Fishery and Biological Characteristics of Jonah Crab (Cancer borealis) in Rhode Island Sound

Jonah crab (Cancer borealis) is a demersal crustacean distributed throughout continental shelf waters from Newfoundland to Florida. The species supports a rapidly growing commercial fishery in southern New England, where landings of Jonah crab have increased more than six-fold since the early 1990s. However, management of the fishery has lagged its expansion; the first Fishery Management Plan for the species was published in 2015 and a stock assessment has not yet been created due to a lack of available data concerning the species’ life history and fishery. Fishing effort in the crab fishery is necessarily tied to fishing effort for American lobster (Homarus americanus), and the structure of this mixed crustacean fishery further complicates efforts to manage each of its component species. In the first manuscript of this thesis, a fishery-dependent sea sampling protocol was developed and implemented in Rhode Island Sound to (1) describe biological characteristics of the Jonah crab in Rhode Island Sound and (2) begin to characterize catch per unit effort in the Jonah crab fishery. With these methods, seasonal patterns in carapace width frequency and sex ratios were described for the commercial trap fishery, along with biological characteristics including shell disease condition, reproductive condition and allometric growth relationships. This study provides a unique description of commercial catch before discards and provides a more comprehensive description of the population than is available with dockside sampling methods. Conducting this type of data collection across the range of the species would substantially expand available data that is considered essential for constructing a stock assessment. The second manuscript describes growth characteristics of Jonah crab in Rhode Island Sound from data gathered during a laboratory study and via sea sampling in the commercial fishery. Description of the growth rates of exploited marine species is essential to understanding the impacts of fishing pressure on these resources and to predicting population abundances. Because crustaceans grow only during discrete molting events, description of growth per molt along with molting probabilities is necessary for estimating absolute growth rates of crustaceans. The results of this project provide characterization of growth per molt for male and female Jonah crabs along with description of molting seasonality for mature males. This thesis provides essential knowledge concerning growth, seasonal catch patterns and basic biological description of the Jonah crab to fishery managers charged with regulating the resource. To date, knowledge of the species has been limited enough to preclude the creation of a thorough stock assessment; these results provide marked progress in meeting the data needs for such an assessment.


LIST OF TABLES
for model parameters)…..26 xii

INTRODUCTION
The Jonah crab, Cancer borealis, is a brachyuran crustacean distributed in northwest Atlantic coastal waters (Haefner, 1977) and a significant contributor to marine trophic webs (Leland, 2002;McKay and Heck, 2008;Stehlik, 1993;Ojeda and Dearborn, 1991). Limited evidence suggests that the species migrates seasonally across the continental shelf (Jeffries, 1966;Krouse, 1980;. Previous studies also found evidence of depth dependence in the size distribution of Jonah crab (Carpenter, 1978;Krouse, 1980). However, there is an overall paucity of data concerning distribution of the species including spatial and temporal dynamics (ASMFC, 2015). Collectively, these results build upon the limited knowledge of the Jonah crab and address data needs for informed management of the resource.

Data collection
At-sea data collection was conducted during normal fishing operations on commercial fishing vessels targeting Jonah crabs in Rhode Island Sound in National Marine Fisheries Service (NMFS) statistical areas 537 and 539 north of 41° N latitude and west of 71° W longitude ( Figure 1). From each trawl, the following fishing effort information was recorded: latitude and longitude, water depth, habitat type (as reported by the vessel operator based on depth sounder readings and local ecological knowledge), bait species, soak time, trap dimensions and escape vent configurations.
When feasible, all traps within a trawl were sampled for catch composition; when census sampling was not possible, a systematically random subset of the traps within a trawl were sampled. In these traps, Jonah crab, American lobster, Atlantic rock crabs (Cancer irroratus), and other bycatch species were quantified.
Biological measurements were collected from a random subset of the Jonah crabs caught within each trawl. A systematic random sample frame was used to select a subset of traps in the trawl, within which all Jonah crabs were sampled. From each of these specimens the following measurements were collected: sex, carapace width, ovigerous condition, presence of sperm plugs in females indicating recent mating (see Elner et al., 1985), and shell disease condition. Carapace width was measured as the widest width of the carapace, in mm, including anterolateral teeth. All carapace measurements were taken with digital Vernier calipers and rounded to the nearest 0.1 mm. A subset of the Jonah crabs caught were collected and weighed to the nearest gram on a digital electronic scale.
Shell disease syndrome is defined as black-spot lesions on the exoskeleton (Vogan et al., 2008). It varies in cause, appearance, and severity between crustacean species but has not been described for Jonah crab. We assessed shell disease on a fourfactor scale adapted from that described for deep sea crab crabs, Chaceon quinquedens, by Sindermann et al. (1989) (see caption of Figure 4). Shell disease was described according to the following factor levels: Absent: no visible dark spots on carapace; Minor: <10% of carapace covered in dark spots; Moderate: 10-50% of carapace covered in dark spots; and Severe: >50% of carapace darkened.
An Onset HOBO Tidbit temperature logger (Model UTBI-001) was deployed on a trap in the general sampling area to capture the region's seasonal temperature pattern. Temperature readings were recorded at a minimum frequency of once every 4 hours. Bottom temperature was averaged over the day of harvest to be used in analysis of temperature effects on catch per trap.

