Author for correspondence: André Kessler Tel: +1 607 254 4219 Email: firstname.lastname@example.org
•Herbivory is thought to be detrimental to plant fitness and commonly results in a metabolic shift in the plant: photosynthetic processes are typically down-regulated, while resource allocation to defenses is increased in herbivore-attacked plants, resulting in fitness costs of induced plant responses.
•Wild tobacco, Nicotiana attenuata, attacked by Tupiocoris notatus mirid bugs becomes resistant against more damaging herbivores through mirid-induced direct and indirect defenses. However, mirid-induced resistance and tissue loss do not result in a reduction of plant fitness. These findings suggest induced metabolic responses allowing the plant to compensate for the lost tissue and resources allocated to defenses.
•While feeding by Manduca sexta larvae results in a strong down-regulation of photosynthesis, we demonstrate a specific induction of elevated photosynthetic activity in N. attenuata leaves by elicitors in mirid salivary secretions. The elevated CO2 assimilation rate is sufficient to compensate for the loss of photosynthetically active tissue and balances the net photosynthesis of infested leaves.
•We discuss the observed increase in the plant’s primary metabolic activity as a mechanism that allows plants to alleviate negative fitness effects of mirid attack and mediates the vaccination effects that result in a net benefit in environments with multiple herbivores.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
In the native tobacco plant Nicotiana attenuata, reduced fitness has been observed after attack by voracious leaf chewers such as tobacco (Manduca sexta) and tomato (Manduca quinquemaculata) hornworms (Kessler & Baldwin, 2004). Attack by hornworm larvae elicits the production of several defensive metabolites and proteins which incur significant physiological costs for their production (Glawe et al., 2003; Zavala et al., 2004). Additionally, hornworm feeding elicits a strong down-regulation of genes involved in nitrogen (N) assimilation, amino acid biosynthesis, and carbohydrate metabolism (Halitschke et al., 2003; Voelckel & Baldwin, 2004a). These changes are consistent with the proposed reallocation of resources from primary to secondary metabolism (Schwachtje & Baldwin, 2008). Furthermore, the redirection of resources is likely to change the source–sink relationships in the damaged tissue, thereby serving as a signal in the feedback loop regulating the plant’s primary metabolic process, of plant photosynthesis. Thus, the reduction in lifetime fitness observed in N. attenuata plants attacked by Manduca hornworms (Kessler & Baldwin, 2004) probably results from the reallocation of resources into defenses in combination with the loss of photosynthetically active tissue.
In contrast to the negative fitness effects of hornworm attack, herbivory by the mirid bug Tupiocoris notatus is fitness-neutral. While causing extensive tissue damage and eliciting a very similar suite of defense responses, mirid attack does not reduce seed production (Kessler & Baldwin, 2004). Furthermore, elicitation by mirid feeding renders the plant more resistant to other, more damaging herbivores such as hornworms. This special case of induced cross-resistance, termed ‘plant vaccination’, leads to a net benefit for mirid-damaged tobacco plants in their native habitats when multiple herbivore species are present (Kessler & Baldwin, 2004). However, it remains unknown how the plant is compensating for the significant loss of tissue to the cell content-feeding mirid and the investment of resources into induced defenses. Interestingly, large-scale microarray analyses comparing the transcriptional responses of N. attenuata to herbivory by the two different herbivores demonstrated that genes involved in secondary metabolism are similarly regulated in mirid- and hornworm-attacked plants, whereas differential responses were primarily observed for genes related to primary metabolism (Voelckel & Baldwin, 2004b). The hornworm-specific reduction in primary metabolism-related gene expression (Voelckel & Baldwin, 2004b) is not observed in mirid-attacked plants, suggesting specific recognition of the different herbivores and differential regulation of metabolic processes exceeding a simple reallocation of resources.
We hypothesize that increased metabolism (i.e. photosynthetic activity) in N. attenuata in response to mirid feeding represents a herbivore-specific, induced mechanism of the plant to compensate for herbivory-mediated costs, such as tissue loss and resource allocation to defense production. According to this hypothesis, we predict an increased photosynthetic activity, specifically induced in mirid-damaged plants. By testing the hypothesis, we provide a missing physiological link between the transcriptional (Voelckel & Baldwin, 2004b) and ecological analyses (Kessler & Baldwin, 2004) of the plant vaccination phenomenon in the N. attenuata study system. Furthermore, we provide evidence for a general plant compensation mechanism to alleviate fitness costs of herbivory through the activation of supplemental resources.
