Selection on larval and adult body size in a marine fish: potential evolutionary responses and effects on population dynamics Public Deposited


Attribute NameValues
Abstract or Summary
  • Many species have complex life cycles in which a dispersive larval stage is followed by a relatively sedentary adult stage. For such species, reproductive output is often high and large variation in survivorship throughout early life-history phases (eggs and larvae) can lead to dramatic fluctuations in recruitment which may in turn drive variation in the abundance of juveniles and adults. Early in the life cycle may therefore be a critical period for both natural selection and population dynamics. On one hand, variability in survival during early stages may provide ample opportunity for selection on early life-history traits. On the other hand, phenotypic variation in early life-history traits and selective mortality may be an important source of variability in population dynamics. Variation in survival of marine fish larvae may be a major driver of variability in benthic population size. However, little is known about how variation in larval phenotype may affect larval survival, and less in known about the evolutionary potential of marine fish larvae. I quantified both environmental and genetic sources of variation in larval traits for a field population of a common Atlantic and Caribbean coral-reef fish, the bicolor damselfish (Pomacentridae: Stegastes partitus). I combined field demographic studies and manipulative experiments in the Bahamas to estimate heritability and quantitative genetic parameters for both larval size and swimming performance – two traits that are associated with early survival. I also compiled published estimates of viability selection on larval size from eight species of fish to estimate the average magnitude of selection on this trait. The initial results of these analyses were somewhat paradoxical. Despite ample heritability (h2 = 0.29 for larval size), and strong selection on larval size (mean selection differential = 0.484), the observed mean larval size is quite far from the estimated phenotypic optimum (0.481 standard deviations greater than current mean size), suggesting that marine fish larvae are on average, maladapted with respect to survival during the larval and juvenile phases. Further analyses focused on potential evolutionary constraints on larval size. First, I estimated trade-offs in individual reproductive output between larval quality and quantity. Mothers that produced larger larvae with greater swimming abilities tended to produce fewer larvae, and these effects explained a large component of the mismatch between mean larval size and the phenotypic optimum for survival. Fluctuation in direct selection on larvae may also partially explain why mean larval size is less than optimal. Evolution of larval size may also be strongly influenced by genetic correlations with body size expressed at later ages. I demonstrated substantial additive genetic covariance between adult asymptotic size and both larval size-at-hatching and swimming performance (0.212 and 0.241 on variancestandardized traits, respectively). Adult asymptotic size was also linked to larval traits via size-dependent maternal effects, in which larger mothers provisioned offspring with more yolk resources. Selection on adult body size may therefore cause a substantial correlated genetic response in larval size that may strongly affect the overall evolutionary trajectory of larval traits. I also examined natural selection on body size and growth form in S. partitus. Using data on size, growth and longevity of individual fish studied at 4 sites over a 7-year period, I analyzed both ontogenetic and spatial variation in the magnitude and direction of viability selection on body size. Selection on asymptotic (adult) size was strong and positive at some sites, but weak and negative at other sites. Moreover, fish that were small as juveniles generally experienced greater survival, even if large adult size conferred survival benefits later in life. Both spatial and ontogenetic reversals in selection on body size would be expected to produce similar reversals in the direction of correlated responses of larvae, thereby altering the evolutionary response of larvae and potentially preventing larval size from evolving toward its optimum value. Although this research identified several potential constraints on the evolution of larval traits, there is still considerable scope for an evolutionary response to selection, especially if selection is consistent and strong. Many marine fishes are subject to size-selective fishing where larger, fast-growing individuals are selectively removed from the population. Such effects are usually strong because fishing mortality rates can greatly exceed natural mortality rates and fishing selectivity and intensity are often constant. Although correlated responses to selection have been hypothesized as potentially important consequences of fishery selection, estimates of quantitative genetic parameters required to predict correlated responses to such selection have been lacking. To my knowledge, my research provides the first estimates of quantitative genetic parameters for larval traits and their links to adult size in a wild population of fish. I used these data to predict how larval size would respond to selection on adults and how evolutionary shifts in larval size would in turn affect population replenishment. My results predict that observed rates of fishery selection on adult marine fishes may decrease average larval size by approximately 0.11 standard deviations after a single generation of selection. Such a reduction in larval size is predicted to reduce survivorship through the larval and early juvenile phases by about 16%. Because the dynamics of many fish populations are highly sensitive to changes in survival of early life stages, the evolution of a higher incidence of low-quality larvae in response to fishery selection may have substantial consequences for the viability of fished populations. Overall, this research indicates that a complex interplay among trait variation, phenotypic selection, and demographic rates may have strong effects on both evolutionary responses and population dynamics. Our understanding of such interactions will be substantially advanced by applying evolutionary quantitative genetics to traditional studies of demography and population dynamics. A combination of these two approaches can yield significant insight into basic evolutionary questions (e.g., why larvae are smaller than expected), as well as applied conservation problems (e.g., predicting correlated responses to fishery selection).
Resource Type
Date Available
Date Copyright
Date Issued
Degree Level
Degree Name
Degree Field
Degree Grantor
Commencement Year
Committee Member
Academic Affiliation
Non-Academic Affiliation
Rights Statement
Additional Information
  • description.provenance : Submitted by Darren Johnson ( on 2009-09-07T00:26:41Z No. of bitstreams: 1 JohnsonDarrenW2009.pdf: 1889081 bytes, checksum: 4d4c62a8c70cacc6b45cec531d1fd5e2 (MD5)
  • description.provenance : Approved for entry into archive by Linda Kathman( on 2009-09-16T23:09:09Z (GMT) No. of bitstreams: 1 JohnsonDarrenW2009.pdf: 1889081 bytes, checksum: 4d4c62a8c70cacc6b45cec531d1fd5e2 (MD5)
  • description.provenance : Made available in DSpace on 2009-09-16T23:09:09Z (GMT). No. of bitstreams: 1 JohnsonDarrenW2009.pdf: 1889081 bytes, checksum: 4d4c62a8c70cacc6b45cec531d1fd5e2 (MD5)
  • description.provenance : Approved for entry into archive by Julie Kurtz( on 2009-09-09T15:57:07Z (GMT) No. of bitstreams: 1 JohnsonDarrenW2009.pdf: 1889081 bytes, checksum: 4d4c62a8c70cacc6b45cec531d1fd5e2 (MD5)


In Administrative Set:
Last modified: 10/20/2017

Downloadable Content

Download PDF

EndNote | Zotero | Mendeley