INTRODUCTION

The impact of herbivores on plants is extensive, removing more than 20 % of annual net primary productivity (^{Agrawal, 2011}). Damage caused by herbivores in agricultural systems has meant great economic losses, comparable to the combined effects of drought and frost (^{Coley et al., 1985}; ^{Oerke, 2006}). It has been estimated that crop losses due to damage caused by arthropods can exceed 15 % annually (^{Mitchell et al., 2016}). Although global productivity levels increased in agricultural systems during the 20th century, so did economic losses due to herbivore attack, despite extensive use of pesticides (^{Oerke et al., 2014}). This context is a challenge to seek sustainable alternatives for agricultural production that depend much less on chemical inputs such as pesticides. An emerging way to address this challenge is to exploit the standing defensive strategies of plants in a broader way (^{Mitchell et al., 2016}).

Changes in the expression of traits before and after herbivore attack, such as compensatory growth or increased chlorophyll content, are associated with tolerance to herbivory (^{Tiffin, 2000}); that is, the defensive strategy that confers plants the ability to reduce the negative effect of herbivores on crop yield once herbivory has occurred (^{Agrawal, 2011}; ^{Simms and Triplett, 1994}; ^{Stout, 2013}). Theoretical models indicate that since plant tolerance attributes generally have no effect on herbivore fitness, they are unlikely to impose a selective pressure on herbivores. This would limit the evolution of herbivore resistance to plant defenses (^{Garrido and Fornoni, 2006}; ^{Weis and Franks, 2006}), pointing out that traits associated with tolerance to herbivory may represent reliable strategies for long-term pest control (^{Mitchell et al., 2016}).

The domestication of plants for agriculture has favored desirable attributes that maximize yield and quality of harvested product in high-input environments (^{Meyer et al., 2012}); however, it has also generated modern varieties with relatively low levels of genetic diversity (^{Doebley et al., 2006}; ^{Khush, 2001}; ^{Wright et al., 2005}). This reduction in genetic variability could constrain the ability of these varieties to tolerate attack by herbivores (^{Bello-Bedoy and Núñez-Farfán, 2011}; ^{Stout, 2013}; ^{Weinig et al., 2003}); thus, it is relevant to study the different plant traits and mechanisms associated with tolerance to herbivory in domesticated plants. In the same vein, it is essential to identify potentially tolerant cultivated varieties, with the aim of integrating them into pest management programs and crop improvement strategies (^{Goggin et al., 2015}).

Planococcus ficus (Signoret) (Hemiptera: Pseudococcidae), popularly known as vine mealybug, is a phloem-feeding insect species and has become a key pest in vineyards around the world (^{Naegele et al., 2020}). Recently, it became an important plague in Southern California and Mexico. The infestation of the vines by this mealybug has meant great losses for the wine market in the United States and in the state of Sonora (Mexico), affecting up to 100 % of the yield in the agricultural region of the Coast of Hermosillo (^{Daane et al., 2006}; ^{Fu et al., 2004}), presents a series of characteristics that make it a serious threat to viticulture. It has a fast life cycle (4-7 generations per year) and a high reproductive rate, with some females laying more than 250 eggs (^{Daane et al., 2006}).

^{Timm and Reineke (2014)} found that grapevine plants respond weakly at the transcriptional level to the attack inflicted by P. ficus; however, the damage induced a lower expression of genes and transcripts related to plant defense, suggesting that other phenotypic traits related to tolerance may help the grapevine to defend itself against this pest. Recently, ^{Naegele et al. (2020)} studied the resistance to P. ficus of ten grapevine lines (species, cultivars and rootstocks), finding significant differences between cultivars and rootstocks in the number of P. ficus (estimated as juveniles, adults and egg sacs), being Cabernet Sauvignon and Chardonnay the most favorable grape cultivars for mealybug growth; however, the effect of P. ficus attack on tolerance-related traits such as plant growth or chlorophyll content of different grapevines cultivars has not been sufficiently studied.

