The Experimental Design

'What are you doing about two weeks from now?'... 'Nothing in particular.

Just beetling around.'

In [44] (also see [27]) we report the results of an experiment in which the adult rate fia are set at the five values indicated in Fig. 3.1, namely, iia = 0.04, 0.27, 0.50, 0.73, and 0.96. Each of these treatments was replicated four times. There was also an unmanipulated control. These cultures were initiated with 100 young adults, five pupae, and 250 small larvae. Each population was contained in a half-pint (237-mL) bottle with 20 g of standard medium and kept in a dark incubator at 32°C. Every 2 weeks the L-, P-, and /l-stages were counted and returned to fresh medium. Dead adults were also counted and removed. We continued this procedure for 36 weeks. At week 12 we randomly assigned four populations to each of the six treatments and commenced the imposition of the selected adult mortalities.

We manipulated adult mortality by removing or adding adults at the time of census in order to make the total number of adults who died during the previous time interval consistent with the desired treatment value of fia. For example, suppose there were 100 ¿4-stage individuals counted at time t and 10 dead adults were counted at time t + 1. If the target mortality for this treatment was //„ = 0.27, then we would remove the 10 dead adults plus an additional 17 adults from the culture. Or if 40 dead adults were observed, then we would add 13 living adults to the culture. In this fashion, we rigorously controlled the parameter ixa.2 If the target death rate resulted in a noninteger number of adults to be added or removed, we rounded the number to the nearest whole integer.

The corn-oil-sensitive (cos) strain of T. castaneum used to obtain the data for the parameterization performed in Chapter 2 is, unfortunately, no longer available. For the experiment reported in [44] it was therefore necessary to use a different genetic strain. In fact, we duplicated the experiment just described using two other strains—the so-called SS strain and the RR strain of T. castaneum. Because of this change (and because

2 To counter the possibility of genetic changes in life-history characteristics, beginning at week 12 all adults returned to the populations after census were obtained from separate stock cultures maintained under standard laboratory conditions.

of the different experimental procedures, protocols, laboratories, etc.) we performed a reparameterization of the LPA model using the data from the new experiment. We also repeated model validation and prediction fit analyses, using the methods described in Chapter 2. The results of these analyses are discussed in Sections 3.1.2 and 3.1.3. As we will see, parameter estimates remain little changed as a result of these reparam-eterizations, as does the sequence of dynamic bifurcations in Fig. 3.1 (which is testimony to the robustness of the LPA model's accuracy in describing the dynamics of the flour beetle populations).

Was this article helpful?

0 0

Post a comment