Introduction

Every spring, American horseshoe crabs (Limulus polyphemus) make spawning migrations into estuaries and embayments along the Atlantic and Gulf of Mexico coasts of the USA. During high tide, a female, often with a male clasped to her carapace in amplexus, moves up to the high water line where she digs a nest and lays her eggs (Shuster and Botton 1985; Brockmann 1990; Weber and Carter 2009). Subsequently, the eggs are externally fertilized by the attached male and/or satellite males (Brockmann 1990; Brockmann et al. 2000). While this general pattern is consistent throughout the horseshoe crab’s range, there are some important differences in the timing of these spawning events in different regions. The goal of this study was to characterize the temporal and spatial patterns of horseshoe crab spawning in the Great Bay Estuary (GBE), New Hampshire.

Currently, due to concerns that horseshoe crab populations might be declining as a result of heavy exploitation (ASMFC 1998), many states conduct surveys to monitor their respective horseshoe crab populations. The protocols used are based on certain assumptions, such as a tendency for more horseshoe crabs to spawn at night and during the new and full moons. The GBE is located near the most northern portion of the horseshoe crab’s geographic range, and it is home to a modest horseshoe crab population (Watson et al. 2009; National Marine Fisheries Service 2010; Schaller et al. 2010; Watson and Chabot 2010; Cheng 2015). However, New Hampshire is one of the few US Atlantic states that does not conduct an annual horseshoe crab spawning survey. Because genetic studies suggest that there are 4 to 11 distinct populations of horseshoe crabs in regions spanning from Mexico to Maine (King et al. 2005; Faurby et al. 2010), and our preliminary studies indicate that some aspects of horseshoe crab spawning in GBE differ from other, better studied, populations (Chabot et al. 2008; Watson et al. 2009; Schaller et al. 2010; Watson and Chabot 2010), this investigation was initiated to better understand the dynamics of horseshoe crab spawning in New Hampshire. One of the ultimate goals is to use these data to design accurate and efficient annual spawning surveys that can be implemented in GBE in the future.

One factor that appears to have a strong influence on the timing of horseshoe crab spawning is water temperature, at least in more northern locations. Horseshoe crabs from Delaware Bay to Maine appear to initiate spawning in April or May, as the water temperatures begin to increase. For example, in Delaware Bay, the movement of spawning females into shallower waters coincided with rising water temperatures (Smith et al. 2010), and a similar correlation has been documented in Pleasant Bay, Cape Cod, Massachusetts (commencing at ∼15.5 °C in late April–May; James-Pirri 2010) and Casco Bay and Taunton Bay, Maine (threshold of ∼10 to ∼12 °C in June; Schaller et al. 2005 and Moore and Perrin 2007, respectively). Finally, we have documented horseshoe crabs in the GBE making the transition from being very sedentary in the winter to actively migrating up the estuary towards spawning beaches, as the water temperature rises above 8–10 °C (Schaller et al. 2010; Watson et al. in press). In this paper, we provide additional evidence demonstrating that spawning occurs earlier in years with warm spring weather and that spawning in the GBE is most prevalent at the warmest beaches.

The overarching objective of this study was to identify the environmental factors that influence the spatial and temporal distribution of spawning horseshoe crabs in the Great Bay Estuary, New Hampshire. Traditional spawning surveys were used to determine the possible influence of tides, water temperature, lunar phases, and weather conditions on horseshoe crab spawning activity at four different locations within the estuary. In addition, horseshoe crabs were tagged and released to gain insight into how often females spawned and how frequently both male and female horseshoe crabs returned to the same beach or different beaches. If, and when, New Hampshire initiates an annual horseshoe crab monitoring program, the data obtained should help make it both more efficient and accurate.

