This procedure minimizes the
intra- and inter-population variation in the rate of development.
The regression analysis of emergence was initially restricted to the
models proposed earlier byTrematerra et al. (1996), but the
3-parameter sigmoid model (y¼a/(1þexp((xb)/c)) provided
better fits for cumulative emergence of populations ofT. castaneum
andO. surinamensis, as also observed by Fragoso et al. (2005), and
the 3-parameter Gompertz model (y¼aexp(exp((xb)/c))
provided better fits for populations ofR. dominica. These models
weretherefore used