114
N. WALOFF
11 000 testa Dempter was able to evaluate the relative importance of
the various predators. These included Heteroptera, Miridae - Asciodema obsoletum (Fieber), Heterocordylw tibialis (Hahn), Orthtylus &no~arpi (Perris), 0. virescens (DougIau and Scott), Heterdoma meriupterca
(Scopoli) ; Anthocoridae - Anthmmis sarothumni (Douglas and Scott),
A . nemorum (L.); Nabidae - Hemicerue apterw (Fabricius); Dermaptera - Forfie& auricularia L. and the red mite Anystis baccarum.
Dempster estimated that a loss of 111 600 immature stages of Phytodecta could be accounted by predators in 1967, and 101 300 in 1968.
These estimates compared well with the numbers that we could not
account for by sterilit'y of eggs and parasitism of immature stages and
which were equal to 140 848 in 1957 and 102 312 in 1958. The total loss
of the progeny in the two years caused by predation waa equal to 78%
and 920/0. The bimodality of the population curves reflected the seasonal abundance of the predatory insects.
The effects of predators were strongly linked with changes in the
habitat. By the end of the study, many bushes were old and dying and
after the severe frosts in the winter of 1 9 5 6 5 6 the habitat was reduced
to half of its original volume. Thus the insect fauna W&B concentratad
and the increased density of prey and the high numbers of predators in
1956 resulted in extremely high mortality, i.e. of 98.98% of the immature stages in t h a t year. The second reduction in the habitat occurred
in the winter of 1957 when the volume of broom was reduced to a
quarter of the original and the pattern was repeated, 9 1 - 6 3 ~ o of the
progeny of Phytodecta being destroyed on the hoat plant.
4. The roles of different mortality factors
These are more clearly seen when the data in Table XI are analysed
by the key factor analysis (Varley and Gradwell, 1960) (Fig. 2). Predation on the immature stages on broom is by far the most effective factor
in causing changes in the population size and runs parallel to the total
mortality within each generation of Phytodecta. As Richards (1963)
pointed out, on any particular quantity of broom there may be a ceiling
for the beetle population determined by predation. When the population increases more eggs and larvae will be discovered by random
searching and predation will be heavier. This particular prey is relatively unimportant, and while the process of predation imposes a check
on its numbers, it cannot in any way affect those of the predators
because there are too many other phytophagoue insects on broom.
Solomon (1064) worked out the coefficients of variation in the percentages of mortality in different stages of Phytdeeta and found that
the least variation occurs in the numbers of autumn beetles surviving
until spring. He suggests that the regulatory fwtor probably operates
N. WALOFF
11 000 testa Dempter was able to evaluate the relative importance of
the various predators. These included Heteroptera, Miridae - Asciodema obsoletum (Fieber), Heterocordylw tibialis (Hahn), Orthtylus &no~arpi (Perris), 0. virescens (DougIau and Scott), Heterdoma meriupterca
(Scopoli) ; Anthocoridae - Anthmmis sarothumni (Douglas and Scott),
A . nemorum (L.); Nabidae - Hemicerue apterw (Fabricius); Dermaptera - Forfie& auricularia L. and the red mite Anystis baccarum.
Dempster estimated that a loss of 111 600 immature stages of Phytodecta could be accounted by predators in 1967, and 101 300 in 1968.
These estimates compared well with the numbers that we could not
account for by sterilit'y of eggs and parasitism of immature stages and
which were equal to 140 848 in 1957 and 102 312 in 1958. The total loss
of the progeny in the two years caused by predation waa equal to 78%
and 920/0. The bimodality of the population curves reflected the seasonal abundance of the predatory insects.
The effects of predators were strongly linked with changes in the
habitat. By the end of the study, many bushes were old and dying and
after the severe frosts in the winter of 1 9 5 6 5 6 the habitat was reduced
to half of its original volume. Thus the insect fauna W&B concentratad
and the increased density of prey and the high numbers of predators in
1956 resulted in extremely high mortality, i.e. of 98.98% of the immature stages in t h a t year. The second reduction in the habitat occurred
in the winter of 1957 when the volume of broom was reduced to a
quarter of the original and the pattern was repeated, 9 1 - 6 3 ~ o of the
progeny of Phytodecta being destroyed on the hoat plant.
4. The roles of different mortality factors
These are more clearly seen when the data in Table XI are analysed
by the key factor analysis (Varley and Gradwell, 1960) (Fig. 2). Predation on the immature stages on broom is by far the most effective factor
in causing changes in the population size and runs parallel to the total
mortality within each generation of Phytodecta. As Richards (1963)
pointed out, on any particular quantity of broom there may be a ceiling
for the beetle population determined by predation. When the population increases more eggs and larvae will be discovered by random
searching and predation will be heavier. This particular prey is relatively unimportant, and while the process of predation imposes a check
on its numbers, it cannot in any way affect those of the predators
because there are too many other phytophagoue insects on broom.
Solomon (1064) worked out the coefficients of variation in the percentages of mortality in different stages of Phytdeeta and found that
the least variation occurs in the numbers of autumn beetles surviving
until spring. He suggests that the regulatory fwtor probably operates