Egg production of Austropandalus grayi ( Decapoda , Caridea , Pandalidae ) from the Magellan region , South America *

Austropandalus grayi is one of three pandalid shrimp species reported from Chilean waters. Here we describe fecundity, reproductive output, and chemical composition of the incubated embryos of A. grayi, providing the first account of such information for a pandalid species inhabiting the southern hemisphere. Ovigerous females (n = 96) were collected during October November 1994 with an Agassiz trawl in waters off Patagonia and Tierra del Fuego, South America. Both egg production (between 50 and 1858 embryos) and early egg volume (x – = 0.069 mm3) of A. grayi were low when compared to values reported from other pandalids. Clutch size increased with maternal size, however, the size of recentlyspawned eggs was not linked to female size. Although the egg mass volume remained almost stable during the incubation period, females lost on average 51.1% of their initially-laid embryos; however, this loss was more than compensated by an 88.4% increase of the egg volume during embryogenesis. Mean reproductive output was relatively low (0.133) and not related to maternal size. The RO-data of A. grayi and other pandalids seem to indicate an increasing energy allocation for reproduction towards higher latitudes, a trend which would be contrary to the results of studies with other crustaceans. Water was the predominant constituent of the chemical composition of developing embryos and increased from 62.2% to 70.2% during the incubation period. At the same time, organic matter declined and mineral content augmented. Generally, our results are in agreement with previous findings with regard to the chemical composition of developing decapod embryos.


INTRODUCTION
Members of the family Pandalidae have a worldwide distribution and occur typically in colder waters of both the northern and southern hemisphere (Holthuis, 1980;Hendrickx and Wicksten, 1989).Representatives of several genera of Pandalidae are currently commercially exploited, and species of Pandalus and Heterocarpus play an important role in the world's shrimp fisheries (Holthuis, 1980;Parsons and Fréchette, 1989).Three species of pandalid shrimps, Heterocarpus reedi Bahamonde, 1955, Plesionika santaecatalinae Wicksten, 1983, and Austropandalus grayi (Cunningham, 1871) inhabit Chilean waters (see Holthuis, 1952;Retamal, 1995).The first species supports the main Chilean crustacean fishery, and several aspects of the fishery biology of H. reedi have been investigated (e.g., Hancock and Henríquez, 1968;Bahamonde and Henríquez, 1970;Roa and Ernst, 1996;Wehrtmann and Andrade, 1998).The second species, P. santaecatalinae, has been found recently in northern Chile (Retamal, 1995), and its life history in Chilean waters remains to be investigated.Referring to A. grayi, virtually nothing is known about the biology and ecology of this shrimp, a shallow water species inhabiting typically the antiboreal region of South America (Holthuis, 1952;Wehrtmann and Carvacho, 1997).A recent study of the decapod fauna of the Magellan zone revealed that A. grayi was the most common caridean shrimp in the study area, occurring generally at considerably less than 100 m water depth (Arntz et al., 1999).
The purpose of this paper is to describe and discuss fecundity, reproductive output, and chemical composition of the incubated embryos of A. grayi.Furthermore, we compare our data with those published for other pandalids from the northern hemisphere, as such a comparison may contribute to our understanding of life history strategies in caridean shrimps.

MATERIAL AND METHODS
Egg-bearing females (n = 96) were obtained from Agassiz trawl catches carried out during October-November 1994 in the frame of the Joint Chilean-German-Italian Magellan "Victor Hensen" Campaign.The cruise covered the Straits of Magellan, the channel system between the Straits of Magellan and the Beagle Channel, and the Beagle Chan-nel.A detailed description of the study area, location of the stations, and sampling gears employed has been published by Arntz and Gorny (1996).The material used for the present study was obtained in the area between 53º42'8' ' S, 70º57'4'' W and 55º09'2'' S, 67º01'6'' W;Arntz et al. (1999) provide information regarding the exact geographical locations where Austropandalus grayi were collected.Ovigerous females of A. grayi were stored immediately in a 10%-buffered formaldehyde solution.In the laboratory, we measured total length (TL; distance between the mid-dorsal anterior margin of the carapace and the distal margin of the telson, excluding setae) and carapace length (CL; distance between the distal part of the eye socket to the distal margin of the carapace) of ovigerous females.The entire egg mass was removed from the females, and the total number of embryos was assessed.Thirty eggs were separated from each batch, and length and width were measured to calculate the egg volume (V) using the formula for oblate spheroids V = 1/6 (π d 1 2 x d 2 ) (see Turner and Lawrence, 1979).The average volume per egg was multiplied with the total number of eggs to estimate the egg mass volume (EMV) of the female.The developmental stage of the embryos per female was determined and divided into three stages according to the following criteria (see Wehrtmann, 1990): Stage I: eggs recently produced; uniform yolk; no eye pigments visible; Stage II: eye pigments barely visible; Stage III: eyes clearly visible and fully developed; abdomen free.To document possible brood mortality during the incubation period, all females were grouped in 2 mm size classes, and their fecundity was estimated according to the regression equations for Stage I -III (see Table 1).Subsequently, estimated egg numbers per size class and developmental stage were compared to obtain estimated values for egg loss.
The removed egg mass per female was divided into three subsamples, and the interstitial water was extracted with filter paper.Aliquots were dried at 60 o C 326 I.S. WEHRTMANN and M.A. LARDIES for approximately 48 hours to obtain dry mass; afterwards, eggs were combusted for 3-5 hours at 500 o C to obtain ash mass.The mass of the organic material was estimated by subtracting ash mass from dry mass.The same methods were applied for the females, except that dry mass was obtained after 72 hours.Based upon the dry mass of both eggs and female, we calculated the stage-independent reproductive output according to the formula provided by Clarke et al.(1991): RO = (wet mass of total egg batch of the female / wet mass of the female without embryos).

