Marine Biology (1999) 135: 321±333 Ó Springer-Verlag 1999

M. Thiel Reproductive biology of a wood-boring isopod, terebrans, with extended parental care

Received: 9 December 1998 / Accepted: 24 June 1999

Abstract The isopod Sphaeroma terebrans Bate, 1866 in size while in the maternal burrow, and juveniles of burrows in aerial roots of the red mangrove Rhizophora similar sizes could also be found in their own burrows. mangle L. The burrows serve as shelter and as a repro- Males did not participate in extended parental care, ductive habitat, and females are known to host their since most of them left the females after copulation. o€spring in their burrows. I examined the reproductive Many females that were born in the summer produced biology of S. terebrans in the Indian River Lagoon, a one brood in the fall and a second during winter/early shallow lagoon stretching for  200 km along the At- spring. Females that were born in the fall produced one lantic coast of Florida, USA. Reproductive isopods were brood during spring/early summer, but then probably found throughout the year, but reproductive activity died. Extended parental care in S. terebrans is short was highest in the fall and during late spring/early compared to other peracarid . It is concluded summer. During the latter periods, large numbers of that this reproductive strategy in S. terebrans serves subadults established their own burrows in aerial roots. primarily to shelter small juveniles immediately after The average numbers of S. terebrans per root were high they emerge from the female body, when their exoskel- during the fall, but decreased during the winter and eton is still hardening and their physiological capabilities reached lowest levels at the end of the summer. Females are still developing. Thus, in S. terebrans, extended pa- reached maturity at a larger size than males, but also rental care probably aids to protect small juveniles from grew to larger sizes than the males. The average size of adverse physical conditions in their subtropical intertidal females varied between 8 and 10 mm, the average size of habitat. males between 6.5 and 8.5 mm. The number of embryos female)1 was strongly correlated with female body length. No indication for embryo mortality during de- Introduction velopment was found. Parental females (i.e. with juve- niles in their burrows) hosted on average between 5 and Extended parental care is known from peracarid crus- 20 juveniles in their burrows (range 1 to 59). Most ju- taceans inhabiting a variety of habitats, such as seagrass veniles found in female burrows were in the manca stage and algal beds (Brearley and Walker 1995, 1996; Aoki and 2 to 3 mm in body length. Juveniles did not increase 1997), shallow soft-bottoms (Mattson and Cedhagen 1989; Thiel et al. 1997), and hard-bottoms (Shiino 1978). The most long-lasting parental care has been found in peracarid species that inhabit burrows, tubes or nests. In Communicated by N.H. Marcus, Tallahassee these species, juveniles may experience parental care for several months (Thiel 1999a). In free-ranging species 1 M. Thiel that do not host their o€spring in a parental dwelling Smithsonian Marine Station, 5612 Old Dixie Highway, but on their body, the parental care period is short, Fort Pierce, Florida 34946, USA usually not exceeding a few weeks (Aoki 1997; Thiel 1999a). For most peracarid species with extended pa- Present address: rental care, the reasons for the break-up of the parent± 1 Facultad Ciencias del Mar, Universidad Catolica del Norte, Larrondo 1281, o€spring association are not known. Campus Guayacan, Parental dwellings may not be stable enough to per- Coquimbo, Chile mit long-lasting coexistence of parents and their o€- Fax: 0056 51 209812 spring. Burrows or tubes in soft-bottoms, and nests e-mail: [email protected] in exposed seagrass or algal beds may be frequently 322 destroyed, resulting in the separation of parents and Careful observations by Messana et al. (1994) only their o€spring. The stability of a parental dwelling could recently revealed that the wood-boring isopod Sphae- thus have a substantial in¯uence on the duration of roma terebrans engages in extended parental care, during parental care. Among the most stable dwellings in the which juveniles live with their parents in so-called family marine environment are those that are established in burrows. Females host their o€spring in their burrows biotic microhabitats such as sponges, ascidians, bra- before these have fully developed their boring capabili- chiopods, bivalves, roots, rhizomes and wood. Several ties (Thiel 1999d). It is not well known how many ju- peracarid (and other) species are known to veniles live in one family burrow, or how large they grow raise their o€spring within biotic microhabitats (Wol€ during this period of extended parental care. In several 1976; Vader and Beehler 1983; Gonzalez and Jaramillo other peracarid crustaceans boring in wood (Limnoria 1991; Conlan and Chess 1992; Messana et al. 1994). spp.: Menzies 1954, 1957), wood-like substances such as These biotic microhabitats may be suciently stable to kelp stipes (Peramphithoe stypotrupetes: Conlan and allow a long period of extended parental care, possibly Chess 1992; Chess 1993), or seagrass blades (Brearley resulting in the occurrence of very large o€spring (Du€y and Walker 1995, 1996), males have frequently been 1996; Thiel 1999b). However, many biotic microhabitats found to cohabit with females. These observations sug- are small (salps, bivalves, ascidians, roots and rhizomes), gest that male boring peracarids may be involved in and thus do not provide much space for large families. parental care to a higher degree than males of other Usually, only one or two parents and a single clutch of peracarid species. It is not unusual to ®nd pairs of fe- juveniles are found in these small biotic microhabitats male and male S. terebrans within one burrow (Estevez (Vader and Beehler 1983; Gonzalez and Jaramillo 1991). and Simon 1975; Estevez 1978; Messana et al. 1994), but Besides sponges, large pieces of wood constitute another it is not known whether males of this species actively biotic microhabitat in the marine environment that participate in parental care or not. In the present study, could accommodate large numbers of closely related I paid ample attention to collect S. terebrans from in- individuals. Wood is utilized both as habitat and as food dividual burrows, to obtain a better understanding of by a variety of di€erent peracarid species (Menzies 1954; the reproductive biology of these wood-boring isopods. Eltringham and Hockley 1961; KuÈ hne and Becker 1964; John 1970; Rotramel 1975; Ribi 1982; Perry and Brusca 1989). Large pieces of wood often host numerous indi- viduals of similar sizes. Due to their destructive activities Materials and methods in arti®cial wood structures, a vast amount of research has been conducted on wood-boring peracarids (see Study area Clapp and Kenk 1963), but surprisingly few details are The study was conducted in the Indian River Lagoon, which ex- available about their reproductive biology. tends 200 km along Florida's Atlantic coast (27°32¢N; 80°20¢W). Sphaeroma terebrans is a wood-boring isopod, whose Sphaeromatid isopods occur in various intertidal and shallow main distribution is in mangrove forests around the subtidal habitats of the lagoon (Kensley et al. 1995). The lagoon was arti®cially modi®ed during the 1950s and 1960s as part of a world, where it preferentially dwells in the aerial roots of program designed to control mosquito populations. The extensive red mangrove trees (see e.g. Harrison and Holdich 1984; modi®cations included the construction of large impoundments Villalobos et al. 1985). This isopod also builds burrows that are temporarily ¯ooded and then drained (De Freese 1995; in other