Climatic Inferences from Dendroecological Reconstructions - Dendroclimatology - page 275

(e.g., western spruce budworm and Douglas-fir tussock moth) generally tend to feed
in stands that have relatively closed-canopy conditions, in moderate to highly mesic
sites, and at mid-elevations. Stands with Abies as a dominant or codominant species,
for example, are more vulnerable than pure Douglas-fir stands. In general, these
kinds of stands are infrequently sampled for dendroclimatic purposes. The classic
western United States drought-sensitive site is an open-canopy conifer stand, in a
xeric site, at the lower forest border, with steep slopes and shallow soils (Schulman
1956 ; Fritts 1976 ) . These are typically the least likely stands to be attacked by west-
ern spruce budworm. High-elevation, near-tree-line stands that are often sampled
for temperature reconstructions also are generally not affected by budworm or other
major defoliators. Although some insect defoliators and stem feeders (e.g., bark bee-
tles) can affect large areas, to the best of our knowledge, past outbreaks have tended
to occur in a relatively small proportion of the range of host species.
In addition to the reasons listed above, episodic or chronic defoliation, or past
insect-induced mortality within stands is usually visually obvious on trees within
the stands, or in the sampled ring-width series (e.g., Fig. 9.6 ; but see Ryerson et al.
2003 for a case study where outbreak signals were not often visually obvious in the
'uncorrected' ring width series). Within stands, past defoliation events are usually
apparent in the presence of trees with old dead tops (spikes), and new dominant
crown leaders, and many dead branches that are not simply the result of great age.
Also, dead trees from past killing events may be present, and in the case of bark
beetles, telltale feeding and egg-laying galleries may be visible on the stems of dead
trees for many years after the event (e.g., Perkins and Swetnam 1996 ) .
In summary, the potential confounding effects on tree-ring series of past insect
attacks—and other disturbances such as fires—are an important consideration that
dendroclimatologists should be more keenly aware of than they generally are.
Precautions should be taken to investigate what is known from documentary sources
and local experts about the potential occurrence of past outbreaks in areas to be
sampled. Dendroclimatologists should have some field and laboratory training and
experience in identifying signs of past outbreaks within stands and on tree-ring
specimens. We encourage greater awareness of this issue by dendroclimatologists.
We also recommend further study of the potential 'contamination' of dendrocli-
matic databases with insect and other disturbance signals. Perhaps a screening of
existing dendroclimatic databases is warranted for potentially unknown insect out-
break signals that may exist in some chronologies from host species in areas with
known outbreaks. However, it is our expectation that if there are any such problem
chronologies, they are likely to be few.
9.2.3 Regional Tree Demography and Climate Effects
Ecologists have long recognized that time series of tree births and deaths are of fun-
damental value for understanding forest and woodland dynamics (Blackburn and
Tueller 1970 ; Henry and Swan 1974 ; Harper 1977 ) . Estimating the age of a tree by
counting its annual rings is probably the best-known tree-ring application in history,