Identification of Mango Cultivars’ Resistance Against Red Spider Mite: Impact of Climate Elements on Resistance Performance

Abstract

The use of resistant plants is recognized as an environmentally friendly measure for mite control. Oligonychus mangiferus, known as the mango red spider mite (MRSM), is a dangerous pest for mango production. To date, the resistance levels of the mango germplasms against the MRSM remain largely unknown. Furthermore, the environmental factors potentially influencing resistance performance have been seldom discussed. To fill those knowledge gaps, this study aimed to identify the resistance level of twelve mango cultivars against the MRSM. Based on three rounds of greenhouse and five seasons of field tests, cultivars with distinct resistant levels were identified. When exploring the climate impact, we found that for the susceptible cultivars, precipitation is the primary external environment factor altering the resistance performance, while temperature presents a secondary effect, and air humidity did not show a significant impact on MRSM resistance. By contrast, MRSM-resistant cultivars were not prone to be affected by changing climate conditions. Furthermore, yield tests indicated that the resistant cultivars can better reduce the yield losses compared with the susceptible ones. This study illustrated the climate element-driven effect on mango tree resistance performance against the MRSM, which can provide insight into insect pest management under changing climate conditions.

1. Introduction

Mango (Mangifera indica L.) is extensively cultivated in tropical and subtropical areas. During 2022, the global mango production reached 57.3 million tons [1]. The outbreak of insect pests and diseases poses a tremendous threat to mango tree growth and fruit development. Oligonychus mangiferus (Rahman and Sapra) (Acari: Tetranychidae), known as the mango red spider mite (MRSM), is a dangerous pest that is found in the main planting areas in many countries such as China (Guangxi, Hunan, Hainan and Taiwan provinces), South and Southeast Asian countries (India, Myanmar, Pakistan, Singapore, and Thailand), and African countries (Egypt, Mauritius and Mozambique). Their infestation causes the leaf chlorosis, significantly affecting plant photosynthesis and reducing the yield and quality of mango fruit [2,3]. Application of acaricides is currently the most popular control method for the MRSM. However, the dense and tall canopy of the tree, especially in those cultivars without dwarfing treatment, makes the acaricide difficult to apply and target. Moreover, the inappropriate use of acaricide may not only pose a significant impact on the natural enemy populations but also result in potential mite resistance problems [4]. There are also some studies about the use of predators to control MRSM [5], and good predation rates were observed in the field; nevertheless, disadvantages like lack of local predators, difficulty in the natural enemy introduction, unstable effectiveness, and acaricide stress hinder the promotion of biological methods in China.

Planting resistant cultivars is acknowledged as an economical, effective, and environmentally friendly method for agricultural mite control. Compared with breeding novel resistant materials, evaluating the resistance level of preserved mango cultivars is much more convenient and efficient. This strategy is even more pragmatic, especially in developing countries as most growers are unwilling to increase the acaricide input. There has been a long history of evaluation or identification of crop resistance to mite species. Several studies have made tremendous efforts to evaluate crop resistance to agricultural mites, which include several typical tropical and subtropical crops. For example, cassava resistance to two-spotted spider mite, known as Tetranychus urticae (Acari: Tetranychidae) [6], citrus resistance to the oriental red mite, Eutetranychus orientalis (Acari: Tetranychidae) [7], and coconut palm resistance to Raoiella indica (Acari: Tenuipalpidae) [8]. Despite the economic importance of the mango tree, most effort has been concentrated on evaluating germplasm resistance to pathogens [9,10,11]; and the resistance mechanism has been illustrated. In contrast, there is a lack of studies demonstrating mango cultivar resistance against MRSM.

Climate conditions, such as temperature, precipitation, and humidity significantly influence pest occurrence [12], which in turn affects the damage severity of the plant as well as subsequent resistance identification. Therefore, multiple-year climate elements should be taken into account to speculate the correlation between pest dynamic and resistance performance. It is necessary not only to screen a stable pest-resistant cultivar but also to plan additional control strategies, i.e., chemical control, to reduce the yield loss by insect pests. However, those studies were seldom conducted to assist the rational utilization of resistant plants in the field. To help fill in these knowledge gaps, we used 12 mango cultivars to conduct greenhouse and field resistance identifications. Moreover, which specific climate elements and how these elements may cause the fluctuation of resistance were delicately investigated. In addition, the competence of each mango cultivar in maintaining the yield was also investigated, based on both acaricide application and acaricide-free conditions. We expect to offer environmentally stable materials for MRSM management or future breeding of resistant mango plants.