Materials and Methods
Plant and insect material
Original seeds of Nicotiana attenuata Torr. ex Wats. (synonymous with Nicotiana torreyana Nelson and Macbr.) were collected from individual plants at an experimental field site in south-western Utah in 1992 (DI Ranch; USA, Halitschke et al., 2000) and plants were inbred for 17 generations. Seeds were soaked in a 1 : 50 aqueous dilution of liquid smoke (B&G Foods Inc., Roseland, NJ, USA) and germinated on a peat moss-based soil mix (Metro Mix 360; Sun Gro, Bellevue, WA, USA). Plants were grown in 1-l pots under the following environmental conditions: 16 h : 8 h and 26 : 24°C, light : dark cycle, 50% relative humidity and 400–600 μmol m−2 s−1 photosynthetically active radiation. Elongated nonflowering plants were used in all experiments. Manduca sexta eggs were purchased from the University of North Carolina State University insectary (Raleigh, NC, USA). A T. notatus colony is maintained in the laboratory on N. attenuata plants.
One freshly hatched M. sexta larva or one T. notatus nymph was allowed to feed on the second stem leaf of elongated plants for 3 d. Insects were enclosed on the leaf with clip cages. Control plants received empty clip cages to control for potential cage effects. Insects and cages were removed for chlorophyll fluorescence and photosynthesis measurements, and a digital picture was taken. We overlaid the images with a digital grid (Adobe Photoshop®) and calculated leaf damage as the proportion of squares of leaf area removed by hornworm feeding or showing yellow mottling symptoms of mirid feeding (Hare & Elle, 2002).
Salivary secretion treatment
Mirid salivary secretions were collected in a sugar solution. A 100-μl droplet of a 40 mM glucose solution was contained between two thin layers of Parafilm® (Menasha, WI, USA) and 10 mirids were allowed to feed on the solution for 12 h. The sugar solution was collected and stored at −20°C until used. One μl of the solution was injected with a 10-μl Hamilton syringe into the spongy parenchyma of the second stem leaf. Control plants received identical treatments with glucose solutions not exposed to mirid feeding to control for the effects of needle damage and glucose application. Oral secretions were collected from third to fourth instar M. sexta larvae, diluted 1 : 1 (v : v) with deionized water (Halitschke et al., 2001), and 1 μl of the diluted oral secretions was injected as described for mirid salivary secretions. Deionized water was injected as a control solution.
Gas exchange measurements
Gas exchange measurements were performed simultaneously with the fluorescence imaging using a Li-Cor 6400 photosynthesis system (Li-Cor, Lincoln, NE, USA). The CO2 concentration and actinic light conditions in the analyzer chamber were adjusted to 400 ppm and 300 μmol m−2 s−1, respectively.
In contrast to the complete tissue removal by the feeding hornworm, mirid feeding damages individual cells and the leaf area-based damage measure probably overestimates the actual loss of tissue/cells. We measured leaf greenness as a proxy of chlorophyll content (SPAD-502; Spectrum Technologies Inc., Plainfield, IL, USA) in a series of leaves (n = 8) with different degrees of mirid damage. Data points were recorded for six different spots (2 mm × 3 mm) on the damaged leaf. The average chlorophyll content in mirid-damaged leaves is negatively correlated with the damaged leaf area (f (x) = 97.3 − 0.60x; r2 = 0.91; P =0.0003). We used the reduction in chlorophyll content as a conservative measure of lost photosynthetic tissue to correct the area of mirid-damaged leaves and to calculate the photosynthetic activity in the remaining tissue. This allowed a more realistic comparison of data from mirid- and hornworm-damaged leaves.