Data Analysis
The relationship between carapace width and weight was defined with a power function using the formula where W represents the weight in g, CW represents the carapace width in mm, and a and b are fitted parameters representing the condition factor and the allometric factor, respectively (Quinn and Deriso, 1999). The power function was linearized and parameters were estimated using least squares regression.
A generalized linear model (GLM) was employed to estimate the effects of the following variables on Jonah crab catch per trap in traps that were configured to target Jonah crab: vessel (as a factor); bottom temperature on the day of trap haul; soak time in days; depth; lobsters per trap; total bycatch excluding lobster; and interactive effects of latitude and longitude. Habitat type and bait species remained constant throughout the study and were therefore excluded from this analysis. Traps were controlled for trap length (i.e., traps that were 4 feet in length rather than the standard 3 feet were excluded) and traps for which counts of all species were not taken were excluded.
Stepwise model selection was used to remove nonsignificant covariates until the selected model was chosen.

Biological catch summary
A total of 8535 male and 930 female Jonah crabs were biologically sampled throughout the study period (Table 1) (Table 2). All females with sperm plugs had recently molted as indicated by a pliability of the shell and overall light color.
Overall, 75.0% of the males caught exceeded the legal minimum size threshold, though this proportion varied throughout the study period ( Figure 2). Large

males (>130 mm) represented the highest proportion of the catch of males in between
August and November. The mean size of males was lowest from November to May.
In general, shell disease presented as black spotting on the exoskeleton without lesions and obvious degradation of the shell; lesions were limited to the most severe cases of shell disease. Shell disease exhibited a seasonal pattern of occurrence in males, with lowest rates of disease occurring in August and September ( Figure 4). The pattern of shell disease occurrence in females appears seasonal as well, with lowest shell disease rates occurring between January and June ( Figure 4). However, female patterns are potentially confounded by seasonal low catch rates of female crabs.

Length-to-Weight Relationship
Analysis of the length-to-weight relationship indicated significant differences in both the condition factor and allometric factor between males and females (p<0.05, Table   4). An allometric factor b=3 indicates isometric growth (Quinn and Deriso, 1999).
This parameter is 2.94 for females and 3.12 for males in the Jonah crab population, indicating negative and positive allometric growth, respectively, although these values are close to isometry. The power function condition factor, a, found by exponentially transforming a from the linearized equation, is 9.42 x 10 -5 for males and 2.06 x 10 -4 for females. Although this coefficient is higher for females than for males, the allometric factor, b, has a greater influence on relative weight-at-length between females and males.

Fishery characteristics and catch per trap
Most common bycatch species in traps targeting Jonah crab were black sea bass (Centropristis striata), unidentified eels, rock crabs (Cancer irroratus), sculpins (Myocephalus spp.), and scup (Stenotomus chrysops) (Table 3). Species composition of bycatch varied throughout the year, but overall catch rates of finfish species were low.
One of the methods fishers use to target Jonah crab rather than lobster is to modify trap escape vents so that crabs are less likely to exit the traps. Traps used throughout this study were outfitted with both rectangular vents measuring 5 ¾" by 1 7 /8"-2" wide and double circular escape vents of 2 5 /8" diameter. When targeting crabs, the  (Table 5). Depth, latitude and longitude, and non-lobster bycatch were not found significant and were excluded from the analysis. The model explained 28.1% of the deviance in Jonah crab catch per trap ( Figure 8).