substrates such as dead wood and marshgrass Larson 1995). The formation of dams and reduced saltwater-ex- change within the impoundments substantially altered the intertidal rhizomes (Estevez 1994). It is generally assumed that the habitats in the lagoon (De Freese 1995). In many parts of the burrow serves as a shelter for the isopod rather than as a lagoon subjected to extensive remodeling, red mangrove trees food resource (John 1970). The size of a burrow gener- (Rhizophora mangle L.) now only grow in a narrow fringe along the ally does not exceed that of its inhabitant, and boring outer side of these impoundments. All collections referred to herein were made along these mosquito impoundments. The collection site activity decreases drastically after the isopod has estab- is 5 km north of the Fort Pierce Inlet. Water temperatures range lished a burrow in whatever substrate is available (John between 15 and 34 °C, and salinities between 17.5 psu (practical 1970). In experimental studies, S. terebrans apparently salinity units) during the winter months and 35 psu during the survived best in waters that contained planktonic algae summer months (SA Reed unpublished data). The oxygen levels in )1 (John 1970). The setation and morphology of the the lagoon water range between 2mll (late spring/summer) to 12 ml l)1 (fall), but usually vary between 5 and 8 ml l)1. The maxilliped and the ®rst three pereopods of S. terebrans water level in the lagoon is highest during the fall, and reaches low strongly suggest that this isopod ®lters its food out of the values during the summer months (Provost 1973). water in a similar manner as its boring congener S. quoyanum (Rotramel 1975). The aerial roots of red mangroves thus provide an ideal habitat for S. terebrans Sampling for several reasons, e.g. (1) the anterior parts of aerial The aerial roots of the red mangrove Rhizophora mangle L., con- roots are very soft (Gill and Tomlinson 1975), and it is taining burrows of Sphaeroma terebrans Bate, 1866, were collected therefore easy to establish burrows in them; (2) aerial monthly along the south-eastern shore of a shallow bay in the roots are frequently ¯ushed by water (during high tide), lagoon. On each sampling date, 40 to 60 roots were collected, and were carefully dissected immediately upon return to the laboratory. and therefore provide an optimal habitat for ®lter- The isopods were faithful to their burrows, remaining in them feeding organisms in a lagoonal environment. during handling, and leaving them only when they were opened 323 during dissection of the roots. All isopods from each burrow were preserved in 4% formalin and later transferred to 70% ethanol. The isopods were stretched out ¯at with forceps, and measured from the tip of the head to the tip of the pleotelson with the aid of a measurement ocular. Females were identi®ed by the presence of oosetigers, oocytes or embryos and males by the presence of penis appendages. Five di€erent embryo stages were distinguished, fol- lowing the categories of Kinne (1954) and Holdich (1968) ± Stage A: embryos shortly after fertilization, still egg-like in appearance; Stage B: limb buds emerge and towards the end of this stage the unpigmented eye becomes visible and the egg membrane is shed; Stage C: the embryo elongates and takes the shape of a comma (also referred to as the ``comma''-stage: see Jensen 1955), limbs develop further and eyes become pigmented; Stage D: appendages develop, pereonal segmentation is increasingly visible, the yolk- mass is largely used up; Stage E: all yolk reserves have been used up, appendages are fully extended, and embryos resemble juveniles in appearance, with a soft and whitish exoskeleton. While many sphaeromatid isopod species are semelparous (see e.g. Shuster 1991), others are reported to be iteroparous, producing two or more broods during their lifetime (see, e.g. Harvey 1969; Charmantier 1974). In order to determine the life history of Sphaeroma terebrans, it is important to know whether females produce one or two broods. In the present study, I recorded all Fig. 1 Schematic drawings of burrows of parental female Sphaeroma females collected on each sampling date and their respective re- terebrans in aerial roots of Rhizophora mangle containing very young productive states; this enabled female cohorts to be distinguished juvenile stages (A), older juvenile stages (B), and subadults remaining and their development to be followed. from a previous cohort (C) On some sampling dates, while dissecting the roots, all isopods from 10 to 27 individual roots were counted to estimate the sea- sonal abundance of Sphaeroma terebrans. Descriptions of the po- )1 sition of isopods in their burrows are given, based on observations isopods root in the winter/spring of 1997/1998 and 3 )1 made during the collection and dissection of maternal burrows. isopods root in August 1998 (Fig. 2). Empty isopod burrows were frequently found, particularly during the summer months. The width of the isopod burrows was strongly correlated with the size of their occupants Results (y = )0.003 + 0.564x, where y = burrow width and x = isopod body length; R2 = 0.776; p £ 0.05) (Fig. 3). General observations on Sphaeroma terebrans Only four small isopods (encircled data points in Fig. 3) in root burrows occupied burrows considerably larger than would have been expected from their sizes. All isopods collected in this study had constructed their burrows in aerial roots of Rhizophora mangle. The numbers of isopods on inhabited roots varied from 1 to Population structure in monthly burrow samples 48 individuals root)1. In roots containing high densities of isopods, burrows were sometimes adjacent, but usu- The isopods reached sexual maturity at a body length of ally physically separated from each other. Most subadult 6 mm for females and  5 mm for males. Females isopods were alone in their burrows. Among adult iso- reached sexual maturity at larger body sizes than males pods, females and males could be found in pairs within and grew to a larger size than males (Fig. 4). Through- one burrow. Parental females hosted their o€spring in out the year, the isopod population was composed of a the terminal end of their burrows. The females were large proportion of sexually mature individuals 6 to situated with their head towards their o€spring and their 12 mm in size. Subadult individuals were most abundant posterior end towards the burrow opening (Fig. 1A). in the months October to December and May to August. When males were present in the burrow of a parental The small numbers of subadult isopods of <5 mm be- female, they were always situated behind the female and tween January and April indicate that during these closest to the burrow opening. Juveniles at the terminal months almost no recruitment occurred (Fig. 4). end of the maternal burrow were usually clustered in a The percentage of female Sphaeroma terebrans paired dense group (Fig. 1A). Fully developed juveniles were with males decreased from almost 70% in September sometimes observed to start their own burrows at the 1997 to 20% in May 1998 (Fig. 5A). This decline in terminal end of the female burrow (Fig. 1B). In some the percentage of paired females covaried with the sex cases a few relatively large juveniles (3 to 5 mm body ratio (males:females), which showed a similar seasonal length) were found in their own burrows, which bran- decrease (Fig. 5A). Interestingly, the percentage of fe- ched o€ from the main burrow behind the female males with embryos in their bodies or with juveniles in (Fig. 1C). their burrows does not appear to have been a€ected by The average number of Sphaeroma terebrans root)1 the percentage of females that were paired in preceding decreased from 9 isopods root)1 in the fall to 4:5 months (cf. Fig. 5A and B). For example, the percentage 324

