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.

321.  A gene for salinity tolerance in tilapia:
         microsatellites are directly selected!
         Microsatellite variation associated with prolactin expression and growth of salt-challenged tilapia. 2002. Streelman, J.T, and T.D. Kocher. Physiological Genomics 9 (1):1-4.
         Tilapia geneticists and many population geneticists should be interested in this paper. The relative growth of individual O. mossambicus X niloticus hybrids in saline water is influenced by variation in the relative expression of two prolactin genes. (Prolactin is a peptide growth hormone which also affects osmoregulation.) The variability in expression is caused in part by variation in the length of a microsatellite repeat embedded in the control region of the prolactin genes.
         This is a very good story: not only is the microsatellite marker decidedly non-neutral, it may be under selection in an environment that has considerable importance in aquaculture and, perhaps, conservation (such as feral mossambicus taking over the seas). The effect of the microsatellite is apparently not caused by linkage, i.e. because the microsatellite is close to the prolactin gene on the chormosome (association overdominance). It is a functional effect. The authors speculate that "repeats of varying length can induce promoter conformations that differ in their ability to bind transcriptional regulators". Fish homozygous for the long microsatellite were only half as big, in half-strength seawater, as fish homozygous or heterozygous for the short microsatellite. There was strong genotype-environment interaction. "Each genotype grew best at a different salinity treatment".
         The authors conclude, "From an evolutionary perspective, our results run counter to the textbook interpretation that dinucleotide microsatellite variation lacks functional consequences. ... [microsatellites] provide a cellular means to alter the amount of gene product among individuals without changing amino acid sequence. The association of microsatellites with a long list of transcription factors and environmentally regulated genes suggests an under-appreciated role in the fine-scale modulation of gene expression." Streelman e-mail: jts3@hopper.unh.edu .

320.  A new relatedness estimator for use when pedigrees aren't available
         An estimator for pairwise relatedness using molecular markers. 2002. Wang, J. Genetics 160:1203-1215.
         When true pedigree data are lacking, microsatellites or other markers may possibly be used instead for practical purposes like estimating heritabilities in the wild, managing fish farms or increasing founder diversity in captive populations (Nov 2001 #261; Jan 2002 #283). This new estimator can be used to estimate inbreeding of individuals as well as the relatedness of pairs of individuals. It is unbiased and appears to offer advantages over previous estimators in its relative insensitivity to problem data. "The new estimator is also robust for small sample sizes and for unknown relatives being included in samples for estimating allele frequencies. Compared to previous estimators, the new one is generally advantageous, especially for highly polymorphic loci and/or small sample sizes." For a recent review of other relatedness estimators see Sep 2001 #227. jinliang.wang@ioz.ac.uk 

319.  Genetic biomarkers for disease and toxic stress in oysters
         Potential indicators of stress response identified by expressed sequence tag analysis of hemocytes and embryos from the American oyster, Crassostrea virginica. 2002. Jenny, M.J., A.H. Ringwood, E.R. Lacy, A.J. Lewitus, J.W. Kempton, P.S. Gross, G.W. Warr, and R.W. Chapman. Marine Biotechnology 4:81-93.
         C. virginica is an important commercial crop and is also a useful indicator of environmental quality in the coastal waters of eastern North America. The animal is susceptible to a variety of water-borne diseases, pollutants and toxic chemicals in low concentrations. This paper describes a search for oyster genes that are expressed -- turned on and manufacturing their products -- in the presence of these harmful agents. (See Sep 2001 #233 for a bit more information about expressed sequence tags.) The idea is to find biomarkers for environmental stress and, more generally, to learn more about what molluscs pretend to be their immune systems. "Several potential biomarkers identified include an antimicrobial peptide [see May 2000 #55], recognition molecules (lectin receptors), proteinases and proteinase inhibitors, and a novel metallothionein."
         The authors conclude that " ... the observation that most of the messages (ca. 70%) encode proteins that are either completely novel or are related to genes of unknown function may be the most interesting finding. If taken at face value, it suggests that a large proportion of metabolic energy is utilized, in oyster embryos and hemocytes, to generate proteins whose functions are unknown." chapmanr@mrd.dnr.state.sc.us 

