Why does early-season herbivory affect subsequent budburst?
The spring emergence of many tree-defoliating insects is closely synchronized with budburst of their host trees (Satchell 1961, Nothnagle and Schultz 1987, Hunter 1991, Yela and Herrera 1993). The time of budburst of a tree is genetically programmed to local climatic conditions (Nienstaedt and King 1969, Owens et al. 1977, Billington and Pelham 1991), as is the emergence of insects feeding on young foliage, which ensures synchrony with budburst (e.g., Wellington 1957). Such synchronization is critical for these insects because the suitability of foliage as food decreases rapidly after budburst (Feeny 1970, Schweitzer 1979, Schneider 1980, Raupp and Denno 1983, Quiring 1992). As a result, trees bursting earlier (Crawley and Akhteruzzaman 1988) or later (Embree 1965, DuMerle 1988, Quiring 1994) than conspecifics are often subjected to lower levels of defoliation by these insects.
It often happens, however, that the time of budburst of a tree defoliated by insects occurs later or sometimes earlier in the following spring than for nearby undefoliated trees (Benz 1974; cited in Gradwell , Heichel and Turner , and Tuomi et al. ). When this happens, the synchrony between budburst and insect emergence is disrupted, and the feeding activity of insects on such trees may be adversely affected. This can result in a reduced level of defoliation on that tree in that year.
Tuomi et al. (1989) discussed two explanations for defoliation-induced delays in budburst. Delayed budburst may be (1) an adaptive, induced defensive response of the tree to reduce further herbivory by early-season defoliators by escaping in time; or (2) a mere physiological consequence of reductions in resources resulting from previous defoliation (resource depletion hypothesis, hereafter referred to as the resource availability hypothesis).
The resource availability hypothesis (RAH) is based on the premise that the date of budburst is inversely related to the quantity of nutrient reserves in the plant. In uneaten shoots, photoassimilates are accumulated during summer and are transferred directly to buds or stored in leaves or stems for bud development and leaf growth in the following spring (Durzan 1968a, b, Tuomi et al. 1989). Consequently, removal of leaf and stem tissue should reduce carbon and mineral reserves for bud development. Bud expansion and subsequent budburst result from cell division and cell enlargement (Kozlowski 1971). Rates of cell division and enlargement are often positively associated with foliar content of water and other nutrients (Zahner 1968, Kozlowski et al. 1991). Tuomi et al. (1989) observed that if a branch of mountain birch, Betula pubescens Ehrh., was defoliated early in the previous season, the growth of leaves on the same branch in the following growing season was much retarded. They suggested that defoliation depleted nutrient reserves on the branch. This explains why leaf growth on the same branch was retarded the following season. However, it does not explain why the time of budburst is sometimes advanced or unaffected, rather than delayed (Bryant et al. 1991).
Such inconsistent effects of herbivory on the time of budburst during the next growing season might be due to changes in branching pattern (i.e., "crown architecture," sensu Halle et al. 1978) as well as to changes in resource availability. In some trees, different types of buds burst at different times (e.g., Powell 1988, Quiring 1993). The crown architecture hypothesis (CAH) states that the mean time of budburst could be advanced or delayed if feeding by insects were to result in a change in the relative proportion of different bud types within such trees, due to bud death or production. Theoretically, the mean time of budburst for any one bud type could be delayed following defoliation due to reductions in the amount of resources available, but the mean time of budburst for a branch could be advanced if there were a sufficiently large increase in the relative proportion of a bud type that bursts earlier. According to both the CAH and RAH, changes in the time of budburst following herbivory may, or may not, reduce subsequent herbivory.
The influence of herbivory on crown architecture and the amount of resources available will vary according to the type and amount of feeding by herbivores (Whitham and Mopper 1985, Haukioja et al. 1990, Mopper et al. 1991, Honkanen and Haukioja 1994, Honkanen et al. 1994, Karban and Niiho 1995). Thus, it would be advantageous to test these hypotheses using insects that feed differently and have different effects on plant architecture.
We examined the RAH and CAH for white spruce, Picea glauca (Moench) Voss, by comparing budburst on unexploited shoots with that on adjacent shoots bearing a stem galler or on adjacent shoots or branches that had been subjected to one of two levels of herbivory by a defoliator or by a defoliator/stem feeder in the previous year. In addition, we evaluated the influence of resource availability on the time of budburst, in the absence of herbivory, in a manipulated field study whereby tree growth was reduced through root pruning and increased through fertilization. Different experiments evaluated responses at the level of shoots, branches, and whole trees.
Description of organisms and predictions
The three insects studied were a defoliator, the spruce budworm, Choristoneura fumiferana (Clem.); a defoliator/stem feeder, the spruce bud moth, Zeiraphera canadensis (Mut. and Free.); and a galler, the spruce gall adelgid, Adelges abietis (L.). All three species are commonly found on young Picea glauca in eastern Canada (Martineau 1984). If newly emerged second-instar larvae of C. fumiferana do not locate newly burst buds, they disperse by silking, an activity associated with high mortality (Morris and Mort 1963, Royama 1984), or they mine old needles until new buds become available (Lawrence et al. 1997). First-instar larvae of Z. canadensis emerge in spring and usually can only successfully colonize buds that burst [less than or equal to]5 d earlier (Quiring 1992). Adelges abietis, which completes two generations per year, overwinters as a nymph and forms a pineapple-shaped gall on developing shoots of spruce trees in spring (Martineau 1984). Because the survival of all three species is linked to tree phenology (Bishoff et al. 1969, Eidt and Cameron 1971, Quiring 1994, Lawrence et al. 1997), delays in budburst following feeding by one species may result in reduced herbivory by all three species.
