The target of selection matters: an established resistance – development-time negative genetic trade-off is not found when selecting on development time.

Trade-offs are fundamental to evolutionary outcomes and play a central
role in eco-evolutionary theory. They are often examined by experimentally
selecting on one life-history trait and looking for negative correlations
in other traits. For example, populations of the moth Plodia
interpunctella selected to resist viral infection show a life-history cost
with longer development times. However, we rarely examine whether the
detection of such negative genetic correlations depends on the trait on
which we select. Here we examine a well-characterised negative genotypic
trade-off between development time and resistance to viral infection in
the moth Plodia interpunctella and test whether selection on a phenotype
known to be a cost of resistance (longer development time) leads to the
predicted correlated increase in resistance. If there is tight pleiotropic
relationship between genes that determine development time and resistance
underpinning this trade-off, we might expect increased resistance when we
select on longer development time. However, we show that selecting for
longer development time in this system selects for reduced resistance when
compared to selection for shorter development time. This shows how
phenotypes typically characterised by a trade-off can deviate from that
trade-off relationship, and suggests little genetic linkage between the
genes governing viral resistance and those that determine response to
selection on the key life-history trait. Our results are important for
both selection strategies in applied biological systems and for
evolutionary modelling of host-parasite interactions.