318.  Fluctuating populations may be larger than we thought
         The effective size of fluctuating populations. 2001. Iizuka, Masaru. Theoretical Population Biology 59:281-286.
         The rates at which a population accumulates inbreeding and loses genetic diversity by random drift are often expressed in terms of the "effective number" of the population. The effective number, Ne, is a theoretical number which is usually smaller than the actual number of animals. It is smaller mainly because of unequal sex ratio, the failure of many animals to breed, non-random mating, and the exaggerated success of a few breeders. These problems lower the effective number in hatcheries and in nature.
         When population size fluctuates from generation to generation the effective population number over the long term is a lot smaller than the ordinary average value. It is usually considered to be the harmonic mean (the reciprocal of the average of the reciprocals) of the individual values. Therefore, when a population occasionally crashes and recovers, the long-term effective population number (the harmonic mean) is actually much closer to the smallest size reached during the crash. Or at least that is what people think. This leads to concern about the genetic health of endangered natural populations and aquacultural populations, which are often reduced to a few breeders whenever a new broodstock is started.
         This theoretical paper by Iizuka turns the usual thinking upside down. He demonstrates (in his theorem #5) that the long-term effective number for inbreeding and drift is actually very close to the largest value reached by a fluctuating population if the sequence of values is highly, and positively, autocorrelated. (Population size is autocorrelated when a "good" year tends to be followed by another relatively good year, a poor year is likely to be followed by another poor year etc.)
         How important is this new insight concerning effective population number? At the very least, it would be interesting to try out the effect of autocorrelation using a sequence of real data. Population sizes of fish generally are positively autocorrelated. Even if they aren't, overlapping generations should induce positive autocorrelation. No e-mail available. The address is: Division of Mathematics, Kyushu Dental College, 2-6-1 Manazuru, Kokurakita-ku, Kitakyushu 803-8580, Japan.

317.  Alien invaders mate with locals in Wisconsin
         Implications of hybridization between introduced and resident Orconectes crayfishes. 2001. Perry, W.L., J.L. Feder, and D.M. Lodge. Conservation Biology 15 (6):1656-1666.
         There are several hundred species of crayfish in North America (80% of the world total) and almost one-third of these species are listed as threatened or endangered. One of the threats to species survival is introduction of non-native species, which not only homogenize species distributions across the continent, i.e. increases within-site diversity while reducing between-site diversity, but can cause local extinction as well. (For an analogous effect on the diversity of N.A. fish see Aug 2000 #95). The authors found that Orconectes rusticus, native to northern Ohio but introduced widely elsewhere in the U.S., is displacing O. virilis and O. propinquus in northern Wisconsin. This is a purely ecological displacement as far as O. virilis is concerned; The authors detected no hybrids of this species.
         The problem facing O. propinquus is more genetic. "Over 6% of the crayfish at one sympatric site were putative F1 hybrids [rusticus X propinquus], 4% were putative F2 individuals ( hybrid × hybrid origin), and 13% were putative backcrosses ( product of hybrid × parental matings). ... Our results suggest that genetic mechanisms play a role in the extirpation of O. propinquus by O. rusticus and are consistent with observations of other researchers suggesting that hybridization with non-native species is common among crayfishes at many other locations. High rates of endemism and widespread introductions of non-native crayfish suggest that invasions and hybridization are a major threat to crayfish biodiversity." wlperry@ilstu.edu 

316.  Markers may be a poor substitute for pedigrees
         The use of marker-based relationship information to estimate the heritability of body weight in a natural population: a cautionary tale. 2002. Thomas, S.C., D.W. Coltman, and J.M. Pemberton. Journal of Evolutionary Biology 15 (1):92.
         It may sometimes be possible to use genetic relationships inferred from microsatellite data as a substitute for pedigree records. Many papers mentioned on this website have explored the idea, including an earlier paper by these same authors (Aug 2000 #88). Apparently it doesn't work very well with the feral Soay sheep of St. Kilda, for which a good pedigree record exists. Several methods of estimating heritabilities from microsatellite-inferred relationships gave unreliable estimates of body weight heritability.
         Even if this turns out to be a general problem with variance component estimation it need not obviate the use of relatedness estimators for other purposes. Note that is a close correspondence between inferred relatedness and that determined by actual pedigrees in thoroughbred race horses (Mar 2002 #303). stuart.thomas@ed.ac.uk 