DISCUSSION
The seasonal trends in sex ratio of Jonah crab commercial catch resemble those described by , who compiled Northwest Atlantic Shelf survey data from the Northeast Fisheries Science Center and found that Jonah crab sex ratios favored females in the fall and males in the winter and spring. Stehlik and colleagues suggested these differences in sex composition may have been due to sex-specific migrations or to changes in catchability related to reproduction. The absence of ovigerous females throughout this study align with findings of previous research; eggbearing females have been rarely encountered and compose a notably low proportion of Jonah crabs sampled in research studies, sometimes leading to the assumption that females are absent from the study area (Haefner, 1977;. However, low activity levels and resultant low catchability have been documented for ovigerous females of other Cancer species (Naylor et al., 1999); thus, the seasonality of spawning and distribution of ovigerous females likely cannot be fully assessed in a fishery-dependent study such as the present one. Sastry and McCarthy (1973) reported ovigerous Jonah crabs in Narragansett Bay in July, but methods used to collect these crabs were not reported. The increased catch of large females in the late summer and fall may be due to post-hatching increases in feeding and activity of mature females.
In Boothbay Harbor, Maine, Krouse (1980) described a seasonal peak in research trap catch per unit effort (CPUE) in the fall which diminished rapidly in the winter and hypothesized this decline to be due to fishing mortality or emigration. The length frequencies of male Jonah crabs in the present study followed a similar pattern.
The mean size of male Jonah crabs was highest in the late summer and fall and decreased through the spring until June (Figure 2). The molting season for mature males in this region occurs during the late spring and early summer (Truesdale et al., in prep). Fishing and natural mortality may cause large male abundance to slowly decline throughout the year until the molting season brings new recruits to large size classes. Alternatively, ontogenetic seasonal migrations may cause these shifts in mean size of males in commercial traps, yet the migration paths are poorly understood (ASMFC, 2015). Seasonal patterns in length frequency distribution may also be influenced by changes in predominant target species (Table 1) wherein trap configuration affects trap size selectivity. These effects were not investigated in the present study.
Length-weight relationships in male and female Jonah crabs indicate that both sexes experience nearly isometric growth. Visually, the growth curves for males and females are quite similar until a carapace width of approximately 95 mm ( Figure 6).
Males have larger claws than females, and this point of divergence most likely results from morphological differences between mature males and females. Size at maturity for Jonah crabs is not well described in southern New England, but previous studies indicate that males reach sexual maturity between 90 and 100 mm and females reach sexual maturity by 80-90 mm in Norfolk Canyon (Carpenter, 1978; Jonah crab catch within this temperature range. These temperatures closely match the preferred temperature range for the species provided in other field and laboratory investigations (Carpenter, 1978;Haefner, 1977;Jeffries, 1966).

Antagonistic interactions between lobsters and Jonah crabs have been
documented in previous studies, and Jonah crabs have been observed avoiding interaction with lobsters by choosing suboptimal substrates (Fogarty, 1978;Richards, 1992). Avoidance of traps containing lobsters is assumed to be the cause of negative correlation between Jonah crab and lobster catch per trap.
Effects of various habitat types on Jonah crab catch per trap could not be tested because fishers targeted Jonah crabs exclusively in flat, muddy sediment during the study. Preference for fishing in this habitat type did not align with previous study of Jonah crabs in Narragansett Bay and Boothbay Harbor, which found that Jonah crab preferred coarse gravel and rocky substrate (Jeffries 1966;Fogarty, 1978;Krouse, 1980). However, this preference for a muddy fishing substrate aligns with research conducted by  on the continental slope of the southeastern United States wherein Jonah crabs were found in highest abundance on soft bioturbated ooze substrates. Further research is necessary to understand spatial and seasonal dynamics of habitat preference by Jonah crabs.
Results of this year-round sampling program indicate seasonal trends in Jonah crab commercial catch length frequencies and sex ratios, along with evidence of the effects of environmental and fishery-dependent parameters on catch per unit effort.
The sea sampling protocol developed here may serve as a template for other fisherydependent sampling programs throughout the range of the species; collection of similar data types across a larger region may aid in describing spatial dynamics of Jonah crab. Given the increasing importance of the fishery, such efforts will aid efforts to understand the impacts of fishery exploitation on the resource.         Table 4 for model parameters).  during the year-long study, suggesting that C. borealis has determinate growth or that there are ontogenetic migrations in the species. The results presented here provide the first published growth data for C. borealis and are highly applicable to management efforts for the species.