2. Materials and Methods

2.1. Mango Cultivars

Twelve mango cultivars (Information can be seen in Table S1) originally derived from the mango germplasm nursery at Sanya Academy of Tropical Agricultural Sciences were used for greenhouse and field tests. Six-month-old mango seedlings were planted with cultivated soil in pots (soil/vermiculite ratio was 3:1) in a greenhouse for MRSM-resistance evaluation. All the seedlings were generated from grafting on root stocks (currently the most prevailing methods of mango propagation in China). The greenhouse conditions were well-controlled by an automatic temperature and humidity control system (Parkoo, Guangzhou Dongao electrical Co., Ltd., Guangzhou, China): light/dark photoperiod (14 h/10 h), temperature (28 ± 1 °C), and relative humidity (75 ± 5%). In addition, mango cultivars for field experiments were five-year-old trees under routine field management.

2.2. Laboratory Rearing of MRSM

MRSM adults were reared on the underside of healthy mango leaves (detached from the five-year-old cultivar Dashehari in the field) under the same conditions as the mango seedling culture mentioned above. A wet (water-saturated) paper strip was used to wrap around the leaf margin to prevent the mite escaping and to maintain the leaf as fresh. The leaves were replaced every 2–4 days before losing freshness.

2.3. Identification of Mango Resistance to MRSM

The severity of leaf damage serves as a basis for identifying mango cultivar resistance to the MRSM (Figure 1A). The MRSM damage symptoms were classified into five scales, depending on the leaf damage percentage. The leaf damage percentage was precisely analyzed using a Leaf Image Analyser (YMJ-E, Daji Co., Ltd., Hangzhou, China) (Figure 1B,C) as described previously [6]. The mite damage index (MDI) was calculated according to the following equation:

$$\mathrm{MDI}={\displaystyle \frac{\sum (\mathrm{S}\times {\mathrm{N}}_{\mathrm{s}})}{\mathrm{N}\times 5}}\times 100$$

where S represents the mite damage scale, Ns represents the corresponding quantity of damaged leaves, N represents the quantity of investigated leaves in total, and 5 indicates the five resistance levels, i.e., highly resistant (HR, 0.0~12.5), resistant (R, 12.6~37.5), moderately resistant (MR, 37.6~62.5), susceptible (S, 62.6~87.5), and highly susceptible (HS, MDI higher than 87.5).

2.4. Greenhouse Identification of Mango Cultivar Resistance to MRSM

Six seedlings were used to identify the MRSM-resistance level of each mango cultivar. All the potted mango seedlings were placed on the ground. For evaluating damage caused by artificial infestations, each cultivar (mites first reared on the detached leaf for two days, then placed on the plants to let the mite move naturally) was inoculated with approximately 50 female adult mites. Six functional leaves from the third leaf below the apical bud of each cultivar were sampled at 8 d post infestation (dpi), and the three most serious infested leaves were used to determine the resistance level according to the method mentioned above. After finishing the identification, the mango seedling was pruned to let it sprout again for the next round of identification. The greenhouse identification was performed for three consecutive rounds (in each round, the newly sprouted seedlings were used).

2.5. Greenhouse Evaluation of Various Mango Cultivars on the Reproduction and Development of the MRSM

The underside leaf (detached) surface of each mango cultivar was inoculated with approximately 50 (one-day-old) female adults. The mortality of the MRSM was recorded every day until 10 dpi (mortality was calculated based on the observed mite quantity (dead and alive mite), in which the escaped mite was not included). Moreover, the one-day-old freshly laid eggs were kept on the leaf until hatched. To investigate the effect on MRSM development, each newly emerged larva was placed on the underside leaf (from the same cultivar) surface grids separated by the water-saturated blotting paper strip. The developmental durations, including eggs, protonymph, and deutonymph of the F₀ generation were investigated every 12 h (50 tested MRSMs for each mango cultivar). In addition, the egg production and hatchability of individual female adults (F₀ generation) were recorded till the female adult died. The oviposition of each female and the hatchability of the F₁ generation were recorded using a microscope. Three biological replicates were conducted for each cultivar.

