Both the length of the growing season and the intensity of herbivory often vary along climatic gradients, which may result in divergent selection on plant phenology, and on resistance and tolerance to herbivory. In Sweden, the length of the growing season and the number of insect herbivore species feeding on the perennial herb Lythrum salicaria decrease from south to north. Previous common-garden experiments have shown that northern L. salicaria populations develop aboveground shoots earlier in the summer and finish growth before southern populations do. We tested the hypotheses that resistance and tolerance to damage vary with latitude in L. salicaria and are positively related to the intensity of herbivory in natural populations. We quantified resistance and tolerance of populations sampled along a latitudinal gradient by scoring damage from natural herbivores and fitness in a common-garden experiment in the field and by documenting oviposition and feeding preference by specialist leaf beetles in a glasshouse experiment. Plant resistance decreased with latitude of origin, whereas plant tolerance increased. Oviposition and feeding preference in the glasshouse and leaf damage in the common-garden experiment were negatively related to damage in the source populations. The latitudinal variation in resistance was thus consistent with reduced selection from herbivores towards the northern range margin of L. salicaria. Variation in tolerance may be related to differences in the timing of damage in relation to the seasonal pattern of plant growth, as northern genotypes have developed further than southern have when herbivores emerge in early summer.

Herbivory can negatively and selectively affect plant fitness by reducing growth, survival and reproductive output, thereby influencing plant population dynamics and evolution. Latitudinal variation in intensity of herbivory is common, but the extent to which it translates into corresponding variation in effects on plant performance is still poorly known. We tested the hypothesis that variation in the fitness-consequences of herbivory mirror differences in intensity of herbivory among three natural populations of the perennial herb Lythrum salicaria along a latitudinal gradient from southern to northernmost Sweden. We documented intensity of herbivory and examined its effect on survival, growth and reproductive output over two years by experimentally removing herbivores with insecticide. The intensity of herbivory and the effects of herbivory on plant fitness were strongest in the southern population, intermediate in the central population and weakest in the northern population. The mean proportion of the leaf area removed ranged from 11% in the southern to 3% in the northern population. Herbivore removal increased plant height 1.5-fold in the southern and 1.2-fold in the central population, the proportion plants flowering 4-fold in the southern and 2-fold in the central population, and seed production per flower 1.6-fold in the southern and 1.2-fold in the central population, but did not affect plant fitness in the northern population. Herbivore removal thus affected the relative fecundity of plants in the three populations: In the control, seed output per plant was 8.6 times higher in the northern population compared to the southern population, whereas after herbivore removal it was 2.5 times higher in the southern population. The results demonstrate that native herbivores may strongly affect the demographic structure of L. salicaria populations and thereby shape geographic patterns of seed production. They further suggest that the strength of herbivore-mediated selection varies among populations and decreases towards the north.

Herbivory can negatively affect several components of plant reproduction. Yet, because of a lack of experimental studies involving multiple populations, the extent to which differences in herbivory contribute to among-population variation in plant reproductive success is poorly known. We experimentally determined the effects of insect herbivory on reproductive output in nine natural populations of the perennial herb Lythrum salicaria along a disturbance gradient in an archipelago in northern Sweden, and we quantified among-population differentiation in resistance to herbivory in a common-garden experiment in the same area. The intensity of leaf herbivory varied >500-fold and mean female reproductive success >400-fold among the study populations. The intensity of herbivory was lowest in populations subject to strong disturbance from ice and wave action. Experimental removal of insect herbivores showed that the effect of herbivory on female reproductive success was correlated with the intensity of herbivory and that differences in insect herbivory could explain much of the among-population variation in the proportion of plants flowering and seed production. Population differentiation in resistance to herbivory was limited. The results demonstrate that the intensity of herbivory is a major determinant of flowering and seed output in L. salicaria, but that differences in herbivory are not associated with differences in plant resistance at the spatial scale examined. They further suggest that the physical disturbance regime may strongly influence the performance and abundance of perennial herbs and patterns of selection not only because of its effect on interspecific competition, but also because of effects on interactions with specialized herbivores.

No abstract