Snails Achatinella Mustelina and Partulina Redfieldii (Achatinellinae)L

An Experimental Study of Growth and Reproduction in the Hawaiian Tree Snails Achatinella mustelina and Partulina redfieldii (Achatinellinae)l

2 3 SHARON R. KOBAYASHI AND MICHAEL G. HADFIELD ,4

ABSTRACT: Hawaiian tree snails of the subfamily Achatinellinae are unique to the Hawaiian Islands and highly endangered in the wild. Achatinellines are arboreal pulmonate gastropods characterized by slow growth and late age at first reproduction. Objectives of the laboratory studies described here were to add to the understanding of growth and reproduction of achatinelline snails. Juvenile Partulina redfieldii (Newcomb) and Achatinella mustelina Migheis were kept in laboratory environmental chambers with conditions set to emulate those in the native habitat of P. redfieldii. The snails were provided with fresh leaves and branches of Metrosideros polymorpha Gaud., a natural substratum for the snails. Laboratory comparisons of P. redfieldii and A. mustelina main tained with a natural diet augmented or not with cultures of native fungi grown on potato dextrose agar revealed that snails of both species grew significantly faster on the augmented diet and that P. redfieldii attained sexual maturity at an earlier age. Comparison of growth of P. redfieldii in the laboratory with similarly sized snails in the field revealed significantly faster growth in the lab oratory animals. There was no significant difference between growth rates of A. mustelina provided with an augmented food supply in the laboratory and similarly sized animals in the field. It is likely that food availability limits growth rate in the field for P. redfieldii, but there is no evidence that growth in the field for A. mustelina is food-limited. However, the natural diet or temperature-humidity requirements of A. mustelina may not have been ad equately met in the laboratory, obscuring laboratory-field comparisons. Par tulina redfieldii, collected from the field as adults and maintained in isolation in the laboratory, produced offspring for at least 4 yr without the oppor tunity to outcross. Fecundity of isolated individuals was comparable with that reported for animals in the field, and there was no indication of fecundity decreasing over time in isolation. In addition, four of five P. redfieldii isolated as juveniles attained apparent sexual maturity at ages of 3.2 to ca. 5 yr. A single offspring was produced by one of these snails, suggesting self-fertilization as one mechanism allowing the species to reproduce for prolonged periods of time in the absence of mates.

BECAUSE GEOGRAPHIC ISOLATION often leads fauna of isolated archipelagos are frequently to evolutionary divergence, the flora and useful in the study of processes of specia tion. Tropical islands are especially interest ing, because many bear a greater variety of I Financial support furnished by the U.S. Fish and habitats than are typically found on tem Wildlife Service, the Cooke Foundation, and a grant from the Conchologists of America. Manuscript accepted 15 perate islands (Clarke and Murray 1969). November 1995. Geographically isolated, the Hawaiian Is 2 University of Hawai'i at Manoa, Department of lands exhibit a great diversity of climates and Zoology, 2538 The Mall, Honolulu, Hawaj'j 96822. habitats ranging from deserts and tundras 3 University of Hawai'i, Kewalo Marine Laboratory and Department of Zoology, 41 'Ahui St., Honolulu, to rain forests and bogs. This combination Hawai'i 96813. of geographic isolation and a diverse array 4To whom correspondence should be addressed. of habitat types has resulted in an un- 339 340 PACIFIC SCIENCE, Volume 50, October 1996