Fig. 2 Sphaeroma terebrans. Average number (‹1 SE) of isopods per aerial root of red mangrove, Rhizophora man- gle; juveniles in burrows of parental females not includ- ed in count

Fig. 3 Sphaeroma terebrans. Width of isopod burrows in aerial roots of red mangrove, Rhizophora mangle,asa function of isopod body length in September 1997; outlying (encircled) data points were excluded from correlation analysis

of females with embryos increased during spring 1998 were found without a male than with a male (Mann± (Fig. 5B) at the same time as the percentage of paired Whitney U-test; p < 0.05) (Fig. 7). Thus, males are females decreased (Fig. 5A). The percentage of parental generally not present when females are caring for o€- females (i.e. those with juveniles in their burrows) fol- spring, and parental care in Sphaeroma terebrans is lowed a similar seasonal trend as the percentage of fe- exclusively the task of the females. males with embryos, but with a lag of  1 mo (Fig. 5B). Throughout the year, 40 to 80% males were paired with females (Fig. 6). Only in August 1998, when the Reproductive potential of females male:female sex ratio increased to 1.76, was a low per- centage of males (22.4%) found in female burrows. In general, after the eggs are fully developed the females Separation of all reproductive females collected between molt to a reproductive stage with fully developed oos- September 1997 and August 1998 into females awaiting tegites. During this molt, the eggs are fertilized, and the fertilization (females with eggs) and females already females thereafter contain embryos. The average female fertilized (females with juveniles or embryos) revealed a size throughout the year varied between 8.5 and 10 mm distinct pattern of male attendance (Fig. 7). For females (Fig. 8A). Females were smallest during the summer with empty gonads or with developing eggs in their go- months. The average size of males displayed a similar nads, there was no signi®cant di€erence between the trend, but they were usually 1 or 2 cm smaller than the number attended by males and the number without females (Fig. 8A). The average number of embryos males in their burrows (Mann±Whitney U-test; female)1 increased slightly during the winter, but de- p > 0.05) (Fig. 7). For females carrying embryos or creased again towards summer (Fig. 8B). Most parental caring for juveniles (parental females), signi®cantly more females hosted an average of between 4 and 22 juveniles 325

