Printer Friendly

Effects of the Biomedical Bleeding Process on the Behavior of the American Horseshoe Crab, Limulus polyphemus, in Its Natural Habitat.

Introduction

The American horseshoe crab, Limulus polyphemus, is an ecologically and economically important species found in bays and estuaries along the Atlantic coast of North America, including the Great Bay Estuary (GBE), New Hampshire. Horseshoe crabs play an important ecological role as bioturbators, as a result of foraging for food (Krauter and Fegley, 1994: Lee, 2010); and their eggs are a vital food source for 425,000 to 1 million migratory shorebirds (Walls et al., 2002; Botton et al., 2010). They are also harvested for use as bait for the eel and whelk fisheries (ASMFC, 1998, 2012) and for their blood, which is used to create Limulus amebocyte lysate (LAL) (Novitsky, 2009). LAL is used in the biomedical industry to test medical devices, vaccines, and pharmaceutical drugs for pathogenic gram-negative bacteria (Novitsky, 2009; Chen and Mozier, 2013).

Currendy, while quotas and regulations have been placed on the bait fishery (ASMFC, 2012), the biomedical fishery remains fairly unrestricted; and harvest rates continue to increase in certain areas (ASMFC, 2013), which could have deleterious effects on populations of this valuable marine species. A quota system and several complete closures of coastal waters have been implemented for harvesting horseshoe crabs for the eel and whelk fisheries (ASMFC, 1998, 2012), and this led to a significant decrease in the commercial harvest levels of horseshoe crabs, from around 2 million crabs in 2000 to 600,000-700,000 crabs in 2014 (ASMFC, 2013). In contrast, the number harvested for the biomedical industry continues to increase, with levels climbing from 340,000 crabs in 2004 to 610,000 crabs in 2012 (ASMFC, 2013).

With the growing demand for LAL as the global population expands, medical advancements improve, and medical needs increase, it is critical to understand the consequences of the biomedical bleeding industry on horseshoe crabs' fitness and population dynamics. The capture process for this industry includes multiple stressors, such as air exposure (time on docks, boats, and trucks) and warm temperatures (on boat decks during the summer or in poorly temperature-controlled facilities and transport vehicles). In addition, the blood extraction process itself can compromise the health of the horseshoe crab, because ~30% of the estimated blood volume of individual horseshoe crabs is extracted (James-Pirri et al., 2012). Female horseshoe crabs are preferentially chosen for this process because of their larger size and subsequently greater blood volume (Rutecki et al., 2004: James-Pirri et al., 2012); this has led to skewed sex ratios in some areas (Leschen and Correia, 2010). Each of the four major biomedical companies that bleed horseshoe crabs has slightly different collection, handling, and bleeding processes; and the extent to which they follow best manufacturing practices (BMPs) likely varies, depending on region. These BMPs include keeping them moist, avoiding bleeding injured animals or those that can be identified as having been bled before, and returning them to their point of capture within 24 hours.

Mortality rates associated with the bleeding process range from 5% to 30% (Rudloe, 1983; Thompson, 1998; Walls and Berkson, 2000, 2003; Kurz and James-Pirri, 2002; Hurton and Berkson, 2006; Leschen and Correia, 2010; Anderson et al., 2013), with a differential mortality rate between sexes (15% mortality in males and up to 29% in females; Leschen and Correia, 2010; James-Pirri, 2012). Sublethal impacts include delayed blood volume recovery, reduced blood protein levels, and behavioral deficits. Specifically, Novitsky (2009) found that in laboratory holding tanks it takes three to seven days for a bled horseshoe crab to regain its total blood volume and up to four months for amebocytes to return to baseline levels. Captive bled animals exhibit significantly lower blood protein values, signifying that biomedical bleeding may have prolonged impacts on horseshoe crab physiology; and bled crabs released back into their natural environment displayed a more random pattern of movements than control animals (Kurz and James-Pirri, 2002; James-Pirri et al., 2012). Finally, Anderson et al. (2013) found changes in the bled horseshoe crab's activity levels, expression of circatidal rhythms, linear and angular movement velocities, and hemocyanin levels.

Although the harvest process used by the biomedical fishery is considered low impact and classified as "minimally harmful to horseshoe crabs" (ASMFC, 2012), the aforementioned detrimental effects could alter population dynamics and could lead to long-term declines (Krisfalusi-Gannon et al., 2018). For example, because females are preferentially bled and because there is a higher mortality rate in females, this could lead to an overall decline in female fecundity and altered sex ratios (Le Moullac and Haffner, 2000; James-Pirri et al., 2005; Leschen et al., 2006; Leschen and Correia, 2010). The sublethal effects of biomedical bleeding on activity levels, expression of tidal rhythms, and movement velocities (Anderson et al. 2013) may disrupt activities such as foraging and spawning and may reduce crabs' ability to find mates and appropriate spawning beaches, thus leading to declines in reproductive output (Powers and Barlow, 1985;Barlow et al., 1986, 2001; Herzog et al., 1996; Barlow, 2001). Finally, extended periods of reduced hemocyanin levels may cause additional respiratory stress and increased susceptibility to infection because hemocyanin plays a major role in immune function and wound repair (Adachi et al., 2005; Coates et al., 2011).

With the exception of a study by Rudloe (1983) on the mortality rates of bled animals and a study by James-Pirri et al (2012) on the impacts of bleeding on horseshoe crab orientation, all relevant studies regarding biomedical bleeding effects on horseshoe crabs have been carried out in the laboratory. Therefore, the major goal of this project was to determine the behavioral and physiological effects that the bleeding process has on horseshoe crabs that are released back into their natural environment. The animals in this study were collected from, and released back into, the GBE. This population of horseshoe crabs has not been previously harvested for biomedical bleeding (ASMFC, 2012), and a great deal is already known about the behavior of horseshoe crabs in this estuary. For example, Schaller et al. (2010) and Watson et al. (2016) found that horseshoe crabs remained in GBE year-round but that they changed depths and locations in the estuary as temperatures changed throughout the year. In the spring (March-April), when water temperatures exceeded 10-11 [degrees]C, animals traveled to shallower areas and moved to spawning beaches at high tides. After spending the summer and early fall scouring the mudflats for food, they moved down the estuary into deeper waters in the late fall to overwinter. If bled horseshoe crabs express these same patterns of behavior, then we can conclude that the bleeding process does not impact them when they are released back into their natural habitat.

In this study, before being released into GBE, designated animals underwent the bleeding process; then all animals, both bled and controls, were fitted with acoustic transmitters to monitor their movements, the times when they were active, and their preferred depths. These data were collected for about two years and were used to discern whether the bleeding process had an impact on spawning activity, the expression of daily and tidal rhythms, overall activity and distances traveled, and seasonal migrations.