Materials and Methods

Study Site

This study took place in Little Bay and Great Bay, located in the Great Bay Estuary, New Hampshire. Six rivers empty into the estuary, which connects with the Gulf of Maine via the Piscataqua River (Fig. 1). Great Bay and Little Bay have very large shallow regions and a deeper, central channel that runs the length of both of them. Shoreline beaches surrounding GBE are composed of sediments ranging from coarse sand to shale-slate; some beaches also consist of Spartina spp. zonation (Short 1992; Jones 2000). Beaches are generally narrow and short due to the rocky intertidal habitats sporadically fringing the estuary (Short 1992). In a typical year, there are large seasonal fluctuations in temperature (−2 to 30 °C) and salinity (∼5 to ∼30), which is typical of high latitudinal temperate estuaries (Short 1992; Watson et al. 2009). The average tidal range of GBE varies from 2.0 to 2.7 m (Short 1992).

Fig. 1
figure 1

Map of horseshoe crab survey locations in the Great Bay Estuary, New Hampshire, USA. The Great Bay Estuary is 16 km inland from the Gulf of Maine and includes Little Bay and Great Bay. Xs denote horseshoe crab survey locations. Great Bay Marina (GBM) and Adams Point (ADAMS PT) are located in Little Bay; Moody Point (MOODY PT) and Sandy Point (Great Bay Discovery Center, GBDC) are located in Great Bay

Full size image

Survey Locations

In 2012, surveys were conducted at the Great Bay Marina in Newington (GBM), Adams Point (where there were three geographically separate beaches) in Durham (ADAMS PT), and the Great Bay Discovery Center at Sandy Point in Greenland (GBDC). In 2013, Moody Point in Newmarket (MOODY PT) was added as a survey location. GBM and ADAMS PT are in Little Bay; GBDC and MOODY PT are in Great Bay (Fig. 1), and for most analyses, data were grouped in this manner.

Horseshoe Crab Spawning Surveys

Horseshoe crabs generally spawn in May and June in the GBE. Prior to May 2012 and 2013 spawning seasons, 25- to 75-m transects, divided into 5-m replicates, were marked by wooden stakes on survey beaches. Survey areas and the number of surveys at each location are shown in Table 1. Surveyors walked the length of each transect and counted all horseshoe crabs (horseshoe crab pairs, single males, and single females) found within 2 m of the high water line (5 m × 2 m transects). Surveyors also noted categorical environmental conditions, including approximate wave heights and precipitation. During 2012, hand-held thermometers were used to record water temperatures when surveys were conducted, and in 2013, temperature loggers (Onset HOBO Pendant® Temperature/Light Data Logger 64K) were placed at GBDC, GBM, ADAMS PT, and MOODY PT to automatically record water temperature every hour during the spawning season. Actual water depth/ tide height data for May and June of 2012 and 2013 were obtained from the “Squamscott River Monitoring Station” located ∼1 km southwest of the GBDC site (data were provided by the National Estuarine Research Reserve Centralized Data Management Office). For certain analyses, the heights of both high tides from each day were averaged and then were paired with horseshoe crab densities from the associated date at the survey location closest to the monitoring station (GBDC). Precipitation data were obtained from the “Great Bay Meteorological Station” located ∼5 km east of the GBDC survey site (also provided by the National Estuarine Research Reserve Centralized Data Management Office).

Table 1 Total area (m2) covered during a survey and the number of surveys conducted for each survey location
Full size table

Spawning surveys were conducted daily at each site, weather permitting, during the peak daytime high tide. During the 5-day periods around the new and full moons (2 days prior, the day of, and 2 days after), surveys were conducted during both peak day high tides (between ∼11:00–14:00 h) and peak night high tides (∼22:00–02:00 h).

Tagging Methods

In both years combined, a total of 442 (∼10 %) of the horseshoe crabs observed during all surveys were tagged with button tags, using the “United States Fish and Wildlife Service (USFWS) Horseshoe Crab Tagging Protocol” (USFWS 2012), in order to determine how often they returned to the same beach and if they spawned at more than one beach. Information on the tags was limited to the USFWS information, contact phone number, and the individual tag number. Typically, some of horseshoe crabs observed during a survey were tagged at the end of the survey. Most recaptures/re-sightings occurred during subsequent surveys. Additional recapture data were obtained from reports provided by USFWS and from the general public.

