These aquaculture- and conservation-oriented commentaries are not abstracts written by the original authors.  They reflect the opinions of someone else -- usually Roger Doyle.  Direct quotations from the papers or abstracts are marked with inverted commas.

426.  Trade-offs in adaptation to toxic stress
         Heritable adaptation and fitness costs in killifish (Fundulus heteroclitus) inhabiting a polluted estuary. 2003. Meyer, J. N. and R. T. Di Giulio. Ecological Applications 13:490-503.
         The authors collected killifish (also known as  mummichogs) from the highly polluted Elizabeth river in Virginia and reared the F1 and F2 generations in the laboratory. Relative to control fish from an uncontaminated site, both generations of Elizabeth River fish were more resistant to the toxicity of Elizabeth River sediment. They were also, unfortunately, less resistant to other stressors including photo-enhanced toxicity and hypoxia, "suggesting that the changes that have conferred resistance to the toxicity of the Elizabeth River sediments carry a cost of reduced fitness in other contexts". See March 2003 #398 for a beautiful study of how the MHC gene complex in this species has evolved to cope with the unusual parasites which are found in a toxic environment. joel.meyer@duke.edu

425.  Heritability can be estimated using microsatellite relatedness, but not very well
         Marker-assisted estimation of quantitative genetic parameters in rainbow trout, Oncorhynchus mykiss. 2003. Wilson, A. J., G. McDonald, H. K. Moghadam, C. M. Herbinger and M. M. Ferguson. Genetical Research 81:145-156.
         Several papers reviewed here have used microsatellite markers to estimate the genetic "relatedness" between individuals in wild or experimental populations, and then proceeded to estimate quantitative genetic parameters, e.g. heritability of growth, using the relatedness information. This approach has advantages in aquaculture, genetic conservation and other situations where pedigrees are impossible to obtain. The authors of this paper have found, unfortunately, that statistical power and precision are not among those advantages.
         They estimated heritabilities of size and spawning traits in three ways (1) conventionally, by using accurate pedigree records constructed from microsatellite data on offspring and parents (Danzmann's PROBMAX program), (2) by using microsatellites to infer full-sib relationships without parental data (Oct 2001 #244) and (3) by using a microsatellite relatedness estimator (see Sept 2001 #227), which acts as the genetic relatedness variable required for the analysis (a regression) without specifying explicitly whether the relatedness is due to full-sib, parent-offspring, half-sib or other relationships in the pedigree. It is the third type of estimator which is easiest to apply and is most enticing for use in poorly controlled populations, such as the start-up of an aquaculture broodstock (Jan 2002 #283).
         As it turns out, the relatedness estimator (Ritland's, Dec 2000 #142) didn't perform very well in the Type 3 quantitative genetic analysis, mainly because of inaccuracies and biases in the estimator itself. The full-sib pedigrees inferred from markers (Type 2 analysis) did much better, but still not as well as the control (Type 1) because of frequent ambiguities in deducing full-sib relationships when half-sibs are also present.  This of course is bad news. The authors do note, however, that the Type 2 and 3 analyses provide genetic information which is qualitatively valid and unlikely to be obtainable in any other way. In fact, as the authors say, "... in this case reconstructing pedigree under the assumption that only full-sib family structure was present resulted in estimates of quantitative genetic parameters very similar to those obtained when the true complex pedigree was known."
         One should note, as well, that there are many types of relatedness estimator in addition to Ritland's estimator. Furthermore, algorithms are being developed which can use markers to infer explicit full-sib relationships nested within half-sibships (e.g. Thomas & Hill 2002. Genet. Res. 79:227-234). awilso00@uoguelph.ca  

424. Low genetic correlation between successive growth periods in vannamei
         Quantitative genetic parameter estimates for size and growth rate traits in Pacific white shrimp, Penaeus vannamei (Boone 1931) when reared indoors. 2003. Pérez-Rostro, C. I. and A. M. Ibarra. Aquaculture Research 34:1-11.
         The rare and useful feature of this experiment is that the growth of individually tagged individuals was measured over successive intervals so that growth and size effects could be separated in the analysis. The experiment generated full-sib families only, and thus the genetic correlations and heritabilities are larger than if half-sibs had been used. Although heritabilities were reasonably high the estimated serial genetic correlation was negative for linear dimensions and only 0.35 for weight, giving little support to the idea that one can save time and space in a vannamei broodstock improvement program by selecting rapidly growing animals early.
         An unusual feature of the analysis was the use of 12-week group weight and density as covariates to statistically remove common environment effects. One cannot help wondering whether serial genetic covariances might also have been reduced by this conditioning of the data (removal of common-cause correlation). aibarra@cibnor.mx   