The first year of growth is crucial in the life cycle of the vines since they must develop a strong root system that allows them to sustain their growth in subsequent years. Poor growth during the first year can increase the time required to fully train vines and the time required to achieve full production potential of the vineyard (^{Bettiga, 2015}); thus, this study was carried out to assess the effect of Planococcus ficus (Signoret) on plant growth of three different grapevine cultivars for winemaking; specifically, leaf chlorophyll content and plant growth (plant height, number of leaves, and leaf area per plant) were compared between infested and non-infested plants. This assessment will help to understand the role of these traits in tolerance to P. ficus infestation of three of the most important varieties of Vitis vinifera L. (Garnacha, Tempranillo and Nebbiolo) grown in Ensenada, Baja California, Mexico.

MATERIALS AND METHODS

RESULTS

The ANOVA showed a significant effect of variety and infestation levels on plant height RCR (Table 1A). Garnacha had significantly lower plant height RCR compared to the Nebbiolo and Tempranillo varieties (33.8 and 38.4 % respectively), which did not differ significantly from each other. Infested plants showed 17.9 % reduction in plant height RCR compared to non-infested plants; however, the analysis also detected a significant effect of variety × infestation on plant height RCR (Table 1A), indicating that the effects of infestation on RCR depended on the variety. Nebbiolo showed a significant (57.3 %) reduction in plant height RCR when it was infested by P. ficus, whereas Garnacha and Tempranillo maintained similar RCR values (Figure 1A).

Figure 1 Averages (± S.E.) of A) plant height RCR for infested and non-infested plants of the three varieties of Vitis vinifera, B) number of leaves RCR for the three varieties of V. vinifera, C) number of leaves RCR of infested and non-infested plants, D) Chlorophyll Content Index RCR and E) mean leaf area RCR of infested and non-infested plants in a common garden experiment with Planococcus ficus. Averages not sharing the same letter differ significantly after a Tukey-Kramer LSD post hoc test. Asterisk denotes significant differences between levels (P ≤ 0.05). 

Variety and infestation treatment had a significant effect on the RCR of the number of leaves (Table 1B). Garnacha variety had a significantly lower number of leaves RCR compared to Nebbiolo (31.2 %), while Tempranillo and Nebbiolo did not differ significantly from each other (Figure 1B). On the other side, infested plants showed a 22.6 % reduction in the number of leaves RCR compared to non-infested plants (Figure 1C). There was a significant effect of variety factor on CCI RCR (Table 1C). Tempranillo variety had a significantly higher CCI RCR than Nebbiolo and Garnacha, which did not differ from each other (Figure 1D). Finally, the infestation treatment also had a significant effect on the RCR of mean leaf area (Table 1D). Infested plants showed a 41.7 % lower value than non-infested plants (Figure 1E).

DISCUSSION

In general, results showed a negative effect of P. ficus infestation on the expression of three plant growth parameters associated with tolerance, plant height, number of leaves and mean leaf area per plant. Interestingly, the effect of P. ficus infestation on plant growth differs between grape varieties. Unlike the Garnacha and Tempranillo varieties, infested Nebbiolo plants showed a significant reduction in plant height compared to non-infested plants, pointing that this variety is less tolerant to the attack by P. ficus. Only chlorophyll content showed an association to the reduction in relative growth rate of infested Nebbiolo plants. The damage caused by this herbivore might impact mechanisms of photosynthetic rates that influence growth rates. The absence of a negative impact of mealybugs infestation on growth rate of Tempranillo and Garnacha plants indicated that these varieties possess mechanisms of tolerance to sap feeding, enabling them to grow at certain level of damage; thus, the differential effect of sap feeding supports the notion that there should be specific management, depending on the level of tolerance to damage.