315.  Graylings may have evolved in response to size-selective fishing
         A century of life-history evolution in grayling. 2001. Haugen, T.O.  and L.A. Vøllestad. Genetica 112-113:475-491.
         This is a study of several populations of graylings (Thymallus thymallus, a sort of trout or salmon) that have been isolated from each other and from a common source for 9 - 22 generations in Norway. The source population still flourishes in one of the lakes studied. Experiments demonstrating that adaptation to new temperature regimes has evolved in these fish have already been published (Jan 2001 #162). The authors have now extended the work to include changes in other phenotypic traits, including size and age at maturity, size-specific fecundity, growth rate, size at swim-up etc. For some of the traits they used preserved scale samples to look at the changing phenotype over the years in the source lake. For other traits they grew fish from different lakes at a single location ("common garden" experiments). For some adult traits they collected samples from the lakes and measured contemporary variation in the phenotype.
         They estimate the selection intensity caused by size-selective fishing (gill-netting) on the basis of a life-history model which maximizes lifetime contribution to the growth of a population with a stable age structure. The evolutionary conclusions in the paper therefore derive from direct genetic observations, i.e. from the common garden experiments, and also from inferences about phenotypic measurements that span a range of time, space and real and hypothetical selection regimes.
         The paper supports the conclusion that size-selective fishing in the lakes has caused rapid evolution towards earlier maturation. An evolutionary change in the growth pattern (faster early growth and slower late growth) and increased size-specific fecundity has also been caused by selective fishing. All these changes are in the direction predicted by theory. The speed and magnitude of the changes show that evolutionary considerations may be important for fisheries management on a decadal time scale. throndh@bio.uio.no 

314.  Swimbladder inflation in bass is not heritable: a useful negative result
         Heritability of swimbladder inflation in striped bass. 2002. Harrell, R.M., W.V. Heukelem, J.M. Jacobs, J.R. Schutz, J.U. Takacs, and D. Jacobs. North American Journal of Aquaculture 64:117-121.
         Striped bass (Morone saxatilis) that don't inflate their swim bladders in a timely way have serious skeletal and other developmental abnormalities, if they survive at all. The authors of this paper conducted a factorial mating experiment to see if the problem is heritable. They found that it isn't. Their estimate of half-sib heritability was essentially zero. The full-sib estimate was a robust 0.35, however, and the authors conclude that "swim bladder inflation is predominantly influenced by dominance, epistasis, or environmental factors".
         They draw the conclusion that there may be genotypic variation for the trait but it isn't additive and will not respond to selection. The authors should be praised for publishing this useful, negative result, which is likely to save people a lot of time and effort. (For a more aggressive approach to the marketing of full-sib data see Feb 2002 #295.) rh116@umail.emd.edu 

313.  Useful additions to the "effective population size" concept
         Relationship of effective to census size in fluctuating populations. 2002. Kalinowski, S.T., and R.S. Waples. Conservation Biology 16 (1):129-136.
         Population biologists and hatchery managers often like to talk about the effective population number even if they are not geneticists. It seems like a simple idea -- to the extent there is such thing in population genetics -- which evidently relates to the size of some sort of population. The effective population is a conceptual one that has tidy properties like random mating and 1:1 sex ratio, and which loses genetic diversity and/or accumulates inbreeding at the same rate as the real one with which it is being compared.
         The ratio of effective number to the real census number, Ne/N is an intuitive measure of how well a population can protect its genetic quality against random degradation. Rather like comparing the mechanical advantage of an ideal "frictionless pulley" to a real pulley to see how well a pulley is made. The effective population size and the frictionless pulley are highly practical, theoretical concepts.
         This paper by Kalinowski and Waples extends the intuitive appeal of the concept by defining several variables for use when the census size of a population fluctuates from generation to generation. These variables are related to the ratio Ne/N and give it some intuitive depth when numbers keep changing from census to census. For example, the meaning given to the variable beta is "a description of how much each real individual contributed to the effective population size" during a period of fluctuation. Their examples and discussions help clarify the meaning of the effective population number in real-world situations. The definitions are algebraic and not tied to any particular genetic model, so there is hope they will continue to apply even in complex situations such as #318, above. steven.kalinowski@noaa.gov 

