Temperature-dependent seminal recovery in the southern king crab Lithodes santolla

Male-biased fishery management can provoke depletion of seminal reserves, which is the primary cause of sperm limitation. Therefore, identifying factors which contribute to the vulnerability to depletion of seminal reserves is a priority. The present study aimed to determine the effect of temperature on the recovery rate of sperm and seminal reserves after their depletion in Lithodes santolla, an important fishery resource in southern Chile. Sperm and seminal reserves were not fully recovered within 30 days. Temperature significantly affected seminal recovery: after 30 days the recovery index increased to 40% and 21% at 9°C and 12°C, respectively. The twice as fast seminal recovery at 9°C may be explained by the zone of origin of the individuals in this study (northern distributional limit), and 12°C may be close to the threshold of temperature tolerance. Lithodes santolla populations subject to intense male-only fisheries may be vulnerable to depletion of seminal reserves and a climate change scenario could additionally aggravate the risk of seminal depletion in L. santolla in its northern distributional limit.


Introduction
Most crustacean fisheries are managed by a male-biased or exclusively male extraction strategy [1]. However, harvesting large males can trigger changes in the mating dynamics of a population, thereby reducing overall reproductive success [2,3]. males derived from their northern distributional limit. Intraspecific latitudinal variability in the reproductive pattern and traits has been documented in crustacean species [31][32][33][34]. Consequently, L. santolla in its northern distributional limit (this study) is exposed to different prevailing environmental conditions (e.g. temperature, photoperiod) which probably have led to variations in the reproductive pattern between the southern and northern limits of its distribution (i.e. extended window of reproduction in northern distributional limit).
As exploitation of this lithodid is regulated by a large male-only management strategy, the target species may be susceptible to depletion of seminal reserves which makes it a suitable model species to identify factors that contribute to the vulnerability to seminal depletion. This is important for fisheries and it is crucial to obtain a greater understanding of males' reproductive biology in L. santolla. As L. santolla inhabits cold seawater environments and has a low energy metabolism, a slow sperm and seminal recovery rate after depletion of seminal reserves is expected. The objective of this study was to determine the effect of temperature on the recovery rate of sperm and seminal reserves after their depletion in L. santolla from its northern distributional limit.