paralleled degree of endemism. It has been the giant African snail, Aehatina fuliea, has estimated that 95% of the terrestrial native proven to be minimally efficient in its in Hawaiian flora and fauna are endemic tended role but is proving to be the primary (Carlquist 1980). Over 250 species of fruit agent of extinction of several native snail flies (Carlquist 1980) and ca. 60 species of faunas (Van der Schalie 1969, Hadfield and passerine birds (Freed et al. 1987) have been Mountain 1980, Clarke et al. 1984, Murray described for the Hawaiian Islands. Among et al. 1988). On 13 January 1981, the entire land molluscs, ca. 800 species have been de genus Aehatinella (comprising some 41 spe scribed (Zimmerman 1948), over 99% of cies), endemic to the island of O'ahu, was which are endemic (Carlquist 1980). placed on the U.S. Fish and Wildlife Service Tree snails of the subfamily Achatinel Endangered Species List, the first taxon linae, family Achatinellidae, are unique to above the level of species to be so listed. The the Hawaiian Islands. Achatinelline snails are recovery plan for the O'ahu tree snails of arboreal, feeding on epiphytic fungi that the genus Aehatinella signaled the necessity grow on native plants. Students of animal of captive propagation as an integral part of evolution have long been attracted to these the recovery process (U.S. Fish and Wildlife snails because of their striking radiations of Service 1992). shell color, banding patterns, and shapes A captive-rearing program for achatinel (Gulick 1905), but only five modern papers line snails was started in 1986 at the Univer have addressed major life-history character sity of Hawai'i. The first goal ofthis program istics of achatinelline snails (Hadfield and is to preserve in the laboratory a represen Mountain 1980, Severns 1981, Hadfield 1986, tative sampling of the genetic diversity of re Hadfield and Miller 1989, Hadfield et al. maining populations of achatinelline snails. 1993). A common suite of life-history traits A second purpose of the captive-rearing is reported for all achatinelline species thus program is to generate sufficient numbers far studied: they are born live and relatively of laboratory-born snails to enable their in large (4-5 mm), grow slowly, become re troduction to the wild when existing hazards productively mature at a relatively late age, to the species are reduced or eliminated (i.e., and have low fecundity compared with most when predators are controlled). To create other terrestrial snails (e.g., Aehatina fuliea laboratory conditions conducive to optimal [Kekauoha 1966], Partu/a spp. [Murray and growth, fecundity, and survivorship, under Clarke 1966], Liguus faseiatus [Voss 1976], standing the basic biology of the target ani Caraeo/us and Po/ydontes spp. [Heatwole mal is essential. To this end, data gleaned and Heatwole 1978]). This achatinelline life from field and laboratory studies enhance history pattern renders these endemic snails our ability to meet the objectives of a captive extremely vulnerable to introduced predators propagation program; for instance, environ and environmental disturbances. mental parameters (temperature, humidity, It was estimated more than 20 yr ago rainfall) are monitored in the field and emu that at least 50% of the Hawaiian terrestrial lated in the laboratory. Laboratory studies, snail species had become extinct since West on the other hand, provide an opportunity to ern colonization of the Islands (Y. Kondo, isolate and investigate environmental param unpubl. ms.). Causes of this massive decline eters that cannot be controlled in the field. range from habitat loss (e.g., agriculture) Although results of ex situ studies may not to shell collectors, and most important in duplicate what is happening in nature, the rcent times, to predation (Hadfield 1986). In data obtained from such studies provide use troduced rats (Rattus spp.) and a carnivorous ful insights into the biological needs and snail, Eug/andina rosea, are quickly decimat capabilities of the animal. ing many remaining achatinelline populations, The research questions addressed in this especially the Aehatinella spp. of O'ahu. Eu study arose from preliminary observations g/andina rosea, introduced into many Pacific of laboratory populations of an achatinelline Islands in an ill-conceived attempt to control species from the island of Moloka'i, Partu- Hawaiian Tree Snails-KoBAYASHI AND HADFIELD 341

/ina redfieldii (Newcomb). Those studies vide insight into the evolution of the snails, suggested that P. redfieldii grew faster in the provide a basis for further investigation, and laboratory than in the field, leading us to enhance the prospect that populations can investigate the basis for the difference. As recover from extreme environmental catas suming that the increased growth rate ob trophes, even when the likelihood of a snail served earlier was not an artifact of a small encountering a mating partner is severely or somehow biased sample, we tested the reduced. hypothesis that growth rate is food-limited in the field for two species of achatinelline snails, P. redfieldii, and a species from O'ahu, Achatinella muste/ina Mighels. Using a MATERIALS AND METHODS strictly controlled laboratory environment, we sought to determine if increased food General Laboratory Setup availability enabled juveniles of these species All animals used in this study were kept to grow at an accelerated rate, and then we in screened containers within environmental compared snail growth rates between the chambers set to emulate field conditions. laboratory and field. We also considered the Temperatures fluctuated between 16°C at alternative hypothesis that optimized envi night and 20°C during the day. The cham ronmental conditions created in the labora bers ran on a 12-hr dark cycle, and a sprin tory, perhaps by providing greater feeding kler system delivered a fine water spray every time, promote increased growth rates. 8 hr for 2-min intervals, 5 days per week. The second preliminary laboratory ob Watering was suspended 2 days of the week servation that stimulated this study was that (Friday and Sunday) to produce a cycle of individuals of Partu/ina redfieldii, isolated in wet and dry periods approximating the natu the laboratory as adults, continued to pro ral situation. All containers (replicate groups) duce offspring for 2 yr in the absence of a were provided each week with freshly col mate. The question addressed here is whether lected, leafy branches of Metrosideros poly there is a cost to reproduction in isolation, morpha Gaud., a tree native to Hawai'i, on such as lowered fecundity or lowered survi which the snails typically live. The sources vorship of offspring. We also hoped to learn of food were (I) the light film of natural whether reproduction in the absence of mat fungi that grow epiphytically on leaves and ing partners results from sperm storage, branches of M. polymorpha and (2) two which would require at least one mating ex strains of fungi, cultured in the laboratory on perience, or, alternatively, from self-fertiliza Difco brand potato-dextrose agar (PDA). tion or parthenogenesis. The original laboratory-cultured fungal Successful captive propagation of endan stocks were isolated from leaf surfaces of gered species such as the achatinelline snails M. polymorpha. relies on in-depth understanding of the ani mals' life histories, and at the same time, in-depth understanding of life histories may Effects ofEnhanced Food Supply on Growth be applicable to management of field pop and Survivorship ulations. Thus, the ultimate objectives of the laboratory studies presented here were All juvenile snails used in this study were to elucidate factors limiting growth and re born in the laboratory to snails collected as production in achatinelline snails and there adults from the field. The 26 mothers of the by to gain insight into how environmental juvenile Partu/ina redfieldii were collected parameters influence growth, age at sexual from The Nature Conservancy's Kamakou maturity, and, by inference, generation time. Preserve on the island of Moloka'i at an ele Discoveries made in the laboratory, such as vation of 1300 m, four between November the ability of achatinelline snails to continue 1986 and July 1987, 12 in June 1989, and a reproducing for years without mating, pro- final 10 in May 1991. The 22 adult snails 342 PACIFIC SCIENCE, Volume 50, October 1996

