© 2004 by ICES/CIEM International Council for the Exploration of the Sea/Conseil International pour l'Exploration de la Mer
Spatial and seasonal variation in reproductive characteristics and spawning of southern calamary (Sepioteuthis australis): spreading the mortality risk
School of Aquaculture, Tasmanian Aquaculture and Fisheries Institute, University of Tasmania Locked Bag 1370, Launceston, Tasmania 7250, Australia
*Correspondence to N. A. Moltschaniwskyj: tel: +61 3 63243802; fax: +61 3 36243804. e-mail: natalie.moltschaniwskyj{at}utas.edu.au.
Southern calamary (Sepioteuthis australis) in Tasmania form spawning aggregations in Great Oyster Bay on the central east coast of Tasmania during spring/summer; these are targeted by commercial fishers. However, it is not known if there are similar aggregations farther south in Tasmania or at other times of the year, mainly because the species lives for less than a year. Therefore, this study describes and identifies differences in reproductive ecology of southern calamary on the east and southeast coasts of Tasmania, by sampling adults and surveying egg masses at inshore sites in both regions. Inshore populations on both coasts showed a similar seasonal trend of large gonosomatic index, reproductive output, and body size, and of greatest abundance during spring, and lowest in autumn. The number of egg masses was higher on the east coast, where mature calamary formed large spawning aggregations during spring and summer. However, there were no such aggregations during winter or autumn. Along the southeast coast, spawning activity was sporadic, resulting in isolated, low density, egg patches deposited over broader areas during spring, summer, and winter. There was no evidence of areas of seagrass or macroalgae associated with large depositions of egg masses at any time on the southeast coast. It appears that, by adopting different spawning behaviour in different locations and seasons, southern calamary may spread the risk of mortality in both space and time. The biological significance of this is unclear, particularly with respect to understanding the mechanisms that drive the development of spawning aggregations. Both spatial and seasonal spawning patterns appear to result from specific use of inshore sites at certain times of the year.
Keywords: egg deposition, reproductive ecology, seasonal patterns, spatial scale, spawning, squid
Received 27 January 2004; accepted 15 June 2004.
 |
Introduction |
|---|
 Top Introduction Material and methodsResultsDiscussionReferences |
|---|
Management of squid stocks using models developed for temperate teleosts is inappropriate owing to the striking differences in life history between the two taxonomic groups. Squid typically have short lives (<1 year) and rapid growth, with aseasonal spawning followed shortly by death. Further, there is considerable plasticity in life history characteristics, such as growth, age, and size at reproductive maturity (Boyle et al., 1995; Moltschaniwskyj, 1995; Pecl, 2001). Temporal variation in life history characteristics at a range of scales of weeks (Jackson and Pecl, 2003), months, and years (Moltschaniwskyj et al., 2003), provides a mechanism to spread mortality risk through time (O'Dor, 1998). This spreading of mortality risk is essential for annual and sub-annual species, because the storage effect in the population attributable to the overlapping of generations is absent (Warner and Chesson, 1985). Spatial risk spreading is evident in oceanic squid (e.g. Illex illecebrosus), which produces floating egg masses that result in offspring of the population experiencing a range of conditions in space (O'Dor, 1998). Spatial risk spreading is also seen in adult chokka squid (Loligo vulgaris reynaudii), which move long distances (typically hundreds of kilometres) between feeding and spawning grounds (Lipi
ski, 1998). Based on age estimates, it is apparent that spawning southern calamary (Sepioteuthis australis) in Great Oyster Bay on the east coast of Tasmania in spring/summer are a product of hatching events during late autumn/early winter (Pecl, 2001; Jackson and Pecl, 2003). Therefore, the question arises as to if and where spawning takes place during cooler months, given the potential importance of such spawning squid to spring/summer spawners. The spawning aggregations in Great Oyster Bay are routinely targeted by fishers, and as a protective measure, the embayment has been managed through short-term closures (Moltschaniwskyj et al., 2002). However, southern calamary are fished farther south along the southeast coast of Tasmania, an area currently unmanaged. These southeastern populations may exhibit different characteristics, and as a result may require a different management plan.