Materials and Methods

Animal collection and tagging

A total of 28 (14 male, 14 female) healthy adult American horseshoe crabs, Limulus polyphemus (Linnaeus, 1758), were hand collected during high tide from a spawning beach on Adams Point, Durham, New Hampshire, in May 2016 (Fig. 1). All captured crabs were brought back to the University of New Hampshire's Jackson E.stuarine Laboratory (JEL) and held in flow-through estuarine-water tanks until they underwent their designated treatment. Half of the animals (7 males; inter-ocular [IO] width = 8.21 [+ or -] 0.61 cm [SD]; 7 females: IO width = 11.56 [+ or -] 0.52 cm) were used as controls, while the remaining half (7 males: IO width = 8.31 [+ or -] 0.47 cm; 7 females: IO width = 11.42 [+ or -] 0.63 cm) were bled according to the industry standard procedures typically followed by the biomedical bleeding facilities, as outlined below (see Bleeding procedure). However, it should be noted that these procedures might vary from facility to facility and from state to state.

After treatments (bleeding or not), all horseshoe crabs were fitted with VEMCO V13AP ultrasonic transmitters (69 kHz, 147-dB low-power output, 13-mm diameter, 48-mm length, 6.5 g in water, estimated battery life of ~530 days; VEMCO, Bedford, Nova Scotia, Canada). The V13AP transmitters were programmed to transmit acceleration and depth data at random intervals about every three minutes. They were also programmed to turn off in December 2016 and then turn back on in March 2017. A transmitter was attached to the dorsal carapace of each individual, using the following method. First, it was superglued into a piece of plastic tubing that had two cable ties attached to it. The cable ties were then affixed to the carapace by using small screws. Finally, duct tape was superglued over the entire harness to ensure that the fixture would not become caught on underwater obstructions. In addition, male claspers were secured in the closed position with cyanoacrylate glue to eliminate their ability to attach to females. This ensured that data from males represented their activities and not those of a female to which they were attached. After the transmitters were attached, the animals were released into the GEE at the same spawning beach where they had been previously collected.

Bleeding procedure

Pre-bleeding treatment. The bleeding process, replicating industry standard procedures, took a total of three days (May 13-15, 2016). Half of the 28 horseshoe crabs (n = 14, 7 males, 7 females) were randomly selected to undergo the bleeding procedure. The animals in this treatment group were evenly distributed between two 50-gallon plastic barrels. HOBO temperature loggers (Onset Computer Corporation, Bourne, MA) were placed in each of the barrels to record temperature. The control animals (7 males, 7 females) remained in the flow-through tanks at JEL( 14.1 [+ or -] 1.4 [degrees]C, mean [+ or -]SD) until transmitters were attached, and then they were released at the same location where they had been collected, which was also the same location where the bled animals were released.

The 50-gallon barrels with the treatment animals were placed outside of JEL in direct sunlight for 4 h, or next to a space heater in the JEL greenhouse (depending on the temperature and ambient sunlight during the selected day), to replicate the duration of time spent on the deck of a boat or a dock prior to transport to biomedical facilities. The average temperature that the animals experienced during this time was 32.6 [+ or -] 2.7 [degrees]C. After the first 4 h, the barrels were placed in the back of a car and were driven for an additional 4 h, to simulate time spent in a truck traveling to a bleeding facility (23.2 [+ or -] 1.7 [degrees]C). After these 4 h, the barrels were placed indoors at JEL for 16 h, to simulate time spent overnight at a bleeding facility (20.7 [+ or -] 0.6 [degrees]C). Finally, after 16 h, hemolymph was extracted as described below.

Hemolymph extraction. Hemolymph was extracted following the procedure of Armstrong and Conrad (2008), with modifications from Anderson et al. (2013). The arthrodial membrane between the prosoma and the opisthosoma of each horseshoe crab was exposed, and the hinge joint was sterilized with 70% ethanol. An 18-gauge syringe needle was inserted into the membrane, and hemolymph was collected in pre-chilled 50-mL conical tubes until 30% of total hemolymph volume had been reached or until the blood flow stopped. Total hemolymph volume for each individual was calculated using the following equation from Hurton et al. (2005):

[mathematical expression not reproducible].

Post-bleeding treatment. To minimize the amount of handling the treated horseshoe crabs experienced, immediately after they were bled transmitters were attached to them, and they were returned to their respective barrels. For the control animals, transmitters were attached at the same time, but they were returned to the flow-through tanks at JEL. The bled animals remained in their barrels overnight to replicate a second night at a bleeding facility (20.5 [+ or -]1.1 [degrees]C). Then the barrels were placed back into a car for 4 h (21.9 [+ or -] 0.9 [degrees]C) to simulate transportation back to the dock, where they would be loaded on vessels and returned to their capture location. Finally, all 28 horseshoe crabs were returned to their collection site at Adams Point and released into the estuary.

A caustic telemetry/tracking

An array of VR2W acoustic monitoring receivers (n =11; 69 kHz; VEMCO) was set up throughout the estuary, ranging from Fox Point to the Great Bay Discovery Center (Fig. 1). The receivers were deployed ~0.5 km apart and were attached to a mooring line suspended ~5-10 m from the bottom, or placed in an empty lobster trap, depending on the depth of the listening station (some areas were only ~2 m deep at low tide). Based on previous range tests, each receiver was capable of detecting a horseshoe crab with a transmitter attached when it was within 200-500 m of the receiver. However, this range varied as a result of currents, turbidity, topography, weather events, and high winds. The transmitters were programmed to take depth (m) and triaxial (x, y, z) acceleration (m [s.sup.-2]) readings at 5 Hz within a 25-s period. Then, every 70-140 s the acceleration data (root mean square of the three axes [m [s.sup.-2] = ([x.sup.2] + [y.sup.2] + [z.sup.2]) averaged over T, time]) obtained during the most recent 25-s period (T), or the depth at that time, were transmitted (depth and acceleration transmissions would alternate). Based on specifications for these transmitters, the proportion of time for which acceleration was measured was only 12% of every 2-5 min. Each time a transmission was received, the receiver would record the date and time and either the acceleration or the depth of the animal.

Receivers were downloaded in VUE software 2.3.0 (VEMCO) every one to two weeks in the spring and summer and about once every month in the fall, and they were removed from the water in the winter after the date the transmitters were programmed to shut off. After each download, if the receiver had multiple detections of different animals, it would be kept in the same location. If the receiver did not contain a viable number of detections, we would move it to a more suitable location. HOBO temperature data loggers were attached to several of the receiver mooring lines to record water temperature every five minutes for the duration of the project. The temperature data loggers were placed near the release site, at JEL, and in the middle of Little Bay. Near the middle of Great Bay. temperature data were collected by a buoy that was deployed by the NOAA National Estuarine Research Reserve System (NERRS). This buoy was located on the outskirts of the deep channel, with the temperature logger ~ 1 m below the surface of the water.

A VR100 acoustic receiver and a VH165 omni-directional hydrophone (VEMCO) were also used to manually track horseshoe crabs. The hydrophone was plugged into the VR100 receiver and was slowly towed behind a research vessel to locate tagged horseshoe crabs. If a horseshoe crab was within range, the geographical position (GPS coordinates), depth, and/or acceleration were logged in the VR100. These data were downloaded after each trip in VUE software and were used to help determine the location of animals and the best positions for deploying receiver stations.

