Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
JOURNAL OF CRUSTACEAN BIOLOGY, 27(4): 616–625, 2007 LARVAL GROWTH OF THE COCONUT CRAB BIRGUS LATRO WITH A DISCUSSION ON THE DEVELOPMENT MODE OF TERRESTRIAL HERMIT CRABS Fang-Lin Wang, Hwey-Lian Hsieh, and Chang-Po Chen (FLW, HLH, and CPC) Research Center for Biodiversity, Academia Sinica, Taipei 115, Republic of China (FLW, correspondence: [email protected]; HLH: [email protected]; CPC: [email protected]) ABSTRACT With intensified harvesting and environment deterioration during the past two decades, a rapid decline in the number of coconut crabs, Birgus latro, which is a protected species listed in the IUCN Invertebrate Red Data Book, has occurred on many islands. Thus, it is important to protect this species by establishing conservation areas and/or replenish natural population by larval cultivation. In this study, the development modes were analyzed and the effect of enriched diet on larval growth and survival were examined. Two types of zoeal development patterns were found. In general, zoeae took 29-33 days to complete five zoeal stages and metamorphose to glaucothoes. However, some zoeae directly metamorphosed from the 3rd zoeal to glaucothoe stage in 25;28 days. Morphologically, these zoeae with accelerated development had thoracic appendages that appeared like the fifth stage zoeae, but with the telson, antennule, and antenna similar to those of the third stage zoeae. When fed Artemia nauplii enriched with nutritious substances, the zoeae had significantly greater survivorship and sizes, particularly at the fourth and fifth zoeal stages. Accelerated development may suggest early adaptation to a terrestrial lifestyle. The adaptation of larval development related to glaucothoe size and zoeal life span is also compared and discussed for eight terrestrial hermit crab species of Coenobitidae. The comparison suggests four adaptive modes of larval development and these are described as mangrove adaptation, larger glaucothoe adaptation, smaller glaucothoe adaptation, and hypersaline adaptation. The selective advantage of each mode may reflect a response to the uniqueness of each specific habitat. MATERIALS AND METHODS INTRODUCTION Larval Sampling The coconut crab Birgus latro Linnaeus, 1767, is the largest terrestrial arthropod and inhabits many coral reef lifted islands throughout Indo-Pacific region, including the Green Island in Taiwan. It is a terrestrial hermit crab, belonging to the monospecific genus Birgus in Coenobitidae (Helfman, 1979). Females release their larvae into the sea in the new moon phase and the larvae remain in the planktonic community for about a month. After the planktonic stage, the glaucothoes settle to the bottom, find and carry a small gastropod shell, and emerge onto the shoreline as a part of terrestrial hermit crab communities (Reese and Kinzie, 1968). When they reach a size of roughly 1 cm across the carapace, coconut crabs give up the shell-carrying habit of their hermit crab ancestors. With intensified harvesting and adverse environment impacts, the number of coconut crabs declined rapidly during the past two decades on many islands. The coconut crab is a world-wide protected species listed in the IUCN Invertebrate Red Data Book since 1983 (Wells, 1983; Bruggren and McMahon, 1988). However, there is not enough data to decide if the coconut crab is an endangered species or not, and therefore it is currently classified as the criteria of ‘data deficient’ in the 2006 IUCN Red List Threatened species. Since the habitat of coconut crabs requires the coastal zones, seashore waters, and land areas, it serves as the flagship species of islands and is a good indicator for monitoring the quality of island ecosystems. The purpose of this study was to establish the methods of larval cultivation and examine the effect of enriched diet on larval growth and survival. In addition, the differences in the adaptation modes of larval development among eight described coenobitid species was compared and analyzed. Two ovigerous coconut crabs carrying dark brown, eyed eggs were captured at the dusk high tide on 30 August 2003 and 4 August 2004. This coincided with the dark phase of the moon as they walked down to seashore to wash their eggs on the south rim of Green Island (228409N; 1218309E), Taiwan. The females were submerged in a vessel containing filtered seawater (45 lm), and their mature eggs immediately hatched to the first stage zoeae when the eggs contacted seawater. The crabs were then released. The next morning these first stages zoeae were brought to the laboratory. Cultivation of Larvae Under Laboratory Condition In 2003, a total of 300 zoeae were reared in 6 aquariums, each contained 50 zoeae with 5 liters of filtered sea water (45 lm). The larvae were fed live freshly hatched Artemia nauplii three times per day from a strain obtained from Great Salt Lake (GSL), Utah, U.S.A. (Parker International, Utah, U.S.A.). Aquarium conditions included gentle bottom aeration, temperature of 25.4 6 0.98C, salinity of 34.6 6 0.4&, and constant 12L:12D light cycle. Half of the volume of sea water was withdrawn from each aquarium just before feeding in order to increase the zoeal opportunity to successfully capture the live diet. After an hour of feeding, the zoeae were transferred into