Tomato Reproductive Success Is Equally Affected by Herbivores That Induce or That Suppress Defenses

Abstract

Herbivory induces plant defenses. These responses are often costly, yet enable plants under attack to reach a higher fitness than they would have reached without these defenses. Spider mites (Tetranychus ssp.) are polyphagous plant-pests. While most strains of the species Tetranychus urticae induce defenses at the expense of their performance, the species Tetranychus evansi suppresses plant defenses and thereby maintains a high performance. Most data indicate that suppression is a mite-adaptive trait. Suppression is characterized by a massive down-regulation of plant gene-expression compared to plants infested with defense-inducing mites as well as compared to control plants, albeit to a lesser extent. Therefore, we hypothesized that suppression may also benefit a plant since the resources saved during down-regulation could be used to increase reproduction. To test this hypothesis, we compared fruit and viable seed production of uninfested tomato plants with that of plants infested with defense-inducing or defense-suppressing mites. Mite-infested plants produced fruits faster than control plants albeit in lower total amounts. The T. evansi-infested plants produced the lowest number of fruits. However, the number of viable seeds was equal across treatments at the end of the experiment. Nonetheless, at this stage control plants were still alive and productive and therefore reach a higher lifetime fitness than mite-infested plants. Our results indicate that plants have plastic control over reproduction and can speed up fruit- and seed production when conditions are unfavorable. Moreover, we showed that although suppressed plants are less productive in terms of fruit production than induced plants, their lifetime fitness was equal under laboratory conditions. However, under natural conditions the fitness of plants such as tomato will also depend on the efficiency of seed dispersal by animals. Hence, we argue that the fitness of induced plants in the field may be promoted more by their higher fruit production relative to that of their suppressed counterparts.

Keywords: costs; fruit; induction; mites; plant defense; reproductive escape; seed; suppression.

FIGURE 1

Relative transcript abundances of defense-related genes in Tetranychus evansi and Tetranychus urticae infested Micro-Tom leaflets at 7 days post-infestation. PI-IIc (A) and PR-1a (B) transcript levels were relative to actin. Uninfested leaflets were used as controls. The bars represent the means ( ± SE) and are scaled to the lowest mean by putting this to 1. Bars annotated with different letters were significantly different according to Fisher’s LSD test (P < 0.05) after ANOVA, n = 5.

FIGURE 2

Micro-Tom tomato infested with spider mites over the course of the experiment [days post-infestation (dpi)]. (A) Uninfested control tomato plants. (B) T. urticae-infested tomato plants. (C) T. evansi-infested tomato plants.

FIGURE 3

Total fruits (A) and red fruits (B) production per plant of spider mites infested Micro-Tom plants in time series post-infestation. The data were fitted to the regression equation: f = a^∗(1-exp(-b^∗x))ˆc. The solid line shows the average total number of fruits for uninfested control plants (closed circles) (A: R² = 0.79, P < 0.0001; B: R² = 0.86, P < 0.0001). The dotted line shows the average total number of fruits for the T. evansi infested plants (closed rectangles) (A: R² = 0.39, P < 0.0001; B: R² = 0.67, P < 0.0001). The dashed line shows the average total number of fruits for T. urticae infested plants (open circles) (A: R² = 0.45, P < 0.0001; B: R² = 0.75, P < 0.0001). Differences in mean fruit production per plant among treatments over time were tested by means of a linear mixed effects (LME) model using the number of fruits and plant as a random factor. Different letters next to the curves indicated overall significant differences among treatments (lsmeans of R, P < 0.05). Asterisk (^∗) indicated the age of plants when infested with mites.

FIGURE 4

Average total number of fruits per plant (±SE) of uninfested Micro-Tom plants in comparison to plants infested with T. urticae or T. evansi. Bars annotated with different letters were significantly different according to Fisher’s LSD test (P < 0.05) after ANOVA with n = 15 per treatment.

FIGURE 5

Average viable seeds per fruit per plant (±SE) of uninfested Micro-Tom plants in comparison to plants infested with T. urticae or T. evansi. Bars annotated with different letters were significantly different according to Fisher’s LSD test (P < 0.05) after ANOVA with n = 15 per treatment.

FIGURE 6

Reproductive fitness (average viable seeds per plant ±SE) of Micro-Tom plants infested with spider mites. Bars indicate the average number of viable seeds per plant of uninfested Micro-Tom plants (control) or plants infested with spider mites (T. urticae, T. evansi). Seeds were considered viable when they germinated. N.S. indicates that no significant differences among treatments were detected using ANOVA (n = 15 plants per treatment).