Discussion
The data presented here demonstrates how exposure of tomato plants to an abiotic stress (salinity), can influence a biotic stress (insect herbivory). Our findings reveal that the bottom-up effects of salinity stress exert discernible influences on H. zea biology, including caterpillar growth rates, tissue selection, and moth oviposition. These effects are potentially mediated by reduced leaf relative water content, leaf total protein content, plant nutrient status, and ionic imbalances, and not entirely via anti-herbivore defense proteins, or the emission of specific plant volatiles. Usually, co-occurrence of multiple stressors can result in either individual dominant, additive, or synergistic effects (Holopainen and Gershenzon, 2010).
Many studies report a positive correlation between water stress and insect herbivory (Thaler and Bostock, 2004). Drought stress is known to drive insect outbreaks in certain cases (Mattson and Haack, 1987; Baoet al. , 2019; Lantschner, Aukema and Corley, 2019). The “plant stress hypothesis” states that insects would find it easier to grow on stressed plants due to the lower ability of plants to defend themselves. A study on leaf miners revealed that leaf miners exhibited accelerated development and shorter developmental times on tomato plants under elevated salinity (Han et al. , 2016). The carmine spider mite (Tetranychus cinnabarinus ) showed faster development on strawberry cultivars under salinity stress (Cakmak and Demiral, 2007). In a no-choice assay, we observed a decrease in H. zeacaterpillar growth rates with an increasing level of salinity in their plant diet (Figure 2 ). We did not observe any changes in leaf area consumption by caterpillars feeding on no salt and salt-treated substrates (unpublished results). This would indicate that lower caterpillar growth rates are not due to lower tissue consumption but are due to other factors. This observation substantiates the “plant vigor hypothesis,” suggesting H. zea prefers plants grown in non-saline conditions, as they would be healthier and exhibit a “higher vigor.” Salt priming of Arabidopsis thaliana seeds led to elevated ROS and glucosinolates, which was correlated to lower performance of the generalist common cutworm caterpillar (Spodoptera litura ) (Xiao et al. , 2019). Another study demonstrated that Spodoptera exigua larvae were less likely to damage tomato plants treated with > 50 mM NaCl, consumed less leaf tissue, displayed decreased pupation rates, and had shorter lifespans (Marsack and Connolly, 2022). It is evident from many such studies that the indirect influence of salinity stress on insects is contingent upon various factors, including the insect feeding guild and plant species.
Several factors, such as the levels of anti-insect herbivore defense proteins, leaf water content, leaf nutrient changes, and ionic toxicity, may collectively contribute to the observed reduction in H. zeacaterpillar growth rates. Dombrowski 2003 reports salt stress-induced accumulation of proteinase inhibitor II (PIN2 ) transcripts and amplified tomato responses to wounding via the wound-signalling geneLOXD transcripts. To investigate the interactive effect of salt stress and insect herbivory on tomato defense responses, we measured the activity of two jasmonic acid (JA)-inducible plant defense proteins: polyphenol oxidase (PPO) and trypsin proteinase inhibitor (TPI) (Tanet al. , 2018). However, in spectrophotometric assays, we did not observe any impact of salt addition on tomato PPO and TPI levels (Figure 6 ). Similarly, a prior study (Winter et al. , 2012) failed to detect salt-stress-induced proteinase inhibitor transcripts or herbivore-induced plant volatiles (HIPVs). It is possible, however, that 200 mM NaCl depicts a highly stressful concentration where plants choose to maintain nutrient homeostasis and growth rather than investing resources in defense metabolites (Abrol, Vyas and Koul, 2012). Investigation of low concentration salt priming could provide more insights into this hypothesis (Xiao et al. , 2019). Gene expression studies showed that GAME4 (glycoalkaloid metabolism gene 4) and AspPI (aspartic proteinase inhibitor) gene expression increases significantly under herbivory and high salt (Supplementary Figure S4) , indicating that salt could prime plants for a higher induced response to insect herbivory through certain defenses such as glycoalkaloids or aspartic proteinase inhibitors. A study in maize showed that although salinity alone did not induce defense gene expression (Zm-Bx1 and Zm-SerPIN ), joint salinity and insect herbivory did (Forieri, Hildebrandt and Rostás, 2016). There are some indications of crosstalk between salinity and insect herbivory pathways, although the precise defenses and genes involved need to be investigated further.
Consistent with prior studies, we observed a substantial reduction in leaf relative water contents following salt treatment (Figure 5 ) (Cuartero and Fernández-Muñoz, 1998; Wang, Haseeb and Zhang, 2021). A study (Lin, Paudel, et al. , 2021) reported M. sextalarvae consumed lower amounts of plant tissue under severe drought stress. M. sexta larvae showed compensatory feeding in later instars by insects feeding more on low-water-containing diets (probably to make up for water deficiencies) (Ojeda-Avila, Woods and Raguso, 2003). While lower water contents in diet could account for lower caterpillar growth rates, additional factors such as altered nutrient availability and ionic toxicity can have additive or synergistic effects.