Material and methods
Physiologically mature similar-sized males of L. santolla ( physiological maturity in the Beagle Channel greater than 75 mm CL, see [35]) were collected in October 2016 (caught with commercial traps) from the Seno de Reloncaví, X region, Chile (41845 0 47.1 00 S; 73805 0 20.1 00 W) which corresponds to the northern limit of distribution of this species. Individuals were transported to the Laboratorio de Ecofisiología de Crustáceos (LECOFIC) of the Universidad Austral de Chile in Puerto Montt. Crabs were acclimated in the laboratory for approximately one month prior to the start of the experiment with flowing seawater, air supply and ad libitum food (Mytilus chilensis). Both temperature conditions (98C and 128C) were conducted simultaneously from mid-November to mid-December 2016.
To simulate depletion of seminal reserves and to ensure standardization of initial condition of males, individuals of L. santolla were stimulated for electroejaculation repetitively on four successive days through short electric shocks of 12 V AC [36]. Electrodes were placed ventrally on the soft section of the opened pleon (figure 1a). After seminal depletion (i.e. no ejaculation after electrostimulation), individuals were maintained in two tanks at 98C and 128C each with flowing seawater (32 psu), air supply (O 2 saturation was guaranteed through bubbling and movement of the water by pumps) and ad libitum food. The first experimental temperature of 9.2 + 0.098C (mean + s.e.) was provided by a chiller. The second experimental temperature of 12.04 + 0.068C (mean + s.e.) was maintained using submersible aquarium heaters with digital controllers. These two experimental temperatures are referred to as 98C and 128C in the text, reflecting the mean sea surface temperatures in the study area, Seno de Reloncaví, during the austral winter (May -August, 98C) and throughout the year (128C) [37]. Individuals in the lower temperature condition were acclimatized for 2-3 days through a gradual temperature decrease until reaching 98C. Temperature was measured every day with a digital thermometer (WTW Cond 330i with sensor WTW TetraCon 325; precision of 0.18C). To identify crabs individually, they were marked with numbered cable ties.
To determine the effect of temperature on the recovery rate of sperm and seminal reserves, individuals were sacrificed (thermal shock 2808C for 15 minutes) after 0, 15 and 30 days. Crabs without electroejaculation were used as controls and sacrificed after 0 and 30 days. After the corresponding experimental period, both vasa deferentia and the hepatopancreas were dissected (figure 1b). The left vasa deferentia were used to estimate their dry weight (seminal material) and to calculate the vasosomatic index (VSI, n ¼ 4). The VSI has been suggested as an indicator of the reproductive condition of males [8,15,38]. The VSI (expressed as a percentage) was calculated: VSI ¼ (2 Â VDW/BDW) Â 100, with VDW being the dry weight of the left vasa deferentia (oven dried for 5 days at 708C and weighed to a precision of 0.0001 g) and BDW being the dry weight of the complete body (oven dried for 5 days at 708C and weighed to a precision of 0.01 g). The weight of the left vasa deferentia was doubled to calculate the VSI of the whole male reproductive system. For the estimation of the VSI, dry weight of crabs without walking legs and chelae was used to increase accuracy.
The right vasa deferentia were used to examine histological cross-sections (n ¼ 4 in each initial group, n ¼ 5 in each recovery group, n ¼ 3 after 30 days at 128C) and to calculate the area covered by sperm (stored in spermatophores) in the vasa deferentia lumen. To prepare histological cross-sections, the right vasa deferentia were fixed in Bouin solution for at least 2 days. Then, samples were sequentially passed through a 50 -70 -80 -96 -100% ethanol series for 30 min each, 100% ethanol -butylic alcohol royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 181700 (1 : 1 v/v) for 30 min and butylic alcohol for 25 min (twice). Samples were embedded in paraffin and serial sections of 6 mm were cut and stained with haematoxylin-eosin. One slice of the middle section of the vasa deferentia, which corresponds to the section where spermatophore reserves are located [35], was used to estimate area covered by sperm in the lumen (figure 2a,b). Photos were taken with a digital camera (Q Imaging Go-3) of histological cross-sections under a light microscope (OLYMPUS BX51) at 10Â magnification for further analysis. Photos were analysed using the image processing software ImageJ 1.52a [39]. The area covered by sperm was identified through adjusting the brightness of the colour threshold and measured. The sperm area was then doubled to calculate the  royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 181700 sperm coverage of the paired vasa deferentia. Sperm area is presented in relation to CL and multiplied by 1000. The hepatopancreas was extracted and the dry weight was determined (n ¼ 5 in each group). The hepatosomatic index (HSI) was calculated: HSI ¼ (HDW/BDW) Â 100. HDW was the dry weight of the hepatopancreas and BDW was the dry weight of the animal without legs. In crustaceans, the hepatopancreas is an organ involved in digestive processes and it is important in the assimilation of energy, storage and mobilization of resources during the moulting process. Thus, the size and weight of the hepatopancreas provide a good indication of somatic resources [40].

Data analyses and statistics
We established a recovery index (RI), which refers to the recovery rate of seminal material expressed as the percentage change in dry weight of the vasa deferentia between initially electroejaculating individuals (n ¼ 5) and after the corresponding experimental period (n ¼ 5 each) in relation to dry weight of the vasa deferentia of the control (i.e. without electroejaculation, n ¼ 4). The dry weight of the vasa deferentia remaining after electroejaculation was not considered. To assess the net change rate of VDW of the control animals (i.e. natural fluctuation of the vasa deferentia without electroejaculation), change rate (in %) of vasa deferentia dry weight after 30 days (n ¼ 5 each and n ¼ 4 in initial group) in relation to the initial vasa deferentia dry weight (n ¼ 4) was calculated for 98C and 128C.
The following parameters were used in the RI in electroejaculated individuals and the net change rate of VDW in the control animals. VDW ¼ vasa deferentia dry weight; E ¼ crab electroejaculated at beginning of experiment; C ¼ control (i.e. without electroejaculation); t 0 ¼ at beginning of the experiment; and t 1 ¼ after corresponding time period of either 15 or 30 days.
and change rate VDW ð%Þ ¼ VDW Ct1 À VDW Ct0 VDW Ct0 Â 100: Both RIs and net change rates of VDW in control individuals were estimated through bootstrapping of the above formulae based on the means of the VDW of each treatment or control (replicates ¼ 1000, using the 'boot' function in the 'boot' package [41,42] in R v. 3.4 [43]). Bootstrapped 95% confidence intervals (Bca method) were generated for the RIs and net change rates of the VDW of each group. RIs and net change rates of the VDW were considered significantly different between groups when the 95% confidence intervals did not overlap. Data were checked for normality using the Shapiro -Wilk test. The Bartlett test was applied to check for variance homogeneity. Planned comparisons of least-squares means (independent t-test: t-value refers to the estimate divided by the s.e.) were performed to detect differences in RIs, change rates of VDW, standardized sperm area in histological sections and HDW between different recovery periods, the two temperature conditions (98C and 128C), and initially electroejaculated and control animals. Planned comparisons were performed with STATISTICA 7.0 (StatSoft, Hamburg). To check for no differences in size of crabs among groups, a one-way ANOVA was performed in R.