Fecundity, egg volume, brood mortality, and reproductive output
The mean size (CL) of ovigerous females analyzed was 17.6 mm (± 3.49 mm standard deviation), ranging from 10.7 mm to 24.6 mm CL.All developmental egg stages were encountered; females with recently spawned embryos were most abundant (Table 2).Females carried a minimum of 50 eggs and a maximum of 1858 eggs, independent of the developmental stage of the embryos (Table 2).The number of recently spawned eggs (Stage I) increased with maternal size (Fig. 1); slopes of the allometric regressions for all three egg stages and maternal size were significantly different (p < 0.05) from zero, however, ANCOVA (Sokal and Rohlf, 1981) did not detect significant differences (p > 0.05) among these regressions.During the incubation period females lost an average of 51.1% of the initially spawned embryos, and brood mortality increased with female size (Table 3).

Chemical composition of developing embryos
Wet and dry mass of recently-laid eggs were 61.15 µg and 21.90 µg, respectively, but while the former increased slightly during embryogenesis, the latter decreased considerably (Table 4).Not surprisingly, water was the predominant component of the developing embryos and increased gradually from 62.2 % to 70.2 %.The organic matter declined and ash content augmented during the incubation period (Table 4).

Egg production
Maternal body size is considered to be the principal determinant of egg production in decapod crustaceans (e.g., Hines, 1982Hines, , 1991;;Corey and Reid, 1991;Lardies and Wehrtmann, 1997), and our results clearly corroborate this tendency.A comparison with other similar-sized pandalid species reveals that the maximum clutch size of A. grayi is the lowest for the species listed in Table 5.It would be interesting to compare also egg volume, egg mass volume, and RO among species, however, such data are unavailable for these similar-sized pandalids.Therefore, it remains to be clarified whether A. grayi allocates less energy (in terms of eggs) for reproduction than other species of the same family with ovigerous females covering a comparable size range.
Young  and constitute only a minor fraction of the early egg volume reported for two Antarctic shrimps, Chorismus antarcticus and Notocrangon antarcticus (see Clarke, 1993a).The comparison with these polar species corroborates the tendency of producing larger eggs at higher latitudes (Thorson, 1950;Clarke, 1992).The volume of Stage I embryos does not increase with maternal size (Fig. 2 and Table 2).This probably implies that there are no maternal size-related differences with regard to energy allocation per embryo in A. grayi, which is in contrast to the results obtained from polar caridean shrimps (Clarke, 1993b).In the four species investigated by Clarke (1993b), female size had a positive influence on egg size.This discrepancy between shrimps from polar and temperate zone may be related to environmental conditions (e.g., temperature, food availability, competition).However, the larger size of recentlyspawned A. grayi eggs from a population inhabiting central-southern Chile (Wehrtmann, unpubl. data) demonstrates intraspecific plasticity in egg size and supports the hypothesis that much of the variability in egg volume among populations may be related to local factors (see Clarke, 1993b).
Intraspecific differences in egg size along a latitudinal gradient have been observed in various decapod species (for caridean shrimps see: Clarke et al., 1991;Gorny et al., 1992;Clarke, 1993b;Lardies, 1995), and A. grayi seems to be another example of such a latitudinal cline in the reproductive biology: a limited data set (Wehrtmann, unpubl.)concerning A. grayi from Putemún, central-southern Chile, suggests that recently-spawned eggs are substantially larger in the Magellan region compared to those produced in central-southern Chile (mean egg volume: 0.037 mm 3 ± 0.0020 mm 3 S.D.; n = 7).Since egg volume is significantly correlated with nutrient content in caridean shrimps (Clarke, 1993a), the above-mentioned difference in early egg volume reflects a real difference in energy investment per embryo.
Austropandalus grayi loses approximately 51.1% of the initially extruded embryos.Some of the egg loss may be attributed to the trawling process (see also Gorny et al., 1992), however, since A. grayi was sampled generally from well above 100 m depth, is seems unlikely that the sampling procedure is responsible for an important portion of the calculated brood mortality.The obtained value of 51.1% is not unusual and lies within the range reported from other caridean shrimps (Corey and Reid, 1991;Kuris, 1991;Pandian, 1994).However, the egg loss observed in A. grayi is more than compensated by a concomitant increase of egg volume by 88.4 %.We assume that the early egg mass almost completely fills the available brood space in A. grayi; during the course of the incubation period, the increasing egg mass temporarily outgrows the attachment area which facilitates the mechanical abrasion of the outer embryos (see Kuris, 1991).Our results regarding the EMV (Table 2) suggest that the egg mortality during embryogenesis does not result in an important decrease of the initially-spawned clutch volume of A. grayi.However, brood mortality in this pandalid shrimp seems to increase with maternal size (Table 3).Kuris (1991) reviewed the size-specific variability of egg loss among crustaceans and found that only half of the data sets analyzed exhibited a constant proportional loss of embryos over the size range of ovigerous females.Thus, our finding concerning the size-specific brood mortality in A. grayi is in accordance with the statement by Kuris (1991) that the relationship between egg mortality and maternal size is highly variable among decapods.