a clean aquarium using a large bore pipette with a 5 mm diameter opening. The aquarium water was half used and half fresh sea water to allow the zoeae to adapt easily to the new environment. Larval developing stages were determined according to the larval morphology described by Reese and Kinzie (1968). Survival rates of zoeae were calculated. The size of individual was measured as total length (TL) and cephalothoracic length (CL; Fig. 1). Total length was measured from the tip of the rostrum to the posterior border of the telson, exclusive of the telson processes. The cephalothoracic length was from rostral tip to posteromedial end of carapace. Effects of Diet Enrichment on Zoeal Development Freshly hatched Artemia nauplii meet the two general criteria for suitable prey items for larval decapods in laboratory studies, they: 1) are of an appropriate size for easy capture and consumption, and 2) contain essential dietary nutrients. However, some studies on larval culture of decapods suggested that Artemia nauplii from GSL, Utah, U.S.A. were deficient in 616 617 WANG ET AL.: BIRGUS LATRO LARVAL GROWTH Table 1. Nutrient components of the enrichment diet (Chuan Kuan Enterprise Co., Ltd., R.O.C.). Components Percentage % Moisture Poly-unsaturated fatty acids (PUFA) (n-3) highly-unsaturated fatty acids ((n-3) HUFA) Docosahexaenoic acid (DHA) Eicosapentaenoic acid (EPA) Phospholipids Vitamins Proteolytic enzymes Carotenoids 1 34 27 11 8 3 + + + Fig. 1. The measurement of the larval size of Birgus latro (modified from Reese and Kinzie’s study in 1968). A, zoeal stage; B, glaucothoeal stage. CL: the cephalothoracic length; TL: the total length for larvae. some essential dietary components (Bookhout and Costlow, 1970). In order to increase the larval survivorship, a diet enrichment experiment was carried out in 2004. The enrichment substances were composed of poly-unsaturated fatty acids (PUFA), (n-3) highly-unsaturated fatty acids ((n-3)HUFAs), eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA), phospholipids, vitamins, carotenoids, and proteolytic enzymes (Table 1; Chuan Kuan Enterprise Co., Ltd., R.O.C.). Freshly hatched Artemia nauplii were enriched with these substances in a concentration of 500 ppm for 12 h under the condition of constant aeration, at a temperature of 24.5 6 0.58C, and a salinity of 34.4 6 0.6&. In 2004, a total of 300 zoeae were reared in 6 aquariums, each contained 50 zoeae with 5 liters of filtered sea water (45 lm). Three aquariums were set as the treatment group and fed on enriched Artemia nauplii, called EA hereafter. The other three aquariums served as the control group and were fed nornal Artemia nauplii diet, called NA hereafter. Zoeae were fed and reared the same procedures followed in 2003 with the temperature and salinity kept at 23.4 6 0.68C and 34.8 6 0.1& respectively. In addition to the survival rates, the larval size increment between two consecutively developing stages was calculated as total length at ‘n þ 1’ stage divided by total length at ‘n’ stage. An overall growth index was computed as the total length at the last survival stage divided by the total length at the first stage. The mean growth index was the average of all larval growth indices. When zoeae metamorphosed to decapodid, glaucothoes, they were removed individually to covered plastic beakers (Fig. 2) which hold 10 ml sea water, a layer of moist sand, and a pile of coral debris which emerged from water surface, creating a microhabitat similar to seashore in nature and allowing glaucothoes to move back and forth between ‘sea’ and ‘land’ environment. The beaker rearing conditions were temperature 24.8 6 0.38C and salinity 34.6 6 0.4&. Glaucothoes were fed pieces of shrimp or clam meat three times daily and provided small vacant gastropod shells (shell mouth: 1.65 6 0.16 mm; shell length: 3.68 6 0.34 mm). Once they began carrying the shell, glaucothoes were transferred into an aquarium simulating an intertidal environment with a slope that the lower side submerged in aerated sea water, the middle portion consisted of moist coral debris, and Fig. 2. The devices used for culturing larvae of Birgus latro. A, a zoearearing aquarium; B, a glaucothoe-rearing glass container holds coral debris which rises out of the water surface; C, an aquarium simulating intertidal environments for rearing glaucothoes. Area A is filled with moderately aerated sea water, Area B with a slope filled with moist coral debris, and Area C is filled with cotton balls containing fresh water. Numerous vacant, minute gastropod shells are also deployed within the aquarium. 