TABLE I SOURCES OF SNAILS USED IN GROWTH AND ISOLATION STUDIES

COLLECTION INFORMATION NO. ASSIGNED OFFSPRING SPECIES DATE LOCATION AGE COLLECTED TREATMENT TREATMENT

P. redfieldii January 1986- Kamakou Preserve Adults 4 None Growth study July 1987 P. redfieldii June 1989 Kamakou Preserve Adults 12 6 snails isolated Growth study, June 1989 isolation study P. redfieldii May 1991 Kamakou Preserve Adults 10 5 snails isolated Growth study February 1992 P. redfieldii June 1989 Below Kamakou Juvenile Isolated May None Preserve 1991 A. mustelina May 1990 Honouliuli Preserve Adults 12 None Growth study

collected in June 1989 and May 1991 and replicate container within the environmental kept in individual containers (detailed in chambers was changed weekly, based on a "Effects of Isolation," below) were mothers random schedule using the same technique of the juveniles utilized in the growth study. described for the treatment container assign The 12 mothers of the juvenile Achatinella ments. Length and weight of each snail were mustelina were collected from The Nature recorded on a bimonthly basis. Conservancy's Honouliuli Preserve, O'ahu, The data for juvenile growth rates from at an elevation of 600 m on 12 May 1990. A field populations, used for comparisons with summary of the sources of snails used in both laboratory animals of similar size, were taken studies is provided in Table I. from long-term field studies conducted by The juvenile snails were separated into Hadfield et al. (1993 and unpubl. data). The two treatment groups. The "food-enhance study populations of A. mustelina are located ment group" was provided with laboratory in Honouliuli Preserve, O'ahu, and the P. cultures of fungi, in addition to branches redfieldii populations are located in Kama and leaves of M. polymorpha. The control kou Preserve, Moloka'i. The populations group received nutrition strictly from the studied within these preserves also served as natural film of epiphytic fungi on leaves and source populations for the snails observed branches of M. polymorpha. Juvenile snails in the laboratory. The methods used in the were assigned randomly to a treatment group mark-recapture studies of field populations and to a container (replicate) within each were described by Hadfield and Mountain treatment group by first marking each snail (1980). with a unique code and then pulling tags with the same set of codes out of a bag. A cursory Effects ofIsolation on Fecundity and Size check of the resultant assignment was made at Birth to ensure that no containers were overly rep resented by the offspring of anyone mother. Juvenile and adult P. redfieldii were kept There were three replicates for each treat in isolation to measure fecundity in the ab ment group of P. redfieldii, each replicate in sence of breeding; apparent age at sexual itially containing six or seven individuals maturity, as indicated by the addition of a with initial shell lengths ranging from 4.31 shell lip (callus at the shell aperture), was also to 20.99 mm. Two replicates were established recorded. Six adult snails were collected from for each treatment group of A. mustelina; the field in June 1989 and maintained in in each replicate initially contained six or seven dividual terraria. Ten additional adults were individuals, ranging from 5.00 to 14.59 mm collected in May 1991 and put into two con in initial shell length. The location of each tainers, each housing five snails. Beginning Hawaiian Tree Snails-KoBAYASHI AND HADFIELD 343

2 February 1992, five of the snails collected period for 10 pairs of P. redfieldii were used in May 1991 were transferred to individual in this comparison; individuals in each pair containers, but were placed together in a sin differed by less than 5% in initial shell length. gle container overnight once a week to pro Seven pairs of A. mustelina were compared; vide an opportunity for mating, to serve as a individuals in each pair differed by less than control group. For the purposes ofthis study, 10% in initial shell length. The time period birth data for offspring of isolated adults between initial and final measurements var were recorded up to 30 November 1992. ied between 3 and 4 months among pairs, but Dates of births, sizes, and weights of all off was approximately equal within pairs (less spring from each adult were recorded. Five than 2 weeks difference). juvenile P. redfieldii were isolated in May 1991. Four of these juveniles were born be ISOLATION STUDY: The survival of off tween 19 June 1989 and 1 February 1990 to spring born to isolated P. redfieldii was ana snails collected as adults in June 1989 from lyzed as a function of birth size and mother Kamakou Preserve. The fifth snail was col using a logistic regression model. This model lected by accident as a juvenile of unknown is a special case of a general linear model for age and unrecorded size in June 1989 near binary response variables. The analysis was the Waikolu Lookout, just below the Kama based on whether offspring lived 120 days or kou Preserve boundary; the size of this snail more, and the response variable was either when first measured in June 1990 suggests an positive or negative. Logistic regression was age of 1.5-2.0 yr when collected. Until the also used to test whether there were signif latter snail was placed in isolation, it was icant container (maternal) effects. kept in a container with two juveniles of A. Linear regression was used to determine mustelina. whether there was a relationship between (1) increasing time in isolation of adult P. red fie/dii and birth size of their offspring and (2) Data Analyses increasing time in isolation and the time in GROWTH STUDY: A nested analysis of vari terval between successive births. ance was employed to test for differences in Polynomial regression (to the fourth order) growth rates between snails in laboratory was employed as a preliminary test, to pursue treatments: growth rates of individual snails circumstantial evidence of a trend between were nested within containers within treat the time interval between successive births ment groups (enhanced food supply versus and birth size of subsequent offspring for iso control). A growth rate for each snail was lated adult P. redfieldii. calculated by dividing the change in length The average birth size for each isolated by the duration of the growth period in days. and nonisolated (control) adult P. redfieldii The length of the growth study was 6.5 was calculated, and an unpaired t-test was months for A. mustelina and 9.5 months for used to determine whether there was a dif P. redfieldii. An analysis of variance was also ference between birth sizes. used to determine whether there was snail size bias introduced as a result of the initial container assignment. Comparisons between a subset of labo RESULTS ratory and field animals were made using Wilcoxon's signed-rank test. To minimize Growth Study growth-curve effects, the animals used in this analysis were selected in a pairwise manner The results of the analysis of variance (one from a laboratory population, one from suggest that the random assignments made the field population) on the basis of similarity for the natural food versus enhanced food in shell lengths at the start of the growth groups did not unduly bias anyone container study. Changes in shell length over the study toward larger or smaller snails. For initial 0.035 0.035