Fig. 4 Sphaeroma terebrans. Number of females, males, and subadults in respective size classes collected from burrows in aerial roots of red mangrove, Rhizophora mangle

in their burrows (range = 1 to 59 juveniles). The A comparison of the reproductive stage of all females numbers of embryos were positively correlated with collected on each sampling date allowed di€erent co- female body length (y = )75.124 + 11.97x; horts to be distinguished and their longevity to be esti- R2 = 0.646; p £ 0.05) (Fig. 9). No signi®cant di€erence mated. Females born in summer 1997 produced eggs in in numbers of embryos female)1 were found between September 1997, and then increased in size while molt- the di€erent embryonic development stages (Kruskal± ing to the reproductive stage (maturation molt). These Wallis = 1.04; p > 0.05; n = 178 egg- or embryo-car- females incubated embryos in their incubating pouches rying females), indicating that embryonic mortality in October/November 1997 (Fig. 11). During the sub- during development was negligible. Most juveniles sequent months (November 1997 to January 1998), they found in female burrows were in the manca stage. Most released their o€spring and began to develop new eggs juveniles left the maternal burrow before the ®rst juve- in December 1997/January 1998. Apparently, after re- nile molt. Few juveniles reached sizes >3 mm in the leasing their ®rst brood, these females did not grow in maternal burrow (Fig. 10). On some sampling dates size during their molt to the resting stage (Fig. 11: fe- (October/November 1997, May 1998), juvenile Sphae- males with eggs in December 1997/January 1998). After roma terebrans <3 mm were also found in their own they had been fertilized, they incubated their second burrows (Fig. 4), indicating that they are capable of brood in February/March 1998, and most probably died establishing burrows at this size. after releasing this brood. Since some very large females 326

Fig. 5 Sphaeroma terebrans. A Percentage of females found paired with a male and sex ratio on each sam- pling date; B percentage of parental females (i.e. hosting juveniles in their burrows) and with developing embry- os in their bodies (Nos. on top abscissa number of fe- males found on each date)

Fig. 6 Sphaeroma terebrans. Percentage of males found paired with a female and sex ratio on each sampling date (Nos. on top abscissa number of males found on each date)

were found with eggs between March and June 1998; (see Fig. 4), recruitment of females during the spring/ some may have produced a third brood during this time summer months appears to occur continuously. A short (see Fig. 11 which shows 3 and 4 large reproductive break in recruitment of newly maturing females is rec- females in April and May 1998, respectively). Females ognizable in May 1998 (Fig. 11) that may separate the born in fall 1997 began producing eggs in February cohort born in the fall/winter from that born in the 1998, and the ®rst of these females incubated embryos in spring (May/June 1998). The ®rst females born in spring March 1998. After releasing their brood, some of these started to produce eggs in June 1998, but most did not females began to produce another brood, but many produce eggs until August 1998. These ``summer'' probably died (Fig. 11: low numbers of females females would persist throughout the subsequent winter, >10 mm with eggs from May to August 1998). Since and produce one brood in the fall and a second in juveniles born in the winter/spring grow very rapidly winter/spring. 327

Fig. 7 Sphaeroma terebrans. Average number of females in respective reproductive stages that were found un- paired and paired with males in their burrows on each sampling date (n = 583 fe- males collected between September 1997 and August 1998). Box-plots show me- dians, 25 and 75 percentiles and ranges (d outliers); Mann±Whitney U-test (n.s. no signi®cant di€erences; *p < 0.05)

Fig. 8 Sphaeroma terebrans. A Average length (‹1 SE) of females and males on each sampling date; B average numbers (‹1 SE) of embry- os and juveniles per brood- ing and parental female, respectively (Nos. on top abscissa number of females and males found on each date)

extended parental care. Juveniles of the same size as Discussion those in maternal burrows could be found in their own burrows. Males were not involved in parental care. In Female Sphaeroma terebrans reproduced and hosted fact, there is little indication that isopod families remain juveniles in their burrows year-round. The size-fre- together for long time-periods, despite the fact that their quency distribution of juveniles found in maternal bur- microhabitat is relatively stable and family burrows rows indicates that juvenile size does not increase during would be able to persist for several months. 328

Fig. 9 Sphaeroma terebrans. Numbers of embryos female)1 as a function of female size; all embryo- bearing females (n = 178) collected between September 1997 and August 1998 were pooled

Fig. 10 Sphaeroma tere- brans. Number of juveniles of various size classes (length, mm) found in bur- rows of parental females; all juveniles from each sampling date were pooled 329

Fig. 11 Sphaeroma tere- brans. Size and reproductive stage of females collected on each sampling date; female reproductive stage is char- acterized by the most devel- oped o€spring stage with which they were found [arrows subadult cohorts maturing to adult stage (ru- dimentary oostegites, developing eggs)] 330