Feeding by larvae of Z. canadensis can cause changes in crown architecture. These larvae consume stem tissue as well as leaves, resulting in destruction of distal portions of shoots when larval densities are high (Carroll and Quiring 1993). Such partial shoot destruction often stimulates the production of basal buds [ILLUSTRATION FOR FIGURE 1 OMITTED]. In open-grown P. glauca, budburst is acropetal: buds in the lower crown burst before those in the upper crown, and within a shoot, medial-lateral buds burst first, followed by distal-lateral and then terminal buds (Quiring 1993). Basal buds are more proximally located than medial-lateral buds and thus should burst earlier. If basal buds burst earlier than other buds on a shoot, partial shoot destruction could, according to the CAH, result in an earlier mean time of budburst at the shoot level. Predictions emanating from the two hypotheses are listed in Table 1.
Effect of defoliation by C. fumiferana
The effects of defoliation by Choristoneura fumiferana on subsequent budburst were studied using 11-yr-old white spruce trees in an unmanaged plantation at Tracy, New Brunswick, Canada (45 [degrees] 33 [minutes] N, 66 [degrees] 55 [minutes] W). This site was chosen because natural populations of C. fumiferana and all other herbivores were very low (i.e., no more than two shoots defoliated per tree). In spring, 1990, second-instar larvae, which overwintered in diapause, [TABULAR DATA FOR TABLE 1 OMITTED] were obtained from a laboratory colony at the Forest Pest Management Institute, Sault Sainte Marie, Ontario, Canada, and were stored in the dark at 4 [degrees] C. When buds began to burst at the study plot, larvae were placed in an environmentally controlled chamber at 20 [degrees] [+ or -] 1 [degrees] C and 70 [+ or -] 5% relative humidity under a photoperiod of 14L:10D. Newly emerged ([less than] 12-h-old) larvae were then placed on four midcrown branches inside sleeve cages made of fine mesh on each of 10 trees. To obtain two defoliation levels of [approximately]50% and 95%, on five of the 10 trees we placed one larva for every four buds, and on the other five trees we placed one larva per bud. After all individuals had pupated, sleeve cages and pupae were removed from branches. In an independent study with similarly aged, open-grown P. glauca (Carroll and Quiring 1994), sleeve cages did not influence shoot and bud development.
In early spring, 1991, we recorded the numbers of burst and unburst buds as well as the phenological stages of all buds on the terminal shoot of each defoliated and adjacent undefoliated branch. An adjacent branch to the right or the left of the defoliated branch was chosen by flipping a coin. Different phenological stages of foliage were categorized as described in Quiring (1992): 1, bud caps tightly attached; 1.5, bud caps attached but small portions of bud are covered only by a thin transparent membrane; 2, bud beginning to burst with [less than]35% of needles visible; 3, 36-65% of needles visible but shoot not flaring; and 4, 66-100% of needles visible on flaring shoot. We measured leaf length and width in October 1991, well after leaf growth had finished, and used their product as an index of the area of 20 needles per shoot. In this, as well as all other experiments carried out in 1991 and 1994, variations in leaf area were attributable to variations in leaf length rather than leaf width ([R.sup.2] [greater than or equal to] 0.90 for regressions between leaf area and leaf length).
Variations in the mean stage of bud development and mean proportion of buds open were not influenced by tree (P [greater than] 0.05, Kruskal-Wallis test). We used the Wilcoxon signed ranks test to compare the mean stage of bud development and mean proportion of buds open for 20 terminal shoots that were undefoliated to those on 20 terminal shoots on adjacent branches that had been subjected to 40-60% defoliation (i.e., one terminal shoot per branch x four pairs of branches per tree x five trees = 20 pairs of shoots). Similarly, the Wilcoxon signed-ranks test was used to compare bud development on 20 undefoliated shoots to 20 shoots on which the budworm had removed [greater than or equal to]90% of the foliage.
Effect of feeding by Z. canadensis and A. abietis
Effects of feeding by moderate-to-high densities of Zeiraphera canadensis and Adelges abietis on Picea glauca were evaluated at a seed orchard at Pokiok, New Brunswick, Canada (46 [degrees] 7 [minutes] N, 67 [degrees] 15 [minutes] W) in 1991 and at Queensbury, a seed orchard located [approximately]15 km east of Pokiok, in 1994. The study was not repeated at Pokiok in 1994 because low densities of both insects precluded adequate replication for all treatments. At Pokiok, 12-yr-old trees were spaced 2 m apart within rows separated by 4 m of grass (Quiring 1993). At Queensbury, 11-yr-old trees were spaced 3 m apart within rows separated by 6 m of grass. Before budburst in spring, we located one undefoliated distal-lateral shoot and an adjacent distal-lateral shoot (2-4 distal-lateral shoots arise from the distal end of a midcrown branch; [ILLUSTRATION FOR FIGURE 1 OMITTED]) on which: 20-40% of the foliage had been naturally removed by Z. canadensis in the previous year, but the stem was not broken (four pairs in each of 10 trees); feeding by Z. canadensis had resulted in stem breakage and destruction of [greater than]50% of the shoot (two pairs in each of 10 trees); or a gall of A. abietis was present (four pairs in each of 10 trees). Because it was difficult to locate unexploited distal-lateral shoots adjacent to ones partially destroyed by Z. canadensis, bud development and leaf size on a broken shoot (i.e., our second category of herbivory) was compared to that on an adjacent unbroken distal-lateral shoot with [less than]20% defoliation. To protect shoots from herbivory, we smeared vaseline on scales at the base of each shoot, where Z. canadensis eggs overwinter, before budburst. In previous studies, this procedure did not influence subsequent bud development (Quiring 1993, 1994). Bud phenology (1991) and final leaf size (both years) were measured as previously described. In 1994, however, buds were examined daily and the date when each bud reached stage 1.5 was recorded as the date of budburst. The mean proportion of open buds and the mean phenological stage of buds were not influenced by tree in 1991 (P [greater than] 0.05, Kruskal-Wallis test). Thus, the Wilcoxon signed ranks test was used to determine whether bud phenology was influenced by herbivory in 1991. Mean contrasts were made to compare bud phenology in 1994 and final leaf size on exploited and unexploited shoots in both years and were tested using ANOVAs. Wilcoxon signed-ranks tests (1991) and paired t tests (1994) were used to compare the phenological stage and date of budburst for each bud type on shoots partially destroyed by Z. canadensis to those for adjacent control shoots.