This study aims to combine the collection of adult calamary to assess reproductive status with large-scale surveys of egg masses along Tasmania's eastern and southeastern coasts, to identify whether spatially and temporally separated populations display varying spawning behaviour and life history characteristics. Although seasonal variation in the population structure of southern calamary has been explored previously (Pecl, 2001), there is a need to combine this information with data on seasonal variation in egg deposition. This study aims to address such broad-scale seasonal differences, in the hope that the output may support development of appropriate management strategies.
|
Material and methods |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Collection of adults
Most samples were obtained from commercial fishers, and complemented with research sampling. Research sampling targeted sites based on anecdotal information from commercial and recreational fishers about when and where large quantities of calamary were caught and/or calamary eggs sighted. Animals were collected using squid jigs, dipnets, and hand-spears. Mature adults constituted the majority of the catch, so all conclusions were drawn from the adult population. Calamary were obtained from October 1999 to April 2002, from two regions in Tasmania; the east (north of latitude 43°05'; Great Oyster Bay to Eaglehawk Neck on the Tasman Peninsula) and the southeast (south of latitude 43°05'; Fortescue Bay on the Tasman Peninsula to D'Entrecastreaux Channel). Over this period, 16–22 sites were fished in each of the two regions.
During 2001 and 2002, collections of adult calamary were made bi-monthly from each region between January and September. Outside this time period, sampling was carried out monthly owing to a seasonal increase in spawning activity. Research sampling consisted of two fishers fishing for 15 min each at every site. If no calamary were caught during that time, then the next site was visited. If calamary were caught at the site, then fishing continued until either 60 calamary were caught or until no calamary were caught over a 15-min period. In some months insufficient numbers were obtained, so additional samples were purchased from commercial fishers, when available. Fishers provided information on where animals were caught and the fishing method they used.
The mantle lengths (ML) and sexes of 2396 calamary were recorded during the course of the study. Total weight and reproductive status (including gonad weight) were measured in 1394 calamary. The stage of reproductive maturity was estimated macroscopically based on size and colour of reproductive organs, using a modified Lipiski scale (Pecl, 2000). Gonosomatic index (GSI) was calculated as gonad weight (ovary and oviduct for females, testis for males) as a percentage of total body weight.
Egg surveys
Subtidal egg surveys were undertaken along both coasts, sites being selected on the basis of anecdotal information and egg sightings by commercial and recreational fishers. In addition, sites that displayed similar characteristics to known spawning sites (i.e. providing suitable substratum for egg deposition and relatively sheltered from wave action) were also targeted. Sites in Great Oyster Bay on the east coast were those used in earlier intensive egg surveys (Moltschaniwskyj and Pecl, 2003). The habitat of the sites ranged from soft bottom covered with seagrass (Amphibolis antarctica and Heterozostera sp.), to reef covered with macroalgae (Ecklonia radiata, Sargassum spp., and Macrocystis pyrifera), all in water <10 m deep.
Surveys of egg masses were carried out with the same periodicity as the adult collections, 7–15 sites being visited in each region in each season. Most sites were surveyed one to three times each season. During 2001 and 2002, timed swims or manta-tows were undertaken at each site, allowing more area to be covered. Where habitat was broken reef and patchy seagrass with large areas of sand, a diver was towed behind the boat on a manta-board for 20 min (c. 500 m). The diver recorded the number of egg masses and estimated their development stage, ranging from I (newly laid) to IV (hatching), on the basis of the external appearance of the strands (Moltschaniwskyj et al., 2002). Sites with dense macroalgae and/or seagrass cover required more intensive searching, so one to two divers completed 20-min swims (covering c. 300–340 m2), searching within the dense vegetation for egg masses. During spring/summer when spawning activity increased and egg masses were frequently encountered, 20-min timed swims were used for most sites.
The density of egg strands provided a measure of spawning intensity, but because of logistical difficulties in counting egg strands in situ, the length of the egg mass was measured (from the attachment point to the distal point of the last egg strand in the mass). The number of strands was then estimated from the length of the mass, using a predictive regression equation (Moltschaniwskyj et al., 2002). If there were fewer than 20 strands in the egg mass, egg strands were counted directly. Each egg mass measured was allocated a development stage, as defined above. To convert the number of egg masses seen on a timed swim to density, the area searched was determined by timing how long it took to search 20 m2 under different densities of egg masses. This then allowed us to estimate how much substratum was searched during a 20-min swim.
