The short-term effects of harvesting on survival and dispersal of insect pr
edators in hay lucerne (Medicago saliva L.) were investigated in a commerci
al crop in Australia. Adults of two coleopterans, transverse ladybird beetl
e (Coccinella transversalis Fabricius Coccinellidae) and pollen beetle (Dic
ranolaius bellulus (Guerin-Meneville) Melyridae), and one heteropteran, spi
ned predatory shield bug (Oechalia schellembergii (Guerin-Meneville) Pentat
omidae), were marked and released into lucerne strips immediately before th
ey were cut for hay. Vacuum sampling of the stubble immediately after harve
st showed that proportional mortalities were 0.16, 0.00 and 0.07 respective
ly. Proportions of the released individuals that were recovered alive were
0.66 for C. transversalis, 0.45 for O. schellembergii and 0.02 for D, bellu
lus. The proportion of predators that were not recaptured from release area
s was greatest for D, bellulus (0.80). To assess dispersal after harvest, 1
2 strips of lucerne were left uncut within the field in which the marked in
sects had been released. Intensive sweep-net sampling of these strips caugh
t marked survivors of all three predators. The numbers of insects caught in
strips declined with increasing distance from the release point. Many inse
ct predators, therefore, survived harvesting and, if uncut refuge strips ar
e provided, the within-field community of biological control agents is larg
ely preserved.The short-term effects of harvesting on survival and dispersa
l of insect predators in hay lucerne (Medicago saliva L.) were investigated
in a commercial crop in Australia. Adults of two coleopterans, transverse
ladybird beetle (Coccinella transversalis Fabricius Coccinellidae) and poll
en beetle (Dicranolaius bellulus (Guerin-Meneville) Melyridae), and one het
eropteran, spined predatory shield bug (Oechalia schellembergii (Guerin-Men
eville) Pentatomidae), were marked and released into lucerne strips immedia
tely before they were cut for hay. Vacuum sampling of the stubble immediate
ly after harvest showed that proportional mortalities were 0.16, 0.00 and 0
.07 respectively. Proportions of the released individuals that were recover
ed alive were 0.66 for C. transversalis, 0.45 for O. schellembergii and 0.0
2 for D, bellulus. The proportion of predators that were not recaptured fro
m release areas was greatest for D, bellulus (0.80). To assess dispersal af
ter harvest, 12 strips of lucerne were left uncut within the field in which
the marked insects had been released. Intensive sweep-net sampling of thes
e strips caught marked survivors of all three predators. The numbers of ins
ects caught in strips declined with increasing distance from the release po
int. Many insect predators, therefore, survived harvesting and, if uncut re
fuge strips are provided, the within-field community of biological control
agents is largely preserved.