crab burrows for egg deposition (Moore 1980; Middaugh et al. 1981). The sandy beach shoreline, like the bulkhead shorelines, consisted of very homogenous structure. Eggs were deposited evenly among the beach stations, with very low variability in egg density. All eight sandy beach stations had eggs present on a similar percentage of sampling days; however, egg density was low across beach stations, likely due in part to limited occurrence of Enteromorpha spp.
Among the shoreline types sampled in the present study, marsh shorelines comprised of well-established invasive P. australis have physical and biological structure most similar to S. alterniflora marsh. Variability in shoreline structure and substrate availability was higher at P. australis shorelines than at hardened shorelines, and this is reflected in the high variability in egg density at the P. australis shoreline. However, total egg density along the P. australis shoreline was more similar to that at hardened shorelines than that at S. alterniflora shorelines. One factor which may have contributed to disparate egg densities between S. alterniflora marsh and P. australis marsh is elevation of the marsh plain. Larger areas of S. alterniflora marsh are inundated at high tide compared to P. australis stands. The higher elevation of established P. australis marshes has been shown to reduce fish movement into the marsh interior (Weinstein and Balletto 1999; Able and Hagan 2003; Able et al. 2003). P. australis also grows less densely than does S. alterniflora at the marsh surface, potentially reducing predation protection for M. menidia embryos.
Spawning site selection is likely impacted by unique current flow and wave action among the shoreline types studied. Dense stands of S. alterniflora potentially provide the greatest amount of wave attenuation in the mid intertidal zone, while hardened shorelines provide the least amount of wave attenuation. Wave energy has been noted to be one of the most important physical factors affecting intertidal organisms on hardened shorelines (Southward and Orton 1954; Denny and Wethy 2001; Jonsson et al. 2006) and may be a factor in M. menidia spawning site selection. Middaugh and Takita (1983) hypothesize that the spawning runs of M. menidia occur in direct response to current velocities. At slack high tide, water movement decreases and sperm is less susceptible to dispersion. By selecting spawning sites with shallow water over suitable substrate, water volume per unit of suitable substrate is reduced, and M. menidia can increase the concentration of milt at spawning sites (Middaugh et al. 1984).
By exhibiting searching behavior and actively selecting a shoreline location for egg deposition, schools of spawning M. menidia (Middaugh et al. 1981; Conover and Kynard 1984) may not reduce total egg deposition in response to encountering P. australis and hardened shorelines. Rather, egg deposition may be largely displaced and concentrated in areas along S. alterniflora shorelines, where preferred combinations of shoreline type and substrate for M. menidia oviposition are found. Whether greater egg densities affect hatching and larval survival is a matter for future research. As the area of hardened shoreline increases in urbanizing estuaries, preferential spawning habitat may eventually reach thresholds below which reproductive activities are critically impaired.
Egg Attachment by Substrate Type
Substrate availability differed at each shoreline type, but Enteromorpha spp. was utilized most frequently, across all shorelines for egg attachment by M. menidia. Over 95 % of eggs collected had been attached on Enteromorpha spp. The variety of other substrates, including Fucus spp., U. lacuta, S. alterniflora stems and detritus, were used significantly less often. Previously cited substrates used for deposition such as sand, trash, and crab burrows were not found to be utilized in the present study. By utilizing the green alga Enteromorpha spp. for egg attachment at a greater percentage than the coverage percentage of Enteromorpha spp. at each shoreline type, M. menidia are displaying a preference for oviposition on the substrate. Various species of Enteromorpha spp. are found throughout the range of M. menidia (Schneider and Searles 1991) and several studies have found filamentous algae to be a preferred substrate to the north (Massachusetts, Conover and Kynard 1984) and south (South Carolina, Moore 1980) of the present study area.
The preference for Enteromorpha spp. by M. menidia as a spawning substrate raises the possibility that areal coverage of this alga is the driving factor in their preference of shorelines for egg deposition. Riprap and riprap-sill shorelines had significantly greater coverage of Enteromorpha spp. than other shorelines, yet they had the third and fourth greatest egg densities of the shoreline types examined. Bulkhead shorelines had greater Enteromorpha spp. coverage than P. australis and beach shorelines, but far fewer eggs were deposited on bulkheads. Thus, although Enteromorpha spp. is the preferred substrate for egg attachment by M. menidia across all shoreline types, shoreline type affects choice of spawning location as well. Were this not the case, 94 % of all eggs would not have been deposited along S. alterniflora shorelines. Likewise, bulkhead shorelines would have had greater egg densities than the P. australis shoreline and far greater egg densities than sandy beach shoreline. Within S. alterniflora shorelines, no significant differences existed in mean rank sums of egg density by Enteromorpha spp. coverage percentage. So, across a range of Enteromorpha spp. coverages (10–90 %), egg density did not differ. As such, in the highly variable environment of S. alterniflora marsh, M. menidia are utilizing their preferred substrate, even as this increases the density of eggs over a smaller area of substrate. This further supports the concept that M. menidia are very precisely choosing locations for egg deposition/attachment.