INTRODUCTION
The Jonah crab, Cancer borealis, is a brachyuran crustacean distributed in continental shelf waters from Newfoundland to Florida (Haefner, 1977). Spatial dynamics of Jonah crab populations are not well described but previous research suggests they are structurally complex, displaying seasonal inshore-offshore movements and apparent sex segregations corresponding to water depth and offshore distance (Jeffries, 1966;Krouse, 1980;. Additionally, previous studies found evidence of depth dependence in the size distribution of Jonah crab (Carpenter, 1978;Krouse, 1980). Small Jonah crabs were observed to be absent from inshore, shallow waters in Maine and Rhode Island where larger individuals were frequently caught (Krouse, 1980). In the Mid-Atlantic Bight, Haefner (1977)  Understanding growth of exploited marine species is essential to quantitative fisheries stock assessments (Chang et al., 2012). However, description of growth in crustaceans is challenging because it is discontinuous, occurring only during periodic molting events. Characterizing absolute growth requires knowledge of the size increase during each molt (molt increment) and the time between molting events (intermolt period). Historically, collecting these parameters has been accomplished via tag-recapture studies, laboratory observation, and cohort analysis of wild populations (Hartnoll,1982). Each of these methods comes with limitations, and none are universally robust across crustacean species. Tag-recapture studies demand substantial resources and ensuring retention of tags through the molt can be difficult (Hartnoll, 1982). Size-frequency analysis relies heavily on assumptions about reproductive timing and migration and becomes less reliable for large and long-lived species (Hartnoll, 1982;Kilada et al., 2017). Laboratory studies are comparatively practical, but holding crustaceans in an artificial environment has been shown to influence growth (Wainwright and Armstrong, 1993;Stone et al., 2003;Somerton et al., 2013).
Direct aging methods have been studied for multiple crustacean species, but these methods are new and are infrequently represented in the literature (Kilada et al., 2017).
In this study, we describe the growth per molt of Cancer borealis and characterize the molting seasonality of mature males of the species. Molt increments were collected by short-term observation of crabs in a laboratory setting, and growth seasonality and molting probabilities were explored via periodic sampling and observation throughout a year-long study period. Comparison with data collected by field sampling on commercial Jonah crab fishing vessels corroborates laboratory findings and supports assumptions made regarding the validity of the data collection methods employed.

Molt increments
Jonah crabs were collected from commercial fishing traps deployed in Rhode Island in flow-through seawater tanks directly sourced with ambient seawater from the mouth of Narragansett Bay. To reduce differences between in situ temperatures and more extreme Narragansett Bay water temperatures, seawater was heated during the coldest winter months and chilled during the peak of summer with the goal of maintaining water temperatures between 5° C and 17° C. Diel light cycles were provided by windows situated near the holding tanks, and artificial light was reduced by blocking out overhead lights with black plastic sheeting. Crabs were fed to satiation once weekly with scup, herring or squid (Stenotomus chrysops, Clupeidae, and Loligo pealei, respectively).
Crabs that molted in the laboratory were isolated in individual compartments and allowed to harden for a minimum of three days before being measured post-molt.
Growth was assessed by measuring carapace width-the widest width of the carapace including anterolateral teeth-before and after molting. Carapace measurements were taken with digital Vernier calipers and rounded to the nearest 0.1 mm.
Growth increments were modeled with ordinary least squares linear regression of post-molt carapace width against pre-molt carapace width (Hiatt, 1948). Interactive effects of year (2016 or 2017), laboratory water temperature at time of molt, sex, laboratory treatment (URI or RIDEM), and statistical area of capture were tested using step-wise model selection to investigate the effects of these variables on growth per molt. Comparison between the two laboratory facilities was intended to represent nonmeasured effects of various laboratory treatments (e.g. flow rate, water quality), and NMFS statistical area was included to investigate differences between Jonah crabs caught closer (statistical area 539) and farther (statistical area 537) from shore.
In addition to the molt increments collected in the laboratory, 24 molt increments (16 male, 8 female) were collected by fishermen or sea samplers from crabs that molted in commercial traps. These newly molted crabs could be matched with their recently shed exuvium based on individual carapace markings and anterolateral tooth structure. The exuvium was measured to obtain pre-molt carapace width. Molt increments collected at sea in this manner were introduced to the selected growth per molt model; at-sea collection versus laboratory collection was incorporated as a factor to test for significant differences in growth per molt between data collection methods.
Effort was made to sample monthly during the study period, but February, April, and May were missed due to weather and logistical limitations. Male and female crabs were collected across the range of size classes (defined as 10-mm size bins) caught in commercial traps; crabs from common size classes were collected from randomly selected traps, and rare size classes were opportunistically collected. Specimens were measured, tagged with individual identification numbers, and kept in flow-through seawater tanks at URI as was done for molt increment collections.
Crabs were held in the laboratory to observe binary molt response-whether or not an individual crab molted within a 30-day period after capture (Chang et al. 2012).
The chosen observation period was intended to minimize laboratory bias, as other crab growth studies found a significant reduction in growth per molt for crabs that were retained in the laboratory more than 30 days before molting (Somerton et al., 2013;Stone et al., 2003).
Crabs that died in the lab before molting were excluded from analysis. Binary molt responses were aggregated by size bin and by month according to the dates and months listed in Table 1. Comparison of relative molting probabilities between months and size classes across the study period was conducted to explore molting seasonality and ontogenetic changes in growth.

