Data from: Multi-trophic occupancy modeling connects temporal dynamics of woodpeckers and beetle sign following fire

In conifer forests of western North America, wildlife populations can
change rapidly in the decade following wildfire as trees die and animals
respond to concomitant resource pulses that occur across multiple trophic
levels. In particular, black-backed woodpeckers (Picoides arcticus) show
predictable temporal increases then declines following fire; this
trajectory is widely believed to be a response to the woodpeckers'
main prey, woodboring beetle larvae of the families Buprestidae and
Cerambycidae, but we lack understanding of how abundances of these
predators and prey may be associated in time or space. Here, we pair
woodpecker surveys over 10 years with surveys of woodboring beetle sign
and activity, collected at 128 survey plots across 22 recent fires, to ask
whether accumulated beetle sign indicates current or past black-backed
woodpecker occurrence, and whether that relationship is mediated by the
number of years since fire. We test this relationship using an integrative
multi-trophic occupancy model. Our results demonstrate that woodboring
beetle sign is a positive indicator of woodpecker presence 1–3 years
following fire, an uninformative indicator from 4-6 years after fire, and
a negative indicator beginning 7 years following fire. Woodboring beetle
activity, itself, is temporally variable and dependent on tree species
composition, with beetle sign generally accumulating over time,
particularly in stands with diverse tree communities, but decreasing over
time in Pinus-dominated stands where faster bark decay rates lead to brief
pulses of beetle activity followed by rapid degradation of tree substrate
and accumulated beetle sign. Altogether, the strong connections of
woodpecker occurrence to beetle activity support prior hypotheses of how
multi-trophic interactions govern rapid temporal dynamics of primary and
secondary consumers in burned forests. While our results indicate that
beetle sign is, at best, a rapidly shifting and potentially misleading
measure of woodpecker occurrence, the better we understand the interacting
mechanisms underlying temporally dynamic systems, the more successfully we
will be able to predict the outcomes of management actions.