Data Analyses

All horseshoe crab counts and categorical environmental conditions gathered from surveys were organized by location and then by date. Statistical analyses were conducted using the software program InStat (GraphPad Software, La Jolla, CA, ver. 3.1a 2009) and Microsoft Excel Analysis ToolPak (Microsoft Office Professional Plus 2010).

The total number of horseshoe crabs per unit area (ind./m2) was used to compare horseshoe crab densities between locations. Only day high tide surveys were used for analyses of temporal patterns of spawning because more day surveys were conducted than night surveys. To determine if more horseshoe crabs mated during the new and full moons, data from the 5-day full and new moon periods were compiled and compared to other 5-day periods in the same month using the Mann-Whitney test.

Day and night densities were compared using the Wilcoxon matched-pairs signed-ranks test. Horseshoe crab densities from all locations were combined for day versus night analyses and for determining correlations between environmental conditions, such as tide height and temperature, and spawning activity. The examination of the relationship between tide height and spawning activity was only carried out using data from the GBDC site. Horseshoe crab densities associated with different categorical wave heights were compared using the Kruskal-Wallis test nonparametric analysis of variance (ANOVA), and changes in horseshoe crab densities on days before, during, after rainfall were compared using the Friedman test nonparametric repeated measures ANOVA. A multiple linear regression analysis of temperature and rainfall with spawning density was performed using only 2012 and 2013 data from the GBDC site.

Tag-recapture information was used to determine (1) whether horseshoe crabs spawned at more than one beach in a given season or in different years, (2) the time intervals between spawning events for a given horseshoe crab, and (3) how often an individual horseshoe crab spawned. Night recaptures were excluded in these analyses for the purpose of consistency, because night surveys were not conducted every day.

Results

Spatial and Temporal Distribution of Spawning Horseshoe Crabs

In 2012 and 2013, we observed 5,325 horseshoe crabs at the four survey locations. Of these, 3,082 were males and 2,243 were females (Table 2). Of the 2,243 females observed, 97 % were paired with a male. In both years, the mean sex ratio was skewed towards males (1.4 males:1 female), with the highest proportion of males to females at ADAMS PT (1.55:1; Table 2). Overall, there were more than twice as many horseshoe crabs observed in Great Bay than in Little Bay (Fig. 2).

Table 2 Total adult horseshoe crabs counts and densities (ind./m2) from all surveys (day and night; see Table 1) and sex ratios (male/female) at each location
Full size table
Fig. 2
figure 2

Graduated bubble map of mean horseshoe crab density (horseshoe crabs per square meter (ind./m2)) for each survey site, along with mean temperatures taken during times of surveys in 2012 (a) and 2013 (b). More horseshoe crabs were observed at sites located in Great Bay in both years. (Maps made by Google Maps 2013 and GeoCommons 2013)

Full size image

The first peak of spawning, when the density of horseshoe crabs exceeded the monthly mean density, took place about two weeks earlier in 2012 than 2013 (Figs. 3 and 4), most likely due to some unusually warm weather in March and April of 2012. In both years, the first peak of spawning activity was followed by two more peaks, which occurred at about the same time at the different study sites; the three spawning peaks were separated by ∼8–12 days (Figs. 3 and 4) and were not closely correlated with the days around the full and new moons (Figs. 3 and 4). There was no statistically significant difference between the number of horseshoe crabs observed on the 5 days surrounding the new and full moons, compared to other 5-day periods in May or June, in either 2012 (p = 0.1897 and p = 0.5898, respectively, Mann-Whitney test) or 2013 (p = 0.2806 and p = 0.1122, respectively, Mann-Whitney test).