Greater egg densities were found to be related to higher tidal height at the S. alterniflora shorelines in the present study. This agrees with many prior M. menidia studies (Middaugh et al. 1981; Middaugh et al. 1984; Conover and Kynard 1984; Tewksbury and Conover 1987). Higher tides allow more preferential shoreline/substrate to be available, at an elevation with preferential temperature and moisture conditions for egg deposition. Observations of the behavior of pre-spawning schools of M. menidia suggest a probable response to the availability of preferred vs. non-preferred spawning substrates (Middaugh and Takita 1983). These investigators observed schools of M. menidia, on occasions of sub-maximum high tide heights, swimming past areas where spawning activity was commonly observed during occasions of higher high tide heights. Because these sites were not inundated, and were thus unavailable for spawning, these schools moved to a “secondary” spawning location lower in the intertidal, just before slack high tide. These observations are consistent with the concept of tidally (on all high tides), rather than lunar (only on spring high tides) spawning intensity concluded by Conover and Kynard (1984). Moon phase was not found to be significantly related to egg density in the present study, supporting a prior laboratory study by Conover and Kynard (1984).
No detrimental atmospheric/water temperature conditions for embryonic survival were found at any of the shoreline types in this study. Maximum atmospheric/water temperatures when eggs were present were never >34.1°C, well below the danger levels for survival indicated by critical thermal maxima tests (Hutchinson 1961). Atmospheric/water temperatures, both means and maximums, were relatively high at the S. alterniflora shoreline in comparison with the other shoreline types examined (Table 3). Whether this difference is primarily due to vegetative cover and retained moisture is not known. These higher mean temperature decrease incubation time, reducing potential predation (Able and Castagna 1975; Conover and Kynard 1984; Middaugh et al. 1983).
Middaugh et al. (1983) found that different substrate types (S. alterniflora stems, Enteromorpha spp., and mud crab burrows) offered M. menidia embryos differing degrees of protection from thermal and desiccation stresses. Surface temperatures and atmospheric moisture were found to be most favorable for M. menidia embryo survival within Enteromorpha spp. mats (13–34°C; no desiccation observed). Less favorable environmental conditions were observed at the base of S. alterniflora stems (maximum of 36°C; significantly lower atmospheric moisture than algal mats) in that South Carolina study. This is important, as substrate choice for oviposition is among the only forms of parental care M. menidia provide for embryos (DeMartini 1999). Given that desiccation is a major source of mortality during incubation for intertidally spawning fishes (DeMartini 1999), oviposition/substrate choice is critical for M. menidia embryo survival. Enteromorpha spp. mats present at S. alterniflora shorelines provide preferable warm substrate temperatures and high atmospheric moisture for M. menidia embryos.
M. menidia employ spawning behaviors to maximize the survival and health of their embryos. Findings from the present study are consistent with previous findings on M. menidia searching behavior which suggest schools of M. menidia actively choose specific sites for egg deposition. Schools of spawning M. menidia appear to select sites for egg deposition primarily based on shoreline type, and then proceed to choose specific locations for egg attachment based on substrate type. Shorelines dominated by S. alterniflora are utilized more frequently than hardened shorelines, sandy beach shorelines, and shorelines dominated by P. australis. The green alga Enteromorpha spp. is utilized more frequently than all other available substrates for M. menidia egg attachment. Ninety-one percent of all eggs collected in the present study were deposited along a S. alterniflora shoreline, attached to Enteromorpha spp. This combination of S. alterniflora shoreline and Enteromorpha spp. provides preferential characteristics (temperature, atmospheric moisture, predation protection) for health and survival of M. menidia embryos. Increasing areas of hardened shoreline structures in estuarine environments may have serious implications for M. menidia and other intertidally spawning fishes by reducing habitat used most frequently for egg deposition.
We wish to thank Adriana Aruajo and Benjamin Ciotti for field assistance. Special thanks are extended to Dr. Patrick Gaffney for help with statistical analysis. This research was supported by NOAA, National Centers for Coastal Ocean Science, Center for Sponsored Coastal Ocean Research (award number NA09NOS4780219) to T.E. Targett.