To assess the mortality rates of eggs in each region, six egg strands were collected at random from 104 and 59 egg masses from the east and southeast coast, respectively, that had embryos developed beyond Stage 20 (Steer et al., 2003). Embryos (3409 from the east, 1840 from the southeast) were dissected from each strand within 8 h of collection and scored as "dead" if they were unfertilized, undergoing abnormal development, or opaque.
Catch per unit effort (cpue)
Commercial catch records accessed from the Tasmanian Department of Primary Industries Water and Environment (DPIWE) database were separated into two regions, based on the location of the recorded fishing blocks. To reduce the inclusion of records for which southern calamary represented an incidental by-catch rather than a target species, landings <10 kg taken with Danish seines, beach-seines, gillnets, dipnets and hand-spears were excluded. As hand-jigging and purse-seining nets are used specifically to target southern calamary, all catches made with these gears were included in the analyses. As cpue is not normally distributed, the arithmetic mean was a poor descriptor of the centre of the distribution, so the geometric mean of kg d–1 was calculated.
|
Results |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Size and sex ratio
Average mantle length differed as a function of region (F = 24.88, d.f. = 1, 2340, p < 0.001) and an interaction between sex and season (F = 9.97, d.f. = 3, 2340, p < 0.001). East coast calamary were consistently 10% larger than the individuals caught in the southeast (east 280.90 ± 1.78 mm, southeast 256.26 ± 5.21 mm), regardless of season or sex. Differences in ML among seasons were consistent between the two regions, and in both sexes the largest calamary were taken in spring and winter, and the smallest in autumn (Figure 1). A decline in average ML from spring to summer was evident in the male component of the population, but not in the female component (Figure 1).
|
There was no difference in the ratio of males to females between regions in any one season; the sex ratio across all seasons and locations was 2.2 males to 1 female. However, there was a difference in the proportion of males to females among seasons within each location (
2east = 24.56, d.f. = 3, p < 0.001; 2southeast = 28.17, d.f. = 3, p < 0.001). In both regions, there was a consistent pattern of proportionally fewer females in spring, though autumn catches in the southeast failed to follow this trend (Table 1).
|
) or less (
) than expected, under the assumption that the sex ratios were independent of season. Arrows are placed in cells that had normalized residuals (difference between observed and expected values) >±2.Reproductive condition
In both regions, across all seasons and both sexes, more than 80% of the calamary caught were mature (Stage 5), females having eggs in the oviduct ready to deposit, and males having spermatophores ready to be passed to the female (Figure 2). There was a significant difference in the proportion of calamary at each maturity stage between the regions (
2 = 78.54, d.f. = 6, p < 0.001). This was due to the southeast population having proportionately more immature (Stage 1) and spent (Stage 6) calamary than the east coast (Figure 2). Spent females were not caught, despite the presence of spent males during summer in both regions.
|
Comparisons of stage frequency distributions across seasons were only possible for the east coast, not because of any scarcity of calamary caught on the southeast coast, but because there were low numbers in some seasons and in some stages, largely because of the dominance of Stage 5 individuals. In the east coast samples, juveniles (Stage 0) and spent (Stage 6) calamary were omitted from the analysis because of their scarcity. The frequency at each reproductive stage differed among seasons (
2 = 708.71, d.f. = 12, p < 0.001). During spring there were proportionally more mature calamary (Stage 5), and fewer at all other maturity stages (Figure 3). During summer, although the proportion of mature calamary remained high, there were relatively more immature (Stage 1) animals. In contrast, during autumn, relatively few calamary were mature (Stage 5), and more were immature or in the early stages of maturity (Stages 1–4; Figure 3). By winter, the population had proportionally more Stage 4 calamary.
|
Seasonal differences in average female gonosomatic index (GSI; Fseason = 108.68, d.f. = 3, 412, p < 0.001) were consistent between the regions (Fregion*season = 1.71, d.f. = 3, 412, p = 0.164). GSI was highest during spring, declined during summer and autumn, and increased during winter (Figure 4). In contrast, the seasonal changes in male GSI differed in the two regions (Fseason*region = 4.63, d.f. = 3, 932, p = 0.003). For east coast calamary, male GSI was similar among seasons, the only difference being between the autumn low and the winter high (Figure 4). In contrast, on the southeast coast GSI was lowest in summer and autumn, and highest in winter and spring (Figure 4).
