Climatic Inferences from Dendroecological Reconstructions - Dendroclimatology - page 274

A notable example of the use of tree rings in evaluations of animal popula-
tion dynamics and cyclic environmental variables are the studies of Sinclair et al.
( 1993 ; Sinclair and Gosline 1997 ) of hare populations, climate, and sunspot cycles
in Canada. Unfortunately, the physical tree-ring basis of the hare population recon-
structions (feeding scars on aspen stems) seems obscure and is not well illustrated or
described in these papers. Some of the quantitative time series analyses are also dif-
ficult to evaluate. More research in this area, particularly with employment of more
rigorous dendrochronological sampling and analysis procedures, may be fruitful in
evaluating previous findings, and shedding light on the role of climate and animal
population dynamics. New and powerful statistical-analytical tools that are increas-
ingly employed by dendrochronologists and paleoclimatologists might be usefully
employed in this endeavor (e.g., univariate and bivariate singular spectrum analysis
(SSA), or wavelet analysis; e.g., Speer et al. 2001 ; Esper et al. 2007 ) .
The periodic and aperiodic behaviors evident in tree-ring-based insect outbreak
reconstructions, and their potential coherence with climate, deserve much more
study. Most insect population dynamics studies have relied on much shorter obser-
vational datasets, which in some cases include only one to a few insect outbreaks
(e.g., Royama 1984 ; Myers 1998 ) or population 'eruptions' (e.g., Ranta et al. 1997 ) .
Tree-ring-based insect outbreak and climate reconstructions for population dynam-
ics research are of potentially great value because numerous population oscillations
over periods of centuries can be identified at numerous locations over regions, and
even continents. We think there are opportunities here for breakthroughs in our
understanding of how climatic variations and oscillations act as potential synchro-
nizing (e.g., the Moran effect) or disrupting factors in plant and animal population
dynamics. Given recent enormous outbreaks of bark beetles in concert with climatic
changes in North America (e.g., Logan et al. 2003 ; Breshears et al. 2005 ) , and Esper
et al.'s (2007) findings about changes in larch budmoth and temperature in the Alps,
the importance of this topic is likely to increase in coming years.
9.2.2.1 Confounding of Dendroclimatic Signals by Insect Outbreaks
One final point regarding insect outbreaks is important in the context of this chapter
and book; i.e., the potential confounding influence of insect outbreaks/defoliation on
dendroclimatic interpretations from host trees. Dendroclimatologists have long been
aware of this as a potential problem (e.g., Morrow and LaMarche 1978 ) , and as a
general practice they seek to avoid sampling sites and trees that have a known history
of past outbreaks (Fritts 1976 ) . A recent paper by Trotter et al. ( 2002 ) evaluated this
issue in some detail in pinyon pine ( Pinus edulis ) stands near Flagstaff, Arizona,
defoliated by pinyon pine needle scale ( Matsucoccus acalyptus ). They concluded
that, indeed, chronic or episodic effects of this insect could alter the dendroclimatic
estimates of drought if one were to use tree-ring width series from affected trees.
It is our opinion that the large majority of tree-ring width series that have been
sampled and measured for dendroclimatic studies in the western United States do
not have a great risk of being confounded with insect outbreak (or chronic insect
feeding) signals. This is because most of the major episodic defoliators of conifers