618 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 27, NO. 4, 2007 Table 2. Duration of the developmental zoeal stages of Birgus latro in these studies compared with Reese and Kinzie’s study. ‘‘—’’ data not available. Duration of each developmental zoeal stage Study periods This study 2003 experiment Typical development Duration (days) Stage-skipping development This study 2004 experiment Measurement items Type of development II III IV V Glaucothoeal stage after hatching Duration of glaucothoes carrying shells under laboratory conditions 4;5 6;7 5;8 4;5 4;7 29-33 — Mean 6 SD 4.8 6 0.7 6.5 6 0.6 (days) n 287 229 Duration (days) Mean 6 SD (days) n Typical development Duration (days) Mean 6 SD (days) n Reese and Kinzie, Stage-skipping 1968 development I 7.1 6 2.6 4.3 6 0.5 6.1 6 0.5 121 8;16 38 — 18 25-28 12.6 6 2.7 4;5 4;5 4.5 6 0.6 4.3 6 0.4 216 118 Duration (days) 5;6 3;5 the higher side composed of cotton balls moistened by fresh water. The intent was to create a gradient of salinity differences to induce molting. — 7 4;8 7;9 5;7 5.4 6 1.4 8.0 6 0.9 5.9 6 0.8 40 20 18 8;9 — — 6;12 7 27-34 — 12;35 18 21.2 6 8.3 6 — — size at the 2nd zoeal stage was larger in TL in 2004 than in 2003 (t-test, P , 0.05). All zoeal stages from the study of Reese and Kinzie in 1968 were significantly smaller than those from this study. In 2003 experiment, the greatest growth index occurred between the second and the third zoeal stages, being 1.18, while the total growth index was 1.54 and the mean growth index was 1.12. In the 2004 experiment, the greatest growth index occurred between the first and the second zoeal stages, being 1.23, while the total growth index was 1.57 and the mean growth index was 1.12. Survival curves of larvae were shown in Figure 4. The fifty percent survival point occurred between the second and the third zoeal stages in both experiments but on the tenth day in the 2003 experiment and on the ninth day in the 2004 experiment. Survival rates of zoeae at the end of each stage in 2003 were 87%, 57.3%, 29.3%, 17.7%, 16% and were 95.3%, 69%, 38%, 10%, and 5.7% in 2004. RESULTS Zoeal Growth and Survival The number of zoeal stages of B. latro was variable in the 2003 experiment but constant in the 2004 experiment. In the 2003 experiment, larvae passed through three zoeal stages in 25;28 days or passed through five zoeal stages in 29;33 days before metamorphosing into the glaucothoeal stage. However, in the 2004 experiment all larvae developed through five zoeal stages in 27;34 days (Table 2). The duration of the fourth zoeal stage in 2004 was longer than that in the 2003 experiment (t-test, P , 0.05), and the other stages did not differ. The external morphology of the larvae at each stage was similar to those described by Reese and Kinzie (1968). The size of each zoeal stage increased after each molting (Table 3; Fig. 3). The size from the 1st to the fifth zoeal stage ranged from 3.03 mm to 4.68 mm in total length (TL) and 1.31 mm to 2.52 mm in cephalothoracic length (CL) in the 2003 experiment, and those from 3.06 mm to 4.79 mm in TL, 1.33 mm to 2.49 mm in CL in the 2004 experiment. The Accelerated Development of Zoeae A total of 13 zoeae, 4.3% of the original number of larvae, metamorphosed to megalopae, glaucothoe, in the 2003 experiment. Among them were 7 individuals (about 54% of all glaucothoes) that metamorphosed directly from the third Table 3. Sizes of Birgus latro larvae reared in this study and the results from Reese and Kinzie (1968) in TL = total length; CL = cephalothoracic length. ‘‘—’’ data not available. Larval sizes (mm; Mean 6 SD)</ENTRY Stages of Zoeae Glaucothoes Study periods Measurement items I II III IV V G This study in 2003 experiment TL CL n TL CL n TL n 3.03 6 0.07 1.31 6 0.16 12 3.06 6 0.09 1.33 6 0.04 10 2.8 6 0.12 15 3.54 6 0.2 1.53 6 0.13 12 3.77 6 0.14 1.63 6 0.06 10 3.4 6 0.03 15 4.19 6 0.1 2.05 6 0.12 12 4.26 6 0.16 2.07 6 0.07 10 3.9 6 0.1 11 4.44 6 0.09 2.42 6 0.13 12 4.5 6 0.17 2.43 6 0.09 10 4.3 6 0.13 15 4.68 6 0.09 2.52 6 0.1 12 4.79 6 0.18 2.49 6 0.09 10 4.6 6 0.2 11 4.06 6 0.08 1.39 6 0.08 13 4.07 6 0.16 1.96 6 0.08 10 4 — This study in 2004 experiment Reese and Kinzie, 1968 WANG ET AL.: BIRGUS LATRO LARVAL GROWTH stage zoeae (Zoeae III) on the 25th-28th days and named as Glaucothoe III herein, whereas the rearing individuals metamorphosed from the fifth stage zoeae (Zoeae V) on the 29th-33rd days and named as Glaucothoe V herein. Before metamorphosis, the thoracic appendages of these Zoeae III individuals looked similar to Zoea V individuals, but their telson, antennule, and antenna appeared similar to those of Zoea III individuals, which would normally molt to the forth stage zoeae (Zoea IV) (Fig. 5). There was no significant difference between glaucothoe III and glaucothoe V in size of TL (4.02 6 0.05 mm and 4.08 6 0.1 mm; ttest, P . 0.05). However, zoeae exhibiting abbreviated development had a significantly shorter zoeal duration than those which emerged from the fifth stage zoeae (26.7 6 1.1 days and 30.5 6 0.8 days; t-test, P , 0.05). Some glaucothoes began amphibious life on the 32nd day in the laboratory in the 2003 experiment. By the 35th day, surviving glaucothoes were increasingly active on coral debris exposed to air and once tried to push a gastropod shell, but did not carry it. All these glaucothoes died without molting to the first juvenile crab stage. Shell-Carrying Glaucothoe There were twenty fifth zoeal stage, 6.7% of the original number of larvae, metamorphosed to glaucothoes in the 2004 experiment. There was no evidence of abbreviated zoeal development. Six glaucothoes succeeded in carrying gastropod shell (Fig. 6) on the 36th day in their amphibious life until dying between 68 days to 76 days from hatching. Effects of Diet Enrichment on Zoeal Development Zoeal size and duration.