Population-dynamic consequences of predator-induced life history variation in the guppy (Poecilia reticulata)

A fundamental goal of population ecology is to quantify how spatial and temporal variation in life history traits translates into variation in population-level parameters such as vital rates and projected growth rate (lambda). Further emphasis has been on the sensitivity of lambda to absolute variation (sensitivities) and relative variation (elasticities) in vital rates, because lambda can be related to the evolutionary fitness of a life history phenotype (r = lnlambda). Furthermore, because density regulation can affect the expression and evolution of life history traits, which in turn can feed back on population dynamics, correctly incorporating the relationship between the two is critical to understanding the evolution and evolvability of life history strategies. Here we ask how life history variation affects projected population growth and evolutionary fitness dynamics in the long-term study system of the Trinidadian guppy (Poecilia reticulata). There are striking differences in the life history traits of guppies that co-occur with two different assemblages of predators and so these populations are a model natural system for examining the fitness consequences of, and selective pressures on, traits that covary predictably. We found that populations experiencing high rates of predation had lower estimates of growth and fitness than did those experiencing low rates of predation (pooled r = 0.02 and 0.17, respectively). But when we incorporated density regulation in low predation sites, observed with field manipulations and modeled with simulations, this difference became negligible. Furthermore, for all populations irrespective of predators, growth rate was most sensitive to changes in neonatal growth and was contributed to maximally by early adult survival. Particularly in light of g-rowing concerns on correctly identifying sensitive life stages in overharvested fauna, addressing these questions in a long-term natural experimental system may provide guidance.