the absence of planktivorous fish, the mere presence of fish chemical cues
(kairomones) can induce earlier maturity at smaller body size (e.g. Tollrian &
Harvell, 1998 ; Stibor & Navarra, 2000 ; Sakwinska, 2002 ; Weetman & Atkinson,
2002 ; Fig.3.1). In contrast, when Cladocera are reared in the presence of chem-
ical cues from smaller, invertebrate, predators such as the phantom midge
larva,Chaoborussp., which are more effective at killing small rather than large
individual cladocerans, the cladocerans grow to a larger size (Stibor & Lu ̈ ning,
1994 ). In both these examples, the cladocerans are responding to environmental
information (predator cues) about selection pressures (size-dependent mortal-
ity) to induce changes in adult size that appear to be adaptive.
The prediction of optimal reaction norms depends strongly on the spatial
structure of the population and the frequency with which a gene affecting the
reaction norm encounters a particular environment out of the full set of environ-
ments experienced by the population (Houston & McNamara, 1992 ; Kawecki &
Stearns, 1993 ). In an isolated habitat, the termrin Eq. (3.1) is the local intrinsic
rate of increase in the habitat. But in meta-populations, where offspring may
encounter different patches from their parents, therterm is not that for a given
patch but the average for the population patches that the offspring may encoun-
ter (‘globalr’). If the average population-growth rate in the patches likely to be
encountered is 0, then the discount term in Eq. (3.1) disappears and the equation
reduces to that for LRS. This fitness measure is used, for example, in reaction
norm models of Berrigan and Koella ( 1994 ) that investigate effects of growth
and mortality on optimal age and size at maturity, and is good news for field
ecologists, as LRS is easier to measure thanr(Kozłowski, 1993 ). Predictions of
life-history theory differ depending on whether local or globalris maximized.
For example, if globalris 0 and is assumed to be maximized by natural selection
(i.e. other strategies haver<0; Kozłowski, 1993 ), then the rate at which the local050100150200250CH FH CL FL
Fish kairomone-food level treatmentMass(μg)Figure 3.1Mean body mass (þ95% confidence limits) at first reproduction ofDaphnia
magna.CH¼control (no fish kairomone)Daphniaat high food (1.5 mg C L^1 ); FH¼Daphnia
treated with fish kairomone at high food; CL¼controlDaphniaat low food (0.5 mg C L^1 );
FL¼Daphniatreated with fish kairomone at low food. Data are extracted from Figure 1 in
Stibor & Navarra ( 2000 ), and redrawn.38 D. ATKINSON AND A. G. HIRST