Life history identi®er would indicate an apparent (but not real) loss of embryos. Kittlein (1991) suspected that high estimates The time required for sphaeromatid isopods to reach of embryonic mortality in S. serratum may have arisen maturity is relatively long (Borowsky 1996), and the from the fact that some late-embryo stages had already development of embryos and juveniles relatively slow. In left the maternal incubating pouches. These consider- the Indian River Lagoon, the Sphaeroma terebrans ations and the present results on embryonic develop- population is dominated by two major cohorts: one is ment in S. terebrans lead me to suspect that embryonic born during spring/early summer, matures in late sum- mortality in sphaeromatid isopods is generally low. mer/early fall, produces two broods (one in fall, one in Because of low embryonic mortality and the large size winter), and dies thereafter; the second cohort is born in that embryos reach in the maternal body, the whole fall, and its members recruit to their own burrows shortly body-cavity of the female is ®lled with embryos towards after. Members of this latter cohort mature during the the end of embryonic development. Feeding is sup- winter, produce their ®rst brood in spring/early summer pressed in females with fully developed embryos (John and, potentially, a second brood during summer. Isopods 1970). This corresponds to observations by Kinne from both cohorts attain an age of 10 mo. Other (1954), who found mostly empty guts in female Sphae- sphaeromatid isopods from temperate waters live con- roma hookeri containing embryos. In other sphaer- siderably longer, e.g. S. serratum in NW Europe lives for omatid isopods, Holdich (1968) and Shuster (1991) 2 yr, S. rugicauda for up to 18 mo (Harvey 1969), and S. noted that the mouthparts of gravid females are degen- hookeri in northern Europe for 18 mo (Kinne 1954). As erated and that females do not feed while incubating in S. terebrans, other sphaeromatids may produce their embryos in their bodies. Thus, while incubating one ®rst brood within the ®rst months of their life (Kinne brood, female S. terebrans cannot accumulate resources 1954; Harvey 1969; Charmantier 1974). The production for the production of a subsequent brood. Conse- of a second brood usually only occurs a considerable quently, oocytes are very small or undetectable in time (3 to 8 mo) after production of the ®rst brood females that have just released one brood from their (S. hookeri: Kinne 1954; S. monodi: Lejuez 1966; S. ser- bodies. As in free-ranging sphaeromatid isopods ratum: Harvey 1969, Charmantier 1974, Kittlein 1991). (S. quadridentatum: Borowsky 1996; S. hookeri: Jensen Many sphaeromatid isopods from temperate northern 1955, S. serratum: Daguerre de Hureaux 1979), follow- areas produce only one brood during their lifetime (S. ing release of the juveniles from their bodies, female rugicauda: Harvey 1969, Betz 1979; Dynamene bidentata: S. terebrans molt to a resting stage. During this resting Holdich 1968, 1971). stage the females accumulate resources and produce new It is puzzling that many females hosted small juve- eggs; this process requires considerable time (several niles in their burrows during the winter months, but that months). Surprisingly, most juveniles leave their mothers almost no recruitment took place between January and before a new brood is produced. During extended pa- April 1998. Possibly, during the cold winter months rental care, the juveniles are located at the head of the (water temperatures 15 to 17 °C), development is very female; it is possible that this constricts the ecient ¯ow slow, and a continuous but low recruitment activity of water through the maternal burrow. Juveniles may during this time period was not detected (see Fig. 2). interfere with the female's feeding activities, culminating in their active eviction by the female. When they leave the parental burrow, juvenile Early and extended parental care Sphaeroma terebrans are about the same size as when they were born. In other peracarid crustaceans, juveniles In Sphaeroma terebrans, development of the embryos can grow considerably during parental care, albeit some takes place in ventral invaginations originating from the juveniles may be forced to leave their parents at smaller brood pouch (also termed ``incubating pouches'': sizes than their siblings (Aoki 1997; Thiel 1999a). Many Charmantier and Charmantier-Daures 1994). The de- juvenile S. terebrans may establish burrows in the vi- veloping embryos attain a considerable size during this cinity of the parental burrow, and some juvenile burrows time, emerging at  2:5 mm body length, as manca even originate from the maternal burrow (also observed stages, from the female's body. Mortality during em- by Messana et al. 1994). Juveniles of other peracarid bryonic development was negligible; this agrees with the species have been observed to establish their ®rst own observations of Jensen (1955) for S. hookeri and Betz burrows in a similar fashion (Corophium volutator: (1980) for S. rugicauda. Other authors found consider- Thamdrup 1935; Limnoria lignorum: Henderson 1924; able di€erences between the numbers of eggs and the L. algarum: Menzies 1957; Lynseia annae: Brearley and numbers of embryos in S. hookeri, referring these losses Walker 1995, 1996; Leptocheirus pinguis: Thiel et al. to mortality during embryonic development (Kinne 1997). Small juveniles that can establish their own bur- 1954; SchuÈ tz 1963). As pointed out earlier (``Results ± rows under parental protection have the advantage of Reproductive potential of females''), female S. terebrans never being fully exposed at the surface of their burrow- grow in size during the maturation molt, i.e. when eggs substrates (wood, kelp stipes, seagrass blades, soft sed- are fertilized. Thus, comparison of females containing iments), where they could easily fall prey to ®sh or eggs with females bearing embryos using size as an crustacean predators. Not all o€spring from one clutch 331 start their own burrows from within the maternal bur- lowi: Thiel 1998, 1999a). While the mangrove roots row, but those that do so probably have considerably provide a stable environment that would allow the higher survival rates than their siblings that build a new prolonged coexistence of parents and o€spring, in the burrow away from the maternal burrow. In experiments case of S. terebrans both females and large