-;; 0.030 0.030 '" -;; ~ 0.025 a ~ 0025 .5 aa ~ 0.020 &a S • a S 0.020 Q::'" a -= 0.D15 a. a -;" i • ::.:: 0.015 <:> • a .. 0.010 • • -= • • i 0.010 ~ ~ ~ ~ ~ " • ~ ..<:> 0.005 • ~ ~~ ~ a ~. " 0.005 • •• 0.000 •• • •• • 0 10 15 20 25 0.000 • .. 8 9 10 II 12 13 14 Initial Size (mm) 15 Initial Size (mm) FIGURE I. Comparison of laboratory-reared Partu /ina redfieldii provided with only the naturally occurring FIGURE 2. Comparison of laboratory-reared Achati food source (control group) and laboratory snails pro nella mustelina provided with only the naturally occur vided with enhanced food supply. Each symbol repre ring food source (control group) and laboratory snails sents one snail. provided with enhanced food supply. Each symbol repre sents one snail.

III-'$;: field 3.0 • laboratory ---S 2.5 S '-' .... -='=ll 2.0

Q,l ....;l= .5 1.5 Q,l ~ '"Q,l

..(oJ -= 1.0 0.5

o N V) 0- N N V) V) 0- 0- l" ~ V) N ~ V) I"- 0- 0- 0- 00 00 ~ ~ ('; '"N N'" M M ..., '" 00 00 00 Length ci - N M M M '" '" =-- Pair 2 3 4 5 6 7 8 9 10 Initial Length (mm) and Pair Number

FIGURE 3. Paired comparison of growth of Partu/ina redfieldii raised in the laboratory with enhanced food avail ability and similarly sized animals in the field. Pairs are arranged by increasing initial shell lengths. Data from 10 pairs were analyzed by Wilcoxon's signed-rank test (P= 0.0069). Hawaiian Tree Snails-KoBAYASHI AND HADFIELD 345 shell length, there were no significant dif laboratory with an enhanced food supply ferences between treatments or containers were not significantly different from those within treatments for either P. redfieldii or A. of similarly sized snails in the field (z = mustelina. -1.859). However, in four out of seven pairs Growth rates of P. redfieldii and A. mus of A. mustelina, field snails actually grew te/ina maintained in the laboratory with an faster than laboratory-reared snails and the enhanced food supply were significantly results are, in fact, only marginally non faster than those of snails maintained with significant (Figure 4). only the naturally occurring food (Figures Six P. redfieldii born in the laboratory 1 and 2). The amount of variation attribut and raised with an enhanced food supply able to container effects for both species was achieved maximum size and developed a negligible. shell lip at the shell aperture. A seventh snail Growth rates ofjuvenile P. redfieldii raised (isolation study), collected from the field as a in the laboratory and provided with only the juvenile, also developed a shell lip. The mean naturally occurring food were significantly age at sexual maturity for the six snails was faster than those of similarly sized snails in 3.35 yr (±0.52) and ranged from 2.40 to the field (z = -2.701) (Figure 3). Growth 3.81 yr. The mean size at sexual maturity was rates of juvenile A. mustelina reared in the 23.82 mm (±0.53) (Table 2).

~ field 3 • laboratory .-.- 2.5 e... t: '-' .c OJ) 2 -c ~ ~ ~ ,,.,. c :~: .- 1.5 ~~ . ~ '"~ ~ ("l"'" ....C

0.5

V) '

FIGURE 4. Paired comparison of growth of Achatinella mustelina raised in the laboratory with enhanced food availability and similarly sized animals in the field. Pairs are arranged by increasing initial shell lengths. Data from seven pairs were analyzed by Wilcoxon's signed-rank test (P = 0.063). 346 PACIFIC SCIENCE, Volume 50, October 1996

TABLE 2 MAXIMUM SIZE AND AGE AT WHICH Partulina redfieldii DEVELOPED A SHELL LIP IN THE LABORATORY

SHELL LIP FORMATION TREATMENT/ SNAIL MAXIMUM CONTAINER (CODE) BIRTH DATE DATE AGE (yr) SIZE (nun)

Enhanced 1 PA 23 June 1989 4 March 1993 3.7 22.42 Enhanced 2 BD 16 November 1990 10 April 1993 2.4 22.70 Enhanced 2 OD 21 July 1989 4 March 1993 3.6 23.16 Isolated YA 19 June 1989 10 April 1993 3.8 23.82 Isolated GD 19 June 1989 II November 1992 3.4 23.60 Isolated MA I February 1990 27 March 1993 3.2 23.25 Isolated OH ? 5 June 1992 ? 22.55

6.0 * 5.5 'II * >* 5.0 .... E E **f*** "**:~*: * -l.5 ~ ** *. ; * *** -l.0 * ** *;. ** *l* 3.5 ** * * * 3.0 * * 2.5

2.0 0 100 200 300 400 500 Interval Since Last Birth (days)

FIGURE 5. Effects of the duration of the interval since the last birth on subsequent birth size of Partulina redfieldii born to isolated mothers.