Sap-feeders remove assimilates from the phloem or individual cells of their host; thus, the attack by sap-feeders is expected to negatively affect photosynthesis and plant growth (^{Gonda-King et al., 2014}; ^{Strauss and Agrawal, 1999}); however, the magnitude of the response of plants to herbivores is complex and it often changes within and between plant species (^{Walling, 2000}). Likewise, variation in the response of domesticated varieties to herbivore damage is common and its responses largely depend on the variety, even when plants are exposed to similar herbivores (^{Chen et al., 2015}). Results of this research reflect this complexity. Significant differences were found between varieties in the relative growth of plant height, number of leaves, and the chlorophyll content index. This is consistent with previous evidence of cultivar variation in shoot physiological and growth parameters in grapevines reported elsewhere (^{Bica et al., 1997}; ^{Borghezan et al., 2012}; ^{Gris et al., 2010}; ^{Gutiérrez-Gamboa et al., 2018}; ^{Malinovski et al., 2014}). In general, results of this study indicate that the exposure of V. vinifera to the vine mealybug P. ficus reduced the relative change rate of plant height (17.9 %), number of leaves (22.6 %) and leaf mean area (41.7 %), indicating that infestation by this herbivore reduces plant growth of grapevines. This agrees with results of ^{Zvereva et al. (2010)}, who found through a meta-analysis that sap-feeders reduced plant growth (~ 29 %), chlorophyll content (~ 27 %) and leaf biomass (~ 28 %) of woody plants; however, considerable reduction in plant height RCR was primarily observed in the Nebbiolo variety, whereas infested and non-infested plants of Garnacha and Tempranillo retained similar values, which indicates that the damage caused by the mealybug infestation on plant growth differed between varieties, pointing that the studied varieties differ in their ability to tolerate mealybug infestation, and that Nebbiolo is less tolerant to the damage by this sap feeding insect. Similar results were obtained by ^{Khederi et al. (2018)} who found that the negative effect of infestation by the mite Colomerus vitis on shoot length and leaf area differed between five Iranian grapevine varieties: Shahani, Sahebi Uroomie, Khalili Bovanat, Rishbaba and Sezdang Ghalat.

Chlorophyll content is another trait associated with tolerance commonly affected by the infestation of sap-feeding herbivores (^{Zvereva et al., 2010}). It has been reported that sap feeding damage causes a reduction of 27 % in the chlorophyll content in leaves of woody plants as compared to non-damaged plants (Zvereva et al., 2010); however, plants do not always show a reduction in chlorophyll content after herbivory, suggesting that maintaining chlorophyll content levels could be adaptive in terms of tolerance to herbivory (^{Hodge et al., 2000}). In this study, mealybug infestation did not affect variation in the chlorophyll content relative change rate of leaves, even after controlling variation attributable to differences amongst varieties; this agree with previous results from a comparative study between damaged and non-damaged domesticated Vitis vinifera plants that found that the damage by this phloem feeder was not enough to induce the expression of transcripts and genes of traits implicated in defense (^{Timm and Reineke, 2014}). Although it might seem counterintuitive, observed absence of a change in chlorophyll levels after leaf damage may have an important role in tolerance to herbivory (Hodge et al., 2000); however, to what extent, sustaining chlorophyll content in the presence of damage by P. ficus may sustain the reduction of compensatory growth traits remains to be further studied for the V. vinifera-P. ficus interaction.

CONCLUSIONS

A negative effect of P. ficus infestation on plant height, number of leaves and leaf area was found. Results of this study contribute to the understanding of the response of domesticated plants to herbivores, providing insights for future studies intended to identify tolerant varieties, as well as potential applications in decision-making for grapevine cultivation in the study region. In vineyards where mealybugs are a persistent problem, farmers must adopt more rigorous monitoring and control measures when growing varieties with low level of tolerance like Nebbiolo. It is also necessary to conduct long-term studies aimed to identify the cultivar-specific physiological mechanisms underlying the response of vines cultivars to the attack by this important herbivore.