312.  Heterozygosity is probably the best way to detect inbreeding using markers
         Heterosis, marker mutational processes and population inbreeding history. 2001. Tsitrone, A., F. Rousset, and P. David. Genetics 159:1845-1859.
         In practical aquaculture and conservation one often wants to know whether a particular population, in a particular situation, is suffering the effects of inbreeding depression. There are various ways inbreeding can arise and it may be very important to know what is happening. In aquaculture, for example, inbreeding can be caused by systematic mating between relatives in a mis-managed breeding program, accidental mating between relatives in a broodstock that is chronically too small, a too-small founding broodstock, etc. Analogous things happen to natural populations.
         At any given time there will be variation in the inbreeding levels of individuals in the population. If inbreeding depression is causing problems for the population as a whole, then there usually will be some degree of correlation between individual fitness (size, fecundity, parasite load etc.) and individual inbreeding. It is easy to check for this correlation if pedigree records exist. If they don't exist, DNA markers might provide an alternative. But how should DNA markers be used for this purpose?
         One solution is to estimate an inbreeding coefficient for individuals using markers to reconstruct, or deduce, the absent pedigree records (e.g. Dec 2000 #142), and then look for a correlation between fitness and the inbreeding coefficient. The second solution is to look for a correlation between the marker heterozygosity (as an inverse indicator of inbreeding) and the fitness of individuals. Such correlations are often found, and the three or four most likely explanations have been debated by geneticists for about 20 years. A third solution is to construct a genotypic index which depends on the mutational dynamics of the marker locus, rather than its heterozygosity, as an indicator of inbreeding. One such index is d2 (Mar 2002 #303) which is the square of the difference in length of the two microsatellite alleles at a marker locus in an individual.
         This paper compares the power of the heterozygosity and the d2 methods to detect inbreeding, using simulations based on the two prevalent models of the mutation process and a variety of population scenarios. In almost every scenario examined, inbreeding was more highly correlated with marker heterozygosity than with marker d2. This is useful to know. It would be useful to compare heterozygosity and d2 in some of the more pathological breeding scenarios one encounters in hatcheries, to see if it holds true there as well. tsitrone@cefe.cnrs-mop.fr 

311.  Beautiful visual display of a good genetic story
         African pastoralism: genetic imprints of origins and migrations. 2002. Hanotte, O., D.G. Bradley, J.W. Ochieng, Y. Verjee, E.W. Hill, and J.E.O Rege. Science 296:336-339.
         This paper about the domestication of cattle in Africa should interest all conservation geneticists, for two reasons. The first is the way the microsatellite data are displayed visually. The authors extracted the first three principal components (PCs) from microsatellite data on more than 50 breeds of indigenous cattle scattered throughout Africa, and plotted the PCs as a "synthetic map" superimposed on an ordinary map of Africa. This sort of thing has been done before but the result here is absolutely stunning. You can understand 10,000 years of genetic history at a glance. The visual display renders the tale told by the authors instantly convincing.
         The second reason is the tale itself. Apparently cattle were first domesticated locally, in the north-east of Africa, and the original breed of Bos taurus was walked southwards all the way to the Cape by pastoralists. This process may have begun about 8,000 years ago. The animals are pictured in pre-historic rock paintings and both they and their living descendents look very strange to my Canadian eyes. The southward diffusion is shown beautifully in the map of the second principal component (PC2).
         Most of the cattle now in Africa are also descended from another type of cattle (B. indicus) which was domesticated in the Near East, and which PC1 shows to have entered East Africa and the Horn, probably by sea trade. Pictorial representation of these animals, which unlike B. taurus are hump-backed, appear in Egyptian tomb paintings after about 2,000 BC. The site of introduction of hump-back cattle and the spread of their paternal genes over the continent is obvious in the map of PC1. The third PC shows that there has also been more recent introduction of B. taurus genes from Europe and the Near East. An elegant paper. The maps with superimposed principal components were drawn with the GIS program ArcView (Spatial Analyst Extension) which is available, for a price, at http://www.esri.com/software/arcview. The program runs on LINUX and MSWindows. o.hanotte@cgiar.org