According to a study in soybean, salt-stressed plants are photosynthetically less productive and would lead to lower insect growth due to their lower nutritional quality (Najjar et al. , 2018). Many plants, in their bid to tolerate salt stress, typically possess the capacity to selectively exclude Na+ and Cl- ions while facilitating water uptake (Munns, 2005). For example, the halophyte Hordeum marinum starts accumulating Na+ and Cl- ions only after exposure to 450 mM NaCl (Garthwaite, von Bothmer and Colmer, 2005). However, a study on transgenic salt-tolerant tomato plants reported salt accumulation in tomato leaves at much lower concentrations (Zhang and Blumwald, 2001). Additionally, another study demonstrated the accumulation of Na+ in older tomato leaves, possibly as a protective mechanism to shield younger leaves from salt stress (Khelil, Menu and Ricard, 2007). A study under low salt conditions (5 mM) showed an increase in Na+, Cl-, C, S, Zn2+, and Cu2+ in A. thaliana shoots (Hongqiao et al. , 2021). In line with these studies, we detected an accumulation of Na+, Zn2+, Ca2+, Mg2+, and Mn2+ ions in salt-stressed leaves (Figure 9, Supplementary Figure S3 ). Notably, the micronutrients Zn2+ and Mn2+ are components of superoxide dismutases (Cu-Zn SOD and Mn-SOD) that catalyze H2O2 formation from superoxide radicals. The observed accumulation of Zn2+ and Mn2+ ions with increasing salt treatment may be indicative of their role in enhancing NaCl tolerance, as demonstrated by prior research in A. thaliana (Wang et al. , 2004; Liuet al. , 2015). Like some other studies, our study also shows that K+ uptake is hampered possibly through high-affinity K+ transporters, whereas Ca2+ and Mg2+ levels increase with salt application, possibly as salt tolerance mechanisms by the plant (Carter et al. , 2005; Fuchs et al. , 2005; Valdez-Aguilar, Grieve and Poss, 2009; Nieves-Cordones et al. , 2010; Koksal et al. , 2016; Akter and Oue, 2018). It is interesting to note that many insects seek salts actively, and the accumulation of sodium can offset the negative effects of plant secondary metabolites on insect herbivores (Joern, Provin and Behmer, 2012; Kaspari, 2020; Marroquin, Holmes and Salazar, 2023). However, a high level of Na+ ions could disrupt the insect’s osmotic balance, interfering with the excretory, nervous, digestive, and respiratory systems (Silver and Donini, 2021).Helicoverpa armigera larvae grew better on an artificial diet with a lower amount of sodium, indicating that higher sodium concentrations could be toxic to insects (Xiao et al. , 2010). We also observed lower caterpillar growth on a salt-containing artificial diet (Supplementary Figure S5 ), which indicates that higher salt levels (Na+ and Cl- ions) in the diet are detrimental to insect growth.
We do not disregard the influence of other chemical defenses or essential nutrients that are not measured here on H. zeacaterpillar performances. For instance, glycoalkaloids such as tomatine increase in salt-stressed plants and could potentially lead to decreased caterpillar growth and deter caterpillar feeding (Friedman, 2002; Hanet al. , 2016; Dong et al. , 2020). Studies also report that salinity affects the production of phenolics, terpenoids, and cyanogenic glycosides in different plants, all of which can affect insect growth (Wahid and Ghazanfar, 2006; Mahmoudi et al. , 2010; Ballhorn and Elias, 2014; Wang et al. , 2015; Marroquin, Holmes and Salazar, 2023). Tomato plants under salt stress would need to significantly increase their alkaloid production within three days for this hypothesis to hold. Interestingly, we observed that salt treatment alone did not lead to upregulation of GAME4 , an upstream gene in the steroidal alkaloid biosynthesis pathway (Dzakovich, Francis and Cooperstone, 2022). We did, however, see a higher induced response ofGAME4 in herbivore-damaged salt-stressed plants, which indicates some salinity-based priming against insect herbivory. Analyzing alkaloid contents would help validate this hypothesis. However, Thaler & Bostock 2004 report that salt stress by itself does not alter JA-inducible responses, which could be why we did not see any differences in gene expression and enzyme levels between salt treatments in no herbivory tissue.
Lastly, it is important to note that the effects of salt stress on insect herbivores can vary depending on the insect species and the severity of the salt stress. Some insect species can be more salt-tolerant than others. Our findings collectively emphasize that in salt-stressed plants, factors such as nutrient imbalances, ionic toxicity, and lower relative water contents (alone or in conjunction) (Martinez et al. , 2012) cause lower H. zea herbivore performances than alterations to anti-insect herbivore defense protein levels.