Influence of root pruning and fertilization
To determine if variations in resource availability, in the absence of herbivory, could explain differences in the time of budburst, we measured the phenological development of buds of Picea glauca on unexploited shoots of young open-grown trees that had been subjected to root pruning or fertilization in a managed stand in northern New Brunswick, Canada (47 [degrees] 25 [minutes] N, 67 [degrees] 51 [minutes] W). All trees in the stand originated from the same provenance, thereby reducing genetically based variation among trees, and were planted in 1984 with a 2 x 2 m spacing. The site was on level terrain and had a consistent soil profile. Thirty plots consisting of 15 trees each were chosen systematically within an area of [approximately]100 X 120 m near the center of the stand, and five plots were randomly assigned to each of six treatments. Each plot contained six experimental trees: nine "buffer" trees were located between experimental trees to reduce interactions among experimental trees. Manipulations and measurements were only carried out on experimental trees, with the exception of fertilization. Fertilizer was applied to all trees in a plot to compensate for the possibility of overlapping root systems of adjacent trees. Any two adjacent plots were separated by two rows of buffer trees to reduce the possibility that trees in one plot would be influenced by treatments in neighboring plots.
Tree growth rate and foliar chemistry were manipulated during spring 1992-1994 (McKinnon et al. 1998). In 1992, trees were treated with: one of three levels (equivalent to 100, 200, or 300 kg N/ha) of ammonium nitrate 34-0-0 (NPK) fertilizer; or one of two levels of root pruning, a linear trench dug 0.75 m (severe pruning, R1) or 1.50 m (light pruning, R2) from the tree base; or were left untreated (controls). Root pruning was carried out by digging a trench on the east and west sides of trees in 1992 and 1994, and on the north and south sides in 1993. The two parallel trenches were dug to a depth of about 25 cm and extended to the edge of the crown ([approximately]2 m long) of each tree. A plastic sheet was placed between cut roots inside trenches to prevent root reconnection. To avoid reductions in tree growth resulting from excessive fertilization, fertilization levels were reduced to 25, 50, and 75 kg N/ha in 1993, and then raised to 50, 100, and 150 kg N/ha in 1994.
The effects of the treatments on tree growth and foliar chemistry are described in McKinnon et al. (1998). Briefly, aspect did not influence shoot growth or foliar chemistry, with the exception of foliar N content, which was lower in south-facing branches. However, fertilization and root pruning resulted in increases and decreases, respectively, in shoot length, tree volume, foliar N, and foliar water content, producing a continuum of growth rates and foliar nutrient status.
To estimate the response, at the whole-tree level, to variation in resource availability, we recorded the developmental stage of buds on an unexploited, south-facing, terminal shoot on the second and sixth whorls of 5-6 trees per plot when [approximately]80% of buds had burst in spring 1993. A similar procedure was carried out in 1994, except that 4-5 trees per plot were observed daily and the date of budburst was recorded for all buds. Final length of an unexploited terminal shoot was measured on whorls 2 and 6 of each tree in fall 1993 and 1994.
Nitrogen and water are the nutrients that most often limit growth of woody plants (Larcher 1980). Foliar water and nitrogen content were determined by collecting current-year distal-lateral shoots [ILLUSTRATION FOR FIGURE 1 OMITTED] from five trees per plot in late summer of 1992 and 1993, and in early spring of 1993 and 1994. One unexploited shoot was collected from a second whorl branch in 1992, and from one second and one sixth whorl branch in 1993 and 1994. Foliage was collected shortly after midday only on clear, sunny days, when effects of water stress are more prominent (Kramer and Kozlowski 1979, Louda and Collinge 1992). Although the buds collected in 1992 and 1993 were approximately in stage 4 (Quiring 1992) of development, cloudy weather in 1994 delayed foliage collection until shoots were 710 cm long. Each sample, consisting of shoots grouped by plot and whorl, was sealed in a plastic bag and transported in ice. Water content for all years was determined by oven-drying foliage at 60 [degrees] C for 72 h and subtracting dry from fresh mass. In 1992 and 1993, total N was determined by colorimetry on a TrAAcs 800 analyzer (Bremner and Mulvaney 1982) using 100 mg of the dried foliage digested in a sulfuric acid and hydrogen peroxide solution. In 1994, 500 mg of the dried foliage was digested in concentrated sulfuric acid, and total N content was determined in a full-injection analyzer (Parkinson and Allan 1975, Hansen et al. 1977).