2.6. Field Evaluation of Mango Cultivar Resistance to MRSM

The field experiment was the perennial epidemic area of the MRSM, which indicates the MRSM population can spontaneously accumulate during the field test period without artificial inoculation. The details of the experiment site were listed as: A flat land (Figure S1A) located in Yazhou district, Sanya city, Hainan Province, China, 109.518° E, 18.259° N, climate characteristic: monsoon–tropical ocean; soil type: sandy clay loam; organic matter percentage: 13.4%; soil pH = 6.54; mineral elements: Mg²⁺ (4.58), Ca²⁺ (5.33), K⁺ (0.84), Na⁺ (7.54), and Cl⁻ (2.05). In addition, to minimize the influence of fertilizer on mango growth, identical fertilizer application was conducted in each season following local farmer’s operation: Compound water-soluble fertilizer (N:P₂O₅:K₂O = 20%:20%:20%, Siyang Biotechnology Co., Ltd., Shandong, China) was applied. Drip fertigation was applied on mid-Januray, March, and May for each season, respectively.

The five-seasons-long field evaluations were conducted from 2018 to 2023. The treatment plot for each cultivar was performed with a randomized complete block design. Grafted mango trees were eight years old, and each plot consisted of one row (5 mango trees, 3.0 m between rows, and 5.0 m between plants) (Figure S1B). In addition, four replicates (plots) were set for each cultivar (Figure S1C), which was 20 trees in total, and the cultivar Dashehari was grown in two rows in the buffer zone. The resistance identification was conducted in each season, from December of the previous year to April of the current year (once a month, on the 16th of each month, totaling five times each season). During the test, 5 plants from each plot were all selected and evaluated, and the 3 most infested mature leaves (with unique and obvious MRSM-infested symptoms) from the bottom, middle, and top canopies of the mango tree were selected for resistance evaluation (9 leaves in total). There was a total of 45 leaves for a plot and 180 leaves for a cultivar. In addition, the damage scale of each sampled leaf and the corresponding leaf quantity were all recorded. Furthermore, during the test, acaricide was not allowed to be sprayed, but the application of germicide azoxystrobin (SIPICAM Co., Ltd., Shanghai, China) was necessary if mango disease occurred. In addition, the field was weeded by herbicides (Glufosinate ammonium, HAWEISI Co., Ltd., Shandong, China) on February each season before the fruit stage. To identify the high-mite-density period in the mango leaves, which may provide a reference for the suitable identification time, an MRSM population dynamic survey was carried out on Dashehari from December of the previous year to November of the current year (five seasons’ survey (from 2018 to 2023), collected on the 16th of each month, totaling twelve times each season). The average mite density was recorded when sampling the leaves (the sampled leaf was the same leaf for mite resistance identification).

2.7. The Yield of Mango Fruit With or Without Acaricide Application

Acaricide application was applied in parallel with the MRSM-resistance identification test, and the same varieties and plot sets were used. Bifenazate (43% suspension concentrate, Qiangsheng Agriculture biotechnology Co., Ltd., Nanjing, China) was used as an MRSM control as it is effective for mite control and harmless to natural enemies [13]. Acaricide applications were conducted monthly from the month of flowering to the month of harvest (on the 16th of each month), which ranged from 4 to 6 times, depending on the cultivar. The yield of each cultivar was measured (total fruit yield including all mature, immature, and ripe fruits) in either MRSM-resistance identification (No acaricide) or acaricide applications tests (Yield test was performed once per season).

2.8. Climate Elements

Local climate elements, such as daily sum of temperature (hourly average of every day) of the month (T_SUM), sum of monthly precipitation (P_SUM), and relative air humidity (RAH) were collected by an automatic weather station (Zhongke Nenghui Technology Co., Ltd., Wuhan, China) approximately four hundred meters away from the field. During the month of conducting the mite dynamic survey, the climate elements mentioned above were recorded daily in the same month, from December 2018 to April 2023.