—The total length of zoeae at the third stage that were fed with enriched Artemia nauplii as food (EA) were significantly larger than those in the control groups fed on normal diets (NA) (t-test, P , 0.05; Fig. 7). The glaucothoes from the experimental group (EA) were also larger than NA ones (t-test, P , 0.05; Fig. 7). The duration of zoeae in EA groups did not differ statistically from that in the control groups. Survival rate and metamorphosis.—The survival rates of EA groups were significantly higher than that of NA ones at the fourth and the fifth zoeal stages (t-test, P , 0.05; Fig. 8). Among 20 glaucothoes, 15 individuals (75% of the total) metamorphosed from EA groups. DISCUSSION Heterochronal Morphology and Reduction in the Number of Zoeal Stages In the typical development of Birgus latro, as zoea progress toward metamorphosis the larval characters gradually become suppressed and then disappear, and the zoea then displays adult-like characters. However, in this study we found a heterochronal development in which 5% of Zoea III individuals possessed a mixture of zoeal and glaucothoeal characters. The morphology of their telson and segmented antennal flagella resembled with those at the third zoeal stage, but their thoracic appendages and setose pleopods belong to those of the fifth stage zoeae that were ready to metamorphose into the glaucothoeal stage. These individ- 619 Fig. 3. Larval size of Birgus latro (mean 6 SD). The symbol ‘*’ means that size differed significantly (t-test, P , 0.05). Zoeal sizes at all stages from Reese and Kinzie’s study were significantly smaller than in this study (t-test, P , 0.05). uals succeeded in metamorphosing to the glaucothoeal stage without going through the fourth zoeal stage. The phenomenon of heterochronal development was also reported by Reese and Kinzie (1968), but no changes in morphological characteristics were described in their study. This novel morphology is non-genetic in origin, since in our study one female produced zoeae that progressed through two different development pathways, one with only three stages and the other with the typical five stages, while another female produced zoeae all undergoing typical five stage development. Moreover, in the study of Reese and Kinzie in 1968 zoeae were reared in isolation in order to determine from which zoeal stage the terminal molt to glaucothoe occurred. They found that 12.5% of Zoea III went through the terminal larval molt to the postlarval glaucothoes on the 21st day after hatching, and 58.3% Zoea IV metamorphosed to glaucothoe after 25-38 days of larval life, but only 8.3% individuals molted to Zoea V, which finally died. These results suggest that the larval development of B. latro has no fixed mode in their evolutionary history and that B. latro is capable of abbreviating its larval developmental sequence from the typical five zoeal stages. In the typical development of Birgus latro, the terminal zoea appeared to be a little further morphologically developed than those having an abbreviated development with three zoeal stages. The former possessed more setae and segments in the telson, antennule, and antenna. By contrast, this phenomenon does not occur in larvae of Coenobita clypeatus (Herbst, 1791) which metamorphosed to glaucothoeal stage from Zoea IV, Zoea V, and Zoea VI. The zoea of C. clypeatus at preceding stages in longer development series may be less developed than those metamorphosed at the same stage in shorter series. This suggests retrogressive development and that Zoea VI from a long series (five or six stages) were less developed than a terminal Zoea IV (Provenzano, 1962). Valentine (1981) stated that unusual environmental stimuli can alter the level of hormone secretion and produce abnormalities similar to those resulting from mutations. These 620 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 27, NO. 4, 2007 Fig. 4. Survival rates of larval Birgus latro reared in this study during 2003 (A) and 2004 (B). Glaucothoe III, V: indicating the megalopae developed from the third stage zoeae or from the fifth stage zoeae. ‘‘abnormal’’ individuals would presumably have no special trouble in reproducing, assuming that they are well adapted. The morphology and lifecycle could thus be propagated. Later genetic changes could then fix the new morphology, even developmental types, in phylogeny. These genetic changes may be favored by micro-evolution because they would stabilize a phenotype with significant adaptive value. Effect of Diet Enrichment on Zoeal Development The results from the 2004 diet enrichment experiment in this study showed that addition of nutritional supplements to food can enhance larval size and survival. The development of decapod larvae can be divided into two phases: an early slow growth phase and a later metamorphic phase, which is characterized by rapid growth and preparation for mor- WANG ET AL.: BIRGUS LATRO LARVAL GROWTH 621 Fig. 5. The mixture of characters from the Zoea III individuals with skipped development in Birgus latro occurring in this study during 2003. A, lateral view of the third stage zoea which metamorphosed to glaucothoe which exhibits thoracic appendages and setose pleopods (shown in arrows) that looked similar to the Zoea V individuals and were clumped by a layer of membrane; B, the zoea molted to the 4th stage had no mixture characters. phological changes associated with metamorphosis (McConaugha, 1985). Growth during the early larval stages consists of a gradual increase in lipid and protein content