Fig. 3
figure 3

Mean horseshoe crab densities (ind./m2) (bars) and temperatures (lines) for May and June 2012 and 2013. Lunar cycles are outlined; dark circles represent new moon periods, open circles represent full moon periods

Full size image
Fig. 4
figure 4

Mean horseshoe crab densities (ind./m2) for the GBDC location (bars) and daily mean high tide depth (lines) for May and June 2012 and 2013. Lunar cycles are indicated as in Fig. 3

Full size image

Influence of Tide Height on Spawning Activity

It has been proposed that peaks in the number of spawning horseshoe crabs occur during days when the tides are higher than usual (Barlow et al. 1986; Brockmann 2003; Brockmann and Johnson 2011). To test this hypothesis in our study location, we compared actual tide heights with horseshoe crab densities at GBDC, which was the location closest to the buoy from which the tide data were obtained, and also the location with the greatest number of spawning adults. Although the daily mean high tide depth varied by about 1 m throughout the spawning season at this location, tides were not necessarily highest on the new and full moons and there was no consistent relationship between tide height and the number of spawning horseshoe crabs (Fig. 4). In 2012, high tide depths were significantly and negatively correlated with horseshoe crab density (p = 0.0014, r = −0.5001, n = 38 surveys; Spearman rank correlation), and in 2013, there was no relationship between high tide depth and horseshoe crab densities (p = 0.1820, r = −0.1862, n = 53 surveys; Spearman rank correlation).

Influence of Temperature on Spatial and Temporal Distribution

One possible explanation for the greater number of spawning horseshoe crabs at the Great Bay locations is that these areas were significantly warmer compared to those in Little Bay (p < 0.001, Mann-Whitney test). Mean temperatures were ∼2 °C warmer in Great Bay than in Little Bay during the 2012 and 2013 surveys combined (18.7 and 16.9 °C, respectively). In addition, increases in water temperature were often associated with peaks in spawning activity (Fig. 3). For example, as temperatures increased around the days of May 5 and May 18, in 2012, and May 7, May 26, and June 20, in 2013, the densities of spawning horseshoe crabs also increased (Fig. 3). As a result, both temperature and horseshoe crab densities were significantly and positively correlated in Great Bay in both 2012 and 2013 (p = 0.016, r = 0.400, n = 36 surveys; p = 0.006, r = 0.355, n = 59 surveys, respectively; Spearman rank correlation).

Day Versus Night Spawning Activity

Day and night surveys were carried out during the 5 days associated with the new and full moons. In 2012, there were significantly more spawning horseshoe crabs during the day than at night (p = 0.016, Wilcoxon matched-pairs signed-ranks test, n = 12 survey dates), but in 2013, there was no significant difference between day and night counts (p = 0.722, Wilcoxon matched-pairs signed-ranks test, n = 39 survey dates; Fig. 5).

Fig. 5
figure 5

Mean horseshoe crab densities (ind./m2) during day and night high tides in 2012 (n = 12 survey dates) and 2013 (n = 39 survey dates). Asterisks (*) denote statistical significance (p = 0.016 for 2012, p = 0.722 for 2013)

Full size image

Influence of Wave Action and Rain on Spawning Activity

Two other factors that have been reported as possibly influencing horseshoe crab spawning activity are wave action and rain (Smith et al. 2002a, b; Watson et al. 2009). At the GBDC and ADAMS PT study sites, more horseshoe crabs were observed spawning on days with intermediate waves than when there were no waves or when waves were greater than 0.3 m (p = 0.0011, p = 0.0052, Kruskal-Wallis test nonparametric ANOVA, respectively; Fig. 6a). Rainfall also had a significant effect on the spawning activity of horseshoe crabs (p = 0.025, Friedman test nonparametric repeated measures ANOVA, n = 16 days; Fig. 6b). There were fewer spawning horseshoe crabs on days with light and heavy rain than days before and after rain events. Applying precipitation amounts (mm) from both 2012 and 2013, there was a negative effect on horseshoe crab densities (p < 0.001, r = −0.437, n = 48 surveys; Spearman rank correlation). When precipitation amounts and temperature are analyzed in conjunction with horseshoe crab densities from both years, there was a significant interaction between rain and temperature with horseshoe crab densities (p = 0.038, r 2 = 13.58 %, n = 48 surveys; multiple regression analysis).