Trees assigned to the different treatments were of similar size and had similar growth rates before manipulations were carried out (McKinnon et al. 1998). Measurements were pooled to produce plot means to avoid pseudoreplication (Hurlbert 1984). The influences of treatment and whorl on mean bud stage (1993) and mean date of budburst (1994) per shoot were evaluated using nonparametric and parametric two-way ANOVAs for fixed effects (Zar 1984), respectively. When analyses indicated that treatment was significant (P [less than or equal to] 0.05), we compared treatments to the control using a nonparametric test (Rhyne and Steel 1965, Daniel 1978) for 1993, or Dunnett's test (Zar 1984) for 1994. One-tailed correlations between the mean bud stage (Spearman) or mean date of budburst (Pearson) and the mean length of unexploited terminal shoots were carried out to determine if longer shoots burst earlier. Similarly, correlations between the mean nitrogen and water content of foliage for each whorl and the mean bud stage (Spearman) or mean date of budburst (Pearson) were carried out to determine if shoots with more of these two resources burst earlier.
Approximately three-fourths of buds at Pokiok and almost all buds at Tracy had burst when bud phenology was examined in 1991. As a result, differences between exploited and unexploited shoots are generally more marked for phenological stage than for the proportion of open buds (Table 2).
Effect of defoliation by C. fumiferana
There were no significant differences in the proportion of open buds or bud phenological stage between undefoliated shoots and those subjected to 40-60% defoliation by Choristoneura fumiferana (Table 2). Budburst was significantly delayed in shoots subjected to [greater than]90% defoliation (Table 2). Surface areas of leaves on partially and almost completely defoliated shoots were 6.9% and 31.6% smaller than those on undefoliated shoots of adjacent branches (Table 3). There was significant variation among trees in leaf area (Table 3).
Effect of feeding by Z. canadensis and A. abietis
Relatively low levels of defoliation (20-40%) by Zeiraphera canadensis did not significantly affect the time of budburst in the following spring (Tables 2 and 4). However, the proportion of open buds and the phenological stage of buds were higher for shoots subjected to [greater than]50% defoliation (with stems partially destroyed by Z. canadensis in the previous season) than for shoots subjected to [less than]20% defoliation (Table 2). Advancement in the mean time of budburst of shoots [TABULAR DATA FOR TABLE 2 OMITTED] [TABULAR DATA FOR TABLE 3 OMITTED] [TABULAR DATA FOR TABLE 4 OMITTED] that were partially destroyed in the previous year was due to the production of basal buds, which burst before all other buds. The mean phenological stages of terminal, distal-lateral, and medial-lateral buds were lowest on shoots with stems that had been partially destroyed [ILLUSTRATION FOR FIGURE 2a OMITTED], indicating that they burst after buds on adjacent shoots. A similar situation occurred in 1994 [ILLUSTRATION FOR FIGURE 2b OMITTED], but in this year, delays in the budburst of terminal, distal-lateral, and lateral buds on previously exploited shoots were offset by the earlier budburst of basal buds, resulting in the same mean date of budburst as for buds on shoots with [less than]20% defoliation (Table 4). The areas of leaves on shoots previously subjected to 20-40% defoliation by Z. canadensis were 5.4% and 8.9% smaller than controls in 1991 and 1994, respectively (Table 3). Reductions in leaf area were 20.4% and 27.5% on shoots partially destroyed by Z. canadensis larvae in 1991 and 1994 (Table 3). A significant proportion of the variation in leaf area, but not in bud phenology, was attributable to the individual tree in some cases (Tables 3 and 4).
Gall formation by Adelges abietis resulted in a small delay in budburst in the following year (Tables 2 and 4). Leaves developing on shoots that had been galled in the previous year had areas 5.0% (1991) and 8.1% (1994) smaller than those of leaves on adjacent shoots without galls (Table 3). Variations in time of budburst and final leaf size on galled and nongalled shoots were not attributable to individual trees (Tables 2-4).
Influence of root pruning and fertilization
Time of budburst per shoot was influenced by both whorl and treatment in 1993 ([F.sub.1,48] = 14.08 and [F.sub.5,48] = 6.13, respectively; P [less than or equal to] 0.0005) and in 1994 ([F.sub.1,48] = 5.50 and [F.sub.5,48] = 3.76, respectively; P [less than or equal to] 0.0232), but not by their interaction in either year (P [greater than or equal to] 0.9010). Bud development was acropetal; buds on whorl 6 developed sooner than those on whorl 2 [ILLUSTRATION FOR FIGURE 3a, b OMITTED]. In general, fertilization resulted in earlier bud development, whereas root pruning delayed bud development [ILLUSTRATION FOR FIGURE 3c, d OMITTED]. Although a post hoc test did not show any significant differences between the control and treatments in 1993, there was a positive correlation (Spearman) between treatment (ordered in terms of growth rate) and the mean phenological stage of trees per plot (r = 0.64, P = 0.0002, n = 30) or per treatment (r = 0.94, P = 0.0048, n = 6). In 1994, due to warm temperatures, budburst occurred during a shorter period than in 1993 and differences between treatments were less marked. Dunnett's test indicated that only the date of budburst of severely root-pruned trees differed significantly from those of fertilized and control trees. Pearson correlations between treatment and mean date of budburst per plot (r = -0.47, P = 0.3496, n = 6) or per treatment (r = -0.23, P = 0.2129, n = 30) were not significant.