2.9. Statistical Analyses

2.9.1. Statistical Analyses for Resistance and Yield Performance

One-way analyses of variance (ANOVA), followed by Tukey’s Honestly Significant Difference (HSD) test was performed to analyze the MDI and the development and reproduction data of the MRSM after a normality and homogeneity of variance check. Spearman’s correlation analysis was conducted to elucidate the correlation between the MDI and climate elements, and Pearson’s correlation analysis was conducted to elucidate the relationship between the MDI and mite density per leaf. Moreover, a generalized linear mixed model (GLMM) was applied to evaluate the multiple effects, such as experimental years (EY), acaricide application (AA), or experimental mango cultivars (EMC), on the MDI or yield. GLMM was conducted using SPSS 19.0 (SPSS Inc., Chicago, IL, USA), and the heterogeneity of variances was calculated by a robust estimation method for standard errors (Huber-White sandwich estimator) [14].

2.9.2. Statistical Analyses of Climate Elements

Kendall’s Coefficient of Concordance (Kendall’s W)

Kendall’s W was used to evaluate the concordance of the greenhouse- and field-resistance identification results as well as to evaluate the climate elements (T_SUM, P_SUM and RAH) of the five seasons’ survey. In addition, the concordance levels were determined according to Kendall’s W values: very low concordance (<0.2), low concordance (0.2–0.4), moderate concordance (0.4–0.6), high concordance (0.6–0.8), and very high concordance (0.8–1.0) [15]. All of these analyses were conducted by SPSS.

Multiple Linear Least-Squares Regression for Climate Data

Multiple linear least-squares regression (MLR) was conducted for each mango cultivar to elucidate whether climate conditions may affect the resistance level. The MDI served as a dependent response variable, and P_SUM, T_SUM, and RAH (Tables S10–S12) were considered as explanatory variables for MLR models, which was set up by using GraphPad Prism version 8.4. The model equation Y (MDI) was shown below:

Y (MDI) = β₀ + β₁ X1 + β₂ X2 + β₃ X3

where X1, X2, and X3 denote P_SUM, T_SUM, and RAH, respectively. The normality of residuals was tested with the Shapiro–Wilk test.

Mann–Kendall Trend Analysis on Climate Data

A Mann–Kendall test (MK test), as a non-parametric trend test method, is commonly applied in trend analysis and mutation detection in all kinds of time series [16]. Two curves, named UF and UB (Supplementary Materials), were generated by using the MK test, in which the intersection of these two curves indicates a mutation time point [17]. In addition, if UFk in the equation (Supplementary Materials) is more than the significance levels, such as α = 0.05 (UFk ≥ Uα/2 = |±1.96|) and α = 0.01 (UFk ≥ Uα/2 = |±2.32|), it represents a high confidence level [18]. In addition, Theil–Sen’s estimator may generate a slope that is used to indicate the monotonic trend as well as the amplitude of change per unit time.

3. Results

3.1. Identification of Mango Cultivar Resistance to MRSM Under Greenhouse Condition

As presented in

Table 1. The mite-resistance level of twelve mango cultivars under greenhouse condition.

No.	Cultivars	Three Rounds of Greenhouse Tests						MDI Average	Overall RL
		1st	RL	2nd	RL	3rd	RL
1	Dashehari	88.89	HS	94.44	HS	95.83	HS	93.06 ± 2.12 a	HS
2	Keitt	90.28	HS	91.67	HS	94.44	HS	92.13 ± 1.22 a	HS
3	Yuexi	90.2	HS	93.06	HS	91.67	HS	91.67 ± 0.80 a	HS
4	Zihua	80.56	S	81.94	HS	91.67	HS	84.72 ± 3.49 b	S
5	India 901	79.17	S	73.61	S	77.78	S	76.85 ± 1.67 c	S
6	Red ivory	69.44	S	65.28	S	63.89	S	66.20 ± 1.66 d	S
7	Hongguang	76.39	S	79.17	S	79.17	S	78.24 ± 0.92 bc	S
8	Kent	54.17	MR	65.28	S	69.44	S	62.96 ± 4.55 d	S
9	Golden Phoenix	50.00	MR	45.83	MR	48.61	MR	48.15 ± 1.22 e	MR
10	Sanya	30.56	R	29.17	R	34.72	R	31.48 ± 1.66 f	R
11	Sunrise	33.33	R	30.56	R	22.22	R	28.70 ± 3.33 f	R
12	Tainong No. 1	9.72	HR	11.11	HR	9.72	HR	10.19 ± 0.46 g	HR