Fig. 6
figure 6

The relationship between mean horseshoe crab densities (ind./m2) (GBDC and ADAMS PT sites), wave height (a), and rain events (b). Horseshoe crab counts were averaged for days before, during, and after rain days (n = 16 rain events). There was a significant effect of wave height and rain on horseshoe crab spawning densities (p > 0.0001 and p = 0.025, respectively). Bars with the same letters are not significantly different from each other. There were no horseshoe crabs present when wave heights were >0.3 m

Full size image

Tag-Recapture Data

Of the 442 adult horseshoe crabs that were tagged (187 females and 255 males), there were 62 recaptures. Some horseshoe crabs were recaptured more than once; thus, the 62 recaptures were from 43 horseshoe crabs, yielding a recapture rate of 9.7 % (Table 3). Males were recaptured more often than females (42 males, 20 females). Most (74.2 %) of the horseshoe crabs were first recaptured within 1–3 days of when they were tagged (Fig. 7). After the first week of high recaptures, the greatest number of recaptures occurred ∼3 weeks later. Only 4.8 % (3 out of 62) horseshoe crabs were recaptured in a different area than where they were tagged, one of which was a female re-sighted ∼4 km further up-estuary during the same year. There were two horseshoe crabs (one male, one female) that were tagged in 2012 and found the following year, and both of them were recaptured in a different location (∼5 km away).

Table 3 Total tagged and recaptured horseshoe crabs for 2012 and 2013
Full size table
Fig. 7
figure 7

Number of horseshoe crabs recaptured (males: n = 42; females: n = 20) at different intervals after they were tagged. Data includes horseshoe crabs that were recaptured more than once. A total of 255 males and 187 females were tagged

Full size image

Discussion

Most investigations of horseshoe crab spawning behavior have been conducted on beaches along the coasts of Florida, the shorelines of Delaware Bay and Cape Cod, Massachusetts. This study was one of the first investigations of the spawning dynamics of the horseshoe crab population residing in the Great Bay Estuary, New Hampshire. Overall, the same trends reported in previous studies were apparent in this population: mating commenced as water temperatures increased in the spring (however, this is not the rule in the most southern populations), there were more males than females surveyed, and certain environmental conditions, such as wave heights, appeared to have a modest influence on spawning activity. However, some differences were also evident. For example, large increases in spawning activity were not always associated with the new and full moons and more horseshoe crabs were not always observed during the night versus the day high tides (Online Resources 1 and 2).

Despite the generally accepted notion that more horseshoe crabs spawn around the new and full moon, compared to other times, there was no clear bias towards increased spawning at these times, or during the highest tides, throughout this study. This is consistent with the findings of Smith et al. (2010) in Delaware Bay but contrasts with reports from Florida and some Massachusetts populations. In Apalachee Bay and Seahorse Key, Florida, horseshoe crab spawning activity was associated with the full moon spring tides (Rudloe 1980; Cohen and Brockmann 1983), and at the beaches along Mashnee Dike, Cape Cod, Massachusetts, increases in spawning activity were correlated with the highest high tides associated with the new and full moons (Barlow et al. 1986). Penn and Brockmann (1994) suggested that the most likely explanation for these observations is that during the new and full moons the tides tend to be higher, which allow horseshoe crabs to exploit more of a spawning beach and find nesting sites where development of eggs is maximized. A subsequent study by Brockmann and Johnson (2011) indicated that, in Seahorse Key, Florida, spawning density was strongly correlated with the largest high tides that were more associated with the spring tides, rather than the moon phases. The GBE has a semi-diurnal tidal cycle, with two high tides and two low tides in a 24-h day, and they vary by about 1 m, depending on whether it is a spring or neap tide. At our study sites, the highest high tide heights did not occur on the days of the new or full moon, but rather 1–2 days later. Nevertheless, there was no clear bias towards increased spawning activity during days surrounding a new or full moon, or the highest tides (Figs. 3 and 4). This could be due to the characteristics of spawning beaches in the estuary, which are generally narrow and small. As a result, during a spring tide, the water can completely submerge a spawning beach, so there is very little room at the high water mark for spawning. However, beaches are larger in Delaware Bay, and peaks of spawning there are also not tightly associated with the spring tides; Smith et al. (2010) observed only 26 % of spawning activity occurred within 1 day of the spring tide, and 78 % of spawning activity occurred within 5 days of a spring tide. Thus, the characteristics of the beaches in GBE cannot be the only explanation for the lack of intense spawning activity around the times of the highest tides.