Isolation Study 14

12 The average number of offspring per snail per year for isolated P. redfieldii was four, and ranged between three and five. A scat ..!!l .~ 8 tergram of birth size versus interval since rf) prior birth suggested that larger than average ...... <:> 6 birth sizes occurred when the interbirth ,=,.. duration was ca. 150 days or longer (Fig ::l= 4 ure 5): there is no statistical evidence of a Z 2 trend, however. Based on a polynomial re gression analysis, birth size cannot be de scribed as a function of interval since prior 60 120 180 240 300 360 420 480 birth (r 2 < 0.091). Age at Death (days) Within the 3.5 yr of this study, 36 (42%) FIGURE 6. Frequency distribution of age at death of the 85 offspring born to the six isolated for Partulina redfieldii born to isolated mothers in the P. redfieldii died. The frequency distribution laboratory. Hawaiian Tree Snails-KoBAYASHl AND HADFIELD 347 of offspring mortality as a function of age a mean of 4.16 mm (±0.46) for offspring of (Figure 6) showed that most (74%) died control snails. Average birth sizes were not within 120 days of birth. Fitting the data to a significantly different between isolated and logistic regression model revealed that the control adult snails, although it should be probability of living longer than 120 days of noted that control snails were monitored for age increased significantly with increasing a shorter length of time and, as a result, had birth size (Figure 7). fewer offspring. There were no significant Birth sizes ranged from 3.05 to 5.66 mm, maternal effects on birth size, and survival of with a mean of 4.53 mm (±0.62) for off offspring born to isolated snails did not vary spring of isolated P. redfieldii (Figure 8); significantly among isolated adult snails. birth sizes ranged from 3.61 to 4.38 mm with An examination of birth size as a function

0.9 0.8 os '"OJ 0.7 Q = 0.6 ~ ... 0.5 0 * Cl. 0.4 0... Q., 0.3 0.2 ** 0.1 * 0 2 3 4 6 Birth Size (mm)

FIGURE 7. Survival to 120 days as a function of size at birth for offspring born to isolated Partulina redfieldii. Off spring were grouped into 10 size classes by birth size. Data were analyzed by logistic regression (P = 0.0082). Survival of offspring did not vary significantly among mothers (P > 0.4), and birth size did not vary significantly among mothers (P> 0.5).

12 ~ ';j J5 10 ...o ... 8 OJ .Q § 6 Z 4

2.75 3.05 3.35 3.65 3.95 4.25 4.55 4.85 5.15 5.45 5.75 Birth Size (mm)

FIGURE 8. Size-frequency distribution of offspring born to isolated Partulina redfieldii. 348 PACIFIC SCIENCE, Volume 50, October 1996 of the duration a parent snail was held in production at an earlier age. Consequently, isolation yielded inconsistent trends (Figure faster generation time is predicted where and 9). For only one of six isolated snails did re when conditions are optimal for growth. It gression data indicate a significant relation is important to verify this finding for other ship between these variables, with birth size species. appearing to increase over time in isolation. The observation that snails provided with The slopes of the regression data display a only a natural diet in the laboratory still positive, albeit nonsignificant, relationship grew faster than similarly sized snails in the between birth size and duration of isolation field indicates that the physical environment for four of the remaining five snails. (other than food quantity or quality) influ The average time interval between births ences growth in P. redfieldii. The mild and was 80 days, and the time intervals between constant climatic conditions created in the births did not appear to change with time in laboratory may allow the animals to feed isolation (Figure 10). Although regression undisturbed for longer periods of time, data indicated positive relationships between whereas in the field extreme weather con interbirth duration and time in isolation for ditions like storms and drought may force three of six isolated snails, the relationships long periods of inactivity. In addition, 12-hr between the variables were not significant for nights only occur during the winter months any of the snails. in Hawai'i, and the nocturnally active ani Since the five juvenile P. redfieldii reported mals may have grown faster because they on here were isolated in May 1991, four de were able to feed for longer periods in a year veloped a shell lip and are presumed to have round 12-hr light: 12-hr dark regime. reached sexual maturity. One year and 3 Partulina redfieldii kept in the laboratory months after the first of these isolated snails and provided with an enhanced food supply formed a shell lip, it produced an offspring, grew faster than both the laboratory control on 1 October 1993. The recorded birth size group and animals in the field, strongly sug was 4.65 mm, a size comparable with that gesting that food is a limiting factor in the reported above for offspring born to the iso field. It has been demonstrated that intra lated snails. specific competition for food can affect juve nile growth rate (Baur 1990a) as well as age at first reproduction and fecundity (Baur 1990b) in other land snails. Rapid pop DISCUSSION ulation growth currently observed in study Because age at sexual maturity is an populations of P. redfieldii in the field important factor in the rate of increase (Hadfield, unpubl. pers. observ.) may result of a population (Bell 1980), the question of in growth of individuals in these populations whether faster growth will lead to a lower being constrained by competition for a lim age at sexual maturity is an important one. ited food resource. We will have the oppor An enhanced food supply in the laboratory tunity to verify this prediction. results in faster growth rates for both P. red Growth in A. mustelina did not increase in fieldii and A. mustelina and a younger age at the laboratory environment; in fact, snails sexual maturity for P. redfieldii. The mean provided with only a natural diet in the lab size at sexual maturity for P. redfieldii in the oratory grew slower than animals in the field. laboratory was within the range observed for The reason for the different responses of A. wild populations, but the mean age at sexual mustelina and P. redfieldii may lie in their maturity in the laboratory was 3.35 yr com dietary requirements (discussed below). Per pared with an estimated 4 yr in the field haps the physical environment in the labora (Hadfield 1986). Sexual maturity thus ap tory (other than food quality or quantity) pears to be more size dependent than age may not have replicated sufficiently that of dependent in this species, and faster growth, the native environment of A. mustelina. The from whatever the cause, should lead to re- climate (temperature and "rainfall") in the 6 6 • • 4 • • ••