over two or more stages. Poor nutritional status during the zoeal stages of the mud crab, Scylla paramamosain (Estampador, 1949) appeared to have an adverse effect on survival of megalopa (Li et al., 1992). Brachyuran larvae also have dietary requirements for long-chain PUFA before completing development and metamorphosis. In a series of studies, larvae of Rhithropanopeus harrisii (Gould, 1841) were fed Artemia nauplii low in the poly-unsaturated fatty acids of 20:5x3 (Bookhout and Costlow, 1970; Johns et al., 1980). These larvae showed a dramatic drop in survival and displayed a number of abnormalities including partial molting, unusual carapace spination, and irregularly formed thoracic appendages during the last zoeal stage. It is evident that sufficient energy reserves from food for B. latro can contribute to larval development and metamorphosis. Some studies have suggested that it may be possible to culture coconut crabs (Chatterjee, 1977) and experiments on their cultivation have been carried out at the Marine Station in Maribago, Philippines (Horstmann, 1976). However, to date there are no reports of successful cultivation. Our Fig. 6. A glaucothoe of Birgus latro carries a gastropod shell. This individual metamorphosed to a glaucothoe from the fifth zoeal stage on the 27th day and began carrying the shell on the 37th day after hatching. findings and culturing techniques may assist to promote mass culture of the zoeae and glaucothoes so that young crabs produced could be used to restock many islands where stocks have seriously declined and thereby contribute to replenishing natural population. Developmental Categories and the Constant or Variable Number of Larval Stages Larval development has been described for 8 of 16 known species of coenobitids. Detailed data are summarized in Table 4. Based on the flexibility in the number of larval stages, larval development in this family can be divided into constant and variable types. The constant type is predominant and has rather fixed ontogenetic growth and development stages. It consistently recurs in a sequence with five zoeal stages that the majority of the decapod species demonstrate (Broad, 1957), but a minimum of two or a maximum of seven stages are also recorded in some species (Table 4). The variable type deviates either in the number or the duration of larval stages from that seen in the preponderance of related species. This variability has often been postulated to be a consequence of environmental and/ or genetic induction (Sandifer and Smith, 1979; Table 4). Fig. 7. Larval sizes of Birgus latro reared under food enriched (EA) or normal (NA) in 2004 experiment. Zoeae of the third stage in EA group were significantly larger than control groups (NA). The glaucothoes from EA groups were also larger than those from NA ones. *: P , 0.05. 622 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 27, NO. 4, 2007 Fig. 8. Cumulative survival rates through the larval stages of Birgus latro reared under food enriched (EA) or normal (NA). The survival rates of the forth and fifth zoeal stages were significantly higher in EA groups than in NA. *: P , 0.05. During the process of molting, larvae become encumbered by an exuvial cast that prevents swimming and feeding. Such a dangerous condition can lead to death either from the mechanical problems or from pelagic predation. Benzie (1982) suggested that a reduction in the number of larval molts and faster development would be strongly selected for up to a point, beyond which further changes of the developmental plan are more fundamental and more difficult to achieve. Development Type and Ecological Adaptation in Coenobitidae The adaptive strategies for most decapod species that progressively become more terrestrial would include a reduction in the number of larval stages and a faster growth rate (Hartnoll, 1965). The shortened development is adaptive because it: 1) reduces critical molt periods, 2) minimizes exposure to suboptimal conditions for growth and development (such as inadequate nutrition or poor hydrographic conditions), 3) reduces exposure to predation, 4) reduces dispersal from specialized and restricted adult habitats, and 5) reduces recruitment mechanisms (Knowlton, 1970; Robert, 1971). On the evolutionary aspect of habitat utilization, Dobkin (1963) conjectured that the type of development may be related to the period of time that different species have evolved for inhabiting land or freshwater environments. Those having existed terrestrially or in freshwater for a greater length of time in evolutionary history would have an abbreviated development, while those relatively new to this type of habitat would be expected to have shortening planktonic stages in their life history. Limited larval dispersal is also a consequence of shortened larval life, in conjunction with responses of both larva and adult crab, to overcome the evolutionary constraints from hydrographic regimes and osmotic regulation. Within a variety of environmental condition, such as semi-terrestrial, fresh water, estuarine, rocky intertidal shores, marine, and hypersaline, development of phenotypic plasticity would itself be selected for as a singularly important feature. The relationships among the glaucothoeal size, zoeal life span, and given habitats for eight coenobitid species for which larval development data were compared (Table 5). The results suggested that there are four modes of larval development (Table 4; Figs. 9, 10) described as follows. Mangrove adaptation mode (M1).