Data analyses. A previously determined threshold value of 0.1 m [s.sup.-2] (Watson et al., 2016) was used to classify an animal as either active or inactive based on accelerometer tag output. Data were lumped into 10-minute bins; and if an animal exceeded the threshold value during any of the minutes in that 10-minute period, the animal was considered to have been active during that 10-minute period. These values were entered into the program ActogramJ to create actograms that could be used to determine the types of rhythms expressed by individual horseshoe crabs (Schmid et al., 2011). Periodograms, using the Lomb-Scargle method, were used to determine when animals expressed significant circatidal (~12.4-h) or circadian (~24-h) rhythms (peaks exceeding [alpha] = 0.001; tidal: 10-14-h range; daily: 22-26-h range; arrhythmic: no significant peaks).

For most analyses, we used data only from animals that were detected for more than seven days in a row in a given month. Three-way ANOVAs were used to test for effects of treatment group (bled and control), years (2016 and 2017), and sex (males and females) on days at large and also on ranges of movements in the estuary. Days at large were calculated as the first day in a season an animal was detected until the last day an animal was detected. Ranges were measured as the distance from the animal's farthest up-estuary position to its farthest down-estuary position. Two-way ANOVAs were used to test for effects of treatment groups and sex on mating and also on depth changes. A MANOVA was used to look at the impacts of sex, treatment groups, and months on percentage of time active and depth in different animals. In all cases, the time active, movements, and rhythms of control animals were compared to the experimental animals. Tukey's honest significant difference post hoc analyses (with a level of significant difference set at P < 0.05) were used to examine differences between means of treatment groups, monthly depths, rhythms expressed at different depths, and sexes. Correlational and single linear regression analyses were used to determine relationships between temperature and years, as well as between activity and depth across months. Unpaired Student's t tests were used to compare depth or activity between treatment groups.

To determine whether horseshoe crabs were approaching spawning beaches, changes in animals' depths and tide heights were examined together, along with their activity. Water depth and tide height data were obtained from the Squamscott River Monitoring Station (data provided by the NOAA Tide Predictions website, https;//tidesandcurrents.noaa.gov/noaatidepredictions.html?id=8422687) and the Great Bay Buoy (data provided through the Northeastern Regional Association of Coastal Ocean Observing Systems. Portsmouth, NH, and University of New Hampshire Center for Coastal Ocean Observation and Analysis, Durham, NH). Horseshoe crabs were considered "spawning" if they showed high activity levels (>0.1 m [s.sup.-2]; Watson et al., 2016) around the times of high tides, commensurate with a movement to a shallower location, during the peak of the spawning period in 2016. Only data from the first two weeks after release were used for these analyses, for two reasons: (1) this time period represented the peak of the spawning season in 2016 in the GBE, and (2) most of the animals were still within the VR2W array and near the spawning beaches where we collected them, which enabled us to obtain the maximum possible amount of data from the most animals before they dispersed. For each animal we determined how many high tides occurred while we were collecting continuous data from them. Then, to determine the spawning percentage, we divided the number of apparent spawning events (identified as described above) by the number of available high tides they experienced.

Results

Days at large and days delected

A total of 28 horseshoe crabs were successfully tracked in the GBE, from May 15 to December 6, 2016 (205 days); and data were also obtained from 23 of these between April 14 and October 4,2017 (191 days). In 2016, animals were at large for an average of 158 [+ or -] 59 days (max = 205, min = 7) and were detected for 84.3 [+ or -] 50 days (max = 180, min = 8). In 2017, the 23 animals were at large for 91 [+ or -] 56 days (max = 172, min = 1) and were detected for 32.4 [+ or -] 41 days (max = 171, min = 0).

Potential mating events

Based on data from New Hampshire Limulus spawning surveys, which are conducted each year, including during this study (Cheng et al., 2016), the mating season for 2016 started on May 9, when temperatures reached 11.2 [degrees]C; it peaked around May 21-26; and it ended around June 10. The mating season for 2017 started on May 16, when temperatures reached 11.3 [degrees]C; it peaked around June 10-16; and it ended around June 20. We used these data to focus our analyses of potential impacts of the bleeding process on mating behavior on the peak spawning period that occurred right after animals were released.

The criteria outlined in Materials and Methods were used to determine whether horseshoe crabs approached spawning beaches to possibly mate (Fig. 2). Out of all the times animals were detected during each high tide within the first week after they were released, females appeared to mate less than males, bled animals less than controls, and bled females much less than control females (Fig. 3; Table A1). More precisely, control females appeared to spawn 4.8 [+ or -] 2.5 times during the first week after they were released, while bled females spawned only 2.0 [+ or -] 0.7 times. Out of the animals that were detected for at least two weeks, there was no difference in apparent mating events between the first and second weeks post-release (Student's paired t test, P = 0.16; data not shown). These analyses were not performed in 2017 because there were not enough animals pre.sent near the spawning areas that were being monitored with VR2W receivers, and they did not appear to approach beaches to mate for a long enough time period to provide sufficient data for a rigorous analysis.

Biological rhythms

2016. Horseshoe crabs from which we obtained sufficient data for biological rhythm analyses (i.e., at least seven days of continuous data) exhibited both tidal and daily rhythms. In June, all animals expressed tidal rhythms (n = 12; r = 12.4 h; Table 1; Fig. 4). This corresponds to the mating season in 2016 in GBE. In the following months they exhibited a combination of arrhythmic, daily, and tidal rhythms (Table 1; Figs. Al, A2). While there were more control animals detected, the majority of both treatment groups expressed clear daily or tidal rhythms from June to October 2016. When we added up all of the months during which a given animal expressed a clear rhythm, control animals were rhythmic during 30 of 32 months, while bled animals expressed clear biological rhythms in 15 of 16 months (Table 1). All animals that expressed daily rhythms ([tau] = 24 h; Fig. Al) were more active during the day than at night, except for Animal 75, which was more active during the night in July and September. There was no significant difference between the depths where animals resided in months that they displayed daily rhythms (4.4 [+ or -] 1.4 m) versus tidal rhythms (3.3 [+ or -] 1.3 m; unpaired Student's t test, P =0.5). This was also true for the relationship between depths and rhythms for animals that switched from one kind of rhythm to another (Fig. A2). There was also no clear relationship between sexes and the expression of different rhythms (Table 1). Finally, four animals switched from a tidal rhythm to a daily rhythm in July, directly after the mating season (Table 1; Fig. A2).

2017. In 2017, these same animals continued to express the same variety of rhythms (Table 1). Since different animals were detected at different times in 2017, there was a lower sample size and a greater distribution of rhythms, so we have chosen to discuss the two years separately. In April, before the start of the mating season, horseshoe crabs exhibited tidal rhythms or were arrhythmic. Interestingly, all of the bled animals (n = 4) were arrhythmic. After April, there was no clear difference between the types of rhythms exhibited by both groups. In June, only 1 animal expressed a tidal rhythm in 2017, while 13 did so in 2016; but there were also fewer animals detected overall. As in 2016, all animals that displayed daily rhythms were more active during the day than at night. In contrast to 2016, there was a significant relationship between the depths occupied by animals and the types of rhythms they expressed. Animals that were arrhythmic resided in deeper water (9.8 [+ or -] 4.5 m) than animals that expressed tidal rhythms (4.7 [+ or -] 2.4 m) or daily rhythms (4.6 [+ or -] 1.1 m; ANOVA, [F.sub.2,21] = 8.22, P = 0.002).

