Figure 3. Relationship between climate effects on willow shoot
growth, and effects of insect herbivores on shoot growth. Tests are the
same as those in Fig. 2a, except that analyses are run separately for
each of the experimental treatments. Abbreviations indicate: no
vertebrate herbivores (N ); only herbivory by ptarmigans
(P ), herbivory by ptarmigans and reindeer (RP ), and the
interaction of each of these treatments with “rejuvenation” (i.e.
pruning of shoots in year 1 of the experiment) (J , PJ , andRPJ , respectively) (Table 1).
When separated by browsing and rejuvenation treatments, we found that
the general pattern of responses of willow shoot growth to climate and
insect herbivores was similar to that in our pooled analysis (Fig. 3,
compare to Fig. 2a). However, the test revealed that the treatments
differ and that relationships within treatments were not uniform. Recall
that there are six different experimental treatments: neither vertebrate
herbivores (N ); only herbivory by ptarmigans (P ),
herbivory by ptarmigans and reindeer (RP ), rejuvenation (i.e.
pruning of shoots at the initiation of the experiment) (J ),
rejuvenation plus browsing by ptarmigans (PJ ), and rejuvenation
plus browsing by ptarmigans and reindeer (RPJ ). For most
treatments, there was still a negative relationship between the effects
of climate and the effects of insect herbivory, except in the case of
rejuvenation treatments, for which the relationship reversed (J, PJ
treatments in Fig. 3). Additionally, the treatments that included
vertebrate herbivores tended to experience more positive effects of
climate on shoot growth, while for treatments that included
rejuvenation, herbivory effects were generally less negative (compare
treatments P vs. PJ, RP vs. RPJ in Fig. 3).