Spawning during high tides is likely a function of both a horseshoe crab’s endogenous circatidal clock that controls their tidal rhythms of activity (Chabot et al. 2004, 2007; Watson et al. 2008), and their ability to sense the water pressure changes associated with the tides (Chabot et al. 2008, 2011; Watson et al. 2009). It has been suggested that more horseshoe crabs mate at higher tides because they are better able to sense them, and thus more horseshoe crabs become active and are synchronized to those tides (Watson et al. 2008; Chabot and Watson 2010; Chabot et al. 2011). This may be particularly true for northern Gulf of Mexico populations in Florida where there is about a 20 % increase in tide height between neap and spring tides (data from Station St. Marks River Entrance; NOAA Tides and Currents 2013). As a result, during the spring tides, the depth detection “threshold” is exceeded for more horseshoe crabs, as suggested by Watson et al. (2008). In contrast, in most northern locations, high tides are large enough so that most horseshoe crabs can probably detect each high tide, and there is less of a tendency to only mate during certain times of the month or moon phase.

It is also generally accepted that spawning predominately occurs at night (Cavanaugh 1975; Rudloe 1980, 1985; Finn et al. 1990; Swan et al. 1991, 1993; Smith et al. 2010). However, this was not observed for the horseshoe crabs in this study or a previous study in the GBE (Watson et al. 2009). In fact, in 2012, there were significantly more horseshoe crabs spawning during the day. Cohen and Brockmann (1983) reported that horseshoe crabs in Seahorse Key, Florida, where predicted day high tides are 0.1–0.3 m greater in height than night high tides, preferred to spawn during the day. This phenomenon was also observed by Barlow et al. (1986) at spawning beaches near Mashnee Dike, Cape Cod, Massachusetts. Subsequently, Brockmann and Johnson (2011) found that the total number of nesting pairs did not differ between day and night high tides, even though they found more unpaired males at night. In GBE, night high tides are ∼0.5 m larger than day high tides, yet there was no clear tendency for more horseshoe crabs to spawn during the night. Powers et al. (1991), Chabot et al. (2007), and Watson et al. (2008) all proposed that horseshoe crabs evolved the ability to increase their visual sensitivity at night so they could effectively mate during either, or both, day or night high tides, which they appear to do in most locations.

Temperature appears to have a strong influence on when and where horseshoe crabs spawn in the GBE. For example, prior to the spawning season in 2012, water temperatures surpassed 11 °C in mid-March, which is not typical for this month in this northern location. This caused horseshoe crabs to begin moving further up the estuary, towards their preferred spawning beaches, in the middle of March, which is 3–4 weeks earlier than usual (Watson et al. in press). As a result, in 2012, based on our own observations and reports from the public, horseshoe crab spawning commenced 2–4 weeks earlier than usual, in late April and early May. In contrast, in 2013, spawning activity started in mid-May, when water temperatures began to surpass 11–15 °C, and continued into June, which is consistent with other populations within this geographic range (Shuster 1979, 1982; Shuster and Botton 1985; Barlow et al. 1986; Schaller et al. 2004, 2005; Moore and Perrin 2007; James-Pirri 2010). In South Carolina, Thompson (1998) also observed that spawning took place earlier than usual when there was an unusually warm spring in March of 1997. Interestingly, when temperatures dropped in April, there was a decline in mating activity, and once temperatures warmed again in May, spawning increased (Thompson 1998). Smith and Michels (2006) also report that in cold years Delaware Bay horseshoe crabs spawn later than in warm years. Therefore, even though the thermal thresholds that lead to increased activity and spawning might differ between populations, there is very strong evidence supporting the hypothesis that the initiation of spawning, and perhaps fluctuations in spawning activity during a season, are very strongly influenced by changes in water temperature.

