THE EVOLUTION OF CARBON ALLOCATION TO PLANT SECONDARY METABOLITES: A GENETIC ANALYSIS OF COST IN DIPLACUS AURANTIACUS

Diplacus aurantiacus contains large amounts of a leaf phenolic resin, an important deterrent to a leaf-eating caterpillar, Euphydryas chalcedona. The resin can also retard water loss during drought. Furthermore, the leaf resin content differs among plants and populations. This study investigates the existence of heritable variation (h² ) in resin production and tests for a genetic correlation (r_G ) between carbon allocation to secondary metabolites and growth rate, as well as with three other vegetative traits. Nine dam and 10 sire plants were chosen randomly at a field site and used to generate 78 full-sib families (19 half-sib families) by crossing all males to all females in a factorial design. Heritability was estimated in two ways, and genetic correlations were estimated by three methods. We found: (1) the heritability of resin production estimated by the regression of offspring on sires was significantly greater than zero (hs2=0.32, P<0.01); (2) the maternal variance in resin content was significantly greater than zero (21.3% of total phenotypic variance); (3) significant negative genetic correlation between resin content and growth rate was observed from two of three methods and was consistent with the phenotypic correlation; and (4) the cost of resin could be assessed quantitatively. The genetic cost of 1 mg in resin is equivalent to 25 mg of dry shoot-biomass growth, but the phenotypic cost is only 2.1 mg. This study indicates that carbon allocation to these secondary metabolites may respond to natural selection, and the phenotypic cost of resin production has a genetic basis in D. aurantiacus. This trade-off suggests that once selection occurs, increased phenolic resin production may result in decreased growth, or vice versa.