1. Laboratory experiments were carried out to analyse various aspects
of spider performance under different diets, focusing mainly on the va
lue of the aphid Rhopalosiphum padi. Aphis fabae and a sciarid midge w
ere considered in single experiments. Laboratory fruit flies were used
as a standard prey for comparison. The spider species used were Erigo
ne atra, Pardosa amentata and P. prativaga. 2. In preference experimen
ts P. amentata ranked prey types in the order: fruit flies > sciarid >
R. padi > A. fabae. 3. In P. prativaga consumption of fruit flies inc
reased with the length of prior starvation period; consumption of aphi
ds was much lower and independent of hunger. 4. First instar spiderlin
gs of P. amentata and Erigone atra were unable to develop on a pure di
et of R. padi; they died without moulting. 5. Egg production in wild c
aught E. atra females stopped after 2-4 egg-sacs on a pure diet of R.
padi, whereas females given fruit flies or fruit flies + aphids produc
ed an average of eight sacs. Egg numbers were high on a pure diet of f
ruit flies, but hatching success was best with fruit flies supplemente
d with aphids. The size of hatched young depended on both parent size
and diet. 6. It was concluded that R. padi represented a low quality p
rey for cereal spiders which obtain very little energy but essential n
utrients from the aphids. It remains to be established whether this ha
s consequences for the prospects of managing spider communities with t
he view of improving their effectiveness against aphids. It is argued
that with a low preference for aphids, improving availability of alter
native prey in more diverse agrosystems may possibly reduce spider pre
dation on aphids.