Temperature also appeared to influence the preferred locations for spawning in the GBE. Throughout the 2012 and 2013 spawning seasons, the average water temperature differed between Great Bay and Little Bay by ∼2 °C and the density of spawning horseshoe crabs was consistently greater in Great Bay than in Little Bay. The warmer waters that exist in Great Bay may also act as a cue, or attractant, for horseshoe crabs migrating up-estuary to spawn (Schaller et al. 2010; Cheng 2015). In Delaware Bay, for example, spawning generally occurs earlier on beaches along the eastern shoreline of the bay than the western shoreline, most likely because these locations warm up sooner in the spring (Smith and Michels 2006).

Horseshoe crabs may have evolved to spawn when water temperatures are optimal for egg development, as well as larval growth and survival (French 1979; Jegla 1982; Jegla and Costlow 1982), although optimal thermal conditions may vary throughout the species’ geographic range. In GBE, most spawning activity was found to occur between ∼15 and ∼20 °C, while in South Carolina and Florida, spawning activity was most prominent at temperatures above 20 °C (Thompson 1998; Brockmann and Johnson 2011). Thompson (1998) suggested that below 20 °C, overall spawning activity and larval development are suspended in South Carolina and this has also been proposed by Brockmann and Johnson (2011) for Florida horseshoe crabs. Given known genetic differences between horseshoe crab populations (King et al. 2005), it would be interesting to determine if the thermal preferences and thermal tolerances of eggs and larvae are innate or simply the result of acclimation to the different thermal regimes.

Collectively, the available data provide strong evidence that temperature influences horseshoe crab spawning activity. Climate change induced increases in water temperature might have profound effects on northern species because higher latitudes have warmed more than the lower latitudes in the past half-century (Root et al. 2003), and the effects are likely to be even more profound in bays and estuaries (Roessig et al. 2004; Brander 2010; Nye 2010). Warmer than normal spring water temperatures could, for example, advance the timing of spawning and, while the consequences of this shift are not obvious in the GBE, in Delaware Bay it could lead to a trophic mismatch between red knots and horseshoe crab eggs (Edwards and Richardson 2004; Durant et al. 2007). Galbraith et al. (2002), Botton et al. (2003), and Niles et al. (2009) all have emphasized that the timing of horseshoe crab spawning, especially in Delaware Bay, determines the success of many species’ migration strategies and that variation in water temperatures could have adverse implications.

Wave energy had an influence on spawning activity in this study, and it has been shown to influence spawning in previous studies as well. For example, wave action caused by wind conditions influenced the spatial distribution of spawning at Delaware Bay beaches (Smith et al. 2002a, b; Swan 2005), with fewer spawning horseshoe crabs in areas that experienced wave heights of greater than 0.3 m. Moreover, when wave action subsided on those beaches, spawning increased. Horseshoe crabs may prefer low-wave energy conditions on narrow beaches because they are associated with fewer risks of being turned over, stranded, or losing an attached male, and these conditions are also associated with reduced disturbance of sediment and nests (Smith et al. 2002b; Brockmann and Johnson 2011). Additionally, moderate waves (<0.3 m) might lead to increased spawning because wave surge helps horseshoe crabs orient to the beach (Rudloe and Herrnkind 1976, 1980).