Ranges of animal movements

All of the horseshoe crabs appeared to remain within the GBE for the duration of the study. The farthest an animal moved from the release site toward the coast was 3.2 km, and the farthest an animal traveled up-e.stuary was 3.4 km. In 2016, the mean annual range of movement (distance from the animal's farthest up-estuary position to its farthest down-estuary position) was 3.3 [+ or -] 1.7 km, and in 2017, it was 3.2 [+ or -] 1.3 km. There was no difference in range for years, between treatments, or between sexes (Table A2).

In 2017, 5 out of the 23 detected animals returned to their release site (Fig. A3). Four of these animals returned during the mating season in May-June. All four were males, three were controls, and one was a bled animal. The fifth animal was a control female that returned to the release site in late July. Because this is generally two to three weeks after horseshoe crabs stop spawning in the GBE (Schaller et al., 2010; Cheng et al., 2016; Watson et al., 2016), it probably was not spawning at that time.

Overall seasonal changes in behavior

The water temperature in the GBE ranged from 9.5 to 25.4 [degrees]C during the time when we were tracking animals, and there was no difference in the mean water temperature between the 2 years of the study (2016: 18.9 [+ or -] 3.6 [degrees]C; 2017: 18.12 [+ or -] 3.2 [degrees]C; P < 0.0001). We also ran a regression comparing the daily temperatures in each year, and they were highly correlated (r = 0.85, P< 0.0001).

The majority (21 of 23) of the animals that were detected in 2017 were first detected near the location where they were last detected in 2016. The 2 animals that were not detected at their exact location from the previous year were still in Little Bay but [+ or -]2.3 km up- or down-estuary from their last location. Therefore, although transmitters were turned off during the winter season, all of the animals appeared to have remained in the same location throughout the winter months.

Seasonal changes in depth and activity. Only animals that were detected and active for at least seven days in a given month were used for depth and activity analyses (2016: n = 21;2017:n = 11). In 2016, there was a significant correlation between activity and depth ([r.sup.2] = -0.35, P = 0.01). with animals being more active when they were in shallower water (Fig. A4). In addition, animals tended to be more active from May to August when water temperatures were warmer ([greater than or equal to]~18 [degrees]C). There was no difference between bled and control animals in terms of the percentage of time when they were active (unpaired Student's t test, P = 0.4). As water temperatures started to decrease in the fall, animals moved deeper and exhibited a lower amount of activity, and bled animals spent significantly more time in deeper water in 2016 than control animals (Fig. 5; Table A3). Finally, 15 of the 21 animals that migrated from Great Bay to Little Bay (toward the coast) moved in late July to early August when temperatures peaked at 22 [degrees]C. The following year, 14 of 15 of those animals moved back into Great Bay when temperatures reached 11.2 [degrees]C in May.

In 2017. only four animals (three males and one female) were detected continuously, and were active, for at least seven days in a month. Based on data from only these four animals, there was a significant difference between month and depth (MANOVA. [F.sub.6,19] = 3.32, P = 0.021) but no significant difference between activity and month (MANOVA. [F.sub.6]= 1.410, P = 0.262).

Summary of seasonal movements. Most of the animals in this study followed the same seasonal trends of activity and depth preferences that were reported in two previous studies in the GBE (Fig. 6; Schaller et al., 2010; Watson et al., 2016). Animals overwintered in deep water and remained there until water temperatures started to exceed ~11 [degrees]C in the spring. At this time, they moved up into the estuary, where it tends to be shallower, and they had a higher level of activity during warmer months of the year. Then in the fall they returned to deeper water as temperatures began to drop. It should also be noted that when water temperatures exceeded 20 [degrees]C, 8 of the animals in 2017 moved out of Great Bay to Little Bay, where the water is a bit cooler, which is exactly what these same animals did at the same time in 2016 (late July to early August).

Between 2016 and 2017, there were no major differences in the seasonal migration trends, or annual ranges, of bled versus control animals (Fig. 6). However, there were some distinct changes in the seasonal movements of some of the bled animals from one year to the next. One noticeable trend was that in May and June of 2017, the bled animals did not approach shallower areas in Great Bay, preferring to remain in deeper channels. Out of all of the animals that were detected in 2017 (regardless of whether they were also detected at the same time in 2016), bled animals remained in deeper waters than controls (unpaired Student's t test, P < 0.001).

Discussion

While some of the impacts of the biomedical bleeding process have been studied in the laboratory and in the field, this seems to be one of the first comprehensive studies designed to determine the behavioral impacts that the bleeding process has on the horseshoe crabs that are released back into their natural environment. The two most obvious impacts we found were that (1) bled females appeared to approach beaches to mate less frequently than control females, and (2) bled animals remained in deeper water than control animals during certain times of the year. However, there was no difference in the annual ranges of bled versus control animals or in their overall distribution in the GBE across months or years.

Mating

Since it is well known that horseshoe crabs mate at high tide at spawning beaches during the spring and early summer (Rudloe, 1980; Cohen and Brockmann. 1983; Barlow et al., 1986; Cheng et al., 2016), it was possible to use a combination of depth, tide, and accelerometer data to determine possible mating events of animals fitted with appropriate acoustic transmitters. We found that control animals probably mated, or at least moved inshore toward mating beaches during the high tides, more often than bled animals during the first tracking season, especially when comparing bled versus control females. These data suggest that the bleeding process might have an immediate impact on the reproductive output of female horseshoe crabs for that season. These data are also consistent with the decreases in activity in bled animals that were documented by Anderson et al. (2013) in previous laboratory experiments. Since female horseshoe crabs are preferentially selected for in the biomedical bleeding process because of their size (James-Pirri et al., 2012), and their mortalities are consistently greater than males (Leschen and Correia, 2010; James-Pirri, 2012; Krisfalusi-Gannon el al., 2018), both the lethal and sublethal impacts of the bleeding process can alter the effective sex ratio and reproductive output of the population. For example, in several areas where horseshoe crabs are harvested, such as Pleasant Bay, Massachusetts, there has already been a 20% decrease in female appearances at spawning beaches (Carmichael et al., 2003; Malkoski, 2010); and egg abundances have also significantly decreased at spawning sites (James-Pirri, 2012). Although Massachusetts regulations do not permit the harvesting of horseshoe crabs during the 5 days including and surrounding the new and full moons during the mating season, this does not always coincide with the peak of the spawning season; and many horseshoe crabs mate during other high tides during the ~ 1.5-month-long spawning season (Smith et al., 2010; Cheng et al., 2016). Therefore, it might be prudent to revisit harvest regulations and consider not harvesting horseshoe crabs at all during the majority of the spawning period.