Chlorophyll fluorescence imaging
Chlorophyll fluorescence was measured with an imaging system described by Zangerl et al. (2002) and the FluorImager® software (Technologica, Frating, UK). The efficiency of photosystem II (PSII) is a crucial factor determining photosynthetic activity (Daley, 1995; Baker et al., 2001; Oxborough, 2004), and a reduction in PSII efficiency has been found to be responsible for the reduced photosynthetic activity in leaves damaged by some chewing herbivore species (Zangerl et al., 2002; Aldea et al., 2005, 2006; Tang et al., 2006). We analyzed and compared the operating and maximum quantum efficiencies of PSII in mirid- and hornworm-damaged leaves. The operating quantum efficiency of PSII was characterized by fluorescence measurements of light-adapted plants: operating fluorescence (F ′) was measured under 300 μmol m−2 s−1 continuous actinic light after 15–20 min stabilization time with 500 μs measuring pulses and maximum fluorescence () was measured with a saturating light pulse (2750 μmol m−2 s−1). Three data points were recorded at 5-min intervals and averaged for each leaf. The operating quantum yield of PSII () was calculated for each measurement:
Similarly, the maximum quantum yield of PSII was calculated from fluorescence measurements on dark-adapted leaves: fluorescence (Fo) and maximum fluorescence (Fm) were measured with identical measuring pulses as described for light-adapted leaves without the continuous actinic light supply. The maximum quantum efficiency of PSII (Fv/Fm) was calculated:
The effects of herbivory on CO2 assimilation, PSII efficiency and stomatal conductance were analyzed by ANOVAs using StatView (SAS Institute Inc., Cary, NC, USA) and treatment means (control, hornworm-, and mirid-feeding) were separated by Bonferroni-corrected post hoc tests. The elicitor function of mirid secretions was characterized in several independent replicates performed with independently grown sets of plants and individual collections of salivary secretions. Paired t-tests were used to compare plants treated with control and secretion-containing solutions (water vs hornworm oral secretions and glucose vs mirid salivary secretions). Additionally, we compared plants injected with the control solution and untreated control plants to account for syringe needle damage effects.
Herbivores differentially affect photosynthetic gas exchange
To test if N. attenuata plants show differential elicitation of photosynthetic activity, we compared the CO2 gas exchange of plants damaged by M. sexta hornworms or T. notatus mirids with that of undamaged control plants. CO2 assimilation in hornworm-damaged leaves was generally reduced (Table 1). After 3 d of continuous feeding, photosynthetic activity was reduced by 16% in the remaining tissue of the damaged leaves (Fig. 1a; ANOVA: F2,18 = 30.241; P <0.0001). Stomatal conductance and intercellular CO2 concentration were not different (ANOVA; P >0.05) in herbivore-damaged leaves compared with untreated control leaves (Table 1). When combined with the tissue removed by hornworm feeding (30%; Table 1), the reduction in photosynthesis resulted in a total loss of almost 50% of the net photosynthetic capacity in hornworm-damaged leaves (Fig. 1b; ANOVA: F2,18 = 58.466; P <0.0001).
Table 1. Damage levels and photosynthesis in hornworm- and mirid-damaged Nicotiana attenuata leaves
1Insects were consistently feeding on the analyzed leaf for the indicated time period.
2Mean ± SE (n =7) is shown; different letters designate statistically significant differences (Bonferroni-corrected post hoc test; P <0.05).
nd, not determined.
Leaf area damage2 (%)
0 ± 0a
11.8 ± 3.0b
5.9 ± 1.5ab
0 ± 0a
30.0 ± 4.2c
19.3 ± 3.7bc
CO2 assimilation2 (μmol m−2 s−1)
10.9 ± 0.2ab
9.9 ± 0.7a
12.2 ± 0.1b
11.5 ± 0.2a
10.2 ± 0.4a
11.2 ± 0.5a
8.1 ± 0.1b
6.8 ± 0.2a
8.2 ± 0.1b
Stomatal conductance2 (mol m−2 s−1)
0.19 ± 0.02a
0.20 ± 0.02a
0.33 ± 0.05a
0.33 ± 0.02a
0.28 ± 0.02a
0.28 ± 0.04a
0.23 ± 0.02a
0.21 ± 0.01a
0.24 ± 0.04a
Intercellular CO2 concentration2 (ppm)
285 ± 12a
305 ± 8a
307 ± 8a
294 ± 16a
308 ± 2a
3001 ± 9a
291 ± 16a
310 ± 12a
321 ± 9a
Photosystem II efficiency ( )
0.53 ± 0.01a
0.51 ± 0.02a
0.53 ± 0.01a
0.53 ± 0.01b
0.50 ± 0.01a
0.54 ± 0.01b
0.57 ± 0.01b
0.53 ± 0.01a
0.59 ± 0.01b
By contrast, we observed no reduction in CO2 assimilation in mirid-damaged leaves (Table 1). After 1 d, when mirid damage was visible on 5% of the leaf area, we observed a 12% increase in net photosynthesis, which, however, was statistically not different from the photosynthetic activity in undamaged leaves (P =0.088). Furthermore, despite the fact that mirids had caused visible damage to 19% of the leaf area, resulting in a 15% loss of leaf chlorophyll content after 3 d of continuous feeding, net photosynthetic capacity in mirid-damaged leaves was not reduced compared with control leaves (Fig. 1b). An increase in photosynthetic activity in the remaining tissue (Fig. 1a) compensates for the lost tissue and thus balances the net assimilation of CO2 of mirid-damaged leaves.