|
Egg masses
Most egg masses were associated with shallow-water (<6 m) seagrass beds, with some attached to the holdfasts of deeper macroalgal species, and others either dislodged or partly buried in sandy sediment. The seagrass Amphibolis antarctica is the dominant shallow-water seagrass on the east coast of Tasmania, extending from Great Oyster Bay to the Mercury Passage (Edgar, 2000), and the dominant substratum for egg deposition on the east coast. In the southeast, where A. antarctica is largely absent, calamary eggs were attached to Caulerpa sp. and Heterozosterous sp. In deeper water (>6 m), eggs were deposited on the macroalgal species Macrocystis pyrifera, Sargassum sp., and Ecklonia radiata.
Egg masses were seen mainly during spring and summer, in both regions. There were no egg masses during autumn, and those found during winter were at a single site in the southeast (Figure 5). Spawning intensity, assessed using egg mass density, differed significantly between the two regions (F = 4.59, d.f. = 1, 102, p = 0.035), but not among the six months (September–February) that egg masses were seen (F = 0.77, d.f. = 5, 102, p = 0.577). The density of egg masses was three times greater on the east coast (mean = 0.33 20 m–2, s.e. 0.05, n = 84) than on the southeast coast (mean = 0.12 20 m–2, s.e. 0.09, n = 30). The failure to detect a difference among the months was due to the extreme variability in the density of egg masses among sites. Coefficients of variation were large, variances in egg-mass density in the regions for each month ranging from 20 to 348% of the mean.
|
Sizes of egg masses ranged from 3 to 1241 strands on the east coast, and from 10 to 619 strands on the southeast coast. However, the size frequency distribution of the egg masses was not significantly different between the regions (
2 = 6.69, d.f. = 4, p = 0.153). Most egg masses in both regions had 200–299 strands.
Approximately 10% of the embryos examined from both regions and across the spring/summer period (September–February) were dead. There was no difference in the proportion of dead embryos between the two regions (2 = 0.01, d.f. = 1, p = 0.982).
Abundance
Catch rate (cpue) as a measure of abundance suggested that calamary biomass was greater on the east coast in all seasons than on the southeast coast (Figure 6). The southeastern biomass was relatively similar in all seasons, although as for the east coast, abundance was greatest in spring (Figure 6).
|
|
Discussion |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Inshore populations of southern calamary in both regions displayed a similar seasonal trend of largest gonosomatic index, reproductive output, and body size, and greatest abundance during spring, and lowest in autumn. High densities of egg masses typically observed on the east coast during spring/summer were largely absent from the southeast coast. The biological significance of this is unclear, particularly in terms of understanding the mechanisms that drive the development of spawning aggregations. Spatial and seasonal patterns in the number of egg masses inshore are critically relevant to the reproductive ecology of southern calamary, because spatial and seasonal aggregative spawning patterns appear to be linked with very specific use of inshore sites at certain times of the year.
There was no evidence of any spatial difference in the reproductive biology of southern calamary that may explain differences in the number of egg masses deposited. During spring, animals on the southeast coast were reproductively mature and had a high GSI, suggestive of egg deposition. Instead, it appeared that southeast populations did not form large spawning aggregations, and that spawning activity was sporadic in space and time, resulting in isolated and few egg masses deposited over a wider area. Consequently, management concerns about fishers specifically targeting spawning aggregations along the southeast coast may be unfounded.
Spatial patchiness over a large geographic region is seen in the Patagonian long-finned squid (Loligo gahi; Arkhipkin et al., 2000). The South African chokka squid also spawns sporadically throughout the year over a large geographic range (Sauer et al., 1992). Movement of southern calamary between the regions is currently unknown, and it is possible that reproductively mature squid may migrate northwards along the coast. Given that populations of southern calamary off mainland Australia (South Australia) and Tasmania are genetically similar (Triantafillos and Adams, 2001), it is probable that populations around Tasmania are also genetically similar, owing to mixing of squid between the regions.