juveniles in which roots containing isolated S. terebrans females (>manca) may be better o€ in their own burrows than with o€spring were suspended continuously in water, in a family burrow. almost all juveniles built their own burrows from within the maternal burrow (Thiel 1999d). It is not entirely clear why females in sheltered laboratory conditions Male role during extended parental care would provide longer parental care for their o€spring in Sphaeroma terebrans than females in the ®eld, but it is possible that interfer- ence from conspeci®cs, or from their congener S. quad- Detailed analysis of the reproductive stage of female ridentatum (excluded from the laboratory experiments) Sphaeroma terebrans has shown that males leave their reduces the duration of parental care in the ®eld (Thiel mates after fertilization and thus are not present when 1999e). Thus, the presence of fewer burrowing conspe- their o€spring develop. This is apparently in contrast to ci®cs may enable females to care for their o€spring more other boring peracarids, in which females and males are extensively than in the presence of many burrowing usually found in pairs. At present, it is not known conspeci®cs. This scenario would represent an interest- whether males in such species (Limnoria sp.: Henderson ing case of a feed-back mechanism whereby the duration 1924, Menzies 1954, 1957; Peramphithoe stypotrupetes: of parental care (and the survival of o€spring) is regu- Conlan and Chess 1992; Lynseia annae: Brearley and lated by the abundance of conspeci®cs. Walker 1995, 1996) are actively involved in extended While extended parental care in Sphaeroma terebrans parental care and, if so, which role they play. These may help to protect small juveniles from predation (as in species inhabit extensive and large burrow systems and other peracarid crustaceans: Thiel 1999c), the physical feed on the tissues of their hosts. Thus, the males may environment of female burrows may also be conducive support the females in ventilating and excavating the to juvenile survival. Early juvenile stages of S. serratum burrows systems and additionally remove waste prod- do not possess a fully developed osmoregulatory ca- ucts, as do terrestrial wood-boring beetles (Robertson pacity (Charmantier and Charmantier-Daures 1994), and Roitberg 1998). As other sphaeromatid isopods and it is likely that the closely related S. terebrans has a (Rotramel 1975), S. terebrans does not appear to feed on similar ontogenetic development pattern. Small juvenile the wood, but obtain their nutrition by ®lter-feeding S. terebrans that were taken from the maternal burrow (Rice et al. 1990). Instead of supporting females, male required several days to establish their ®rst own burrow S. terebrans might be in their way, and therefore not (Thiel 1999d). During this period they would be exposed tolerated in the female burrows during extended paren- at the surface of the aerial roots, potentially experiencing tal care. Male Limnoria sp. and P. stypotrupetes appar- the drastic salinity variations of their intertidal habitat. ently remain with the female, i.e. are monogamous; Isopods with well-®tting burrows (see Fig. 3) can store consequently, o€spring from di€erent clutches would all small amounts of water in them (my own observations) be from the same male. Thus, in these species the males possibly helping the isopods to survive exposure to may tolerate the juveniles from previous clutches be- sudden changes in salinity. Females hosting their small cause these most probably represent their own o€spring. o€spring at the terminal end of their burrows could thus In Sphaeroma terebrans the female possibly expels the eciently shelter them from osmotic stress. In S. serra- male. However, it is also possible that the male actively tum, osmoregulatory capacity increases substantially leaves the female after fertilization. Many aerial man- between the ®rst and second juvenile stage (Charmantier grove roots in the study area contain several mature and Charmantier-Daures 1994). It has also been re- females, and there may be a good chance of a male ported that during the ®rst days of their free-living ®nding another receptive female nearby. In S. terebrans, stages, wood-boring isopods exhibit a distinct change in females may thus reside in their burrows for long time color from white to brown (Henderson 1924); this was periods while males may move from female burrow to also noted in S. terebrans (my own observations). This female burrow, as already postulated by Venkata color change could result from calci®cation of the exo- Krishnan and Nair (1973). In another sphaeromatid skeleton, and might correspond to the increasing isopod, Paracerceis sculpta, males occupy spongocoels osmoregulatory and boring capabilities of the juveniles. to which they attract females (Shuster 1987, 1992). Thus, only during extended parental care can juvenile Following fertilization of the ®rst females, males attempt S. terebrans fully develop the physiological potential to attract additional mates while guarding ``their'' that will enable them to survive in their own burrows. sponge-territories, thereby protecting the ®rst females Other peracarid species that care for their o€spring in and their developing o€spring. Thus, in this harem- a burrow or tube may provide extended parental care for forming isopod, males indirectly engage in extended much longer time periods than Sphaeroma terebrans, parental care. In S. terebrans, the females that ®t tightly (e.g. Peramphithoe stypotrupetes: Conlan and Chess into their burrow openings (compared to the usually 1992, Chess 1993; Leptocheirus pinguis and Casco bige- smaller males) can presumably block the burrow 332 entrances more successfully than the males. It appears (eds) Recent developments in biofouling control. Oxford & IBH that in this wood-boring isopod there is no need for Publishing Co. Pvt Ltd, New Delhi, India, 97±105 Estevez ED, Simon JL (1975) Systematics and ecology of Sphae- males to attend their mates and o€spring, and they roma (Crustacea: ) in the mangrove habitats of Florida. consequently leave after copulation. In: Walsh GE, Snedaker SC, Teas HJ (eds) Proceedings of the International Symposium on Biology and Management of Acknowledgements I particularly thank S. Reed for assistance Mangroves. Institute of Food and Agricultural Sciences, Uni- during sampling of the isopods, and the other sta€ members of the versity