J 2 , r- = .139 r- = .052 P = .1714 p=.4132

1/1/88 1/1/90 1/1/92 1/1/94 1/1/90 1/1/92 1/1/94 a b

6 6

,-, • • # • E •• • • •• E 4 4 • • • "-' • • ••• • ~ • .-N rJJ 2 2 2 2 .c r = .030 r = .003 p =.5706 p = .8591 .-~-'"' 1/1/88 111/90 1/1/92 1/1/94 1/1/89 1/1/90 1/1/91 1/1/92 C d

6 • 6 • • • •• • •• 4 • 2 2 2 2 r = .001 r = .792 p =.9217 p =0002

1/1/88 1/1/90 1/1/92 1/1/94 1/1/88 1/1/90 1/1/92 1/1/94 e f

Days in Isolation

FIGURE 9. Size of offspring as a function of time in isolation for six isolated Parlu/ina redfie/dii. Each graph rep resents one mother; each symbol represents one birth. 500 ? 500 ? r- = .043 r-=.104 400 p = .4787 400 P =.2607

300 300 •

200 200 • \00 • •• \00 •• ••• ••• 0 • • 0 5 10 \5 5 10 15 a b

'IJ .c 500 2 500 2 r = .002 r = .001 P = .8837 =.9391 .-~-'"" 400 400 P

=~ 300 300 ~ • ~ 200 200 ~ • ~- \00 • 100 .....'IJ • ~ • • • • Q • • •• • • 0 5 10 15 0 5 10 15 C d

500 2 500 2 r = .043 r = 8.88e-6 400 • P = .4970 400 p =.9935 300 300 • 200 • 200 \00 \00 •• ••• •• • • • • 0 5 10 15 0 5 \0 15 e f Order of Births