Note: MDI and RL denote mite damage index and resistance level, respectively. Different letters represent statistical differences in the MDI among cultivars. As annotated by letters in the ’average’ column, the analysis was conducted based on one-way analyses of variance (ANOVA), followed by Tukey’s Honestly Significant Difference (HSD) tests (p < 0.05).

and Tables S2–S4, different cultivars exhibited significantly different MDIs. The three rounds of resistance evaluation showed consistent results, except for the cultivars Kent and Zihua; Kent was identified as ‘moderately resistant (MR)’ in the first round while it was susceptible (S) in the second and the third rounds, but the overall resistance level of the three rounds’ identification was susceptible (S). Similarly, Zihua was susceptible in the first and second round but was highly susceptible (HS) in the third round, and the overall resistance level was susceptible. In addition, Kendall’s coefficient of concordance demonstrated that the results of three rounds’ greenhouse identification results showed low concordance; Kendall’s W value was 0.256, p = 0.510. In summary, three cultivars (Dashehari, Keitt, and Yuexi) were identified as HS, five cultivars (Zihua, India 901, Red ivory, Hongguang, and Kent) were identified as S, and two cultivars (Sanya and Sunrise) were identified as resistant (R). In addition, only one moderately resistant (MR) (Golden Phoenix) and one highly resistant (HR) cultivar (Tainong No. 1) were identified.

3.2. The Impact on Reproduction and Development of the MRSM While Fed on Different Mango Cultivars

The mortality, reproduction, and development of the MRSM were examined while fed on different mango cultivars. The results showed that the mortality of the MRSM was significantly different among the twelve cultivars. The MRSM fed on HS cultivars exhibited very low mortalities (Cumulative mortality within 10 dpi was lower than 10%). Furthermore, mites fed on R cultivars, i.e., Sanya and Sunrise, presented very high mortalities (ranging from 52.45 to 67.26%); in addition, the HR cultivar Tainong No. 1 presented the strongest lethal effect on MRSM, where the mortalities significantly increased and reached 100% death at 8 dpi (Figure 2A).

Different mango cultivars presented significantly distinct capacities in inhibiting MRSM development and reproduction. The average egg quantities on HS, S, MR, R, and HR cultivars were about 27.62 ± 4.23, 18.14 ± 4.21, 14.38 ± 2.72, 9.60 ± 1.63, and 5.5 ± 1.81 (Figure 2B). In addition, the egg hatching rates of MRSMs fed on HS, S, MR, R, and HR cultivars were 96.33 ± 2.78%, 90.83 ± 5.32%, 65.50 ± 4.72%, 39.50 ± 12.11%, and 36.33 ± 4.73%, respectively (Figure 2C). On the contrary, feeding on HR and R cultivars significantly prolonged the development duration, as the developmental duration from egg to deutonymph was between 18.56 d and 16.27 d, which was much longer compared with the HS cultivars (9.10 d on average) (Figure 2D). These results demonstrated that resistant cultivars can significantly hinder the reproduction and development of the MRSM.

3.3. Field Identification of Mango Cultivars Resistance to MRSM

As presented in Figure 3 and

Table 2. Kendall’s coefficient of concordance (W) of mango cultivar resistance level against MRSM under greenhouse, field conditions and their concordance.

	Greenhouse Concordance	Field Concordance	Greenhouse-Field Concordance
Total, N	12	12	12
Kendall’s W	0.256	0.424	0.278
Test statistic	1.348	20.333	1.333
Degree of freedom	2	4	1
Asymptotic sig.	0.510	0.002 **	0.248

Asterisks denote significant difference in the correlation (** p < 0.01).