of Florida, Gainesville, pp 286±304 Smithsonian Marine Station for providing ideal working condi- Gill AM, Tomlinson PB (1975) Aerial roots: an array of forms and tions. I thank S. Shuster and R. ZuÈ hlke for detailed comments on functions. In: Torrey JG, Clarkson DT (eds) The development the manuscript. Financial support for this study was received in and function of roots. Academic Press, London, pp 238±260 form of a Smithsonian Marine Station Postdoctoral Fellowship. Gonzalez M, Jaramillo E (1991) The association between Mulinia This is Contribution #475 of the Smithsonian Marine Station. edulis (Mollusca, Bivalvia) and Edotea magellanica (Crustacea, Isopoda) in southern Chile. Revta chil Hist nat 64: 37±51 Harrison K, Holdich DM (1984) Hemibranchiate sphaeromatids (Crustacea: Isopoda) from Queensland, Australia, with a References world-wide review of the genera discussed. Zool J Linn Soc 81: 275±387 Aoki M (1997) Comparative study of mother±young association in Harvey CE (1969) Breeding and distribution of Sphaeroma (Crus- caprellid amphipods: is maternal care e€ective? J Crustacean tacea: Isopoda) in Britain. J Anim Ecol 38: 399±406 Biol 17: 447±458 Henderson JT (1924) The gribble: a study of the distribution fac- Betz K-H (1979) Sphaeroma hookeri Leach und S. rugicauda Leach tors and life-history of Limnoria lignorum at St. Andrews, N.B. (Isopoda, Flabellifera): Taxonomie, Biologie und Speziesprob- Contr Can Biol 2, part 1 (14): 309±327 lem. Abh Verh naturw Ver Hamburg 23: 195±231 Holdich DM (1968) Reproduction, growth and bionomics of Betz K-H (1980) Sphaeroma rugicauda Leach, 1814 und S. hookeri Dynamene bidentata (Crustacea: Isopoda). J Zool, Lond 156: Leach, 1814 (Isopoda, Flabellifera): Weibchen-GroÈ ûe und 137±153 Embryonalstadien in einer rugicauda/hookeri-Mischpopulation. Holdich DM (1971) Changes in physiology, structure and histo- Eine vergleichende Untersuchung. Mitt hamb zool Mus Inst 77: chemistry occurring during the life-history of the sexually di- 211±216 morphic isopod Dynamene bidentata (Crustacea: Peracarida). Borowsky B (1996) Laboratory observations on the life history of Mar Biol 8: 35±47 the isopod Sphaeroma quadridentatum Say, 1818. Crustaceana Jensen JP (1955) Biological observations on the isopod Sphaeroma 69: 94±100 hookeri Leach. Vidensk Meddr dansk naturh Foren 117: 305± Brearley A, Walker DI (1995) Isopod miners in the leaves of two 339 Western Australian Posidonia species. Aquat Bot 52: 163±181 John PA (1970) Observations on the boring activity of Sphaeroma Brearley A, Walker DI (1996) Burrow structure and e€ects of terebrans Spence Bate, a wood boring isopod. Zool Anz 185: burrowing isopods (Limnoriidae) in southwestern Australian 379±387 Posidonia meadows. In: Kuo J, Phillips RC, Walker DI, Kensley B, Nelson WG, Schotte M (1995) Marine isopod biodi- Kirkman H (eds) Seagrass biology. Proceedings of an Interna- versity of the Indian River Lagoon, Florida. Bull mar Sci 57: tional Workshop Rottnest Island, Western Australia. Univer- 136±142 sity of Western Australia, Nedlands, pp 261±268 Kinne O (1954) Eidonomie, Anatomie und Lebenszyklus von Charmantier G (1974) Le cycle biologique de Sphaeroma serratum Sphaeroma hookeri Leach (Isopoda). Kieler Meeresforsch 10: (Fabricius, 1787) (Crustacea, Isopoda, Flabellifera) dans l'etang 100±120 de Thau (Golfe de Lion). Vie Milieu 24 (Ser A): 285±300 Kittlein MJ (1991) Population biology of Sphaeroma serratum Charmantier G, Charmantier-Daures M (1994) Ontogeny of os- Fabricius (Isopoda, Flabellifera) at the Port of Mar del Plata, moregulation and salinity tolerance in the isopod crustacean Argentina. J nat Hist 25: 1449±1459 Sphaeroma serratum. Mar Ecol Prog Ser 114: 93±102 KuÈ hne H, Becker G (1964) Der Holz-¯ohkrebs Chelura terebrans Chess JR (1993) E€ects of the stipe-boring amphipod Peramphithoe Philippi (Amphipoda, Cheluridae). Z angew Zool (Beih) 1: 1±141 stypotrupetes (Corophioidea: Amphithoidae) and grazing gas- Larson VL (1995) Fragmentation of the land-water margin with tropods on the kelp Laminaria setchellii. J Crustacean Biol 13: the northern and central Indian River Lagoon watershed. Bull 638±646 mar Sci 57: 267±277 Clapp WF, Kenk R (1963) Marine borers; an annotated bibliog- Lejuez R (1966) Comparaison morphologique, biologique et raphy. Oce of Naval Research, Washington, DC ge ne tique de quelques espeÁ ces du genre Sphaeroma Latreille Conlan KE, Chess JR (1992) Phylogeny and ecology of a kelp- (Isopodes ¯abellifeÁ res). Archs Zool exp ge n 107: 468±667 boring amphipod, Peramphithoe stypotrupetes, new species Mattson S, Cedhagen T (1989) Aspects of the behaviour and (Corophioidea: Amphithoidae). J Crustacean Biol 12: 410±422 ecology of Dyopedos monacanthus (Metzger) and D. porrectus Daguerre de Hureaux N (1979) Etude du cycle biologique annuel Bate, with comparative notes on Dulichia tuberculata Boeck d'une population de Sphaeroma serratum (Isopode, Flabellifere, (Crustacea: Amphipoda: Podoceridae). J exp mar Biol Ecol 127: ) du Bassin d'Arcachon. Crustaceana 37: 85±99 253±272 De Freese DE (1995) Land acquisition: a tool for biological di- Menzies RJ (1954) The comparative biology of reproduction in the versity protection in the Indian River Lagoon, Florida. Bull wood-boring isopod crustacean Limnoria. Bull Mus comp Zool mar Sci 57: 14±27 Harv 112: 363±388 Du€y JE (1996) Eusociality in a coral-reef shrimp. Nature, Lond Menzies RJ (1957) The marine borer family Limnoridae (Crusta- 381: 512±514 cea, Isopoda). Part I: Northern and Central America: system- Eltringham SK, Hockley AR (1961) Migration and reproduction of atics, distribution and ecology. Bull mar Sci Gulf Caribb 7: the wood-boring isopod, Limnoria, in Southampton Water. 101±200 Limnol Oceanogr 6: 467±482 Messana G, Bartolucci V, Mwaluma J, Osore M (1994) Preliminary Estevez ED (1978) Ecology of Sphaeroma terebrans Bate, a wood observations on parental care in Sphaeroma terebrans Bate 1866 boring isopod, in a Florida mangrove forest. Unpublished PhD (Isopoda, Sphaeromatidae), a mangrove wood borer from Ke- thesis. University of South Florida, Tampa nya. Ethol Ecol Evolut (Spec Iss) 3: 125±129 Estevez ED (1994) Inhabitation of tidal salt marshes by the estu- Perry DM, Brusca RC (1989) E€ects of the root-boring isopod arine wood-boring isopod Sphaeroma terebrans in Florida. In: Sphaeroma peruvianum on red mangrove forests. Mar Ecol Prog Thompson M-F, Nagabhushanam R, Sarojini R, Fingerman M Ser 57: 287±292 333