Commercial fishery biological sampling
On the dates listed in Table 2, a sea sampler was deployed on commercial fishing vessels targeting Jonah crab in NMFS statistical areas 539 and 537 as described above.
A random subset of the traps hauled throughout each fishing day were sampled, and all crabs within selected traps were processed. For each sampled crab, the carapace width was measured with digital Vernier calipers and rounded to the nearest 0.1 mm.
Molt condition was recorded for each crab on a 3-factor scale: just molted, shell soft to the touch (1); recently molted with brittle shell, abdominal segments bend when pressure is applied (2); and hard shell, abdominal segments do not give with pressure (3).

Molt increments
Coefficients for the linear relationship between pre-molt and post-molt size were found to be significantly different between males and females ( Figure 3). Year, lab temperature at time of molt, statistical area of capture, and lab treatment were not found significant and were not incorporated into the chosen model. Growth per molt was not found to be significantly different between crabs that molted in traps and those that molted in the laboratory, so increments collected at sea were incorporated into the model. The regression for males was described as PostCW=1.22*PreCW+5.47 and female growth per molt was described as: PostCW=0.94*PreCW+23.31 where PostCW represents the post-molt carapace width in mm, and PreCW represents the pre-molt carapace width (R 2 =0.96).
All molt increments from crabs that were held in the laboratory for more than Female Jonah crabs occurred in commercial traps sporadically throughout the year and sample sizes were highly inconsistent between months, precluding description of molting seasonality. Alternative sampling methods may be employed in the future to allow for description of female molting seasonality.

Commercial fishery biological sampling
Carapace width and molt conditions were described for 9465 crabs in total. The mean carapace width for females was 104.9 mm (s.d.=11.3, n=930), and the mean carapace width for males was 126.1 mm (s.d.=9.0, n=8535) ( Figure 6).
Male molt condition exhibited a strong seasonal trend, with newly molted individuals being caught exclusively in late May, when 16.1% of males caught were soft (molt condition=1). Brittle shell males (molt condition=2) appeared as a substantial portion of the commercial catch in June and July (6.9-14.1%) ( Table 2).
Catch of females was highly variable throughout the year, and female molt conditions were not included in this analysis due to seasonally low sample sizes.