FIGURE 10. Interval (in days) between parturitions for six individual Partulina redjieldii in isolation. Each graph represents one mother; each symbol represents one birth. Hawaiian Tree Snails-KoBAYASHI AND HADFIELD 351 laboratory more closely resembled that of strained by poor food quality and minimum Kamakou Preserve at 1300 m than that of adult size is limited by the large birth size of Honouliuli Preserve at 600 m, which is gen offspring, it follows that age at sexual matu erally warmer and drier. rity will be protracted. It is possible that the quality or quantity of The adult P. redfieldii isolated at the be food provided to A. mustelina in the labora ginning of these studies continued to produce tory was not the equivalent of that available offspring nearly 4 yr later, and there was no in the field and thus resulted in slower growth indication that they will cease to do so in the rates in the laboratory. It has been demon near future. Birth rate and birth size remain strated that variation in lipid, carbohydrate, constant and comparable with these charac protein, and fiber can affect growth in ter ters estimated for field populations of this restrial gastropods [summarized by Ireland species (Hadfield 1986). Although average (1991)], and when providing natural and liv size at birth in the laboratory is smaller than ing food, it is difficult to quantify the food that estimated for field populations (Had resources available to the animals. field 1986), survivorship of snails in the first Although sexual maturity may be attained year class is comparable with that estimated at an earlier age in the laboratory when con for field populations (Hadfield, unpubl. pers. ditions are optimal for growth, there are un observ.). doubtedly genetic constraints on both growth The mechanism that allows P. redfieldii rate and earliest age at which sexual maturity to reproduce for prolonged periods in the can be attained for achatinelline snails. Even absence of mates is poorly understood. The under optimal conditions, growth (and, con birth of offspring to a virgin P. redfieldii pro sequently, age at sexual maturity) is much duced in the laboratory proves that these slower in P. redfieldii than has been recorded snails must be capable of either self-fertiliza for most other terrestrial snails. For exam tion or parthenogenesis. Self-fertilization is ple, in Greece, Bradybaena fruticum (Muller) more likely, because it has been documented hatches at 2 mm and becomes sexually ma in a number of gastropod groups. Fresh ture within 1-2 yr at 19 mm (Staikou and water Helisoma spp. are capable of self Lazaridou-Dimitriadou 1990). Polydontes fertilization, but with fecundity significantly acutangula (Burrows) from Puerto Rico depressed (Lobato Paraense and Correa hatches at 22 mm and becomes sexually ma 1988). Fecundity of virgin Arianta arbusto ture in a year at 130 mm (Heatwole and rum (L.), a terrestrial gastropod, was lower Heatwole 1978). Ovoviviparous Partula spp. compared with nonisolated individuals (Baur from Mo'orea are 3-3.35 mm at birth and 1988). However, fecundity measurements for within a year attain a shell length of 15-16 reproductively isolated Punctum pygmaeum mm and sexual maturity (Murray and Clarke (Draparnaud) (Baur 1989) and Rumina de 1966). In Hawai'i, eggs of the introduced col/ata (L.) (Selander and Kaufman 1973) species Achatina fulica are 5 mm in diameter were not significantly different from those of and snails were reported to be sexually ma outcrossing individuals. The prevalent mode ture (exceeding 60 mm in length) within a of reproduction among some species of ter year (Kekauoha 1966), and Liguus fasciatus restrial slugs may be either obligate-out (Muller) from Florida hatches at 7 mm and crossing or self-fertilizing, varying from pop is sexually mature in 4 yr at 48 mm (Voss ulation to population. In other species of 1976). In an effort to explain the unusual life terrestrial slugs, all individuals are faculta history characteristics of achatinelline snails, tively self-fertilizing, depending on the avail Hadfield and Mountain (1980) speculated ability of mates (McKraken and Selander that growth rate is constrained by the poor 1980). nutrient quality of epiphytic fungi; they also Long-term sperm storage has been docu hypothesized that large birth size in achati mented in a number of terrestrial gastropods nellids is driven by interspecific competition and cannot be ruled out as an explanation for food and/or space. If growth rate is con- for continued reproduction by P. redfieldii 352 PACIFIC SCIENCE, Volume 50, October 1996 maintained in isolation for 4 yr. Arianta casional migration) maintain greater genetic arbustorum is capable of storing sperm for variation than that of panmictic populations, 2 yr; Cepaea nemoralis (L.) stores sperm for possibly due to directional selection within as long as 3 yr; and Helix aspersa (Muller) each microhabitat. Genetic analyses of snail has been reported to store sperm up to 4 yr populations in the field would verify this (summarized in Baur 1988). Parthenogenesis prediction and serve to guide translocation is a less likely explanation for solitary re efforts. In addition, the propensity of achati production by P. redfieldii simply because it nelline snails to self-fertilize may indicate is a very rare mode of reproduction among that fewer individuals are necessary to estab pulmonates (Duncan 1975, Fretter and Gra lish a viable population than would be possi ham 1964). ble if these snails were obligate outcrossers. If the self-fertilization hypothesis holds The fewer snails required to start a viable true, the consequences of this mode of re population means greater availability of food production on heterozygosity and the in resources per snail. It is conceivable, then, to breeding coefficient of both laboratory and create conditions conducive to rapid growth wild populations may be substantial (Hillis of populations in selected areas, enhancing 1989). Greater understanding of the preva the chances of species survival. lent mode(s) of reproduction and how it affects the genetic composition of popula tions is essential for accurately determining parameters such as minimum viable pop ACKNOWLEDGMENTS ulation size that are used in conservation management. This work would not have been possible A thorough understanding ofdemographic without generous financial support from the characters affecting growth and reproduction U.S. Fish and Wildlife Service and the Cooke is important to efforts aimed at promoting Foundation, and a grant from the Con recovery of endangered species. An enhanced chologists of America. We are grateful for food supply led to faster laboratory growth the expertise of L. A. Freed in reviewing the manuscript and for many useful suggestions. in both P. redfieldii and A. mustelina and also A. Taylor provided substantial guidance with enabled P. redfieldii to attain sexual maturity the data analyses. Special thanks to S. E. at a younger age, therefore reducing gen Miller for his invaluable assistance from the eration time. It follows that, if the longevity project's inception and to D. Hopper and L. of early maturing snails is comparable with Hadway for all their help in the laboratory. that of snails that mature later, a younger Access to the Palikea and Kamakou field age at first reproduction will translate to a sites was made possible by The Nature Con longer reproductive life and higher lifetime servancy of Hawai'i. fecundity (thereby reducing the possibility of bottleneck effects [Lovejoy 1977], a particular concern for small populations). In the field, it is not uncommon for a few trees to be LITERATURE CITED densely populated by achatinelline tree snails, while most of the surrounding native vegeta BAUR, A. 1990a. Experimental evidence for tion seems acceptable but uninhabited. Thus, intra- and interspecific competition in two practical application of the results presented species of rock-dwelling land snails. J. here might include translocating small num Anim. Ecol. 59: 301-315. bers of snails from densely populated trees ---. 1990b. Intra- and interspecific influ to adjacent uninhabited trees, potentially ences on age at first reproduction and fe lowering intraspecific competition and max cundity in the land snail Balea perversa. imizing the rate of increase of populations. Oikos 57: 333-337. Using computer simulations, Lacy (1987) BAUR, B. 1988. Repeated mating and female found that subdivided populations (with oc- fecundity in the simultaneously hermaph- Hawaiian Tree Snails-KoBAYASHI AND HADFIELD 353