—The hermit crabs, which adapt to swamp environments such as mangrove areas, have a development mode characterized by the largest glaucothoe size (4.38 mm), the least number of zoeal stages, and the shortest duration of zoeal life span. This mode is called herein ‘‘mangrove adaptation mode (M1)’’. An example species is the Australian land hermit crab, Coenobita variabilis (McCulloch, 1909) that undergoes abbreviated development and reaches the glaucothoeal stage in only six days at 308C after two non-feeding zoeal stages (Harvey, 1992). The adults are distributed intertidally and up to 100 m above tide levels. They are often found behind mangroves, sheltering from the heat of the day under rocks or logs. These mangroves extend from Exmouth Gulf to North Queensland in the Northern Australia, forming fringing forested zones usually with scrub communities along Table 4. The development modes of species of Coenobitidae reared under laboratory conditions. The symbol ‘*’ reveals the major zoeal stage that metamorphoses into the glaucothoeal stage. Adaptation mode Mangrove adaptation (M1) Larger glaucothoe adaptation (M2) Smaller glaucothoe adaptation (M3) Hypersaline adaptation (M4) Development category Number of zoeal stage Terminal zoeal stage Coenobita variabilis constant 2 2nd Coenobita capives constant 5 5th Coenobita purpureus constant 5 5th Coenobita compressus Coenobita clypeatus Birgus latro variable variable variable Coenobita rugosus constant 4,5 4,5,6 3,5 3,4,5 5 4th, 5th* 4th, 5th*, 6th 3rd, 5th* 3rd, 4th* 5th Coenobita scaevola constant 7 7th Species Adult habitat Beach and vicinity, usually near mangroves Beach and vicinity, commonly found on the shore to 600 m or more inland Beach and vicinity, commonly found on the shore to 600 m or more inland Beach and vicinity Beach and inland Inland Beach and vicinity, usually inhabits the shore Beach only Reference Harvey, 1992 Shokita and Yamashiro, 1986; Nakasone, 1988 Nakasone, 1988 Brodie and Harvey, 2001 Provenzano, 1962 This study; Reese and Kinzie, 1968 Shokita and Yamashiro, 1986 Al-Aidaroos and Williamson, 1989 623 WANG ET AL.: BIRGUS LATRO LARVAL GROWTH Table 5. The number of zoeal stages, zoeal life span, length of glaucothoe, and growth indices of Coenobitidae recorded to date. Species Coenobita C. variabilis C. compressus C. clypeatus C. rugosus C. purpureus C. cavipes C. scaevola Birgus B. latro B. latro Zoea III!G (this study, 2003) B. latro Zoea V!G (this study, 2003) B. latro Zoea V!G (this study, 2004) Number of zoeal stages Zoeal life span (days) Total length of glaucothoe (mm) Mean growth index Total growth index 2 4;5 4;6 5 5 5 5;8 19;43 16;25 20;31 24;31 22;34 4.38 4.33 4.30 4.00 4.28 4.33 1.02 1.17 1.17 1.18 1.18 1.22 1.02 1.86 1.85 1.96 1.91 2.18 7 54;80 3.94 1.14 2.16 Harvey, 1992 Brodie and Harvey, 2001 Provenzano, 1962 Shokita and Yamashiro, 1986 Nakasone, 1988 Shokita and Yamashiro, 1986; Nakasone, 1988 Al-Aidaroos and Williamson, 1989 3;4 3 24;38 25;28 4.00 4.02 1.13 1.12 1.64 1.58 Reese and Kinzie, 1968 This study 5 25;33 4.08 1.12 1.54 This study 5 27;34 4.07 1.12 1.57 This study landward and seaward margins and river and tidal creek edges (Hopley et al., 1975). Such habitats are subjected to extreme or unpredictable fluctuations in salinities and abbreviated development would be advantageous (Lucas, 1971). Larger glaucothoe adaptation mode (M2).—The hermit crabs with a large glaucothoe size (4.28-4.33 mm), but smaller than that in M1, mode exhibit 4 or 6 zoeal stages rather than the typical five zoeal stages and have a wider variation in zoeal life span, ranging from 16-43 days. This mode is called ‘‘the larger glaucothoe adaptation mode (M2)’’. These species include Coenobita purpureus (Stimpson, 1858), and C. cavipes (Stimpson, 1858), which both have the typical Fig. 9. Comparison of the glaucothoel size and duration of the zoeal life stage in eight coenobitid species. Four adaptive modes are recognized as mangrove adaptation (M1), larger glaucothoe adaptation (M2), smaller glaucothoe adaptation (M3), and hypersaline adaptation (M4). Reference five zoeal stages and adults living in the region from the shore to 600 m or more inland (Nakasone, 1988), C. compressus (H. Milne Edwards, 1837), a beachfront species that has inconsistent, 4-5 stages (Brodie and Harvey, 2001), and C. clypeatus, an inland species, with 4-6 stages (Provenzano, 1962). It is worth pointing out that C. cavipes has the greatest growth increment (mean growth index, 1.22; Table 5) at each intermolt and metamorphoses into a larger glaucothoe. This feature also occurs in other M2 species in which the intermolt increment is greater and metamorphosis is to a larger glaucothoe (Figs. 9, 10). Smaller glaucothoe adaptation mode (M3).—The hermit crabs that metamorphose at smaller glaucothoe size (4.00- Fig. 10. The relationship between mean growth index and total growth index. The species in M2 category seem to have a faster larval growth rate and thus become a larger glaucothoe through the four or five zoeal stages. 