Results
Mean CL of males was 111.7 + 5.6 mm (range ¼ 98 -122 mm). All values in the Results section are means + standard deviations. No significant difference was detected in size among individuals of all experimental groups (one-way ANOVA, F 7, 31 ¼ 0,481, p ¼ 0.841; n ¼ 39).
Within 30 days after electroejaculation, seminal reserves of L. santolla were not fully recovered at any of the experimental temperatures ( figure 4). The RI at 98C increased significantly between 15 and 30 days. After 30 days, the RI was significantly larger in seawater at 98C compared to 128C. Also, the change rate of VDW in the control after 30 days was significantly larger at 98C than at 128C (figure 5).

Discussion
Sperm and seminal reserves of Lithodes santolla were not fully recovered within 30 days after depletion through electroejaculation, either at 98C or at 128C. However, temperature significantly affected   Figure 6. Standardized sperm area of paired vasa deferentia of Lithodes santolla in initially electroejaculated crabs after the corresponding time periods at 98C and 128C (n ¼ 4 in each initial group, n ¼ 5 in each recovery group, n ¼ 3 after 30 days at 128C). Box plot indicates mean, first and third quartiles and 95% confidence interval of the mean (whiskers). Blue horizontal line represents mean standardized sperm area of initial crabs without electroejaculation +95% confidence interval.
royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 181700 recovery of seminal stock, and L. santolla replenished seminal reserves faster at 98C than at 128C. Effect of temperature on sperm recovery was not that pronounced and a trend of increase and decrease of the standardized sperm area at 98C and 128C, respectively, was detected. Lithodes santolla can be fished at sites between 58C and 128C along its range of distribution. The zone of origin of the individuals in this study is located close to the northern limit of distribution of the species where crabs reproduce in shallow waters and mean temperatures present during mating are 98C (austral winter) and 128C (throughout the year). Temperature had a significant effect on the recovery rate of seminal reserves in L. santolla after 30 experimental days. The faster seminal material recovery rate in seawater of 98C than 128C after 30 days may be explained by the close proximity of the crabs in this study to their northern distributional limit. Hall & Thatje [44] concluded that deep-water representatives of the Lithodinae subfamily were excluded from waters exceeding temperatures of 138C. The experimental temperature of 128C may be close to their threshold of temperature tolerance and therefore, crabs may be energetically less efficient. Similarly, 1-year-old juveniles of L. santolla can tolerate seawater between 68C and 158C; however, their preferred temperature is 98C [45] with a maximum growth and moulting rate [46,47]. The HDW has doubled during 30 days. At 128C, crabs started to accumulate energy in the hepatopancreas already after 15 days but then did not immediately invest energy in the reproductive system during the duration of the experiment. Probably at 98C, energy of the hepatopancreas was converted into seminal material more efficiently.
Temperature-dependent seminal recovery and seasonal variability in the vertical distribution of temperature in the zone of study [48 -53] suggest the importance of bathymetrical migrations which could be associated with moving to the optimum temperature to promote recovery of seminal reserves. Seasonal migrations associated with reproduction are a characteristic of lithodid crabs [26] and are well described in lithodid species from the Northern Hemisphere [26]. By contrast, little details are available on these migratory movements in the Southern Hemisphere. Observations of migrations of L. santolla exist mainly from the Magellan region and the Atlantic Ocean in Argentina where male and female crabs have been observed to migrate to shallow waters in October and November to reproduce where mating takes place in December and subsequently both sexes return to deeper waters [54]. Timing and magnitude of the pattern of migration in L. santolla from northern parts of its distribution is unknown. However, migration of L. santolla in the Seno de Reloncaví and interior waters of Chiloé to deeper and cooler waters after reproduction could be substantial to recover seminal material stock in a more efficient form.
Effect of temperature on the recovery of sperm was not that pronounced (non-significant changes) as seminal recovery; however, a trend of increase and decrease in the standardized sperm area in the vasa deferentia lumen was detected at 98C and 128C, respectively. We suggest that the estimation of area royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 181700 covered by sperm in the present study has restrictions due to small sample size and possibly little statistical power.