Provost MW (1973) Mean high water mark and use of tidelands in Thamdrup HM (1935) BeitraÈ ge zur OÈ kologie der Wattenfauna auf Florida. Fla Scient 36: 50±65 experimenteller Grundlage. Meddr Kommn Danm Fisk-og Ribi G (1982) Di€erential colonization of roots of Rhizophora Havunders (Ser: Fiskeri) 10: 1±125 mangle by the wood boring isopod Sphaeroma terebrans as a Thiel M (1998) Reproductive biology of a deposit-feeding amphi- mechanism to increase root density. Pubbl Staz Zool Napoli pod, Casco bigelowi, with extended parental care. Mar Biol 132: (I: Mar Ecol) 3: 13±19 107±116 Rice SA, Johnson BR, Estevez ED (1990). Wood-boring marine Thiel M (1999a) Duration of extended parental care in marine and estuarine in Florida. Unpublished Report, Florida amphipods. J Crustacean Biol 19: 60±71 Sea Grant College Program, University of Florida, Gainsville Thiel M (1999b) Host-use and population demographics of the (Extension Bull No. 15) ascidian-dwelling amphipod Leucothoe spinicarpa ± indication Robertson IC, Roitberg BD (1998) Duration of paternal care in for extended parental care and advanced social behaviour. J nat pine engraver beetles: why do larger males care less? Behavl Hist 33: 193±206 Ecol Sociobiol 43: 379±386 Thiel M (1999c) Extended parental care in marine amphipods. II. Rotramel G (1975) Filter-feeding by the marine boring isopod, Maternal protection of juveniles from predation. J exp mar Biol Sphaeroma quoyanum H. Milne Edwards, 1840 (Isopoda, Ecol 234: 235±253 Sphaeromatidae). Crustaceana 28: 7±10 Thiel M (1999d) Parental care behavior in the wood-boring isopod SchuÈ tz L (1963) Die Beziehung zwischen Ei-, Embryonenzahl und Sphaeroma terebrans. Crustacean Issues (In press) KoÈ rpergroÈ ûe der Weibchen einiger Peracarida aus dem Nord± ThielM(1999e)JuvenileSphaeromaquadridentatuminvadingfemale- Ostsee-Kanal. Zool Anz 171: 291±302 o€spring groups of Sphaeroma terebrans. J nat Hist (In press) Shiino SM (1978) Tanaidacea collected by French scientists Thiel M, Sampson S, Watling L (1997) Extended parental care in on board the survey ship Marion Dufresne in the two endobenthic amphipods. J nat Hist 31: 713±725 regions around the Kerguelen Islands and other Subantarctic Vader W, Beehler CL (1983) Metopa glacialis (Amphipoda, islands in 1972, ¢74, ¢75, ¢76. Sci Rep Shima Marineland 5: 1± Stenothoidae) in the Barents and Beaufort Seas, and its asso- 122 ciation with the lamellibranchs Musculus niger and M. discors. Shuster SM (1987) Alternative reproductive behaviors: three dis- Astarte 1983 (12): 57±61 crete male morphs in Paracerceis sculpta, an intertidal isopod Venkata Krishnan R, Nair NB (1973) Observations on the brood from the northern Gulf of California. J Crustacean Biol 7: 318± size and brood composition of timberboring sphaeromatids of 327 Cochin waters. Zool Anz 191: 119±122 Shuster SM (1991) Changes in female anatomy associated with the Villalobos CR, Cruz GA, Cruz RA (1985) Notas sobre la biologia reproductive moult in Paracerceis sculpta, a semelparous isopod de Sphaeroma terebrans Bate, 1866 (Sphaeromatidae: Isopoda) crustacean. J Zool, Lond 225: 365±379 en el manglar de Pochote, Provincia de Puntarenas, Costa Rica. Shuster SM (1992) The use of arti®cial sponges as breeding habitat Brenesia 24: 287±296 in Paracerceis sculpta (Holmes) (Crustacea: Isopoda) J exp mar Wol€ T (1976) Utilization of seagrass in the deep sea. Aquat Bot Biol Ecol 165: 75±89 2: 161±174