Examination of relationships between bud phenology and foliar nutrient content or shoot length also indicated that the time of budburst is directly related to the amount of resources available to trees. Bud stage was positively related and mean date of budburst was negatively related to the mean length of unexploited shoots of trees in 1993 (r = 0.55, P = 0.0010, n = 30) and 1994 (r = -0.46, P = 0.0042, n = 30). The time of budburst was generally more closely related to foliar nutrient content in the previous fall than it was to foliar chemistry during shoot elongation. In 1993, bud stage was positively related to the foliar N content of foliage in fall 1992 (r = 0.58, P = 0.0005, n = 30); influence of the water content of foliage collected in fall 1992 was not significant, although water content of foliage collected in spring (r = 0.74, P [less than] 0.0001) 1993 was, and the mean date of budburst in 1994 was negatively related to the N and water content of foliage collected in fall 1993 (r = -0.32, P = 0.0412, n = 29; and r = -0.32, P = 0.0319, n = 29, respectively).
Watson (1995) emphasized that studies that focus on the timing of plants' developmental phenologies relative to the timing of herbivore attack are needed to produce predictive theories for insect-plant interactions. Our study shows that herbivores feeding on newly burst buds can cause advances, delays, or no effects on the timing of budburst in the next growing season, depending on their mode, and the amount, of feeding. Our study also indicates that the parameter most often used to explain herbivore performance, plant nutrient status, is correlated with plant phenology, a factor that is often overlooked in field studies.
Changes in the time of budburst of white spruce in the spring following feeding by insect herbivores were attributable to changes in crown architecture and resource availability, a result replicated in time and space. In the absence of changes in crown architecture, delays in budburst were directly related to the amount of biomass lost to herbivores. Moderate levels of defoliation by Zeiraphera canadensis and Choristoneura fumiferana did not result in delayed budburst in the following year, but complete defoliation by the latter did. A concomitant reduction in leaf size, which was directly related to the amount of previous herbivory, also supports the resource availability hypothesis (RAH).
The RAH correctly predicted the inverse relationship between the time of budburst and the amount of resources available for bud development, estimated by shoot length and the nitrogen and water content of foliage, observed in experiments that manipulated the amount of resources available to trees. Our conclusion that the time of budburst is inversely related to resource availability is supported by results of a laboratory study on young Douglas-fir seedlings, in which Lerdau et al. (1995) noted that plants subjected to the highest and lowest fertilization regimes burst first and last, respectively.
Partial destruction of Picea glauca shoots by Z. canadensis resulted in a similar (1994) or earlier (1991) mean time of budburst in the following year, because the production of basal buds compensated (in 1994) or more than compensated (in 1991) for the effect of depleted resources. If fewer basal buds had been produced, the mean time of budburst on partially destroyed shoots might have been later than that of shoots subjected to [less than]20% defoliation. Thus, feeding by Z. canadensis larvae had opposite effects on the time of budburst of white spruce shoots in the following spring, depending on whether or not basal buds were produced. These results are similar to recent observations that feeding on different tissues produces opposite results in terms of subsequent plant susceptibility to herbivores (Haukioja et al. 1990, Karban and Niiho 1995).
Defoliation can alter the architecture of shoot systems (Whitham and Mopper 1985) and influence intra-tree variability in leaf and shoot growth (Danell et al. 1985, Tuomi et al. 1989, Mopper et al. 1991, Honkanen and Haukioja 1994, Honkanen et al. 1994). Consequently, Tuomi et al. (1990) suggest that it might be useful to distinguish between plant responses resulting from "developmental shifts" related to alternative developmental programs and those resulting from factors that modify the "expression" of a given program. According to this terminology, the variation in budburst due to the induction of basal bud production could result primarily from a developmental shift. Conversely, the effects of resource availability on budburst and leaf size of a given bud type may represent a modification of the expression of the developmental program of that bud type.
The present study with P. glauca reemphasizes the importance of the modular structure of plants (Harper 1977, Tuomi et al. 1983, Halle 1986, Haukioja et al. 1990, Sprugel et al. 1991) on insect-plant interactions (Whitham and Slobodchikoff 1981). In P. glauca, the effect of localized resource deficiencies caused by C. fumiferana was evident at the branch level, with respect to both time of budburst and leaf size. A similar phenomenon occurred at the shoot level on trees exploited by Z. canadensis and Adelges abietis, where adjacent shoots subjected to different levels of herbivory had different phenological responses in the next growing season.
Commonly, different shoots within the crown of an open-grown P. glauca tree are subjected to various levels of defoliation by Z. canadensis and several other defoliators, partial shoot destruction by Z. canadensis, and galling by A. abietis and several other gallers (Martineau 1984). Because changes in resource availability or in crown architecture caused by such herbivory are localized, feeding by these insects should increase intra-tree heterogeneity in the time of budburst in the following spring. The period of time when suitable foliage is available on a tree is increased in the year following herbivory because budburst will be delayed or advanced on exploited branches or shoots, but not on adjacent unexploited ones. Results from our fertilization study, combined with those from a recent study investigating the influence of herbivory on source-sink relationships in another conifer (Honkanen et al. 1994), indicate that increases in intra-tree heterogeneity in the time of budburst in the year following herbivory may be greater than our data indicate. Honkanen et al. (1994) demonstrated that defoliation can result in increased and decreased growth in pine shoots below and above, respectively, the one defoliated. Thus, it is possible that some unexploited shoots below those subjected to herbivory will be larger and will burst earlier in the following year.
Our study showed that the combined effects of resource depletion and changes in crown architecture can also increase the length of the phenological window for bud colonization within individual shoots. When the distal portion of shoots was destroyed by Z. canadensis, buds were suitable for colonization both earlier and later during the next spring, than were buds on unexploited adjacent shoots, due to the production of basal buds and to delays in bursting of lateral, distal-lateral, and terminal buds, respectively.