Males are more frequently detected at mating beaches than females (Loveland and Botton, 1992). Furthermore, females usually only spawn several times within one week and do not return until the next year, whereas males retum to these beaches more often (Rudloe, 1980; Brockmann and Penn, 1992: Leschen et al., 2006). This trend was also evident in this study, further supporting the hypothesis that the skewed sex ratios observed near most mating locations are due to behavioral differences between males and females and are not indicative of the actual sex ratios of the local populations.

In this study, we restricted our analyses of mating behavior impacts to just 2 weeks during the peak of the mating season, allowing for ~ 14-15 potential mating events per week. Control animals appeared to mate, on average, three to four times a week, whereas bled animals mated one to two times per week. For control animals, this meant that they were likely to be mating during 22%-29% of the potential high tides, while bled animals appeared to be mating during only 7%-14% of them. Although these numbers are consistent with the results reported by Watson and Chabot (2010) and Brousseau et al. (2004), they are slightly lower. This difference could be due to two main factors. First, because our animals were released midway into the spawning season and because we examined only two weeks of data, some of our subjects might have already mated, or attempted to mate, several times before they were captured. Second, environmental factors could have affected spawning. During our spawning surveys, in both years, we noted that there was (1) heavy rain, (2) increased detritus on beaches on days immediately following large storms, and (3) overcast conditions. All of these factors appear to be correlated with fewer mating animals. A similar pattern was observed during spawning surveys in the GBE in 2012 and 2013, when periods of heavy rain were correlated with periods of decreased spawning activity (Cheng el al., 2016).

Behavioral rhythms

Horseshoe crabs possess endogenous clocks (Chabot et al., 2007) that allow them to anticipate changes in environmental factors, specifically changes in tides, so they can synchronize their spawning and foraging activities to particular phases of the tides (Cohen and Brockmann, 1983; Barlow et al., 1986; Watson and Chabot, 2010). Any disruptions to the rhythms controlled by these clocks could have negative implications for reproduction and survival. In 2016, all animals that expressed activity in June had tidal rhythms. This is likely because this was the spawning season, when they approach mating beaches at high tide (Rudloe, 1980; Shuster and Botton, 1985; Barlow et al., 1986). During the following summer, animals expressed both tidal and daily rhythms or became arrhythmic; and then in October no animals expressed tidal rhythms. This same seasonal transition was reported in previous studies, and it is likely due to the fact that as winter approaches, crabs move into deeper waters and prepare for overwintering (Chabot and Watson, 2010; Watson et al., 2016).

In 2016, animals that had daily rhythms were more active during the day than at night. It has been previously argued that horseshoe crabs spawn more and increase their activity at night (Cavanaugh, 1975; Rudloe, 1980, 1981; Barlow, 1983; Barlow et al., 1986; Finn et al., 1990; Swan et al., 1991, 1993; Smith et al., 2010). In the GBE, this hypothesis has not been supported by previous telemetry studies or spawning surveys; in contrast, horseshoe crabs have actually been shown to be significantly more active during the day (Watson et al., 2009; Watson and Chabot, 2010). The data from this study are consistent with this pattern of activity.

In 2017, there appeared to be some impact of the bleeding process on the expression of biological rhythms, as indicated by the fact that more bled animals were arrhythmic (four of nine animals) in 2017 than control animals (zero of three animals). For example, Animal 70, a bled female, remained arrhythmic throughout all of the months that it was active, including during the spawning season. However, the animals that were arrhythmic were also deeper and mostly manifested these behaviors in April, before the start of the spawning season, which is not too unusual. They usually do not begin to express clear rhythms until temperatures increase in late spring (10-11 [degrees]C; Schaller et at., 2010). Nevertheless, two other bled animals were also arrhythmic in May and June during the spawning season, and this could be a result of having been bled, especially considering that all of the horseshoe crabs expressed tidal rhythms in June in 2016 but only two of five bled animals in May and June of 2017. However, it should also be noted that only one of three control animals expressed a tidal rhythm in 2017, so other factors could have played a role. Therefore, it is possible that the bleeding process could have delayed and extended impacts on the expression of rhythms and depth preferences, and thus on both foraging and mating behaviors. However, we have no explanation for how the bleeding process could lead to behavioral impacts that manifest themselves about a year after the procedure. Clearly, more studies are necessary to confinn that bled animals tend to reside in deeper water and approach mating beaches less often than control animals many months after the procedure.

Movements and migrations

The 28 animals that were tracked in 2016 appeared to remain within the GBE, along with the 23 that were detected in 2017. In fact, 21 of the 23 animals detected in 2017 were first detected at the same location where they were last detected in 2016. This is consistent with previous findings in the GBE (Schaller et al., 2010), as well as in other embayments, such as Pleasant Bay, Massachusetts (James-Pirri, 2010). Horseshoe crabs are thought to be philopatric to the embayments where they spawn, and in New England specifically, horseshoe crab populations appear to be more localized and do not seem to migrate offshore (Baptist et al., 1957; Botton and Ropes, 1987; James-Pirri et al., 2005; Moore and Perrin, 2007). Therefore, if horseshoe crabs are harvested from a particular area, it is unlikely that they would be replenished from adults or larvae from a different area. In addition, if they are removed from one area, bled, and returned to a different area, they might not have the ability to adjust their migration and spawning behaviors to the new region.

There were no clear differences in the annual range of movements or seasonal migrations between control and bled animals. These annual ranges were similar to those found in Cape Cod, Massachusetts, and Maine embayments (Kurz and James-Pirri, 2002; James-Pirri et al., 2005) and followed the same patterns that have been described in other New England estuaries and in the GBE. In the spring, when the water temperature reached 10-11 [degrees]C (Schaller et al., 2010), the horseshoe crabs moved up-estuary to shallow areas near mating beaches. In 2017, this was clearly shown in our data because the transmitters turned on before the mating season. There was a clear trend of animals moving from deeper waters where they stayed in April to shallower waters in May, when the temperature threshold was reached. Then, after mating, when the estuary was the warmest in July and August, animals moved down-estuary; and they eventually overwintered at the farthest down-estuary portion of their annual range. As previously mentioned, in 2017, when our transmitters re-activated, 21 of the 23 animals were in the same location where they were last detected in December 2016, when the transmitters were de-activated. These seasonal movements are thought to be driven by temperature preferences, with animals seeking warmer water during the spawning season, then moving toward the coast when the water temperature rises above ~20 [degrees]C, and finally seeking deep water in the fall, which tends to be warmer in the winter (Schaller et al., 2010). These thermal preferences could be an important factor driving mating behaviors because temperature plays a large role in egg development (French, 1979). Moreover, recent studies have demonstrated that horseshoe crabs can detect changes in water temperature and that, when given a choice, they prefer slightly warmer water (Cheng, 2015).