Throughout the 2012 and 2013 spawning seasons, there were many periods of heavy rain, and during these events spawning activity decreased. A reduced density of horseshoe crabs during rain events has also been seen in previous studies in the GBE (Watson et al. 2009). In a large semi-enclosed estuary, like GBE, heavy rains can cause large decreases in salinity (Short 1992). Even though horseshoe crabs have been found in salinities as low as 7 (McManus 1969), the costs of osmoregulating (Robertson 1970; Mangum et al. 1976) might reduce their tendency to mate and, if they deposited eggs during these times, conditions could be detrimental for development (Jegla and Costlow 1982; Sugita 1988; Ehlinger and Tankersley 2003). During this study, even though rain events caused minor changes in salinity, there was a negative relationship between the amount of rainfall (mm) on days when surveys were conducted and the density of horseshoe crabs. However, temperatures on rainy days were significantly cooler than days before the rain, but not the days after (p = 0.026, repeated measures ANOVA, Bonferroni multiple comparisons post-test), and a multiple regression analyses indicated that there was a significant interaction between rain and temperature on horseshoe crab spawning activity. Specifically, temperature has the significant effect within this interaction. Thus, at least in this study, at this location, the impacts of small rain events on spawning activity were most likely due to the associated decreases in water temperature, although this remains to be more rigorously determined.

The sex ratios we observed in GBE were skewed towards males, which is consistent with spawning horseshoe crab populations in other locations (Rudloe 1980; Cohen and Brockmann 1983; Smith et al. 2002a; Carmichael et al. 2003; James-Pirri et al. 2005; Brockmann and Johnson 2011). These male skewed sex ratios during spawning could be the result of differences in the behavior of males and females. For example, we obtained more male than female recaptures during the 4-week period after horseshoe crabs were tagged. Brockmann and Penn (1992) reported that females usually nest and lay eggs for multiple tidal cycles during 1 week and do not return again until the following year, while males return repeatedly to the beaches throughout the entire spawning season, and Brockmann and Johnson (2011) have suggested that this is the cause of the male skewed sex ratios. While our data are consistent with their hypothesis and with their observations of male behavior, we also recorded several cases where females were recaptured ≥5 days after they were tagged (5, 7, 11, and 22 days).

Following the high incidence of tag returns within the first few days after they were released, the next peak of recaptures, which were predominately males, occurred ∼2–3 weeks later. This time interval corresponded with the second peak of spawning activity. It is possible that subsets of horseshoe crabs, or the same individual horseshoe crabs, mate at ∼2–3-week intervals. Historically, the explanation for the timing of peaks in spawning activity has been that they coincided with the new and full moons. However, in this population of horseshoe crabs, where peak spawning events were not clearly linked to the highest tides or the phases of the moon, there must be another explanation for this 2–3-week interval. Leschen et al. (2006) found that even though females in Pleasant Bay, Massachusetts returned to the beaches more frequently during the week when they were tagged, some were also recaptured 2 and 4 weeks later. Furthermore, their egg clutch size data suggested that they were only laying part of their total load of eggs during each spawning event. Similar results have also been seen in the Delaware Bay population (Smith et al. 2002a, 2010) and our data are consistent with this view. In the future, it would be worth investigating why it might take horseshoe crabs ∼2 weeks to recover from several days of laying eggs and prepare for the next bout of spawning.

In the US Atlantic states, annual horseshoe crab monitoring programs conduct surveys during the times of the full and new moon high tides and some programs only conduct nighttime surveys, despite having a semi-diurnal tidal cycle at that particular location (Swan et al. 2012). Smith and Michels (2006) have emphasized that accurate assessments of the temporal and spatial distributions of horseshoe crabs depend on using methodology that is specific to a given location. Based on the findings of this study, such a customized approach would likely yield the most accurate and useful data for the Great Bay Estuary. We would recommend that in the GBE, in order to be both accurate and efficient, annual surveys should commence when the water temperature exceeds 11 °C and take place each day during just the daytime high tides. Furthermore, although tag-recapture methods in this study were primarily used to determine the location and frequency of horseshoe crab returns to spawning beaches, in the future they should be expanded so that sufficient data are obtained to make it possible to estimate the size of the horseshoe crab population in the Great Bay Estuary.