Herbivore-specific changes in photosystem II efficiency
Hornworm damage caused an 6 and 8% reduction of the operating quantum efficiency () in the remaining undamaged leaf tissue after 2 d (ANOVA: F2,9 = 6.140; P =0.0208) and 3 d (ANOVA: F2,18 = 12.053; P =0.0005) of continuous feeding, respectively (Table 1). The reduction resulted from damage or down-regulation of PSII, because the maximum efficiency (Fv/Fm; measured in dark-adapted leaves) was also significantly reduced (ANOVA: F2,9 = 8.919; P =0.0073) in hornworm-damaged leaves (0.82 ± 0.004) compared with undamaged control leaves (0.84 ± 0.003). By contrast, operating and maximum PSII efficiencies in mirid-attacked leaves were similar to those in control plants (Table 1). Furthermore, we observed a differential effect of hornworm and mirid feeding on the relationship between PSII efficiency and CO2 assimilation rate (Fig. 2). Control and hornworm-damaged plants show a linear positive correlation between the two parameters, which has been reported for plants under different physiological conditions (Genty et al., 1989; Krall & Edwards, 1992; Tang et al., 2006). Conversely, in mirid-damaged plants, CO2 assimilation was negatively correlated with the operating efficiency of PSII (Fig. 2).
Increased photosynthetic activity elicited by mirid salivary secretions
To dissect the impact of damage caused by stylet penetration from the introduction of potential elicitors during mirid feeding, we collected salivary secretions of mirids and compared the effect of the secretions and control solutions on photosynthesis in N. attenuata leaves. Injection of mirid salivary secretions consistently increased photosynthetic activity by up to 11% compared with the injection of a secretion-free sugar solution (Fig. 3; paired t-test; P <0.0001). The injection of all other solutions, including the control sugar solution, pure water, and diluted M. sexta oral secretions, slightly reduced (2–4%) photosynthetic CO2 assimilation compared with control plants (Fig. 3), but this reduction was not statistically significant (ANOVA, F2,18 =0.599; P =0.5601) and was probably caused by the mechanical damage from needle penetration during injection.
Previous findings (Kessler & Baldwin, 2004) as well as the data presented here suggest plant compensatory responses that can be specifically induced by the attacking herbivore. Three main mechanisms have been proposed to explain how plants can compensate for herbivore damage: (1) (over)compensatory responses can result from bud and meristem dormancy release when damage of the main stem breaks the dormancy of lateral meristems (Lennartsson et al., 1998; Agrawal, 2000; Tiffin, 2000); (2) plants may change the allocation of resources from established reserves after herbivore damage (Stowe et al., 2000; Tiffin, 2000; Schwachtje et al., 2006); and (3) plants may up-regulate primary metabolism to increase the accessible resource pool (Strauss & Agrawal, 1999; Agrawal, 2000). We never observed damage to the dominant meristem nor the systemic activation of lateral meristems in response to mirid damage (A. Kessler and H. Halitschke, pers. obs.), suggesting that dormancy release is not likely to be a mechanism of compensation in our study system. Reallocation of resources after herbivory has ceased has been demonstrated in N. attenuata as a mechanism to reduce fitness consequences of hornworm damage (Schwachtje et al., 2006). However, the fitness-neutral response to mirid feeding (Kessler & Baldwin, 2004) is evident while the insects are actively feeding on the plant, suggesting that reallocation of resources alone cannot explain the compensatory seed set. It rather suggests an increase of the accessible resource pool as a result of mirid-induced changes in primary metabolism.
