During 1991-1994 we tested whether martens (Martes americana) selectiv
ely used postfire seres in the Alaskan taiga and whether selection cou
ld be explained by differences in marten hunting behaviour, habitat, p
rey abundance, or demography. Forest seral stages included early-succe
ssional tall shrub - sapling (1985 burn), midsuccessional dense tree (
1966 burn), and mature coniferous (100-115 years old). Most studies of
marten-habitat relationships from lower latitudes suggest that marten
s require coniferous forest and avoid open areas. We found that marten
s did not select forest cover types or burn features at the stand scal
e (within the home range). However, marten abundance was greatest in t
he 1985 burn, which had the lowest canopy cover but the highest coarse
woody debris density, autumn arvicoline biomass, and winter hunting i
ntensity (index to foraging suitability). Martens in the 1985 burn wer
e predominantly juvenile, with few adult (greater than or equal to 2 y
ears old) females present. We hypothesize that our study area of predo
minantly early -midsuccessional forest was a ''sink'' for immature and
transient martens dispersing from surrounding mature forest. If our h
ypothesis is correct and applies elsewhere in the taiga, then fur trap
ping for martens in recent burns could be a productive yet conservativ
e harvest strategy. Future research should focus on habitat requiremen
ts of parturient females.