We reviewed 95 studies (published from 1972 to 1997) that examined relation
ships between timber harvest and populations of songbirds and cavity-nestin
g birds. We critique the way in which studies have been conducted, evaluate
their usefulness to forest managers, and suggest new directions of study,
The number of bird-forestry studies conducted increased throughout our revi
ew period and most appeared in The Journal of Wildlife Management (24%) and
U.S. Department of Agriculture Forest Service technical publications (19%)
. More studies (32%) have occurred in the northeastern United States than e
lsewhere and most have examined effects of clearcutting (53%). Researchers
typically collect data on all bird species, especially songbirds (78%), usi
ng common sampling protocols such as point-count surveys, line transects, a
nd spot-mapping techniques to assess relative avian abundance (55%) and den
sity(32%). Few studies (13%) measured avian demographic parameters such as
nest success or survivorship. Most studies (68%) lasted only 1-2 years; onl
y 7 (7%) lasted >4 years. Most studies (27%) had only one replicate/treatme
nt. Research on effects of timber harvest on bird populations has been limi
ted to mensurative (observational) studies in which treatment effects canno
t be inferred statistically. Most research is correlational (84%) and does
not address cause-and-effect relationships. Incorporating experimental trea
tments to provide pre- and post-timber-harvest comparisons is rare (16%). F
uture research should: 1) be more long-term; 2) incorporate rigorous experi
mental designs in which treatments are assigned randomly and better replica
ted; and 3) although difficult, measure parameters related to avian fitness
and population viability. Rather than only documenting observed patterns,
researchers need to focus on identifying causal mechanisms that can be tran
slated into meaningful management recommendations to enhance conservation o
f forest avifauna.