A lower cpue on the southeast coast suggests that populations of squid are smaller, and therefore if they aggregate, such aggregations may be smaller and less obvious to fishers than those on the east coast. Unfortunately, biomass estimated from cpue may be regionally biased, especially if squid are forming large spatially predictable aggregations in the east. Moreover, biomass estimated from the cpue of species with aggregating behaviour and where those aggregations are targeted by fishers is problematic, because depletion of the aggregation may be masked by new individuals moving into the area (Carvalho and Nigmatullin, 1998; Lipiski et al., 1998).
Spawning aggregations on the east coast are mainly on beds of the seagrass Amphibolis antarctica (Moltschaniwskyj and Pecl, 2003), whose southern limit does not extend into the sites surveyed in the southeast. Egg masses along the southeast coast were attached to the seagrass Heterozostera tasmanica, and the macroalgae Macrocystis pyrifera, Ecklonia sp., and Sargassum sp. (Steer, 2004). Therefore, differences in spawning behaviour between the regions explored may be due to the use and availability of suitable or preferred habitat for egg deposition.
A consequence of non-aggregating spawning by southern calamary populations on the southeast coast would be to spread the risk of mortality through space. This will ensure that widely dispersed offspring will experience different environmental conditions, and as a result, the genetic make-up of the population would reflect the diversity of spatial selection pressures (O'Dor, 1998). A further consequence of non-aggregating behaviour during spawning on the southeast coast is that such behaviour effectively provides a refuge from fishing. The spatial and temporal predictability of spawning aggregations on the east coast allows effective and efficient targeting of squid by fishers. This has led to predictions that the fishery is overexploited (Moltschaniwskyj et al., 2003). In contrast, dipnets and spears used at night by fishers along the southeast coast (Moltschaniwskyj et al., 2003) target feeding, not spawning, calamary.
Based on age estimates for southern calamary in Tasmania, animals live for <12 months, and the adults in the spring/summer spawning aggregations are the product of spawning activity in the previous autumn/early winter (Pecl, 2000; Jackson and Pecl, 2003). However, in inshore areas of the east and southeast coasts of Tasmania, densities of egg masses were high only during spring/early summer. Many females caught during autumn and winter were reproductively mature, and although GSI was low in autumn, it increased during winter in both regions. Both of these parameters suggest the capability of some squid in the population to reproduce. However, given that there were few egg masses inshore, particularly during winter, it seems that aggregating spawning was not occurring at the same sites as during the spring/summer spawning period.
Although there was an ephemeral use of spawning areas among seasons, this was not due to the loss of habitat, because the seagrass Amphibolis does not suffer die-back in winter. It is not clear what cues female squid use to deposit egg strands in certain areas, or why the inshore habitats were used for egg deposition only during spring. Visual cues for egg deposition are important for many cephalopods (Hanlon and Messenger, 1996), females often adding eggs to an existing egg mass. The presence of eggs is an important cue for female oviposition (Arnold, 1962), and an increase in agonistic behaviour of males (King et al., 1999). Therefore, it is possible that the population structure and biomass of squid on the spawning beds are influenced by the quantity of eggs present.
Clearly, during spring, squid populations in different regions display different spawning behavioural strategies (aggregating vs. non-aggregating). Moreover, reproductively mature squid in winter show different spawning behaviour, and do not make use of the spring spawning grounds. Southern calamary, like the chokka squid (Lipiski, 1998), may use both spatial and temporal strategies to ensure genetic diversity in spawning aggregations. Therefore, plasticity in spawning behaviour by southern calamary may ensure the successful persistence of young of the year, by exposing juveniles to mortality risk across space and time.
|
Acknowledgements |
|---|
This research was supported through funding from the Fishery Research Development Corporation awarded to NAM (FRDC 2000/121). We thank the fishers who allowed us to measure their catch, the technical staff at the Marine Research Laboratories for assistance in the field, Gretta Pecl for constructive comments and discussion during writing of the manuscript, and Marek Lipi
ski and an anonymous referee for valuable critiques of an early draft. |
References |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
-
Arkhipkin A., Laptikhovsky V.V., Middleton D.A.J. (2000) Adaptations for cold water spawning in loliginid squid: Loligo gahi in Falkland waters. Journal of Molluscan Studies 66:551–564.