The extent to which changes in the time of budburst observed in the present study influence the susceptibility of P. glauca shoots to the three early-season herbivores in subsequent years remains to be determined. Increases in intra-tree heterogeneity in the time of budburst can increase and decrease tree susceptibility to herbivory by mobile as opposed to immobile herbivores, respectively (Tuomi et al. 1989, Quiring 1993, Carroll and Quiring 1994). Due to larval mobility, we believe that increases in the length of the phenological window during the season following high levels of feeding by Z. canadensis or C. fumiferana will make P. glauca more, rather than less, susceptible to both of these herbivores. Partial stem destruction may render white spruce particularly susceptible to Z. canadensis, not just because of an increase in the length of the phenological window for colonization of buds within trees and within individual shoots, but also because of reduced dispersal costs, due to the location of basal buds next to the scales in which eggs are laid. In the field, bud moth females lay more eggs on heavily exploited trees (Quiring and Butterworth 1994), further supporting this hypothesis. A test of this hypothesis will be reported elsewhere.
In our study, feeding by the only immobile herbivore, overwintering nymphs of A. abietis, which do not move after inserting their stylers in the host plant in summer, produced the smallest response by the tree, in terms of delayed budburst. Although statistically significant, the delay in budburst of gall-bearing shoots was so small that it probably has little effect on subsequent gall formation. The very small effect of galling on budburst may be due to the galls' ability to act as a sink, thereby potentially drawing resources from nearby shoots (Larson and Whitham 1991).
The influence of previous herbivory on the phenology of host plants in the following growing season is a commonly overlooked factor that may have a significant effect on variations in herbivore abundance and distribution in space and time in other systems. For example, many phytophagous insects cause extensive defoliation during outbreaks (Barbosa and Schultz 1987). If such defoliation does not influence bud production, the quantity and/or quality of foliage not only may be reduced in subsequent growing seasons, as previously reported (e.g., Baltensweiler et al. 1977), but also young foliage may only be available at a relatively later date.
In conclusion, changes in resource availability and crown architecture can explain advances and delays in budburst following either herbivory or nonherbivore-mediated changes in resource availabilityu The generality of these hypotheses awaits further studies that will investigate within-tree variations in budburst at the bud levelu
We thank J. Meating for permission to use his study plot at Tracy; the Forest Pest Management Institute for providing spruce budworm larvae; E. Bauce and H. Kfause for help with foliar analyses; J. Kershaw for statistical advice; G. Powell and T Royama for helpful discussion; and C. Cloutier, T Hoffmeister, A. Hunter, J. Kershaw, R. McGregor, R. Monson, Y. Pelletier, K. Raffa, M. Roberts, B. Roitberg, T Royama, S. Sopow, J. Sweeney, J. Tuomi, T Whitham, and two anonymous reviewers for comments on an earlier version of this manuscriptu The cooperation of the Department of Natural Resources and Energy of New Brunswick and the Canadian Forest Service is also gratefully acknowledgedu This project was funded by a research grant from the National Science and Engineering Research Council of Canadau
Baltensweiler, W., G. Benz, P. Bovey, and V. Delucchi. 1977. Dynamics of larch bud moth populationsu Annual Review of Entomology 22:79-100.
Barbosa, P., and J. C. Schultz. 1987. Insect outbreaks. Academic Press, Toronto, Ontario, Canada.
Benz, G. 1974. Negative Ruckkoppelung dutch Raum- und Nahrungskonkurrenz sowie zyklische Veranderung der Nahrungsgrundlage als Regelprinzip in der Populationsdynamik des Grauen Larchenwicklers, Zeiraphera diniana (Guenee) (Lep., Tortricidae). Zeitschrift fur angewandte entomologie 76:196-228.
Billington, H. L., and J. Pelham. 1991. Genetic variation in the date of budhurst in Scottish birch populations: implications for climate change. Functional Ecology 5:403-409.
Bishoff, M., S. B. Ewert, and W. Thalenhorst. 1969. Untersuchungenuber die Abhangigkeit der Befallsstarke der Gallenlaus Sacchiphantes abietis (L.) vom Austreibetyp der Fichte. Zeitschrift fur angewandte entomologie 64:65-85.
Bremner, J. M., and C. S. Mulvaney. 1982. Nitrogen-total. Pages 595-624 in A. L. Page, editor. Methods of soil analysis. Part 2. Second edition. Agronomy Monographs Number 9. American Society of Agronomists, Madison, Wisconsin, USA.
Bryant, J. P., I. Heitkonig, P. Kuropat, and N. Owen-Smith. 1991. Effects of severe defoliation on the long-term resistance to insect attack and on leaf chemistry in six woody species of the southern African savannau American Naturalist 137:50-63.
Carroll, A. L., and D. T. Quiring. 1993. Influence of feeding by Zeiraphera canadensis (Lepidoptera: Tortricidae) on growth of white spruce: larval density-damage and damage-shoot production relationships. Journal of Applied Ecology 30:629-639.
Carroll, A. L., and D. T. Quiring. 1994. Intra-tree variation in foliage development influences the foraging strategy of an early season folivore. Ecology 75:1978-1990.
Crawley, M. J., and M. Akhteruzzaman. 1988. Individual variation in the phenology of oak trees and its consequences for herbivorous insectsu Functional Ecology 2:409-415.
Danell, K., K. Huss-Dannell, and R. Bergstrom. 1985. Interactions between browsing moose and two species of birch in Sweden. Ecology 66:1867-1878.
Daniel, W. W. 1978. Applied nonparametric statistics. PWS-Kent Publishing, Boston, Massachusetts, USA.