DISCUSSION
We present growth increment information for Jonah crab through laboratory and field-based estimates. The sexual dimorphism in growth per molt of the Jonah crab aligns with previous study of the growth of the species, although this data is limited to a single unpublished study with a small sample size (Ordzie and Satchwill, 1984). In Dungeness crabs, molt increments are similar between males and females until sexual maturity is reached, at which point males exhibit higher proportional growth than females (Wainwright and Armstrong, 1961). Differences in growth per molt between the sexes after maturity has been shown for many other crab species (Gerhart and Bert, 2008;Hartnoll, 1982;Moriyasu et al., 1987), and is attributed to differential investment in reproduction (Hartnoll, 1982). Few studies have described size at maturity for Jonah crabs. According to available literature, females reach sexual maturity by 80-90 mm in Norfolk Canyon (Carpenter, 1978;. Moriyasu et al. (2002) found that males reach gonadal maturity at 69 mm on the Scotian Shelf, while Carpenter (1978) found males to reach sexual maturity between 90 and 100 mm in Norfolk Canyon. Most of the molt increments collected in this study were from crabs molting to sizes exceeding these maturity thresholds, and the clear dimorphism in growth per molt may be attributed to this divergence in reproductive investment between males and females. Because few juvenile crabs of either sex were observed, ontogenetic shifts in growth per molt could not be assessed here.
The 30-day observation time limit employed in this study is assumed to sufficiently minimize laboratory bias by limiting observed growth events to crabs that were likely preparing for ecdysis at the time of capture. Molt stages of crustaceans have been classified into multiple intermolt and pre-molt stages so that molt timing can be predicted for marine crustaceans (Drach, 1939). Pre-molt stages D0-D4 describe successive levels of active preparation for ecdysis, and Stage D1' is the first stage after which a subsequent molt event is inevitable (ie. physiological preparation for molting cannot be halted) (Miller and Hankin, 2004). Studies of premolt setal development stages have estimated the time to molt from Stage D1' as 60 days for Cancer magister (Miller and Hankin, 2004); 18-40 days (depending on temperature treatment) for Homarus americanus (Aiken, 1973); and 6-9 weeks for Chionocetes opilio (O'Halloran and D'Or, 1988 (Table 2). Similarly, the growth per molt observed in the laboratory is assumed to be representative of in situ growth because any biases introduced in the laboratory are minimized by sufficiently limiting holding time.
Alignment between lab-collected molt increments and increments measured from crabs that molted in traps supports this assumption (Figure 7). When incorporated into regression analysis, no significant difference was found between increments collected at sea and those collected in the laboratory.
Juvenile growth rates have not been described for Cancer borealis and the age at recruitment for the species is unknown. However, quantification of the growth per molt and molting seasonality of adult male C. borealis provides a potential opportunity to predict recruitment in subsequent years. For instance, male crabs at 95.4 mm are the smallest we would expect to recruit into the commercial fishery at their next molt, becoming larger than the legal minimum size of 121.65 mm (ASMFC, 2015). If an annual molt is assumed for crabs of this size and mortality can be estimated, subsequent year recruitment prediction may be possible. Pre-recruit surveys may thus be a practical means of monitoring and predicting the magnitude of the Jonah crab resource (Caputi et al., 2014).
The small sample sizes for small size classes in the monthly observation study were a consequence of collecting specimens using fishery-dependent methodssmaller individuals are less susceptible to commercial crab traps, which are outfitted with escape vents to allow small individuals to exit. The sample sizes within each size bin are somewhat representative of their relative abundance in commercial traps, with deviations due to deaths during the observation period. The anomalously large sample size in the 90 mm size bin for June may be caused by pre-molt behavior. The traps from which the crabs were collected were set in flat, muddy habitat, and would be expected to provide some of the only available shelter. Crabs are most vulnerable to predation and cannibalism after molting (Romano and Zeng, 2016;Ryer et al., 1997, Sotelano et al., 2016, and previous studies have shown that shelter availability reduces post-molt mortality for crustaceans (Marshall et al., 2005;Zhang et al., 2018).
We hypothesize that Jonah crabs about to molt seek shelter in traps. This is the proposed explanation for having high representation in the 90 mm size class for June only, when 13 of 14 of the sampled individuals molted. The inherent bias that this behavior introduces to the molting probabilities is recognized; however, the relative probabilities between size classes are still expected to be reflective of the population in situ. Thus, decreasing molt probabilities with size is accepted as a characteristic of the wild population. Similarly, observations of molting seasonality would not be affected by this behavior; the relative molt frequency between months supports a clear seasonal pattern notwithstanding the potential bias of trap-seeking behavior.
Monthly molt probability observations are interpreted to indicate an annual molt period in the summer for male Jonah crabs between 90 and 110 mm. Molting events become much less common in male crabs larger than 110 mm ( Figure 5), aligning with observations of other brachyuran crabs for which the intermolt period increases as crabs grow larger (MacKay and Weymouth, 1934;Hartnoll, 1982). Molt events in crabs larger than 120 mm were quite rare: over the course of the entire study, only one crab larger than 120 mm (120.6 mm) molted in the laboratory. The Jonah crab may thus be a species that experiences a terminal molt, or determinate growth-individuals molt to a certain size or through a predetermined number of instars, rather than continuing to molt indefinitely (see Hartnoll, 1985). Length frequencies in the commercial fishery ( Figure 6) lend support to this hypothesis. Using the Hiatt model that was fitted to molt increments, a male crab at 120 mm would be expected to grow to 151 mm during its next molt. Crabs larger than 120 mm composed 77.67% of the male crabs caught in the commercial fishery (6629/8535) yet crabs larger than 150 mm composed only 0.37% of males caught in the commercial fishery (32/8535). The discordance in abundance between these size classes in the commercial fishery indicates that C. borealis may experience determinate growth. Ontogenetic migrations from the study area provide an alternative to this terminal molt hypothesis, and are supported by the available literature showing size segregation in Jonah crab populations (Carpenter, 1978;Krouse, 1980). These distributions are not well characterized across the range of the species, and further investigation of movement patterns is necessary to evaluate these hypotheses for the absence of molting in large male Jonah crabs.