423. Big, dominant, male cod reduce effective population size
         Male reproductive competition in spawning aggregations of cod (Gadus morhua, L.). 2002. Bekkevold, D., M. M. Hansen and V. Loeschcke. Molecular Ecology 11:91-102.
         Groups of cod were spawned experimentally and the parentage of the offspring sorted out using microsatellite markers. "Our results show that multiple males contributed sperm to most spawnings but that paternity frequencies were highly skewed among males, with larger males on average siring higher proportions of offspring. It was further indicated that male reproductive success was dependent on the magnitude of the size difference between a female and a male." The reproductive success was highly non-normal (non-Poisson), which would reduce the effective population size of this "broadcast spawner" well below the census size. Note Jan 2003 #385, in which another broadcast spawner, the New Zealand red snapper, had a ratio of effective to population sizes of around 0.00001. dorte.bekkevold@biology.au.dk   

422. Microsatellites measure inbreeding depression, but not very well
         Microsatellite measures of inbreeding: a meta-analysis. 2003. Coltman, D. W. and J. Slate. Evolution 57:971-983.
         The two hypotheses about natural populations being tested simultaneously here are, (1) individual variation in life history and morphological traits is associated with variation in inbreeding and (2) microsatellite data can be used as a surrogate for pedigree records to estimate the inbreeding of individuals within populations.
         Published and unpublished studies on many taxa were pulled together in this meta analysis. The Pearson correlation coefficient "r" was used to measured the size of the inbreeding/phenotype relationship. Two measures of genetic diversity were used, ordinary multi-locus heterozygosity and d2 (the mean of the squared difference in repeat units between the two alleles at each locus).
         Results are interesting. The association between trait variation and inbreeding inferred from microsatellite variation  is not very large but is larger for life history traits than for morphological traits. This is consistent with the idea that much of the variation in traits related to fitness involves dominance and other non-additive interactions between genes (see Feb 2000 #10).  Ordinary multilocus heterozygosity usually showed stronger association  than d2, suggesting that it may be a better indicator of (short term) inbreeding. See May 2002 #312 for a theoretical paper which reaches similar conclusions, and #316 for another data-based study. The authors of the present paper also found a strong "publication bias" in favor of significant results which leads to an overestimate of  inbreeding effects in the literature.  d.coltman@sheffield.ac.uk   

421. More evidence that Baltic salmon populations can resist disease
         Comparative susceptibility of two races of Salmo salar (Baltic Lule river and Atlantic Conon river strains) to infection with Gyrodactylus salaris. 2003. Dalgaard, M. B., C. V. Nielsen and K. Buchmann. Diseases of Aquatic Organisms 53:173-176.
         The extreme susceptibility of Norwegian and Scottish commercial salmon stocks to the Gyrodactylus skin parasite lead to major losses of aquaculture production. Two salmon strains in the Indals river (Sweden) and Neva river (Russia) are already well known to be relatively much more resistant than the eastern Atlantic stocks. This paper reports on a comparative study of another Baltic stock from the Lule river in Sweden, which also turns out to be less susceptible than Atlantic coast salmon to colonization by Gyrodactylus. mbd@kvl.dk 

 420. Genetic background modulates the effect of inbreeding on parasite load
         Test of synergistic interaction between infection and inbreeding in Daphnia magna. 2003. Haag, C. R., O. Sakwisk and D. Ebert. Evolution 57:777-783.
         Inbreeding increases parasite load. Right? Many papers cited here have suggested this, e.g. Feb 2000 #9, Nov 2000 #137, Apr 2001 #184, Aug 2002 #340, May 2003 #409. It turns out, though, that genetic background may be sometimes be more important than inbreeding per se.
         The cladoceran Daphnia magna is in many ways a good model organism for genetic studies because of its rapid generation time and ability to propagate either clonally or sexually (both outcrossing and selfing) for many successive generations. These traits were used here to look at the effect of inbreeding on resistance to two species of microsporidean parasite. The test was powerful and direct: the outcome of clonal completion between the descendents of selfed and outcrossed siblings. As expected, in control experiments where the parasites were not present, selfed clones almost always lost out to sibling outbred clones.
         What is very surprising is that this was not true in the presence of the parasites! Sometimes the selfed clones did better. Furthermore, the effect was not always the same for the two parasite species. So resistance to the parasites was quite specific, presumably depending on which particular genes were sampled when the clones were produced. Susceptibility to particular parasites was is not the result of general inbreeding depression.
         "This indicates that, contrary to the hypothesis that parasites generally lead to a decreased performance of inbred genotypes, their effect may depend on the genetic background of the host as well as on the parasite species, and suggests that inbreeding can lead to reduced or increased resistance to parasites." See April 2001 #180 for a related paper which showed genetic variation in parasite resistance in natural populations of Daphnia, and May 2003 #402 for similar variation among inbred lines of Drosophila. christoph.haag@unifr.ch   

