DYNAMICS OF F I E L D POPULATION O F P I N E LOOPER
29 1
September density (Fig. 37-A). The regression coefficient (0.33) differs
significantly from zero.
To test this relationship further we computed the regression of nymphal density over larval September density (Fig. 37-B). As already
stated, without density governed effects this should result in a regression coefficient equal to unity. Here, however, it is 0.68 and differs
significantly from unity (confidence interval of regression coefficient
b j= 3 . s, ; 0.46 < b < 0.90).
As to the causation of this density governed mortality we can only
make suppositions, because our field evidence with reference to this
point is too scanty. Disease and predation by birds are the first agents
to be mentioned. The response of birds would be functional in this case
(Solomon, 1949; see also p. 2943, and might be due to the fact that the
polyphagous predators concentrate on eating prey species which are
temporarily numerous. An effect of this type has heen found by Varley
and Gradwell (1963) in the winter moth, Operopiitera brumata, which
are preyed upon by the larvae of a beetle and some mammalian predators. In Dutch pine woods titmice with their varied diet concentrate
on the more numerous prey species, thus giving rise to a density
dependent predation over at least part of the density range of these
species (Tinbergen, 1960; Tinbergen and Klomp, 1960).
The density dependence of the mortality in this age-interval is more
powerful than that of the fecundity. It can be derived that a density
deviation up induces roughly a net rete of reproduction R z l/Ta
(see p. 288). As a result of this the population would also tend to
apprciach asymptotically to the mean level, but moxe rapidly than under
the sole influence of the density governed fecundity, the successive densities being e.g. 5p, 2.9p, 2.013, 1 . 6 ~ .
. . .
The question arises concerning the power of the combined regulatory
effect of these density governed processes. This question can be
approached roughly by the following reasoning. Both the mortality of
advanced larvae and the fecundity are affected by larval density.
Assuming that the population is in balance at the mean level (p. 288),
being about 10 larvae/m2, then according to Figs. 26 and 37 the mean
fecundity is 182 eggs and the larval mortality amounts to 64.5%
(k4-5 = 0-43). Then R = 1 = 0.355 x 182 x s, and s = 0.0155, being
the fraction surviving the other mortality factors.
If the population increases to 293 = 20, then fecundity decreases to
163, larval mortality increases to 72% (kdP5 = 0.53), and R = 0.28
If the population declines to 1/2p = 5, then fecundity increases to
193, larval mortality decreases to 53.5% (k4-5 = 0.32), and R = 0-465
x 163 x 0.0155 = 0.71
1 / 4 2 .
x 193 x 0.0155 = 1.39
1/+J(1/2).
29 1
September density (Fig. 37-A). The regression coefficient (0.33) differs
significantly from zero.
To test this relationship further we computed the regression of nymphal density over larval September density (Fig. 37-B). As already
stated, without density governed effects this should result in a regression coefficient equal to unity. Here, however, it is 0.68 and differs
significantly from unity (confidence interval of regression coefficient
b j= 3 . s, ; 0.46 < b < 0.90).
As to the causation of this density governed mortality we can only
make suppositions, because our field evidence with reference to this
point is too scanty. Disease and predation by birds are the first agents
to be mentioned. The response of birds would be functional in this case
(Solomon, 1949; see also p. 2943, and might be due to the fact that the
polyphagous predators concentrate on eating prey species which are
temporarily numerous. An effect of this type has heen found by Varley
and Gradwell (1963) in the winter moth, Operopiitera brumata, which
are preyed upon by the larvae of a beetle and some mammalian predators. In Dutch pine woods titmice with their varied diet concentrate
on the more numerous prey species, thus giving rise to a density
dependent predation over at least part of the density range of these
species (Tinbergen, 1960; Tinbergen and Klomp, 1960).
The density dependence of the mortality in this age-interval is more
powerful than that of the fecundity. It can be derived that a density
deviation up induces roughly a net rete of reproduction R z l/Ta
(see p. 288). As a result of this the population would also tend to
apprciach asymptotically to the mean level, but moxe rapidly than under
the sole influence of the density governed fecundity, the successive densities being e.g. 5p, 2.9p, 2.013, 1 . 6 ~ .
. . .
The question arises concerning the power of the combined regulatory
effect of these density governed processes. This question can be
approached roughly by the following reasoning. Both the mortality of
advanced larvae and the fecundity are affected by larval density.
Assuming that the population is in balance at the mean level (p. 288),
being about 10 larvae/m2, then according to Figs. 26 and 37 the mean
fecundity is 182 eggs and the larval mortality amounts to 64.5%
(k4-5 = 0-43). Then R = 1 = 0.355 x 182 x s, and s = 0.0155, being
the fraction surviving the other mortality factors.
If the population increases to 293 = 20, then fecundity decreases to
163, larval mortality increases to 72% (kdP5 = 0.53), and R = 0.28
If the population declines to 1/2p = 5, then fecundity increases to
193, larval mortality decreases to 53.5% (k4-5 = 0.32), and R = 0-465
x 163 x 0.0155 = 0.71
1 / 4 2 .
x 193 x 0.0155 = 1.39
1/+J(1/2).