Our results of incomplete sperm and seminal recovery during 30 days correspond with the previous description of a long spermatogenesis in this species from a different population in the Beagle Channel [27] and in the red king crab Paralithodes camtschaticus which inhabits the Northern Hemisphere [55,56]. Direct comparison of the previously described period of spermatogenesis in L. santolla and annual pattern of sperm fullness in the vasa deferentia [27] is avoided, due to large latitudinal differences between the study zones and probably varying reproductive traits [33]. Similarly to the results of our recovery rates, in the spiny king crab P. brevipes and in the stone crab Hapalogaster dentata, sperm numbers in the vasa deferentia are not fully recovered even after 28 and 20 days, respectively, after depletion [16,17]. By contrast, in large males of the brachyuran swimming crab Callinectes sapidus, weight of the vasa deferentia is recovered relatively fast and recovery requires 9-20 days after two consecutive matings [7]. The majority of experiments investigating sperm recovery rate in the context of sperm depletion were not conducted under stable experimental temperatures, which makes it difficult to compare results among them (seawater temperature ranges: 20.6-4.38C [16], 10-19.98C [17], 0-298C [7]). By contrast, in aquaculture research, temperature affects the sperm replenishment period in the shrimp Litopenaeus setiferus after sperm depletion through electroejaculation (8 days at 258C, 7 days at 308C and 6 days at 338C [19]). In another tropical penaeid, spermatophores are regenerated after manual extrusion in 16 days (Penaeus brasiliensis: 278C [18]).
Apart from fluctuating temperature conditions in some studies, it is difficult to make interspecific comparisons due to inconsistency in the method applied to deplete seminal reserves (electroejaculation versus varying number of successive natural matings) and the type of information reported related to the male reproductive organ (sperm number in the vasa deferentia-vasa deferentia weight-macroscopic visual inspection). The two methods we used to estimate recovery are complementary, as they refer to distinct components of the reproductive system. Dry weight of the vasa deferentia represents the total of sperm and seminal fluids, while the standardized sperm area relates solely to the relative sperm amount. We suggest that the most appropriate method to be applied in further studies depends on the specific research question. In the present study, electroejaculation was a useful, fast and standardized method to artificially induce ejaculation and deplete sperm and seminal material stocks. While our manipulative approach with electroejaculation provides a good basis, it would be favourable in a next step to determine the amount of ejaculate delivered during one mating and the number of possible successive matings for a better understanding of the male reproductive biology and interpretation of our results. Details about lithodid mating ability are described only in P. camtschaticus and P. brevipes which transfer only portions of their sperm reserves during one mating [4,57]. Lithodid crabs are polygamous. For example, the male king crab P. camtschaticus is able to mate with up to seven females and result in full egg clutches [57]. However, the maximum possible mating frequency of males of L. santolla is unknown [58]. In P. brevipes, an increase in the mating frequency reduces the sperm number ejaculated, the percentage of spawning females and female fertilization rate. After the second successive mating of small males (less than 100 mm CL), partial or null fertilization occurs in females [4]. Crustaceans with slow seminal recovery in a maleonly fishery which have mated repeatedly might have depleted seminal reserves especially as the reproductive season progresses. Temporal variation of sperm reserves has been observed in P. brevipes and the proportion of depleted males increased throughout the reproductive season [4].
Lithodes santolla is distributed over a wide range extending over 198 latitudes in the Pacific [22,23,59]. Along the coast of Chile, the sea surface temperature shows a gradient and is decreasing from north to south [60]. Therefore, L. santolla is exposed to regional differences in seawater temperature. The effect of temperature on the seminal recovery rate may generate variations in the reproductive potential of males. Males located close to the northern distributional limit have a slower seminal recovery rate and after successive matings may deplete their reserves faster in contrast to individuals inhabiting more southern or deeper habitats with a temperature closer to 98C. Considering the predictions of ocean warming over the next decades, this fact could have an even greater impact on the reproductive potential of males in the future [61]. While our results represent a first step towards improving knowledge on male reproductive traits, we highlight the importance of further research on aspects of reproduction of L. santolla in its northern distributional limit to allow a sustainable crustacean fishery in the future.

Conclusion
Summarizing the aspects of the reproductive biology of L. santolla, such as absence of a sperm storage organ coupled with slow sperm and seminal recovery, suggest that L. santolla populations subject to royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 181700 intense male-only fisheries may be generally vulnerable to depletion of seminal reserves. Temperature modulated the time necessary to recover seminal reserves. Seminal recovery was faster in seawater of 98C than that of 128C, indicating that different seminal replenishment rates in relation to latitude and depth might exist. Especially individuals inhabiting the northern limit of distribution might be susceptible to seminal depletion, in contrast to those located in southern or deeper habitats. Considering that the seminal recovery rate was slower in seawater of 128C, a climate change scenario could additionally aggravate the risk of seminal depletion in L. santolla in its northern distributional limit.
Data accessibility. Data analysed in this study are available online in the electronic supplementary material. Authors' contributions. K.Pr., L.M.P. and K.Pa. designed the study; K.Pr. and K.Pa. carried out the laboratory work; K.Pr.