Consistent with earlier studies (Watanabe & Kitagawa, 2000; Zangerl et al., 2002; Aldea et al., 2005, 2006; Tang et al., 2006), we found reduced CO2 assimilation in the remaining tissue of hornworm-damaged leaves, resulting in a 50% reduction of net photosynthesis (Fig. 1). By contrast, an increase in photosynthetic activity in mirid-damaged leaves compensated for the lost tissue and resulted in net photosynthetic CO2 assimilation similar to that in undamaged control plants (Fig. 1). These findings are consistent with predictions of the increased primary metabolism hypothesis (Agrawal, 2000). Increased photosynthesis alone may not be sufficient to increase resource availability and may require the increased acquisition of nutrients as well as adjusted assimilate transport and allocation. However, a strong correlation between photosynthesis and growth rate has been reported in N. attenuata (Giri et al., 2006) and N. tabacum grown under conditions of high N availability (Fichtner et al., 1993) or saturating light conditions (Jiang et al., 1994). Light and N availability are rarely limiting resources in the nutrient-rich soils of freshly burned desert habitats in which N. attenuata plants typically grow (Lynds & Baldwin, 1998). Thus, the increase in photosynthetic activity in mirid-damaged leaves could supply the energy and C resources required to compensate for the tissue loss. This herbivore-specifically induced compensation response provides a likely physiological mechanism for plant vaccination as a fitness-neutral form of cross-resistance previously observed in N. attenuata (Kessler & Baldwin, 2004).
The efficiency of PSII is a crucial factor determining photosynthetic activity (Daley, 1995; Baker et al., 2001; Oxborough, 2004) and a reduction in PSII efficiency is responsible for the reduced photosynthetic activity in leaves damaged by chewing herbivores (Zangerl et al., 2002; Aldea et al., 2005, 2006; Tang et al., 2006). We observed a similar reduction in PSII efficiency in hornworm-damaged leaves (Table 1), which is consistent with the down-regulation of photosynthesis-related genes in response to M. sexta herbivory (Halitschke et al., 2003). By contrast, the mirid-induced change in CO2 assimilation was not mediated by PSII efficiency (Fig. 2), indicating an herbivore-specific regulation of additional photosynthetic processes. Differences in stomatal conductance resulting from the different types of damage do not account for the differences in photosynthetic activity (Table 1). It is more likely that herbivore-specific cues elicit changes in photosynthetic pathways comparable to the differential regulation of photosynthesis- and carbohydrate metabolism-related genes observed in gene expression studies (Voelckel & Baldwin, 2004b).
In N. attenuata, induced responses to herbivory involve herbivore-specific changes in both primary and secondary metabolism, and these changes have profound effects on plant fitness and interactions with the associated arthropod community (Kessler & Baldwin, 2004). Differential changes in primary metabolism could explain why some herbivore species significantly reduce plant fitness while others have neutral or positive effects. However, our results inevitably evoke the question of why N. attenuata plants do not maintain higher constitutive photosynthesis rates if that would result in increased fitness. The herbivore-induced increase in photosynthesis may reduce or eliminate the plants’ ability to respond to additional environmental stresses, which may result in fatal physiological damage with nonadditive fitness effects of multiple stressors (Chapin et al., 1987). One of the most likely environmental conditions with potentially high impact on N. attenuata performance in the Great Basin Desert habitat is drought stress in combination with high light and temperature conditions. Reduced C metabolism and down-regulation of photosynthesis are commonly observed in response to drought stress, and the plasticity serves as an important mechanism to protect the photosystem from damage (Chaves et al., 2002).
Moreover, a physiological link between increased photosynthesis and seed production as observed in our study system does depend on a complex metabolic network. Herbivore-induced changes have been observed for several primary metabolic processes, including assimilation : respiration ratio (Schmidt et al., 2009), assimilate allocation, transport, and storage (Schwachtje et al., 2006; Babst et al., 2008). It remains unclear, however, to what extent herbivore-specific elicitation can orchestrate the multiple underlying plant signaling cascades that mediate changes in primary metabolism. However, complex changes in secondary metabolism are characteristic of herbivore-induced plant responses, and have been demonstrated to be orchestrated by specific herbivore-derived elicitors (Mattiacci et al., 1995; Alborn et al., 1997, 2007; Halitschke et al., 2001; Schmelz et al., 2006). The ecological limitation(s) and physiological constraints maintaining less-than-maximal photosynthetic activity in undamaged plants have to be addressed to fully understand the ecological consequences and coevolutionary trajectories as well as the potential applicability of tolerance-enhancing mechanisms in agriculture.
We thank Ian Baldwin, Anurag Agrawal, Jed Sparks, Carsten Kühlheim and Robert Raguso for comments on an earlier version of the manuscript. The study was funded by the National Science Foundation (NSF-IOS 0950225) and Cornell University, Ithaca, NY, USA.