Although there was no significant difference in the annual range of movements between bled and control animals, or in the overall seasonal migration trends, there were some distinct changes in the seasonal movements of some animals from one year to the next. Most of the control animals followed similar patterns, but there were several bled animals that had very different migration routes between 2016 and 2017. One noticeable trend was that in May and June, the bled animals did not approach shallower areas in Great Bay but remained in the deeper channels. This trend is consistent with data showing that bled animals remained deeper throughout the year in 2016 and that bled females mated less than control females. In other words, some of the immediate impacts of bleeding that we observed in 2016 continued in 2017. There are two potential reasons why these bled animals remained in deeper water. First, in the laboratory, disrupted orientation and a random direction of movements have been observed in bled horseshoe crabs (Anderson et al., 2013). Moreover, in one of the previous field studies comparing the impact of biomedical bleeding on movement patterns in Cape Cod, it was found that the bled group tended to have more random movements than control animals (Kurz and James-Pirri, 2002). This disorientation could prevent horseshoe crabs from locating spawning beaches and could also explain why the bled animals remained deeper and farther away from spawning beaches. Second, bled animals may not have been as motivated to spawn, and, therefore, they did not move toward spawning beaches as often as controls. This lack of motivation could be due to the fact that bleeding influenced their energy utilization and shifted it from reproduction to stress remediation. For example, Hu et al. (2011) showed that starvation of two Asian horseshoe crab species led to a decrease in their respiration rates and ammonia excretion rates, which was correlated with a decrease in scope for growth.

Interestingly, 5 out of the 23 animals in 2017 returned to their original release site where they had been captured while spawning. Four were control animals, and one had been bled. In a telemetry study completed in Delaware Bay, 77% of animals did not return to the same beach to spawn (Smith et al., 2010), so we did not necessarily expect all of the animals to show up at their original spawning location in 2017. Therefore, while the 4 : 1 ratio of controls to bled animals is interesting, it is possible that this relates not to an inability of bled animals to find the same spawning beaches from year to year but to the fact that most animals do not return to the same spawning beach each year.

Overall impacts

Overall, there were some immediate impacts of the bleeding procedure on mating behaviors and the expression of biological rhythms, and there also appeared to be a tendency for bled animals to reside in deeper water during both the first and second years of the study. If bled animals, especially females, have alterations in their biological rhythms and mating behaviors, it is likely to further alter the sex ratio on spawning beaches, reduce reproductive output, lower population levels, and decrease the fitness and survival of this keystone species. Further studies should investigate the immediate effects of the bleeding process on horseshoe crabs during the mating season to obtain a more precise and accurate assessment of their behavior in the vicinity of mating beaches. Also, the orientation of bled animals in their natural habitat should be examined because one laboratory study and one field study indicate that some bled horseshoe crabs do not orient in the same manner as normal adult horseshoe crabs. Taken together with the results of many previous investigations, it appears as if the biomedical bleeding process might have some behavioral impacts that could impact the sustainability of harvested Limulus populations; and, thus, there is a need for further investigations into possible improvements that might reduce these effects.

Acknowledgments

This research project could not have been completed without the help of Dave Shay and the staff the University of New Hampshire Jackson Estuarine Laboratory. We would also like to thank the graduate and undergraduate students at the University of New Hampshire and Plymouth State University for all of their hard work and support. This study was supported by a Leslie S. Hubbard Marine Program Endowment grant and Marine Biology Graduate Program grant to MO and a New Hampshire Sea Grant (R/HCE-4) to CC and WW.

Literature Cited

Adachi, K., H. Endo, T. Watanabe, T. Nishioka, and T. Hirata. 2005. Hcmocyanin in the exoskelelon of crustaceans: enzymatic properties and immunoiocalizalion. Pigment Cell Res. 18: 136-143.

Anderson, R. L., W. H. Watson HI, and C. C. Chabot. 2013. Sublethal behavioral and physiological effects of the biomedical bleeding process on the American horseshoe crab, Limulus polyphemus. Biol. Bull. 225: 137-151.

Armstrong, P., and M. Conrad. 2008. Blood collection from the American horseshoe crab. Limulus polyphemus. J. Vis. E.xp. 20: 958.

ASMFC (Atlantic States Marine Fisheries Commission). 1998. Fishery Management Report No. 32 of the Atlantic States Marine Fisheries Commission. Interstate Fishery Management Plan for Horseshoe Crab. Arlington, VA.

ASMFC (Atlantic States Marine Fisheries Commission). 2012. Review of the Fishery Management Plan in 2011 for Horseshoe Crab (Limulus polyphemus). Interstate Fishery Management Plan for Horseshoe Crab. Arlington. VA.

ASMFC (Atlantic States Marine Fisheries Commission). 2013. Horseshoe Crab Stock Assessment Update. Atlantic States Marine Fisheries Commission. Washington, DC.

Baptist, J. O. R., J. Smith, and W. Ropes. 1957. Migrations of the Horseshoe Crab Limulus polyphemus in Plum Island Sound. Massachusetts. Special Scientific Report--Fisheries No. 220. U.S. Fish and Wildlife Service. Washington. DC.

Barlow, R. B. 1983. Circadian rhythms in the Limulus visual system. J. Neurosci. 3: 856-870.

Barlow, R. B. 2001. Circadian and efferent modulation of visual sensitivity. Prog. Brain Res. 131: 487-503.

Barlow, R. B., J. M. Hitt, and F. A. Dodge. 2001. Limulus vision in the marine environment. Biol. Bull. 200: 169-176.

Barlow, R. B., Jr., M. K. Powers, H. Howard, and L. Kass. 1986. Migration of Litnulus for mating: relation to lunar phase, tide height, and sunlight. Biol. Bull. 171: 310-329.

Botton, M. L., and J. W. Ropes. 1987. Populations of horseshoe crabs Limulus polyphemus on the northwestern Atlantic and continental shelf. Fish. Bull. 85: 80.5-812.

Botton, M. L., R. A. Tankerslcy, and R. E. Loveland. 2010. Developmental ecology of the American horseshoe crab Limulus polyphemus. Cuir. Zool. 56: 550-562.

Brockniann, H. J., and D. Penn. 1992. Male mating tactics in the horseshoe crab Limulus polyphemus. Anim. Behav. 44: 653-665.

Brousseau, L. J., M. Selafani, D. R. Smith, and D. B. Carter. 2004. Acoustic-tracking and radio-tracking of horseshoe crabs to assess spawning behavior and subtidal habitat use in Delaware Bay. N. Am. J. Fish Manag. 24: 1376-1384.

Carmichael, R. H., D. Rutecki, and I. Valiela. 2003. Abundance and population structure of the Atlantic horseshoe crab Limulus polyphemus in Pleasant Bay. Cape Cod. Mar Ecol. Prog. Ser. 246: 225-239.

Cavanaugh, C. M. 1975. Observations on mating behavior in Limulus polyphemus. Biol. Bull. 149: 422. (Abstract).

Chabot, C. C., and W. H. Watson III. 2010. Circatidal rhythms of locomotion in the American horseshoe crab Limulus polyphemus: underlying mechanisms and cues that influence them. Curr Zool. 56: 499-517.

Chabot, C. C., S. H. Betournay, N. R. Braley, and W. H. Watson HI. 2007. Endogenous rhythms of locomotion in the American horseshoe crab, Limulus polyphemus. J. Exp. Mar. Biol. Ecol. 345: 79-89.

Chen, L., and N. Mozier. 2013. Comparison of Limulus amebocyte lysate lest methods for endotoxin measurement in protein solutions. J. Pharm. Biomed.Anal. 80: 180-185.