roditic land snail Arianta arbustorum. In HEATWOLE, H., and A. HEATWOLE. 1978. vertebr. Reprod. Dev. 14: 197-204. Ecology of the Puerto Rican camaenid ---. 1989. Growth and reproduction of tree-snails. Malacologia 17(2): 241-315. the minute land snail Punctum pygmaeum HILLIS, D. M. 1989. Genetic consequences of (Draparnaud). J. Molluscan Stud. 55: 383 partial self-fertilization on populations of 387. Liguus fasciatus (Mollusca: Pulmonata: BELL, G. 1980. The costs of reproduction Bulimulidae). Am. Malacol. Bull. 7(1): 7 and their consequences. Am. Nat. 116: 12. 45-76. IRELAND, M. P. 1991. The effect of dietary CARLQUIST, S. 1980. Hawai'i: A natural his calcium on growth, shell thickness and tory. SB Printers, Inc., Honolulu. tissue calcium distribution in the snail CLARKE, B., and J. MURRAY. 1969. Eco Achatina fulica. Compo Biochem. Physiol. logical genetics and speciation in land A Compo Physiol. 98(1): 111-116. snails of the genus Partula. BioI. J. Linn. KEKAUOHA, W. 1966. Life history and pop Soc. 1: 31-42. ulation studies of Achatinafulica. Nautilus CLARKE, B., J. MURRAY, and M. S. JOHNSON. 80: 39-46. 1984. The extinction of endemic species by LACY, R. C. 1987. Loss of genetic diversity a program of biological control. Pac. Sci. from managed populations: Interacting 38: 97-104. effects of drift, mutation, immigration, se DUNCAN, C. J. 1975. Reproduction. Pages lection and population subdivision. Con 309-366 in V. Fretter and J. Peake, eds. servo BioI. 1(2): 143-158. The pulmonates, Vol. 1. Academic Press, LOBATO PARAENSE, W., and L. R. CORREA. London. 1988. Self-fertilization in the freshwater FREED, L. A., S. CONANT, and R. C. snails Helisoma duryi and Helisoma tri FLEISCHER. 1987. Evolutionary ecology volvis. Mem. Inst. Oswaldo Cruz Rio 83(4): and radiation ofHawaiian passerine birds. 405-409. Trends Evol. Ecol. 2(7): 196-203. LOVEJOY, T. E. 1977. Genetic aspects of FRETTER, V., and A. GRAHAM. 1964. Chapter dwindling populations, a review. Pages 4, Reproduction. Pages 127-164 in K. M. 275-279 in S. Temple, ed. Endangered Wilbur and C. M. Yonge, eds. Physiology birds. University of Wisconsin Press, of Mollusca, Vol. 1. Academic Press, New Madison. York. McKRAKEN, G. F., and R. K. SELANDER. GULICK, J. T. 1905. Evolution, racial and 1980. Self-fertilizing and monogenic strains habitudinal. Carnegie Inst. Washington in natural populations of terrestrial slugs. Publ. 25: i-xii, 1-269. Proc. Natl. Acad. Sci. U.S.A. 77: 684-688. HADFIELD, M. G. 1986. Extinction in Ha MURRAY, J., and B. CLARKE. 1966. The in waiian achatinelline snails. Malacologia heritance of polymorphic shell characters 27(1): 67-81. in Partula (Gastropoda). Genetics 54: HADFIELD, M. G., and S. E. MILLER. 1989. 1261-1277. Demographic studies on Hawai'i's endan MURRAY, J., E. MURRAY, M. S. JOHNSON, gered tree snails: Partulina proxima. Pac. and B. CLARKE. 1988. The extinction of Sci. 43: 1-15. Partula on Moorea. Pac. Sci. 42(3-4): HADFIELD, M. G., and B. S. MOUNTAIN. 150-153. 1980. A field study of a vanishing species, SELANDER, R. K., and D. W. KAUFMAN. Achatinella mustelina (Gastropoda, Pul 1973. Self-fertilization and genetic pop monata), in the Waianae Mountains of ulation structure in a colonizing land Oahu. Pac. Sci. 34: 345-358. snail. Proc. Natl. Acad. Sci. U.S.A. 70(4): HADFIELD, M. G., S. E. MILLER, and A. H. 1186-1190. CARWILE. 1993. The decimation of en SEVERNS, R. M. 1981. Growth rate determi demic Hawaiian tree snails by alien pred nations of Achatinella lila, a Hawaiian ators. Am. ZooI. 33: 610-622. tree snail. Nautilus 95(3): 140-144. 354 PACIFIC SCIENCE, Volume 50, October 1996

STAIKOU, A., and M. LAZARIDOU-DIMI VAN DER SCHALIE, H. 1969. Man meddles TRIADOU. 1990. The life cycle, population with nature-Hawaiian style. Biologist dynamics, growth and secondary produc 51(4): 136-146. tion of the snail Bradybaena fruticum. J. Voss, R. S. 1976. Observations on the ecol Molluscan Stud. 56: 137-146. ogy of the Florida tree snail, Liguus fas U.S. FISH AND WILDLIFE SERVICE. 1992. Re ciatus (Muller). Nautilus 90 (2): 65-69. covery plan for the O'ahu tree snails of the ZIMMERMAN, E. C. 1948. Insects of Hawaii. genus Achatinella. U.S. Fish and Wildlife Vol. I, Introduction. University of Ha Service, Portland, Oregon. wai'i Press, Honolulu.