419. A crucial conservation parameter can be estimated
         The power of experiments for estimating relative reproductive success of hatchery-born spawners. 2003. Hinrichsen, R. A. Canadian Journal of Fisheries and Aquatic Sciences 60:864-872.
         The genetic effect of  stock supplementation depends on how successful the hatchery fish are at breeding in the wild, relative to wild fish. This paper tries to establish how hard it is to estimate relative success, in the real world.
         The author simulates a salmon population which conforms to the familiar Ricker stock-recruitment relationship. The wild component of the population is supplemented to varying extents with hatchery-bred fish, and the question asked is how much work is required to prove that the reproductive success of hatchery females is different than wild ones. Or more precisely, what kind of experiment will maximize "the probability of rejecting the null hypothesis that the reproductive successes of hatchery- and wild-born spawners are equal when they are not. This probability is defined as power".
         Variables included in the simulation included "the length of experiment, schedule of hatchery-born spawner introduction, stock productivity, spawner abundance, interannual variance in production, fraction of spawners and recruits sampled, and true values of [relative reproductive success] ." Likelihood functions were derived for analyzing the simulated data.
         Results are encouraging. "The power analysis suggests that reliably estimating reproductive success of hatchery-born spawners relative to their wild-born counterparts is possible with just a few brood years of spawner–recruit data, and that choosing a population that has large spawner numbers and high productivity may decrease the number of brood years required."
         The author adds an important caveat about experiments in the highly-politicized world of stock supplementation, aquaculture escapes etc. "In all experiments of which I am aware for salmon populations. (including the experiment analyzed in this paper), relative reproductive success of hatchery-born spawners is not measured against reproductive success of wild-born spawners without supplementation, which, I believe, is the appropriate baseline. This is the baseline necessary to understand how a population would perform in the absence of supplementation."
        See #416 below, Nov 2000 #130 for an experimental study of relative reproductive success, and Feb 2002 #287 for inferences based on microsatellite time series in stocked rivers. hinrich@seanet.com  

418. DNA test for local ancestry in a re-introduction program
         . 2003. Hofkin, B. V., A. Wright, J. Altenbach, K. Rassmann, H. M. Snell, R. D. Miller, A. C. Stone et al. Conservation Genetics 4:105-108.
         This is an interesting use of museum DNA samples to solve a practical conservation problem. More than 150 years after Darwin's visit the Galapagos islands are still a hotbed of evolution and evolutionary research. One of the islands, Baltra, lost its population of land iguanas in the late 1940s, and there is an intention to start the population up again. But where should the replacement stock come from? Another island which was stocked from Baltra decades ago would seem a logical source, but there are suspicions that this island may also have been stocked from other sources. (Apparently it was decided that any increase in evolutionary potential which might be achieved by using mixed stock for a re-introduction would not be a Good Thing relative to the potential loss of local adaptation.)
         So, using DNA from museum specimens collected from several Galapagos islands way back in 1905, the authors were able to identify animals with cytochrome B haplotypes which are "unambiguous Baltra origin ....These results provide scientific criteria for the ecological restoration of these endangered reptiles." The unambiguous Baltra origin is presumably through the female line only. snell@unm.edu  

417. The trout marker map gets bigger and more dense
        A consolidated linkage map for rainbow trout (Oncorhynchus mykiss). 2003. Nichols, K. M., W. P. Young, R. G. Danzmann, B. D. Robison, C. Rexroad, M. Noakes, R. B. Phillips et al. Animal Genetics 34:102.
         The authors report the updating of their linkage map, which represents the state-of-the-art for map for  species of aquacultural importance.  The markers now total "1359 genetic markers and the sex phenotype including 799 EcoRI AFLPs, 174 PstI AFLPs, 226 microsatellites, 72 VNTR, 38 SINE markers, 29 known genes, 12 minisatellites, five RAPDs, and four allozymes". The markers fall into 30 linkage groups.
         It is interesting that AFLP markers may be growing more useful for this sort of work than microsatellites even though they are dominant markers (see also March 2002 #310, May 2003 #404). thorglab@wsu.edu   

