Herbivore Resistance In Betula Pendula: Effect Of Fertilization, Defoliation, And Plant Genotype

Ecology, Jan, 2000 by Pia Mutikainen, Mari Walls, Jari Ovaska, Markku Keinanen, Ritta Julkunen-Tiitto, Elina Vapaavuori

PIA MUTIKAINEN [1,2,5]

MARI WALLS [1]

JARI OVASKA [1,3]

MARKKU KEINANEN [4]

RITTA JULKUNEN-TIITTO [4]

ELINA VAPAAVUORI [3]

Abstract. Plant resistance to herbivores is affected both by genetic and environmental factors. The carbon-nutrient balance hypothesis (CNB) explains environmentally induced variation in both constitutive and delayed herbivore-induced resistance (DIR) in terms of variation in soil fertility and light regime. The CNB hypothesis predicts that an increase in the availability of nutrients (e.g., fertilization) decreases both constitutive and induced resistance against herbivores. We tested the relative roles of plant genotype, defoliation, and soil fertility in determining herbivore resistance of cloned silver birch Betula pendula Roth saplings. As indicators of insect and mammalian resistance we conducted bioassays with a geometrid moth, Epirrita autumnata (Borkhausen), and counted the resin droplets on the shoot of the saplings, respectively. In addition, we measured rapid induced resistance (RIR) against the insect herbivore. Finally, we analyzed leaf secondary chemistry to investigate the correlations of sec ondary chemicals with the level of resistance measured using the performance of E. autumnata.

With respect to the constitutive resistance against an insect herbivore, our results support the CNB hypothesis; the larvae of E. autumnata had a higher relative growth rate and pupal mass on fertilized saplings compared to nonfertilized saplings, i.e., the fertilized saplings had a lower resistance level. However, the relative growth rate of E. autumnata was significantly decreased by defoliation only when the larvae were grown on fertilized saplings. The number of resin droplets increased due to fertilization and, in fertilized saplings, following defoliation, but these responses were highly determined by the genotype of the sapling. Altogether, the results on resin droplets are not in accordance with the CNB hypothesis.

The concentration of condensed tannins correlated negatively with E. autumnata growth rate and pupal mass in both fertilization levels, whereas the concentration of total nontannin phenolics correlated positively with the E. autumnata growth rate in nonfertilized saplings. In addition, the concentration of myricetin glycosides correlated negatively with the pupal mass of E. autumnata, whereas the correlations between E. autumnata performance indices and other groups of flavonol glycosides were either significantly positive (kaempferol glycosides) or nonsignificant (quercetin glycosides). Further, the concentration of 3,4'-dihydroxypropiophen one 3-glucoside (DHPPG) correlated positively with the magnitude of induction in E. autumnata growth rate and pupal mass in fertilized saplings, where the significant induction in resistance occurred. The correlations of secondary chemistry and E. autumnata performance indices suggest that the constitutive level of resistance of B. pendula against E. autumnata is mainly determined by the concentration of condensed tannins, whereas the induced resistance is determined by the concentration of nontannin phenolics, such as flavonol glycosides and DHPPG.

We observed significant differences among the clones in their insect and mammalian resistance (i.e., genetic basis for the resistance), which indicates that resistance can evolve as a response to herbivory. However, fertilization explained a higher proportion of variance in insect performance indices than the genotype of the plant, whereas the opposite was true for the amount of resin droplets, which we used as an indicator of mammalian resistance.

Key words: Betula pendula; carbon/nutrient balance; clonal variation; defoliation; Epirrita autumnata; fertilization; herbivore resistance; plant-herbivore interactions; resin droplets; secondary chemicals.

INTRODUCTION

Phenotypic variation in plant resistance to herbivory is caused by genetic and/or environmentally induced differences among plant individuals (for a review, see Marquis [1992]). The maintenance of genetic variation in resistance is a dilemma because if resistance is beneficial in terms of plant fitness, plant populations should evolve towards the highest level of resistance and lose additive genetic variation in resistance (Fisher 1930). However, genetic variation in resistance traits within populations has been observed in numerous studies (e.g., Hanover 1966, Berenbaum et al. 1986, Fritz 1990, Kennedy and Barbour 1992, Zangerl and Berenbaum 1997). The relative contributions of genetic and environmental factors, and their interaction, to the variation in resistance have not been extensively studied until recently (Maddox and Cappuccino 1986, Maddox and Root 1987, Fritz 1990, Karban 1992, Kennedy and Barbour 1992). In his review, Karban (1992) concluded that plant genotype has a stronger effect on resistance to herbivores (i.e., on herbivore populations) than the interaction between genotype and environmental factors.

Environmentally induced variation in plant resistance may be due to both biotic (e.g., previous herbivory) and abiotic factors. The growth--differentiation balance (GDB, Loomis 1932, Lorio 1986, Herms and Mattson 1992) and carbon-nutrient balance (CNB, Bryant et al. 1983, 1988, 1993, Tuomi et al. 1984) hypotheses explain environmentally induced variation in both constitutive and herbivore-induced resistance with variation in growth-limiting environmental factors, such as soil fertility and light regime. According to the CNB hypothesis, the balance between carbon and nutrients is changed in nutrient-limited plants, whose growth is limited more than photosynthesis, leading to the accumulation of carbon-based secondary metabolites. These metabolites (e.g., phenolics) may reduce plant quality as food for herbivores and thus contribute to the resistance against herbivores (e.g., Scriber and Slansky 1981). In plants not suffering from nutrient limitation, the carbon is used for growth and resistance is decreased. The CNB hypothesis also provides an explanation for DIR in deciduous plants. The within-plant balance between carbon and nutrients is disturbed by defoliation which removes more nutrients than carbon, again followed by the accumulation of carbon-based secondary metabolites and the expression of DIR. Consequently, the CNB hypothesis predicts that an increase in the availability of nutrients (e.g., fertilization), coinciding with the defoliation mitigates the induction of resistance and, consequently, the herbivores perform better. Shading, by decreasing the amount of carbon available for secondary metabolites, also decreases the level of induced resistance. Studies that have tested the CNB hypothesis in deciduous trees in relation to DIR have obtained contrasting results. For example, Bryant et al. (1993) found strong support for the CNB hypothesis as an explanation for DIR in Betula resinifera--Rheumaptera hastata system in Alaska, whereas the results of Ruohomaki et al. (1996) from Betula pubescens ssp. tort uosa-Epirrita autumnata system do not seem to support the CNB hypothesis. In addition to the DIR observable in leaf chemistry and performance of insect herbivores, the CNB hypothesis has been tested in relation to the resinous substances on the bark of young B. pendula which have been shown to determine the resistance against mammalian herbivores (Rousi et al. 1991, Rousi et al, 1993). These studies have also given contrasting results on the effect of fertilization on resistance: resistance against hares was not decreased (Rousi et al. 1991) whereas resistance against voles was decreased due to fertilization (Rousi et al. 1993). Taken together, the explanatory power of CNB hypothesis seems to depend on both the plant and herbivore species and group of secondary chemicals in question (Reichardt et al. 1991, Gershenzon 1994, Hartley et al. 1995, Kyto et al. 1996). Further, most studies on CNB hypothesis in relation to induced resistance have concentrated on the across-year induction of resistance, i.e., on DIR (e.g., Bryant et al. 1993, Ruohomaki et al. 1996). However, Hunter and Schultz (1995) found out that fertilization prevented induction in oak within one season, and suggested that the effects of fertilization on induction of resistance may occur at more than one temporal scale.