624 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 27, NO. 4, 2007 4.08 mm), possess fewer zoeal stages (3-5 stages) than typical and have less variation in zoeal life span, ranging from 20-38 days are placed in the ‘‘smaller glaucothoe adaptation mode (M3)’’. These species include B. latro (in this and Reese and Kinzie’s studies in 1968), and Coenobita rugosus (H. Milne Edwards, 1837). Coenobita rugosus is a beach inhabitant and that consistently has five fixed stages (Shokita and Yamashiro, 1986). Birgus latro, which adult is the largest terrestrial crustacean and the most inland distributed hermit crab, has a non-fixed number of zoeal stages with a tendency toward shorter zoeal stages and overall zoeal life without the trade-off of a smaller glaucothoe size (Fig. 9). Although B. latro evolves as a terrestrial species, it still must to rely on marine environment for planktonic larval development, dispersal and salt balance as the hermit crab members of M2 group which predominantly live on coastal forest. Starting from a normal five zoeal stages, heterochronal development evolved in Coenobitidae, with both inter- and intra-species variation in zoeal development patterns. These patterns were either accelerated or prolonged by changing molting times in concordance with their habitat conditions, particularly salinities levels, in addition to factors related to terrestrial or suboptimal conditions for growth. The mangrove adapted species, C. variabilis, which undergoes abbreviated development may reveal the influence of freshwater. Birgus latro, and M2 species, and other M3 species exhibit strategies allowing them to metamorphose into larger glaucothoe to help against harshness in intertidal zone or to decrease molting times and planktonic duration. The hypersaline inhabitant, C. scaevola, which produces smallest glaucothoes, seems to accommodate well to its more marine habitat. Hypersaline adaptation mode (M4).—The hermit crabs which metamorphose with the smallest glaucothoe size (3.94 mm) have the greatest number of zoeal stages (7 stages) and the longest zoeal life span (54-80 days; Table 5) are categorized as the hypersaline adaptation mode (M4). This mode includes only the Red Sea species, Coenobita scaevola (Forskål, 1775). Larval development of C. scaevola exhibits a condition of ‘‘mark-time molting’’ that is a kind of delayed development in which the zoeal stage may enter a sequence of molts with very little morphological change taking place, although each succeeding instar may show some slight increase in size (Al-Aidaroos and Williamson, 1989). When compared with other land hermit crabs, C. scaevola does not follow the linear relationship in growth indices where the larger the value of the mean growth index, the larger the total growth index. This may be attributed to the large number of zoeal stages for which C. scaevola is associated with its exclusive beach habitat. C. scaevola does not have a larger mean growth index, but, it exhibits a relatively large total growth index after completing six molts. Although it has larger terminal zoeae than in M2 and M3 species (Fig. 10), it metamorphoses into the smallest glaucothoe (3.94 mm; Table 5) of all the hermit crabs reported. Coenobita scaevola, lives close to the edge of sea where it is exposed to high salinity. Although this species is not always submerged in seawater, it always keeps seawater in its shell. This species performs osmoregulation where salinities are greater than 25&, whereas only a weak regulation is found at lower salinities (Spaargaren, 1977). With the energy intensive demand for osmoregulation in higher salinity water, C. scaevola may produce smaller eggs and thus rely on longer pelagic zoeal life to reach a size competent for successful glaucothoe metamorphosis. Al-Aidaroos and Williamson (1989) reported that C. scaevola went through seven zoeal stages, but this was based on only two observed glaucothoes. The variation in the number of zoeal stages may not be detectable and further detailed study is needed to confirm this observation. These four developmental modes seem primarily related to the salinity regimes occupied, but the difference between M2 and M3 groups is unclear. Future thorough study is needed to clarify this issue. ACKNOWLEDGEMENTS This work was partially supported by the National Science Council, Research Center for Biodiversity, Academia Sinica and National Taiwan University. We thank very specially Dr. Daniel Sheehy for his comments and suggestions on the manuscript. REFERENCES Al-Aidaroos, A., and D. I. Williamson. 1989. Larval development of the land hermit crab Coenobita scaevola (Forskål, 1775) (Crustacea: Anomura: Coenobitidae) reared in the laboratory. Journal of Natural History 23: 111-128. Benzie, J. A. H. 1982. The complete larval development of Caridina mccullochi, Roux, 1926 (Decapoda, Atyidae) reared in the laboratory. Journal of Crustacean Biology 2: 493-513. Bookhout, C. G., and J. D. Costlow. 1970. Nutritional effects of Artemia from different locations on larval development of crabs. Helolander wiss Meeres 20: 435-442. Broad, A. C. 1957. Larval development of Palaemonetes pugio Holthius. Biology Bulletin 12: 144-161. Brodie, R., and A. W. Harvey. 2001. Development of the terrestrial hermit crab, Coenobita compressus, in the laboratory. Journal of Crustacean Biology 21: 715-732. Burggren, W. W., and B. R. McMahon. 