Cheng, H. 2015. The environmental influences on American horseshoe crab (Limulus polyphemus) behavior and distribution in Great Bay Estuary, New Hampshire U.S.A. M.S. thesis. University of New Hampshire, Durham.

Cheng, H., C. C. Chabot, and W. H. Watson HI. 2016. Influence of environmental factors on spawning of the American horseshoe crab (Limulus polyphemus) in the Great Bay Estuary. New Hampshire. USA. Estuaries Coasts 39: 1142-1153.

Coates, C. J., S. M. Kelly, and J. Nairn. 2011. Possible role of phosphatidylserine-hemocyanin interaction in the innate immune response of Limulus polyphenus. Dev. Comp. Immunol. 35: 155-163.

Cohen, J., and H. Brockmann. 1983. Breeding activity and mate selection in the horseshoe crab, Limulus polyphemus. Bull. Mar Sci. 33: 274-281.

Finn, J. J., C. N. Shuster, and B. L. Swan. 1990. Limulus Spawning Activity on Delaware Bay Shores 1990. Finn Tech Industries. Cape May. NJ.

French, K. A. 1979. Laboratory culture of embryonic and juvenile Limulus. Pp. 61-71 in Biomedical Applications of the Horseshoe Crab (Limulidae), E. Cohen, ed. Alan R. Liss, New York.

Herzog, E. D., M. K. Powers, and R. B. Barlow. 1996. Limulus vision in the ocean day and night: effects of image size and contrast. Vis. Neurosci. 13: 31-41.

Hu, M., Y. Wang, S. T. Tsang, S. G. Cheung, and P. K. S. Shin. 2011. Effect of starvation on the energy budget of two Asian horseshoe crab species: Tachypleus Iridentatus and Carcinoscorpius rotundicauda (Chelicerata: Xiphosura). Mar Biol. 158: 1591-1600.

Hurton, L., and J. Berkson. 2006. Potential causes of mortality for horseshoe crabs (Limulus polyphemus) during the biomedical bleeding process. Fish. Bull. 104: 293-298.

Hurton, L., J. Berkson, and S. Smith. 2005. Estimation of total hemolymph volume in the horseshoe crab Limulus polyphenms. Mar. Freshw. Behav. Physiol 38: 139-147.

James-Pirri, M. J. 2010. Seasonal movement of the American horseshoe crab Limulus polyphemus in a semi-enclo.sed bay on Cape Cod, Massachusetts (USA), as determined by acoustic telemetry. Curr. Zool. 56: 575-586.

James-Pirri, M. J., K. Tuxbury, S. Marino, and S. Koch. 2005. Spawning densities, egg densities, size structure, and movement patterns of spawning horseshoe crabs Limulus polyphemus within four coastal embayments on Cape Cod, Massachusetts. Estuaries 28: 296-313.

James-Pirri, M. J., P. A. Veillette, and A. S. Leschen. 2012. Selected hemolymph constituents of captive, biomedically bled, and wild caught adult female American horseshoe crabs [Limulus polyphemus). Mar. Freshw. Behav. Physiol. 45: 281-289.

Krauter, J. N., and S. R. Fegley. 1994. Vertical disturbance of sediments by horseshoe crabs (Limulus polyphemus) during their spawning season. Estuaries 17: 288-294.

Krisfalusi-Gannon, J., A. Waleed, K. Dellinger, L. Robertson, T. E. Brady, M. K. M. Goddard, R. Tinker-Kulberg, C. L. Lepley, and A. L. Dellinger. 2018. The role of horseshoe crabs in the biomedical industry and recent trends impacting species sustainability. Front. Mar. Sci. 5: 1-13.

Kurz, W., and M. J. James-Pirri. 2002. The impact of biomedical bleeding on horseshoe crab, Limulus polyphemus, movement patterns on Cape Cod. Massachusetts. Mar. Freshw. Behav. Physiol. 35: 261-268.

Le Moullac, G., and G. Haffner. 2000. Environmental factors affecting immune responses in Crustacea. Aquaculiure 191: 121-131.

Lee, W. J. 2010. Intensive use of an intertidal mudflat by foraging adult American horseshoe crabs Limulus polyphemus in the Great Bay Estuary, New Hampshire. Curr. Zool. 56: 611-617.

Leschen, A. S., and S. J. Correia. 2010. Mortality in female horseshoe crabs (Limulus polyphemus) from biomedical bleeding and handling: implications for fisheries and management. Mar. Freshw. Behav. Physiol. 43: 135-147.

Leschen, A. S., S. P. Grady, and L Valiela. 2006. Fecundity and spawning of the Atlantic horseshoe crab, Limulus polyphemus, in Pleasant Bay, Cape Cod, Massachusetts. USA. Mar Ecol. 27: 54-65.

Loveland, R. E., and M. L. Botton. 1992. Size dimorphism and the mating system in horseshoe crab, Limulus polyphemus. Anim. Behav. 44: 907-916.

Malkoski, V. 2010. Massachusetts 2010 Compliance Report to the Atlantic States Maiirie Fisheries Commission--Horseshoe Crab. Massachusetts Division of Marine Fisheries, New Bedford, MA.

Moore, S., and S. Perrin. 2007. Seasonal movement and resource-use patterns of resident horseshoe crab (Limulus polyphemus) populations in a Maine, USA, estuary. Estuaries Coasts 30: 1015-1026.

Novitsky, T. J. 2009. Biomedical applications of Limulus amebocyte lysate. Pp. 315-330 in Biology and Con.sen'ation of Horseshoe Crabs, J. T. Tancredi, M. L. Botton, and D. R. Smith, cds. Springer. New York.

Powers, M. K., and R. B. Barlow, Jr. 1985. Behavioral correlates of circadian rhythms in the Limulus visual system. Biol. Bull. 169: 578-591.

Rudloe, A. 1980. The breeding behavior and patterns of movement of horseshoe crab, Limulus polyphemus, in the vicinity of breeding beaches in Apalachee Bay, Florida. Estuaries 3: 177-183.

Rudloe, A. 1981. Aspects of the biology of juvenile horseshoe crabs. Limulus polyphemus. Bull. Mar Sci. 31: 125-133.

Rudloe, A. 1983. The effect of heavy bleeding on mortality of the horseshoe crab, Limulus polyphemus. in the natural environment. J. Invertebr. Pathol. 42: 167-176.

Rutecki, D., R. H. Carmichael, and L Valiela. 2004. Magnitude of harvest of Atlantic horseshoe crabs, Limulus polyphemus. in Pleasant Bay, Massachusetts. Estuaries 27: 179-187.

Schaller, S. Y., C. C. Chabot, and W. H. Watson III. 2010. Seasonal movements of American horseshoe crabs Limulus polyphemus in the Great Bay Estuary. New Hampshire (USA). Curr. Zool. 56: 587-598.

Schmid, B., C. Helfrich-Forster, and T. Yoshil. 2011. A new ImageJ plugin "ActogramJ" for chronobiological analyses. J Biol. Rhythms 26: 464-467.

Shuster, C. N., Jr., and M. L. Botton. 1985. A contribution to the population of horseshoe crabs, Limulus polyphemus (L.), in Delaware Bay, Estuaries 8: 63-372.