DuMerle, P. 1988. Phenological resistance of oaks to the green oak leafroller, Tottrix viridana (Lepidoptera: Tortricidae). Pages 215-226 in W. J. Mattson, J. Levieux, and G. Bernard-Dagan, editors. Mechanisms of woody plant defenses against insects. Springer-Verlag, New York, New York, USA.
Durzan, D. J. 1968a. Nitrogen metabolism of Picea glauca. I. Seasonal changes of free amino acids in buds, shoot apices, and leaves, and the metabolism of uniformly labelled 14C-l-arginine by buds during the onset of dormancy. Canadian Journal of Botany 46:909-919.
-----. 1968b. Nitrogen metabolism of Picea glauca. II. Diurnal changes of free amino acids, amides, and guanidino compounds in roots, buds, and leaves during the onset of dormancy of white spruce saplings. Canadian Journal of Botany 46:9021-9928.
Eidt, D.C., and M.D. Cameron. 1971. Delayed budbreak and spruce budworm survival. Bimonthly Research Notes 27:28-29.
Embree, D. G. 1965. The population dynamics of the winter moth in Nova Scotia, 1954-1962. Memoirs of the Entomological Society of Canada 46:1-57.
Feeny, P. 1970. Seasonal changes in oak leaf tannins and nutrients as a cause of spring feeding by winter moth caterpillar. Ecology 51:565-581.
Gradwell, G. 1974. The effect of defoliators on tree growth. Pages 182-193 in M. G. Morris and F. H. Perring, editors. The British oak. Classey, Farringdon, UK.
Halle, F. 1986. Modular growth in seed plants. Philisophical Transactions of the Royal Society of London B 313:77-87.
Halle, F., A. S. Oldeman, and P. B. Tomlinson. 1978. Tropical trees and forests: an architectural analysis. Springer-Verlag, New York, New York, USA.
Hansen, E. H., F. J. Krug, A. K. Ghose, and J. Ruzicka. 1977. Rapid determination of nitrogen, phosphorus and potassium in fertilizers by flow injection analysis. Analyst 102: 714-722.
Harper, J. L. 1977. Population biology of plants. Academic Press, New York, New York, USA.
Haukioja, E., K. Ruohomaki, J. Senn, J. Suomela, and M. Walls. 1990. Consequences of herbivory in the mountain birch (Betula pubescens ssp. tortuosa): importance of the functional organization of the tree. Oecologia 82:238-247.
Heichel, G. H., and N. C. Turner. 1976. Phenology and leaf growth of defoliated hardwood trees. Pages 31-40 in J. F. Anderson and H. K. Kaya, editors. Perspectives in forest entomology. Academic Press, London, UK.
Honkanen, T., and E. Haukioja. 1994. Why does a branch suffer more after branch-wide than after tree-wide defoliation? Oikos 71:441-450.
Honkanen, T., E. Haukioja, and J. Suomela. 1994. Effects of simulated defoliation and debudding on needle and shoot growth in Scots pine (Pinus sylvestris): implications of plant source/sink relationships for plant-herbivore studies. Functional Ecology 8:631-639.
Hunter, A. 1991. Traits that distinguish outbreaking and non-outbreaking macrolepidoptera feeding on northern hardwood trees. Oikos 60:275-282.
Hurlbert, S. H. 1984. Pseudoreplication and the design of ecological field experiments. Ecological Monographs 54: 187-211.
Karban, R., and C. Niiho. 1995. Induced resistance and susceptibility to herbivory: plant memory and altered plant development. Ecology 76:1220-1225.
Kozlowski, T. T. 1971. Growth and development of trees. Volume I. Seed germination, ontogeny and shoot growth. Academic Press, New York, New York, USA.
Kozlowski, T. T., P. J. Kramer, and S. G. Pallardy. 1991. The physiological ecology of woody plants. Academic Press, Toronto, Ontario, Canada.
Kramer, P. J., and T. T. Kozlowski. 1979. Physiology of woody plants. Academic Press, New York, New York, USA.
Larcher, W. 1980. Physiological plant ecology. Springer-Verlag, Berlin, Germany.
Larson, K. C., and T. G. Whitham. 1991. Manipulation of food resources by a gall-forming aphid: the physiology of sink-source interactions. Oecologia 88:15-21.
Lawrence, R. K., W. J. Mattson, and R. A. Haack. 1997. White spruce and the spruce budworm: defining the phenological window of susceptibility. Canadian Entomologist 129:291-318.
Lerdau, M., P. Matson, R. Fall, and R. Monson. 1995. Ecological controls over monoterpene emissions from Douglasfir (Pseudotsuga menziesii). Ecology 76:2640-2647.
Louda, S. M., and S. K. Collinge. 1992. Plant resistance to insect herbivores: a field test of the environmental stress hypothesis. Ecology 73:153-169.
Martineau, R. 1984. Insects harmful to forest trees. Ministry of Supply and Services Canada, Ottawa, Ontario, Canada.
McKinnon, M. L., D. T. Quiring, and E. Bauce. 1998. Influence of resource availability on growth and foliar chemistry within and among young white spruce trees. Ecoscience 5:295-305.
Mopper, S., J. Maschinske, N. Cobb, and T. G Whitham. 1991. A new look at habitat structure: consequences of herbivore-modified plant architecture. Pages 260-280 in S. S. Bell, E. D. McCoy, and H. R. Mushinsky, editors. Habitat structure. Chapman and Hall, London, UK.
Morris, R. F., and D. G. Mott. 1963. Dispersal and the spruce budworm. Pages 180-187 in R. F. Morris, editor. The dynamics of epidemic spruce budworm populations. Memoirs of the Entomological Society of Canada, Volume 31.