1988. Biology of the land crabs. Cambridge University Press: New York. Chatterjee, S. K. 1977. Wildlife in the Andaman and Nicobar Islands. Tigerpaper 4: 2-5. Dobkin, S. 1963. The larval development of Palaemonetes paludosus (Gibbes, 1850) (Decapoda, Palaemonidae), reared in the laboratory. Crustaceana 6: 21-41. Estampador, E. P. 1949. Studies on Scylla (Crustacea: Portunidae). I Revision of the genus. Philippines Journal of Science 78: 95-108. Forskål, P. 1775. Descriptions Animalium, Avium, Amphibiorum, Piscium, Insectorum, Vermium. 19 þ xxxii þ 164 pp., Copenhagen. Gould, A. 1841. Report on the Invertebrata of Massachusetts, comprising the Mollusca, Crustacea, Annelida, and Radiata. Folsom, Wells, and Thurston, Cambridge. Hartnoll, R. G. 1965. Notes on the marine grapsid crabs of Jamaica. Proceedings of the Linnean Society of London 176: 113-147. Harvey, A. W. 1992. Abbreviated larval development in the Australian terrestrial hermit crab Coenobita variabilis McCulloch (Anomura: Coenobitidae). Journal of Crustacean Biology 12: 196-209. Helfman, G. S. 1979. Coconut crabs and cannibalism. Natural History 88: 77-83. Hopley, D., S. T. Langsford, J. C. Schofield, M. A. J. Williams, R. P. Bourman, N. K. Chick, and E. D. Gill. 1975. 1973-1974 Research on Quaternary Shorelines in Australia and new Zealand a summary report of the ANZAAS Quaternary Shorelines Committee. Australian Quaternary Newsletter. Horstmann, U. 1976. Some aspects on the culture of the coconut crab (Birgus latro). National Research Council of the Philippines Bulletin 60: 8-13. WANG ET AL.: BIRGUS LATRO LARVAL GROWTH Johns, D. M., M. E. Peters, and A. D. Beck. 1980. International study on Artemia. VI. Nutritional value of geographical and temporal strains of Artemia: effects on survival and growth of two species of Brachyuran larvae, pp. 291-304. In, G. Persoone, P. Sorgeloos, D. A. Roels, and E. Jaspers (eds.), The brine shrimp, Artemia. Knowlton, R. E. 1970. Effects of environmental factors on the larval development of Alpheus heterochaelis Say and Palaemonetes vulgaris Say (Crustacea: Decapoda: Caridea), with ecological notes on larval and adult Alpheidae and Palaemonidae. Ph.D. thesis. University of North Carolina. Li, S., C. Zeng, H. Tanh, G. Wang, and Q. Lin. 1999. Investigations into the reproductive and larval culture of the mud crab (Scylla paramamosain): A research overview, pp. 121-124. In, C. P. Keenan (ed.), Mud Crab Aquaculture and Biology. Proceedings of an international scientific forum held in Darwin, Australia, 21-22 April 1997. ACIAR Proceedings No. 78, ACIAR, Canberra, Australia. Linnaeus, C. 1767. Classis V Insecta. Systema Naturae per Regna tria Naturae 1(2): 533-11327. Lucas, J. S. 1971. The larval stages of some Australian species of Halicarcinus (Crustacea, Brachyura, Hymenosomatidae) I. Morphology. Bulletin of Marine Science 21: 471-490. McConaugha, J. R. 1985. Nutrition and larval growth, pp. 127-154. In, A. M. Wenner (ed.), Crustacean growth: Larval growth, Crustacean Issues 2. A. A. Balkema, Rotterdam. McCulloch, A. R. 1909. Studies in Australian Crustacea. No. 2. Records of the Australian Museum 7: 305-314. Milne Edwards, H. 1837. Histoire naturelle des Crustacés, comprenant l’anatomie, la physiologie et la classification de ces animaux. Paris: Librariare Encyclopedique de Roret. Vol. 2, pp. 532. Nakasone, Y. 1988. Larval stages of Coenobita purpureus Stimpson and Coenobita cavipes Stimpson reared in the laboratory and survival rates and growth factors of three land hermit crab larvae (Crustacea: Anomura). Zoological Science 5: 1105-1120. 625 Provenzano, A. J., Jr. 1962. The larval development of the tropical land hermit crab Coenobita clypeatus (Herbst) in the laboratory. Crustaceana 4: 207-228. Reese, E. S., and R. A. Kinzie. 1968. The larval development of the coconut or robber crab Birgus latro (L.) in the laboratory (Anomura, Paguridae). Crustaceana Supplement 2: 117-144. Roberts, M. H., Jr. 1971. Larval development of Pagurus longicarpus Say reared in the laboratory II. Effects of reduced salinity on larval development. Biology Bulletin 140: 104-116. Sandifer, P. A., and T. I. J. Smith. 1979. Possible significance of variation in the larval development of palaemonid shrimp. Journal of Experimental Marine Biology and Ecology 39: 55-64. Shokita, S., and A. Yamashiro. 1986. Larval development of the land hermit crab, Coenobita rugosus H. Milne Edwards and C. cavipes Stimpson reared in the laboratory. Galaxea 5: 267-282. Spaargaren, D. H. 1977. On the water and salt economy of some decapod crustaceans from the Gulf of Aqaba (Red Sea). Netherlands Journal of Sea Research 11: 99-106. Stimpson, W. 1858. Crustacea Anomura. Prodromus descriptionis animalium evertabratorum, quae in Expeditione ad Oceanum Pacificum Septentrionalem a Republica Federata missa, Cadwaladaro Ringgold et Johanne Rodgers ducibus, observavit et descripsit, Pars 7. Proceedings of the Academy of Natural Sciences of Philadelphia 10: 225-252. Valentine, J. W. 1981. Emergence and radiation of multicellular organisms, pp. 229-257. In, J. Billingham (ed.), Life in the Universe Cambridge: MIT Press. Wells, S. M., R. M. Pyle, and N. M. Collins. 1983. Coconut or robber crab. I.U.C.N. Invertebrate Red Data Book, I.U.C.N.; Gland, Switzerland. 632 pp. RECEIVED: 12 October 2006. ACCEPTED: 23 March 2007.