416. Sea-ranched trout may be swamping a wild population
         Lack of molecular genetic divergence between sea-ranched and wild sea trout (Salmo trutta). 2003. Palm, S., J. Dannewitz, T. Järvi, E. Petersson, T. Prestegaard and N. Ryman. Molecular Ecology 12:2057-2071.
         This interesting microsatellite and allozyme study of a supportive breeding program in the Dalälven river (Sweden) demonstrates that the hatchery has a profound genetic impact on the wild stock -- one-way gene flow may be around 80% per generation!
         A key feature of the supplementation program is that only fish which are produced in the hatchery and succeed in returning to the river are used as spawners in the hatchery. Thus domestication selection for reproducing in the hatchery operates in tandem with selection for survival in the wild. Although some returning hatchery fish also spawn naturally in the river no wild genes have been introduced into the hatchery stock (which was founded with Dalälven fish) since the late 1960s. It has been a completely closed stock with a sea-ranching selection component, for about seven generations.
         Other studies have demonstrated genetic and even morphological differences between wild and hatchery Dalälven trout. This study, on the other hand shows, that the difference between stocks in a single year is no greater than the year-to-year variation within the two stocks; i.e. there is no persistent stock difference. The authors infer from their analyses that hatchery and wild fish have about the same reproductive success in the wild (see #419 above). This is important and would appear to be very good news from a genetic conservation perspective since it implies that the hatchery should not be reducing the fitness of the wild component of the population, despite the fact that natural reproduction is not selected during the sea-ranching phase (at least, not directly; selection for wild survival may counter selection for hatchery reproduction).
         The authors point out, however, that after 30 years of intensive supportive breeding with this high level of introgression, the "wild" Dalälven fish can hardly be the same as they were when the program started. The same point is made in #419 above. stefan.palm@popgen.su.se  

415. Adaptation to acid environment is partly through the mother
         Geographic variation in acid stress tolerance of the moor frog, Rana arvalis. I. Local adaptation. II. Adaptive maternal effects. 2003. Räsänen, K., A. Laurila and J. Merilä. Evolution 57:352-362.
         The authors compared early life history traits of frogs collected from highly acidified environments with frogs from well buffered, neutral environments, and also looked at the reciprocal crosses between the two types of population. Sure enough, when tested in an acid environment, the offspring of animals from acid environments had superior survival and  hatchling size and weight. An interesting finding in this study (also see Part II, ibid 363–371) is a strong maternal effect on survival, through the structure and composition of the gelatinous egg capsule. Unfortunately, the experimental design could not separate environmental from heritable maternal effects.
         Another example of a maternal effect which is important in aquaculture and which may involve a heritable maternal effect is the correlation between egg size and offspring growth. That particular maternal effect, egg size, should be taken into account in selection programs where growth is the objective, especially if there is also a tradeoff between female growth and female investment in egg size. katja.rasanen@ebc.uu.se  

414. White shrimp genetically uniform over its range
         Population genetic analysis of white shrimp, Litopenaeus setiferus, using microsatellite genetic markers. 2003. Ball, A. O. and R. W. Chapman. Molecular Ecology 12:2319-2330.
         The white shrimp is endemic to the east coast of North America, from the Carolinas into the Gulf of Mexico.  This microsatellite study (6 loci) found that the shrimp in the Atlantic were slightly but significantly different from those in the Gulf. Some weak local differentiation within these broad groupings was also observed but the authors conclude that this is largely "short-term sampling effects rather than persistent genetic differences". Deviations from Hardy Weinberg distributions were observed but attributed to null alleles, not gametic disequilibrium due, e.g., to population fluctuations.
         On the whole, there is considerable genetic homogeneity over the whole range, probably because of the pelagic spawning and planktonic dispersal phases of the life cycle. "The major findings of this study were the consistency of allele frequencies over time, sporadic differentiation among neighbouring locations indicative of the effects of high reproductive variance, and slight but significant differentiation between the Gulf of Mexico and the Atlantic."
         Litopenaeus setiferous appears to shows less regional differentiation than Penaeus monodon or P. stylirostris. The authors estimated long-term effective population sizes in the millions, but the sporadic, local fluctuations are intriguing and may have adaptive and evolutionary implications (e.g. Nov 2001 #259). balla@mrd.dnr.state.sc.us  

413.  A useful shrimp population marker
         Application of mitochondrial control region in population genetic studies of the shrimp Penaeus. 2003. Chu, K. H., C. P. Li, Y. K. Tam and S. Lavery. Molecular Ecology Notes 3:120-122.
         "This study reports a primer set for amplifying a partial fragment of about 610 bp in the fast mutating mitochondrial control region in shrimps of the genus Penaeus [merguiensis] (Decapoda: Penaeidae). ... The results indicate that the mitochondrial control region provides more informative sites and reveals more haplotypes [than other mt sequence and RFLPs] , making it most useful for evaluating genetic variations within and between populations of Penaeus species." kahouchu@cuhk.edu.hk   

412. Identify any species with one primer?
         Novel universal primers establish identity of an enormous number of animal species for forensic application. 2003. Verma, S. K. and L. Singh. Molecular Ecology Notes 3:28.
         "This study describes a polymerase chain reaction (PCR)-based approach, which .... is able to reveal whether the source of the sample is human or animal, and, if animal, which of the 221 animal species included in the study, simply by using one set of novel primers... The primers described in this study universally amplify a specific segment of mitochondrial cytochrome b sequence from a sample of unknown origin and delineate its identity to the level of family, genus and species...." lalji@ccmb.res.in