Arnold J. M. (1962) Mating behavior and social structure in Loligo pealii. Biological Bulletin, Marine Biological Laboratory, Woods Hole 123:53–57.
Boyle P.R., Pierce G.J., Hastie L.C. (1995) Flexible reproductive strategies in the squid Loligo forbesi. Marine Biology 121:501–508.[CrossRef]
Carvalho G.R. and Nigmatullin C.M. (1998) Stock structure analysis and species identification. In Rodhouse P., Dawe E., O'Dor R. (Eds.). Squid Recruitment Dynamics: The Genus Illex as a Model, the Commercial Illex Species and Influences on Variability(FAO, Rome) pp. 105–124.
Edgar G.J. (2000) Australian Marine Life: the Plants and Animals of Temperate Waters(Reed Books, Australia) 544 pp.
Hanlon R.T. and Messenger J.B. (1996) Cephalopod Behaviour(Cambridge University Press, Cambridge) 232 pp.
Jackson D.A. and Pecl G.T. (2003) The dynamics of the summer-spawning population of the loliginid squid (Sepioteuthis australis) in Tasmania, Australia – a conveyor belt of recruits. ICES Journal of Marine Science 60:290–296.
King A. J., Adamo S. A., Hanlon R. T. (1999) Contact with squid egg capsules increases agonistic behavior in male squid (Loligo pealei). Biological Bulletin, Marine Biological Laboratory, Woods Hole 197: p. 256.
Lipiski M.R. (1998) Cephalopod life cycles: patterns and exceptions. South African Journal of Marine Science 20:439–447.
Lipiski M.R., Durholtz M.D., Underhill L.G. (1998) Field validation of age readings from the statoliths of chokka squid (Loligo vulgaris reynaudii d'Orbigny, 1845) and an assessment of the associated errors. ICES Journal of Marine Science 55:240–257.
Moltschaniwskyj N.A. (1995) Multiple spawning in the tropical squid Photololigo sp: what is the cost in somatic growth? Marine Biology 124:127–135.[CrossRef]
Moltschaniwskyj N.A. and Pecl G.T. (2003) Small-scale spatial and temporal patterns of egg production by the temperate loliginid squid Sepioteuthis australis. Marine Biology 142:509–516.
Moltschaniwskyj N., Pecl G., Lyle J. (2002) An assessment of the use of short-term closures to protect spawning southern calamary aggregations from fishing pressure in Tasmania, Australia. Bulletin of Marine Science 71:501–514.[Web of Science]
Moltschaniwskyj N., Pecl G., Lyle J., Haddon M., Steer M. (2003) Population dynamics and reproductive ecology of the southern calamary (Sepioteuthis australis) in Tasmania. FRDC Final Report 2000/121. 158 pp.
O'Dor R.K. (1998) Can understanding squid life-history strategies and recruitment improve management? South African Journal of Marine Science 20:193–206.
Pecl G. (2000) Comparative life history of tropical and temperate Sepioteuthis squid in Australian waters. Ph.D. thesis, James Cook University, Townsville.
Pecl G. (2001) Flexible reproductive strategies in tropical and temperate Sepioteuthis squids. Marine Biology 138:93–101.[CrossRef]
Sauer W.H.H., Smale M.J., Lipiski M.R. (1992) The location of spawning grounds, spawning and schooling behaviour of the squid Loligo vulgaris reynaudii (Cephalopoda: Myopsida) off the Eastern Cape coast, South Africa. Marine Biology 114:97–107.
Steer M. (2004) Embryonic development and early life history of the southern calamary, Sepioteuthis australis (Quoy & Gammaird, 1832). Ph.D. thesis, University of Tasmania, Launceston.
Steer M.A., Moltschaniwskyj N.A., Jordan A.R. (2003) Embryonic development of southern calamary (Sepioteuthis australis) within the constraints of an aggregated egg mass. Marine and Freshwater Research 54:217–226.[CrossRef][Web of Science]
Triantafillos L. and Adams M. (2001) Allozyme analysis reveals a complex population structure in the southern calamary Sepioteuthis australis from Australia and New Zealand. Marine Ecology Progress Series 212:193–209.[Web of Science]
Warner R.R. and Chesson P.L. (1985) Coexistence mediated by recruitment fluctuations: a field guide to the storage effect. American Naturalist 125:769–787.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||