length), lightly pigmented, wingless beetles with
greatly reduced elytra and eyes. The group is most
diverse in South America, where there are five de-
scribed genera and over 50 described species. Only
one genus, Amblyopinus, is found in Central America,
and two species occur in Monteverde.
Amblyopinines were believed to be obligate, blood-
feeding ectoparasites, based on the presence of adults
with their heads tightly attached in the fur of their
hosts, skin damage occurring when they are removed,
large strongly sclerotized mandibles, and the presence
of blood in the guts of beetles. Studies in Monteverde
have shown that amblyopinines are not parasites (Ashe
and Timm 1986,1987a, 1987b). We live-trapped hun-
dreds of rodents, studied the distribution of the am-
blyopinines they hosted, kept hosts and beetles to-
gether in observation chambers and examined their
interactions, and did host choice experiments to de-
termine if the beetles could distinguish among differ-
ent species of rodents as potential hosts (Ashe and
Timm 1986, 1987a,b).
The beetles are host-specific: Amblyopinus emargi-
natus showed a strong preference for the neotropical
Rice Rat (Oryzomys albigularis)', A. triptoniwas found
on two closely related hosts (Peromyscus nudipes and
Reithrodontomys creper, Muridae). Fewer than 1% of
the beetles were found on the "wrong" host. This level
of host specificity is typical of many parasitic insects
and is consistent with the hypothesis of parasitism for
these beetles.
We confirmed observations made by P. Hershkovitz
in 1952 that the host takes no notice of the beetles,
even when they crawl across its face, and no skin
damage was found that could be attributed to the
beetles. The skin of mice with heavy infestations of
beetles appeared healthy. When we transferred beetles
to a species of mouse on which the beetles did not
normally live, the mouse became irritated by their
presence and would scratch at the beetles until it was
able to remove and kill them.
The beetles leave the host during the day when the
host becomes inactive and roam throughout the nest.
They climb back onto the mouse and take up their
characteristic position behind or between the ears
when the mouse becomes active at nightfall. The
mystery of their feeding habits was solved when we
observed the beetles eating fleas and other true ecto-
parasites in the nest during the day. Rather than be-
ing ectoparasites, the beetles were actually "ectocom-
mensals," and likely mutualistic, with the hosts. The
host primarily provides transport from nest to nest in
which the beetles feed on ectoparasites. The blood
found in the guts of some beetles probably came from
secondary ingestion of blood from the guts of true ecto-
parasites when the beetles ate them. The hosts appear
to actively tolerate the presence of the beetles in their
fur, perhaps because their presence is beneficial.
If the beetles are predators only on ectoparasites,
why are they so host-specific? Any mouse that toler-
ates their presence and has ectoparasites would seem
to be a suitable host. Where are the immature stages
of the beetles? Is this behavior typical of other ambly-
opinines, or have others made the transition to true
parasitism? What characteristics of rodents and
beetles have driven the evolution of this interaction?Staphylinids associated with ants. Some staphylinids,
especially species in the subfamily Aleocharinae,
have been successful at invading the nests of social
insects, particularly those of ants (150 staphylinid
genera) and termites (over 100 genera). Termites are
uncommon in Monteverde, and no termitophilous
staphylinids have been reported. Of the staphylinid
genera associated with ants ("myrmecophiles"), more
than 100 are found in association with army ants in
both the Old and New World.
Species of Tetradonia and related genera are usu-
ally found in association with the edges of raiding
columns or swarms of army ants, or in their refuse
piles, where they attack injured ants or steal prey from
the ants. They avoid contact with the ants and repel
the attackers by using defensive chemical secretions
from an abdominal gland (Akre and Rettenmeyer
1966, J. Ashe, pers. obs.). Other taxa (e.g., Ecitodonia
species of which are not myrmecoid) are common in
refuse piles of Eciton bivouacs; they are found run-
ning in the ant columns when the colony moves to a
new bivouac. Others (e.g., Ecitophya, Ecitomorpha,
Ecitochara), which have more myrmecoid bodies, are
with Eciton army ants in the raiding columns. Spe-
cies ofEcitosus, which have very myrmecoid bodies,
have been reported in the columns of Neivamyrmex
(Akre and Rettenmeyer 1966). The most unusual
myrmecophiles in Monteverde are the myrmecoid
members of the tribe Leptanillophilini that occur
with Labidus army ants. Species of Labidoglobus and
Mimonilla are virtually unpigmented, have lost their
wings and eyes, have reduced the elytra to a vestigial
nub, and are difficult to distinguish from the ants
with which they run. Beetles with such highly de-
rived myrmecoid body forms are rarely attacked by
the ants.Staphylinids associated with other substrates. Fungi
that produce macroscopic fruiting bodies offer a va-
riety of resources to insects: mycelia (fungal fibers),
asexual spores, sexual spores, portions of macroscopic
fruiting bodies, and other arthropods. Examples of
large, strikingly colored, predaceous staphylinids
associated with fungi in Monteverde include all spe-110 Insects and Spiders