Smith, D. R., L. J. Brousseau, M. T. Mandt, and M. J. Millard. 2010. Age and sex specific timing, frequency, and spatial distribution of horseshoe crab spawning in Delaware Bay: insights from a large-scale radio telemetry array. Curr. Zool. 56: 563-574.

Swan, B. L., W. R. Hall, and C. N. Shuster. 1991. Limulus Spawning Activity on Delaware Bay Shores on 25 May 1991. Delaware National Estuarine Research Reserve. Dover.

Swan, B. L., W. R. Hall, and C. N. Shuster. 1993. Limulus Spawning Activity on Delaware Bay Shores on 5 June 1993. Delaware National Estuarine Research Reserve. Dover.

Thompson, M. 1998. Assessments of the population biology and critical habitat for the horseshoe crab. Limulus polyphemus. in the South Atlantic Bight. M.S. thesis. University of Charleston, Charieslon, SC.

Walls, E. A., and J. M. Berkson. 2000. Effects of blood extraction on the survival of the horseshoe crab, Limulus polyphemus. Va. J. Mar. Sci. 51: 195-198.

Walls, E. A., and J. M. Berkson. 2003. Effects of blood extraction on horseshoe crabs (Limulus polyphemus). Fish. Bull. 101: 457-459.

Walls, E. A., J. Berkson, and S. A. Smith. 2002. The horseshoe crab, Limulus polyphemus: 200 million years of existence, 100 years of study. Rev. Fish. Sci. 10: 39-73.

Watson, W. H. HI, and C. C. Chabot. 2010. High-resolution tracking of adult horseshoe crabs Limulus polyphemus in a New Hampshire estuary using fixed array ultrasonic telemetry. Curr. Zool. 56: 599-610.

Watson, W. H. III, S. Y. Schaller, and C. C. Chabot. 2009. The relationship between small- and large-scale movements of horseshoe crabs in the Great Bay estuary and Limulus behavior in the laboratory. Pp. 131-147 in Biology and Consen'alion of Horseshoe Crabs, J. T. Tanacredi, M. L. Botton, and D. R. Smith, eds. Springer, New York.

Watson, W. H. III, S. K. Johnson, C. D. Whitworth, and C. C. Chabot. 2016. Rhythms of locomotion and seasonal changes in activity expressed by horseshoe crabs in their natural habitat. Mar. Ecol. Prog. Ser. 542: 109-121.

Appendix
Table A1 Two-way ANOVA results from likely mating events for horseshoe
crabs tracked in 2016 comparing different sexes (males and females) and
treatments (bled and controls)

                     Likely mating events
Source           df     MS       F      P

Sex               1  2568.405  6.019  0.024
Treatment         1  3800.770  8.906  0.008
Sex x treatment   1  2449.484  5.740  0.027
Error            19   426.746
Total            23

Significant differences are shown in bold (P < 0.05). MS, mean square.

Table A2 Three-way ANOVA results comparing the ranges of horseshoe
crabs tracked in 2016 and 2017 comparing different years (2016 and
2017), sexes (males and females), and treatments (bled and controls)

                                Range
Source                  df    MS     F      P

Sex                         1.555  0.613  0.438
Treatment                   0.475  0.187  0.667
Year                        0.010  0.004  0.950
Sex X treatment             2.842  1.120  0.296
Sex X year                  0.153  0.060  0.807
Treatment x sex             0.017  0.007  0.936
Year x sex x treatment      4.583  1.807  0.186
Error                   40  2.537
Total                   48

MS, mean square.

Table A3 MANOVA results from depth and activity data for horseshoe
crabs tracked in 2016 comparing different sexes (males and females),
treatments (bled and controls), and months (May-November)

                                         Depth and activity
Source                             df     MS      F      P

Month                    Depth     6    36.683  3.947  0.008
                         Activity  6   309.357  0.895  0.515
Sex                      Depth     1    30.426  3.274  0.084
                         Activity  1     0.047  0.000  0.991
Treatment                Depth     1    63.789  6.863  0.016
                         Activity  1   184.634  0.534  0.472
Month x sex              Depth     5     2.676  0.288  0.915
                         Activity  5    38.196  0.111  0.989
Month x treatment        Depth     6     7.608  0.819  0.567
                         Activity  6    86.332  0.250  0.954
Sex x treatment          Depth     1     1.157  0.124  0.728
                         Activity  1    70.64   0.204  0.656
Month x sex x treatment  Depth     1     0.141  0.015  0.901
                         Activity  1    13.061  0.038  0.848
Error                    Depth     22    9.295
                         Activity  22  345.489
Total                    Depth     44
                         Activity  44

Significant differences are shown in bold. (P < 0.05). MS, mean square.


MEGHAN OWINGS (1), CHRISTOPHER CHABOT (2), AND WINSOR WATSON III (1,*)

(1) Department of Biological Sciences, University of New Hampshire, Durham, New Hampshire 03824; and (2) Department of Biological Sciences, Plymouth State University, Plymouth, New Hampshire 03264

Received 19 September 2018; Accepted 30 January 2019; Published online 15 April 2019.

(*) To whom correspondence should be addressed. Email: win@unh.edu.

Abbreviations: BMP. best manufacturing process; GBE, Great Bay Estuary; 10, interocular; JEL. Jackson Estuarine Laboratory; LAL, Limulus amebocyte lysate.

DOI: 10.1086/702917
Table 1 Rhythms expressed by individual horseshoe crabs for each month
in 2016 and 2017

                                         2016
Animal no.  Treatment  Sex  June  July  August  September  October

69          B          F    T     T
70          B          F    T     T             T
71          B          F    T     A
39          B          M    T
73          B          M                T
74          B          M                T
75          B          M          T/D   D       D
85          B          M                D
87          B          M    T     T
37          C          F    T     T/D   T       T          A
38          C          F    T     T
76          C          F    T     D     D
77          C          F          T     T
80          C          F    T     T/D   T       T
37          C          F    T     T/D   T       T          A
38          C          F    T     T     D
68          C          M    T     T     T                  D
36          C          M    T     T     T                  D

                              2017
Animal no.  April  May  June  July  August  September

69
70          A      A    A           A
71
39
73                 T
74          A      T                D
75          A                               D
85          A      A                        D
87
37
38
76
77                 D                D
80                 D          D
37          T      D    T     D
38
68
36

A, arrhythmic; B, bled: C. control; D. daily rhythm: T, tidal rhythm;
T/D, both tidal and daily rhythms. Empty cells indicate that data were
not sufficient to identify a type of rhythm.
COPYRIGHT 2019 University of Chicago Press
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2019 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Owings, Meghan; Chabot, Christopher; Watson, Winsor, III
Publication:The Biological Bulletin
Geographic Code:1USA
Date:Jun 1, 2019
Words:9993
Previous Article:Maintenance of a Genetic Cline in the Barnacle Balanus glandula.
Next Article:Impacts of Salt Stress on Locomotor and Transcriptomic Responses in the Intertidal Gastropod Batillaria attramentaria.
Topics:

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters