Induced Volatile Emissions, Photosynthetic Characteristics, and Pigment Content in Juglans regia Leaves Infected with the Erineum-Forming Mite Aceria erinea

: Persian walnut ( Juglans regia L., Juglandaceae), one of the essential nut crops, is affected by different diseases, including mite attacks which result in gall and erineum formation. As the proportion of leaf area covered by mite galls or erineum is typically relatively low, the impact on tree photosynthetic productivity is often considered minor, and no pest control management is usually suggested. However, the effect of erineum-forming mites on walnut photosynthesis might be disproportionately larger than can be predicted from the leaf area impacted. In the present study, we studied how the foliage photosynthetic characteristics, pigment contents, and stress-induced volatile organic compounds scaled with the severity of infection varied from 0% (control trees) to 9.9%, by erineum-forming mite Aceria erinea in J. regia . Both leaf net assimilation rate (up to 75% reduction) and stomatal conductance (up to 82%) decreased disproportionately, increasing infection severity. Leaf total chlorophyll and β -carotene contents also decreased with infection severity, although the reduction was less than for photosynthetic characteristics (28% for chlorophyll and 25% for β -carotene). The infection induced signiﬁcant emissions of green leaves volatiles (( Z )-3-hexenol, ( E )-2-hexenal, ( Z )-3-hexenyl acetate and 1-hexanol), monoterpenes and the homoterpene 3-( E )-4,8-dimethyl-1,3,7-nonatriene, and these emissions scaled positively with the percentage of leaf area infected. These results collectively indicate that erineum-forming mite infection of walnut leaves results in profound modiﬁcations in foliage physiological characteristics that can signiﬁcantly impact tree photosynthetic productivity. L.C., Ü.N., and D.M.C.; writing—review and editing, Ü.N., L.C., and D.M.C.; visualization, L.C.; supervision, L.C., D.M.C.; project administration, L.C., D.M.C.; funding acquisition, L.C., D.M.C.


Introduction
Persian walnut (Juglans regia L., Juglandaceae) is a deciduous tree species native to central Asia. It is one of the most important nut crops with a wide area of cultivation, including Europe, North America, Central, and East Asia, North Africa, New Zealand [1]. In Europe, it is primarily grown in central to southern countries, although cultivars selected for superior cold hardiness can survive in the south part of Northern Europe [2]. The most extensive J. regia plantations are in Turkey, France, Romania, Serbia, and Hungary. As a tree species with a sequenced genome, it becomes one of the model species for physiological studies in woody plants [3]. positively with the infestation severity by petiole gall aphids in leaves of P. × petrovskiana.
In [17], in gall wasp-infected leaves of Quercus robur, constitutive isoprene emissions decreased in all cases with increasing infection severity. In contrast, the emission of monoterpenes was increased in leaves infected by Neuroterus albipes, Cynips divisa, and C. quercusfolii wasps and reduced in leaves infected by N. anthracinus [17]. In another study, monoterpene emissions were greater in leaves of Tilia cordata infected by Eriophyes tiliae gall-forming and E. exilis erineum-forming mites, while monoterpene emissions decreased in Alnus glutinosa leaves infected by E. inangulis gall-forming mites [20]. These studies provide encouraging evidence that alterations in constitutive and induced emissions might be used to diagnose the type and severity of infections of gall-and erineum-forming arthropods.
In J. regia, the gall-and erineum-forming mites all belong to Eriophyoidea: Aceria brachytarsus, A. avanensis, A. erinea, and A. tristata [10]. Walnut blister mite (Aceria erinea Nalepa) is a mite that causes erinea resembling yellowish blisters on the upper surface on walnut leaves [11,33]. The blisters correspond to hollows on the leaf underside; the hollows are covered with a dense whitish or pale brown mat of hypertrophic trichomes harboring mites [34] for a description of erineum structure. Although having a spectacular visual appearance, the A. erinea-induced erinea typically covers a low proportion of leaf area, less than 10%. Still, it is likely that leaf physiological characteristics are altered by erinea to a much greater degree than expected based on the degree of infection.
In this study, we hypothesized that A. erinea infections (1) reduce photosynthesis characteristics and photosynthetic pigment concentration, and elicit volatile stress emissions in leaves of J. regia in an infection severity-dependent manner, and (2) that the relative alterations in leaf traits are greater than changes in the proportion of area infected. This study demonstrates that even moderate infections of A. erinea signals strongly reduced leaf photosynthetic activity and enhanced production of stress volatiles that can be employed to predict the reduction in foliage photosynthetic productivity.

Study Site and Plant Material
The study takes place in Arad, Romania (46 • 17 N, 21 • 17 E, elevation~60 m above sea level) close to Mures river in the middle of July 2020. For July 2020, the average minimum air temperature was 14.6 • C, and the maximum air temperature was 27.8 • C (National Meteorological Administration. Available online: https://www.inmh.ro (accesssed on 28 June 2020)). The Juglans regia trees chosen for performing the measurements were 20-30 years old and 20-25 m tall. South-exposed leaves with varying degrees of visual infection symptoms were selected to measure gas-exchange parameters, and volatile organic compound (VOC) emission rates with the leaves attach to the tree. Control non-infected leaves were measured in nearby healthy trees. After that, all leaves have been collected, used to assess the degree of infection followed by extraction of pigments.

Foliage Gas-Exchange Measurements
In the field, photosynthetic characteristics of attached leaves with different degrees of infection were measured using a portable gas-exchange system GFS-3000 with 8 cm 2 leaf chamber (Waltz, Effeltrich, Germany) as described in [35] under the following environmental conditions: 65% air humidity in the chamber, 400 µmol mol −1 CO 2 concentration, 1000 µmol m −2 s −1 (10% blue and 90% red LED light) the incident light intensity, and 25 • C leaf temperature. After the chamber's enclosure, the leaf was maintained until stomata opened and steady-state CO 2 and water vapor exchange rates were reached. The stabilization period was typically 20-30 min. Once in the steady-state, foliage gas exchange rates were logged. For calculation of the values of stomatal conductance to water vapor (g s ) and net assimilation (A) the same procedure described in [36] has been used.
Other environmental conditions were maintained as above. At each light intensity, the gas exchange rates were stabilized until the steady-state values were observed. Lightresponse curves were measured in triplicate for control and leaves with approximately 3.5% and 5.1% infection. The light response curves were fitted by the rectangular hyperbola of Smith [37] as in [38], and dark respiration rate (R d ), initial quantum yield (α) and maximum net assimilation rate (A max ) were obtained.

Leaf Pigment Analysis
Pigment extraction was accomplished according to the method defined in [39], followed by the HPLC analysis described in [40].

Volatile Sampling and GC-MS Analyses
An air sample pump 210-1003 MTX (SKC Inc., Houston, TX, USA) with the constant flow was used to sample volatile organic compounds (VOC) via the outlet of the gasexchange system using the same procedure and multibed stainless steel cartridge as described in [41]. The flow rate for the volatile collection was 200 mL min −1 , and volatiles from the 4 L chamber air was collected. The adsorbent cartridges were analyzed for different volatile organic compounds as green leaf volatiles (GLV) and terpenes using a Shimadzu TD20 automated cartridge desorber integrated with a Shimadzu 2010 Plus GC-MS instrument (Shimadzu Corporation, Kyoto, Japan) following the method of [41,42]. Volatile emission rates were calculated according to Niinemets et al. [43]

Estimations of the Leaf Area and the Degree of Infection
The leaves were scanned at 200 dpi, and the leaf area was infected. The projected leaf area was estimated using custom-built software "Leaf Area Measurement" (www. plant-image-analysis.org, accessed on 5 May 2021) ( Figure 1 for representative examples of control and infected leaflets). In representative leaves with two different degrees of infection, leaf gas exchange's parameters light-response curves were also recorded with the same system (GFS-3000, Heinz Walz GmbH, Effeltrich, Germany). The light-response curve measurements were conceded in the following sequence (light intensities in µmol m −2 s −1 ): 0  50  100  200  300  400  500  600  800  1000  1200  1500  1800.
Other environmental conditions were maintained as above. At each light intensity, the gas exchange rates were stabilized until the steady-state values were observed. Lightresponse curves were measured in triplicate for control and leaves with approximately 3.5% and 5.1% infection. The light response curves were fitted by the rectangular hyperbola of Smith [37] as in [38], and dark respiration rate (Rd), initial quantum yield (α) and maximum net assimilation rate (Amax) were obtained.

Leaf Pigment Analysis
Pigment extraction was accomplished according to the method defined in [39], followed by the HPLC analysis described in [40].

Volatile Sampling and GC-MS Analyses
An air sample pump 210-1003 MTX (SKC Inc., Houston, TX, USA) with the constant flow was used to sample volatile organic compounds (VOC) via the outlet of the gas-exchange system using the same procedure and multibed stainless steel cartridge as described in [41]. The flow rate for the volatile collection was 200 mL min −1 , and volatiles from the 4 L chamber air was collected. The adsorbent cartridges were analyzed for different volatile organic compounds as green leaf volatiles (GLV) and terpenes using a Shimadzu TD20 automated cartridge desorber integrated with a Shimadzu 2010 Plus GC-MS instrument (Shimadzu Corporation, Kyoto, Japan) following the method of [41,42]. Volatile emission rates were calculated according to Niinemets et al. [43]

Estimations of the Leaf Area and the Degree of Infection
The leaves were scanned at 200 dpi, and the leaf area was infected. The projected leaf area was estimated using custom-built software "Leaf Area Measurement" (www.plantimage-analysis.org, accessed on 5 May 2021) ( Figure 1 for representative examples of control and infected leaflets).

Statistical Analysis and Data Handling
Altogether 20 leaves with different degrees of infection and five control leaves were measured for foliage gas-exchange characteristics and volatile emissions. Initially, we

Statistical Analysis and Data Handling
Altogether 20 leaves with different degrees of infection and five control leaves were measured for foliage gas-exchange characteristics and volatile emissions. Initially, we tested them using the null hypothesis (parametric one-way ANOVA) whether there were any differences between different tree leaves. Because there was no difference and it was demonstrated that the tree effect was not significant, all data were pooled for the regression analyses. Linear and non-linear regressions (Michaelis-Menten type equation: y = a·x/(b + x)) were used to explore the relationships of foliage physiological characteristics and pigment contents with the degree of infection. The parameters of the light response curves among control and differently-infected leaves (n = 3 in each case) were compared by parametric ANOVA followed by the Tukey post hoc test for pairwise comparison. All statistical analyses were conducted with GraphPad Prism version 9.1.2 for Windows (GraphPad Software, San Diego, CA, USA), and the statistical tests were considered significant at p < 0.05.

Photosynthesis Characteristics of Leaves Infected with A. erinea
The infection with A. erinea induced a drastic decrease of average net assimilation rate from 14.8 ± 1.0 µmol m −2 s −1 in control leaves to around three µmol m −2 s −1 in heavily infected leaves (Figure 2a). Altogether 20 leaves with different degrees of infection and five control leaves were measured for foliage gas-exchange characteristics and volatile emissions. Initially, we tested them using the null hypothesis (parametric one-way ANOVA) whether there were any differences between different tree leaves. Because there was no difference and it was demonstrated that the tree effect was not significant, all data were pooled for the regression analyses. Linear and non-linear regressions (Michaelis-Menten type equation: y = ax/(b + x)) were used to explore the relationships of foliage physiological characteristics and pigment contents with the degree of infection. The parameters of the light response curves among control and differently-infected leaves (n = 3 in each case) were compared by parametric ANOVA followed by the Tukey post hoc test for pairwise comparison. All statistical analyses were conducted with GraphPad Prism version 9.1.2 for Windows (GraphPad Software, San Diego, Ca, USA), and the statistical tests were considered significant at p < 0.05.

Photosynthesis Characteristics of Leaves Infected with A. erinea
The infection with A. erinea induced a drastic decrease of average net assimilation rate from 14.8  1.0 µ mol m −2 s −1 in control leaves to around three µ mol m −2 s −1 in heavily infected leaves (Figure 2a).

Modification of Light Responses of Photosynthesis by A. erinea Infection
Analysis of the light response curves further confirmed the major reduction in photosynthesis rate. The maximum net photosynthetic rate was higher for control plants (16.41  0.10 µmol m −2 s −1 ) than for infected plants (11.94  0.28 µmol m −2 s −1 for average degree of infection of 3.5% and 10.66  0.27 µmol m −2 s −1 for 5.1% infection, p < 0.05 for comparison with the control, Figure 3a). The initial quantum yield decreased from 0.0324  0.0005 μmol μmol -1 in control leaves to 0.0083  0.0003 μmol μmol -1 in leaves with 5.1% infestation (p < 0.05). However, differences in the dark respiration were minor among control and infected plants (Figure 3a). Even a minor infection (0.9% degree of infection) reduced leaf assimilation rate by 10-15% (Figure 2a). Stomata conductance to water vapor declined with the infection rate from 151.6 ± 2.3 mmol m −2 s −1 in control leaves to values of 23-25 mmol m −2 s −1 in leaves with more than 7% infection (Figure 2b).

Modification of Light Responses of Photosynthesis by A. erinea Infection
Analysis of the light response curves further confirmed the major reduction in photosynthesis rate. The maximum net photosynthetic rate was higher for control plants (16.41 ± 0.10 µmol m −2 s −1 ) than for infected plants (11.94 ± 0.28 µmol m −2 s −1 for average degree of infection of 3.5% and 10.66 ± 0.27 µmol m −2 s −1 for 5.1% infection, p < 0.05 for comparison with the control, Figure 3a). The initial quantum yield decreased from 0.0324 ± 0.0005 µmol µmol −1 in control leaves to 0.0083 ± 0.0003 µmol µmol −1 in leaves with 5.1% infestation (p < 0.05). However, differences in the dark respiration were minor among control and infected plants (Figure 3a).  Similar to the net assimilation rate, stomatal conductance increased with the increasing light intensity (Figure 3b). The difference in stomatal conductance between control and infected leaves was particularly large at a lower light intensity of 0-400 μmol m −2 s −1 than at higher light intensity of 400-1800 μmol m −2 s −1 (Figure 3b). The ratio of stomatal conductance at light intensities of 400 μmol m −2 s −1 and 1800 μmol m −2 s −1 was lower for control (15.13 ± 0.15) than for infected leaves (2.79 ± 0.35 for leaves with 5.1% infection, p < 0.05).

Responses of Foliage Chlorophylls and β-Carotene Contents to A. erinea Infection
Both chlorophyll a and b contents reduced intensely with the area infected by the A. erinea, e.g., the average chlorophyll content decreased from 742  4 mg m −2 in non-infected leaves to 554 mg m −2 in the leaves with a considerable infestation (Figure 4a). The chlorophyll a/b ratio varied between 2.25 and 2.75 (average ± SE = 2.49 ± 0.17) for all leaves sampled and did not depend on the degree of infestation (p > 0.05).   (Figures 2, 5, and 6). The data have been fitted using linear regression. Similar to the net assimilation rate, stomatal conductance increased with the increasing light intensity (Figure 3b). The difference in stomatal conductance between control and infected leaves was particularly large at a lower light intensity of 0-400 µmol m −2 s −1 than at higher light intensity of 400-1800 µmol m −2 s −1 (Figure 3b). The ratio of stomatal conductance at light intensities of 400 µmol m −2 s −1 and 1800 µmol m −2 s −1 was lower for control (15.13 ± 0.15) than for infected leaves (2.79 ± 0.35 for leaves with 5.1% infection, p < 0.05).

Responses of Foliage Chlorophylls and β-Carotene Contents to A. erinea Infection
Both chlorophyll a and b contents reduced intensely with the area infected by the A. erinea, e.g., the average chlorophyll content decreased from 742 ± 4 mg m −2 in noninfected leaves to 554 mg m −2 in the leaves with a considerable infestation (Figure 4a). The chlorophyll a/b ratio varied between 2.25 and 2.75 (average ± SE = 2.49 ± 0.17) for all leaves sampled and did not depend on the degree of infestation (p > 0.05).  Similar to the net assimilation rate, stomatal conductance increased with the increasing light intensity (Figure 3b). The difference in stomatal conductance between control and infected leaves was particularly large at a lower light intensity of 0-400 μmol m −2 s −1 than at higher light intensity of 400-1800 μmol m −2 s −1 (Figure 3b). The ratio of stomatal conductance at light intensities of 400 μmol m −2 s −1 and 1800 μmol m −2 s −1 was lower for control (15.13 ± 0.15) than for infected leaves (2.79 ± 0.35 for leaves with 5.1% infection, p < 0.05).

Responses of Foliage Chlorophylls and β-Carotene Contents to A. erinea Infection
Both chlorophyll a and b contents reduced intensely with the area infected by the A. erinea, e.g., the average chlorophyll content decreased from 742  4 mg m −2 in non-infected leaves to 554 mg m −2 in the leaves with a considerable infestation (Figure 4a). The chlorophyll a/b ratio varied between 2.25 and 2.75 (average ± SE = 2.49 ± 0.17) for all leaves sampled and did not depend on the degree of infestation (p > 0.05).  (Figures 2, 5, and 6). The data have been fitted using linear regression.  (Figures 2, 5 and 6). The data have been fitted using linear regression.
In addition, the emission of characteristic stress terpenes, the monoterpene (E)-βocimene, and DMNT also scaled positively with the infection percentage ( Figure 6).

The Influence of Aceria erinea Infection on Photosynthetic Characteristics of Juglans regia
The general response of plants to infection by different pathogens includes reducing leaf net assimilation rate [22,25,28,44]. According to previous studies, our data also demonstrated a substantial decrease in the net assimilation rate for infected plants. Notably, the relative reduction in net assimilation rate observed in A. erinea-infected J. regia leaves was much greater (up to 80%) than the degree of infection (up to 10%, Figure 2a). This indicates that erineum-formation led to profound physiological modifications in both erineum-infected and non-infected leaf areas. Furthermore, it suggests that the impact of gall-and erineum-forming mites can be much more significant than generally thought based on a relatively small fraction of the area infected in individual leaves.
The relative reduction in leaf net assimilation rate in erineum-infected leaves can be due to stomatal and non-stomatal (reductions in leaf photosynthetic capacity) factors. In the present study, the relative decrease in net assimilation rate was related to reduced stomatal conductance ( Figure 2b) and leaf water loss. An analogous negative influence of gall-and erineum-forming mites on water loss and photosynthesis has been found in Vitis vinifera cultivars infected with erineum-forming mite Colomerus vitis [45], Acer saccharum trees galled by the mite Vasates aceriscrumena [46], in Alnus glutinosa trees infected by Eriophyes inangulis, and Tilia cordata trees infected by E. exilis [20]. The relative reduction in stomatal conductance in pest-infected leaves could be explained by impaired stomatal function in gall-and erineum-covered leaf regions and surrounding leaf tissues due to pathogenic infection or formation of necrotic areas [17,28,47]. However, stomatal conductance is typically reduced less than the net assimilation rate, resulting in reduced leaf water use efficiency [17,48], as observed in our study.
The evidence of a greater reduction in net assimilation rate than in stomatal conductance suggests that the infection also reduced leaf photosynthetic capacity. Typically, leaf photosynthetic capacity is significantly reduced in the infected leaf regions, reflecting altered leaf anatomy and reduced photosynthetic pigment content [46], as observed in our study (Figure 4). A significant reduction in leaf chlorophyll content has been shown in the horn-shaped galls induced by a cecidomyiid on Copaifera langsdorffii leaves [49] and leaves of Cinnamomum tamala with Aceria doctersi galls [50]. The decrease in chlorophyll contents has been connected with a dilution of pigments by cell hypertrophy in leaves with galls [51,52]. However, in our study, the relative decrease in pigment content in J. regia at given infection severity was much less than in the net assimilation rate (cf. Figures 2 and 4).
In fact, there is evidence that the mites might also affect the sink-source relationships (by interaction with others sinks within integrated physiological units) during leaf growth and thereby alter photosynthetic biomass accumulation per unit leaf area and whole leaf photosynthesis rate [53]. Such a sink-source impact has been demonstrated in Populus × petrovskiana infected by petiole gall aphids; the aphid galls are formed on petioles, and thus, the aphids do not directly affect leaf lamina, but nevertheless, leaf photosynthesis rate was strongly reduced in leaves with infested petioles (Ye et al. 2019). It is plausible that the disproportionately more significant reduction in net assimilation rate at the given degree of infection in J. regia is also associated with altered leaf development and decline of photosynthetically active biomass per unit area in infected leaves.

Emission of Volatile Organic Compounds from Leaves Infected with A. erinea
The green leaf volatiles (GLV, C5 and C6 alcohols, aldehydes, and ketones) are specific elicitor molecules emitted from damaged plant membranes, usually after herbivory, pathogen, or oxidative stresses [28,29,[54][55][56]. In the case of A. erinea infection in J. regia, GLV emissions is amplified by increasing the degree of infection (Figure 4a). Similar scaling of GLV emissions with the degree of infection of gall-or erineum-forming arthropods has also been observed in other studies [17,20,25]. However, quantitatively, the rate of GLV emission observed in these studies was less, except in Quercus robur leaves infected by Neuroterus anthracinus wasp galls where the rates of GLV emission were similar to J. regia leaves with A. erinea infection of 1-2% [17] (Figure 5).
A sustained GLV emission could be a direct response to the immediate sites of impact during mite feeding, reflecting the free fatty acids discharge from damaged plant membranes and activation of lipoxygenase at the sites of feeding within the erinea [57,58]. However, once formed, the primary GLV can translocate to other leaf regions with the transpiration stream and can be derivatized before the release [59]. In the case of infection with A. erinea, GLV emissions were dominated by 1-hexanol, suggesting that derivatization of a part of primary aldehydes could have occurred in non-damaged leaf parts. Analogously, a large share of derivatized GLV was found in Quercus robur leaves infected with Neuroterus albipes and Cynips spp. [17] and in Alnus glutinosa infected by Eriophyes inangulis mites and Tilia cordata infected by E. tiliae mites [20].
The leaves of J. regia do not emit monoterpenes in non-stressed conditions, but this species has glandular trichomes on the leaf surface, and leaf essential oil extractions contain different mono-and sesquiterpenes [60,61]. Thus, a certain fraction of monoterpene release in A. erinea-infected mites might be originated from terpene released upon breakage of glandular trichomes. However, given that the erinea harboring the mites are primarily covered by non-glandular hypertrophic trichomes, the contribution of glandular trichomes to terpene release in mite-infected leaves is likely minor. On the other hand, de novo synthesis of monoterpenes is typically elicited in response to pest and parasite attacks as a part of the induced suite of traits involved in direct and indirect defense (for review, see [62,63]). As a part of the direct defense, terpenes can be toxic to infecting organisms, including mites [63,64]. Volatile terpenes also play a compelling role in indirect defense by attracting predators of plant-attacking insects and mites [63,65,66]. In the case of J. regia leaves, there is only limited information on arthropod-induced terpene blends. Infestation of J. regia leaves by Meliboeus ohbayashii ssp. primoriensis larvae induced emissions of monoterpenes limonene, sabinene, β-phellandrene, α-pinene, and β-pinene and sesquiterpenes germacrene D, (E,E)-α-farnesene [67]. In our study, J. regia leaves infected with A. erinea emitted monoterpenes α-thujene, (E)-β-ocimene, and 3-carene and its oxidized derivative 3-caren-2-ol, and total monoterpene emissions scaled with the percentage of the leaf area infected (Figure 4b). In addition, the emissions of two characteristic terpenes de novo synthesized in stressed leaves, (E)-β-ocimene and the homoterpene DMNT, which also scaled positively with the degree of leaf area damage ( Figure 6), suggesting that the positive scaling of total terpene emission with the degree of infection in our study (Figure 4b) primarily reflected de novo synthesis of terpenes. The significant differences in the induced terpene emission blends from J. regia infected by the mite A. erinea in our study and the buprestid insect M. ohbayashii ssp. primoriensis [67] suggest that the induced terpene composition provides informative insights for identifying the infecting organism. We suggest that further work with different walnut-infecting organisms is needed to characterize the modification of the emission blend and the degree of elicitation of volatile emissions as determined by various leaf-attacking arthropods.

Conclusions
Our study demonstrates the infestation of walnut leaves with Aceria erinea strongly affects the plant's metabolic and physiological processes. In addition to visual symptoms, the infection led to a decline in photosynthetic characteristics, emission of volatile organic compounds, and a decrease in chlorophylls and carotene contents. All chemical and physiological modifications were quantitatively related to the percentage of the leaf area infected. In contrast, the decline in photosynthetic characteristics was much more significant than could be predicted based directly on the infected leaf area. We conclude that A. erinea infection leads to systemic alterations in J. regia leaf physiological activity.