Nienstaedt, H., and J. P. King. 1969. Breeding for delayed budbreak in Picea glauca (Moench) Voss: potential frost avoidance and growth gains. Pages 61-80 in Proceedings: Second World Consultation on Forest Tree Breeding. Food and Agriculture Organization of the United Nations, Rome, Italy.
Nothnagle, P. J., and J. C. Schultz. 1987. What is a forest pest? Pages 59-80 in P. Barbosa and J. C. Schultz, editors. Insect outbreaks. Academic Press, San Diego, California, USA.
Owens, J. N., M. Molder, and H. Langer. 1977. Bud development in Picea glauca. I. Annual growth cycle of vegetative buds and shoot elongation as they relate to date and temperature sums. Canadian Journal of Botany 55:2728-2745.
Parkinson, J. A., and S. E. Allen. 1975. A wet oxidation procedure suitable for the determination of nitrogen and mineral nutrients in biological material. Communications in Soil Science and Plant Analyses 6:1-11.
Powell, G. R. 1988. Shoot elongation, leaf demography and bud formation in relation to branch position on Larix laricina saplings. Trends in Ecology and Evolution 2:150-164.
Quiring, D. T. 1992. Rapid change in suitability of white spruce for a specialist herbivore, Zeiraphera canadensis, as a function of leaf age. Canadian Journal of Zoology 70: 2132-2138.
-----. 1993. Influence of intra-tree variation in time of budburst of white spruce on herbivory and the behaviour and survivorship of Zeiraphera canadensis. Ecological Entomology 18:353-364.
-----. 1994. Influence of inter-tree variation in time of budburst of white spruce on herbivory and the behaviour and survivorship of Zeiraphera canadensis. Ecological Entomology 19:17-25.
Quiring, D. T., and E. W. Butterworth. 1994. Genotype and environment interact to influence acceptability and suitability of white spruce for a specialist herbivore, Zeiraphera canadensis. Ecological Entomology 19:230-238.
Raupp, M. J., and R. F. Denno. 1983. Leaf age as a predictor of herbivore distribution. Pages 91-124 in R. F. Denno and M. S. McClure, editors. Variable plants and herbivores in natural and managed systems. Academic Press, New York, New York, USA.
Rhyne, A. L., and R. G. D. Steel. 1965. Tables for a treatments versus control multiple comparisons sign test. Technometrics 7:293-306.
Royama, T. 1984. Population dynamics of the spruce budworm Choristoneura fumiferana. Ecological Monographs 54:429-462.
Satchell, J. E. 1961. Resistance in oak (Quercus sp.) to defoliation by Tortrix viridana L. in Roudsea Wood National Nature Reserve. Annals of Applied Biology 50:431-442.
Schneider, J. C. 1980. The role of parthenogenesis and female aptery in microgeographic, ecological adaptation in the fall cankerworm, Alsophila pometaria Harris (Lepidoptera: Geometridae). Ecology 6:1082-1090.
Schweitzer, D. F. 1979. Effects of foliage age on body weight and survival of larvae of the tribe Lithophanini (Lepidoptera: Noctuidae). Oikos 32:403-408.
Sprugel, D. G., T. M. Hinckley, and W. Schaap. 1991. The theory and practice of branch autonomy. Annual Review of Ecology and Systematics 22:309-334.
Tuomi, J., P. Niemela, I. Jussila, T. Vuorislao, and V. Jormalainen. 1989. Delayed budbreak: a defensive response of mountain birch to early-season defoliation? Oikos 54:87-91.
Tuomi, J., P. Niemela, and S. Siren. 1990. The panglossian paradigm and delayed inducible accumulation of foliar phenolics in mountain birch. Oikos 59:399-410.
Tuomi, J., J. Salo, E. Haukioja, P. Niemela, T. Hakala, and R. Nannila. 1983. The existential game of individual self-maintaining units: selection and defense tactics of trees. Oikos 40:369-376.
Watson, M. A. 1995. Sexual differences in plant developmental phenology affect plant-herbivore interactions. Trends in Ecology and Evolution 10:180-182.
Wellington, W. G. 1957. Individual differences as a factor in population dynamics: the development of a problem. Canadian Journal of Zoology 38:289-314.
Whitham, T. B., and S. Mopper. 1985. Chronic herbivory: impacts on architecture and sex expression of pinyon pine. Science 228:1089-1091.
Whitham, T. B., and C. N. Slobodchikoff. 1981. Evolution by individuals, plant-herbivore interactions, and mosaics of genetic variability: the adaptive significance of somatic mutations in plants. Oecologia 49:287-292.
Yela, J. L., and C. M. Herrera. 1993. Seasonality and life cycles of woody plant-feeding noctuid moths (Lepidoptera, Noctuidae) in Mediterranean habitats. Ecological Entomology 18:259-269.
Zahner, R. 1968. Water deficits and growth of trees. Pages 191-254 in T. T. Kozlowski, editor. Water deficits and plant growth. Volume 2. Academic Press, New York, New York, USA.
Zar, J. H. 1984. Biostatistical analysis. Second edition. Prentice Hall, Englewood Cliffs, New Jersey, USA.
|Printer friendly Cite/link Email Feedback|
|Author:||Quiring, D.T.; McKinnon, M.L.|
|Date:||Jul 1, 1999|
|Previous Article:||Induced responses to herbivory in wild radish: effects on several herbivores and plant fitness.|
|Next Article:||Blade abandonment/proliferation: a novel mechanism for rapid epiphyte control in marine macrophytes.|