In petri dish-based choice assays, H. zea caterpillars avoided plants treated with salt (Figure 3 ). The initial behavior of the insect caterpillars revealed a random selection of leaves for feeding, as evident by the “first establish” parameter. This indicates that cut leaf-disc-based volatile or visual cues do not influence insect behavior. However, as time progressed, a discernible shift occurred, with caterpillars increasingly favoring leaves treated with lower salt concentrations for consumption. This caterpillar feeding preference was seen in caterpillar leaf area consumption as well as the “first finish” parameter. The observed change in caterpillar feeding preference complemented caterpillar growth rates, suggesting that caterpillars preferred to feed on no-salt-containing diets, which would result in a higher growth rate. Insect preference for plants under lower salinity is also observed in other studies (Quais et al. , 2020; Ali et al. , 2021; Marsack and Connolly, 2022).
Like the caterpillars, we observed a clear preference of H. zeamoths for no-salt-treated plants (Figure 4 ). Conversely, a study on monarch butterflies showed no difference in female oviposition between no-salt and salt-treated plants (Mitchell et al. , 2019). The “optimal oviposition theory” or “mother knows best hypothesis” suggests that female moths exhibit a preference for laying their eggs on plant hosts that will maximize the fitness and performance of their offspring (Jaenike, 1978; Valladares and Lawton, 1991; Courtney and Kibota, 2017). Female moths utilize various parameters to determine suitable host plants for oviposition. For many noctuid lepidopterans, olfactory cues have been identified as crucial in locating and selecting host plants (Jost and Pitre, 2002; Rojas, Virgen and Cruz-López, 2003; McCallum et al. , 2011). These cues include volatile compounds emitted by plants, which can either attract or repel moths based on their preferences and adaptations (De Moraes, Mescher and Tumlinson, 2001; Bruce, Wadhams and Woodcock, 2005). For example, isothiocyanates act as oviposition stimulants for the diamondback moth Plutella xylostella (Renwick et al. , 2006), and pyrrolizidine alkaloids stimulate oviposition in the cinnabar moth (Tyria jacobaeae ) (Macel and Vrieling, 2003). Studies report that salt stress changes the emission of three terpenes, (Z)-beta-ocimene, 2-carene, and β-phellandrene, which have been quantitatively correlated with salt concentration (Tomescu et al. , 2017). Green leaf volatiles such as (Z)-3-hexenal and terpenoids like β-phellandrene and α-pinene have been implicated in attracting lepidopteran moths for oviposition (Huanget al. , 2020; Agbessenou et al. , 2022; Sisay et al. , 2023). Like the abovementioned studies, we also observed decreased levels of β-phellandrene as well as 2-carene, α-phellandrene, β-caryophyllene, and α-humulene in salt-treated plants (Figure 10 ). Insect responses to plant volatiles depend on the insect species, the specific volatile compound in the volatile blend, and the concentration of relevant volatiles. Decreased constitutive emissions of β-phellandrene might cause lower H. zea moth oviposition on salt-treated plants. We did not observe any differences in total volatiles emitted per gram of fresh weight between stressed and unstressed plants, as corroborated by some other studies (Loreto and Delfine, 2000; Teuber et al. , 2008; Forieri, Hildebrandt and Rostás, 2016). However, the total volatiles emitted by the whole plant could be a more ecologically relevant unit.
Furthermore, moths may also rely on visual and tactile cues, including leaf shape, color, and texture, as indicators of plant quality for egg-laying (Rojas and Wyatt, 1999; Jakobsson et al. , 2017; Keerthi et al. , 2023). Plant physical characteristics, such as the presence and density of trichomes and leaf hairs, can likewise influence moth oviposition choices. For example, oviposition by P. xylostella on A. thaliana decreases with an increasing density of leaf trichomes (Handley, Ekbom and Ågren, 2005). The presence of glandular trichomes on a wild potato species (Solanum berthaultii ) acts as a deterrent to oviposition by the potato tuber moth (Phthorimaea operculella ) (Malakar and Tingey, 2000). However, some studies have reported that moths tend to lay eggs on plant parts exhibiting higher trichome densities and concentrated defense chemicals (Thomson, 1987). Additionally, the nutritional composition of host plants, including nitrogen levels, sugar content, and secondary metabolites, plays a vital role in shaping moth oviposition decisions (Heisswolf, Obermaier and Poethke, 2005; Chen, Ruberson and Olson, 2008; Gripenberg et al. , 2010). We did not observe any changes to leaf trichome density under salt stress (unpublished results). However, salt stress, leading to lower water availability, may potentially increase the concentration of chemicals in trichomes. Trichomes could also lose turgor and decrease in size, thus influencing moth oviposition. The influence of trichome-level changes on H. zea oviposition decisions cannot be ignored and remain to be investigated.
Lastly, moth host finding and consequently oviposition can also be altered due to local changes in relative humidity (Godfrey and Holtzer, 1991; Wolfin et al. , 2018). Salt stress would lead to stomatal closure and reduced transpiration, which would in turn reduce local humidity. We observed statistically non-significant decreases in leaf relative humidity under salt treatment for 5 days (preliminary data). These differences, although statistically non-significant, could still be ecologically significant for an ovipositing H. zea moth.