Chemical composition of developing embryos
Water is the major constituent of the developing egg of A. grayi, and the water content increases during embryogenesis from 62 to 70 %.These values are similar to those reported for a variety of other decapod crustaceans which release planktonic larvae (e.g., Pandian, 1970;Valdés et al., 1991; EGG PRODUCTION OF AUSTROPANDALUS GRAYI 329  Lardies and Wehrtmann, 1997).The overall percent increase of water during the incubation period (16%) is surprisingly low when compared to other decapods such as Macrobrachium rosenbergii (74%: Clarke, 1990), Betaeus emarginatus (106%: Lardies and Wehrtmann, 1997), Homarus gammarus (202%: Pandian, 1970), and Necora puber (148%: Valdés et al., 1991), and is relatively close to the water uptake reported for Crangon crangon (17%: Pandian, 1967), Nauticaris magellanica (27 -30%: Wehrtmann and Kattner, 1998), Betaeus truncatus (12 -19%: Lardies, 1995), and Callinectes sapidus (20%: Amsler and George, 1984).Pandian (1970Pandian ( , 1994) ) discussed the ecophysiological consequences of water imbibition and demonstrated that marine crustaceans depend heavily on the surrounding environment for water and salt.However, it is still unclear why the egg water content of certain marine decapod species increases substantially during embryonic development whereas other species with a similar initial percent water content show only a slight increase of this component.Mineral ashes constitute 2.6 % of the dry mass of newly-spawned A. grayi eggs (Table 4) which is similar to the ash values reported for early eggs of Crangon crangon (1.4%: Pandian, 1967) and Homarus gammarus (2.7%: Pandian, 1970).However, most decapods seem to extrude embryos with a considerably higher portion of mineral ashes (e.g., Pandian, 1967;Clarke, 1993a;Amsler and George, 1984;Lardies andWehrtmann, 1996, 1997).It is speculated that the differences in ash contents in recently produced embryos of different decapods is related to the degree of calcification of the embryonic and larval exoskeleton (see also Clarke, 1990).

Reproductive output
Compared to other pandalid species, the RO of A. grayi is lower than in Pandalus borealis and P. montagui, but higher when compared to the Chilean deep-water shrimp Heterocarpus reedi (Fig. 3).Although a single low-latitude data point for H. reedi strongly influences the relationship presented in Fig. 3, the results seem to indicate an increasing RO toward higher latitudes among pandalid shrimps which is contrary to the results of studies with other caridean shrimps (Clarke, 1987 and references cited therein): polar species of Pandalidae, Hippolytidae and Crangonidae showed a significantly lower RO than temperate species.Referring to intraspecific variability, the fact that A. grayi from central-southern Chile produces considerably smaller but approximately the same amount of eggs (Wehrtmann, unpubl. data) than similar-sized females inhabiting the southern tip of America may also suggest that the RO is lower at lower latitude.This would further corroborate the trend among pandalids of an enhanced energy allocation for reproduction at higher latitudes (Fig. 3).However, due to the lack of RO data from other pandalids, such a trend is only suggestive but not conclusive.More data from pandalid species inhabiting different climate zones are indispensable to elucidate the possible relation between reproductive investment and latitude.Clarke (1987)).Data regarding P. borealis refer to different populations (Clarke et al., 1991).RO-value for H. reedi refers to a population from northern Chile (Wehrtmann and Andrade, 1998).
FIG. 3. -Comparison of reproductive output among different pandalid species in relation to latitude (independent of their location in the northern or southern hemisphere).Indicated values are based upon dry mass with the exception of P. montagui (wet mass data according toClarke (1987)).Data regarding P. borealis refer to different populations(Clarke et al., 1991).RO-value for H. reedi refers to a population from northern Chile(Wehrtmann and Andrade, 1998).

TABLE 2 .
-Austropandalus grayi: size (CL) of ovigerous females, and clutch size, egg volume and brood mass during the incubation period; S.D. = standard deviation; n = number of observations.

TABLE 3 .
-Percent brood mortality in A. grayi per maternal size group based upon pairwise comparisons of egg numbers in embryonic stages I and III.Regresssion equations for Stage I and III were used to estimate average egg numbers per size group; TL = total length of ovigerous females.

TABLE 5 .
-Comparison of clutch size and maternal size among four pandalid shrimps covering a similar size range; CL = carapace length.