, most cultivars displayed fluctuating resistance among the tested seasons. When we further compared the resistance levels among different tested months, we found that variation more or less existed in each cultivar (Tables S5–S9). Nevertheless, when considering the overall evaluation of resistance performance, the majority of tested cultivars showed identical resistance levels in field identification compared with the greenhouse test, except for Zihua (S in the greenhouse but HS in the field), India 901 (S in the greenhouse but HS in the field), Golden Phoenix (MR in the greenhouse but S in the field) and Sanya (R in the greenhouse but MR in the field) (Table S5). To sum up, Dashehari, Keitt, Yuexi, Zihua, and India 901 were identified as HS cultivars, and there were four S cultivars, including Red ivory, Hongguang, Kent, and Golden Phoenix; in addition, a single MR cultivar (Sanya), a single R cultivar (Sunrise), and a single HR cultivar (Tainong No. 1) were identified. Furthermore, the five-seasons-long field evaluation identifications showed an identical phenomenon wherein the resistance level of mango cultivars were elevated from susceptible to resistance, the number of mango leaves that corresponded to high damage scale would decrease, while the leaf quantity that corresponded to low damage scale increased (Tables S6–S10). In addition, concordance analysis demonstrated that the results of the five seasons’ greenhouse identification results showed moderate concordance (Kendall’s W value was 0.424, p = 0.002). In addition, low concordance was found between greenhouse and field results (Kendall’s W value was 0.278, p = 0.248) (The references should be kept. Besides, SI Appendix section from Supplementray File is not necessary to move to this paper.

3.4. The Climate Element Dynamic in the Five Planting Seasons

The precipitation (P_SUM) was relatively low in first two planting seasons (2019 and 2020) but increased during the last three seasons (2021 to 2023) (Figure 4A). This study revealed a non-significant (p > 0.05) increasing trend (positive MK Tau value) in the P_SUM by 1.02 mm yr⁻¹ (values for MK Tau and Sen’s slope are shown in Figure 4A). The MK mutation analysis showed that there were four mutation time points for P_SUM, which were December 2018, February 2020, April 2020, and April 2023 (Figure 4B). For the T_SUM, the MK trend on the investigated period had shown a non-significant increase (positive values, Sen’s slope was 0.91 °C yr⁻¹) (Figure 4C), and the only mutation instance was in January 2020 (Figure 4D). The RAH also showed non-significant (p > 0.05) increasing trend (positive MK Tau value) with Sen’s slope of 0.25% yr⁻¹ (Figure 4E) but presented multiple mutation time points (mutation existed in each planting season) (Figure 4F). Moreover, Kendall’s coefficient of concordance demonstrated that the P_SUM showed low concordance (Kendall’s w value was 0.280, p = 0.231), T_SUM showed moderate concordance (Kendall’s w value was 0.524, p = 0.012), and RAH showed high concordance (Kendall’s w value was 0.712, p = 0.007) levels throughout the five seasons (

Table 3. Kendall’s coefficient of concordance (W) of the weather elements of the five seasons.

	TSUM	PSUM	RAH
Total, N	5	5	5
Kendall’s W	0.524	0.280	0.712
Test statistic	16.480	5.600	14.240
Degree of freedom	4	4	4
Asymptotic sig.	0.012 *	0.231	0.007 **

* p < 0.05, ** p < 0.01.

).

3.5. The MRSM Density Dynamic Under the Changing Climate Elements

Climate elements can significantly influence the MRSM density (population) of the mango tree leaf. In general, in a given planting season, the T_SUM (Table S11) and P_SUM (Table S12) were lower from December (previous year) to April (current year), compared with the rest of the months; by contrast, the RAH was relatively stable across each planting season (Table S13). Hence, the high mite density as well as the occurrence peak of the MRSM also appeared in this period from December (previous year) to April (current year) (Figure 5). Accordingly, the MRSM-resistance evaluations were also conducted within this period. In addition, Pearson’s correlation analyses stated that the mite densities of the five seasons were all significantly negatively correlated to P_SUM, while only the T_SUM of 2018–2019 and 2022–2023 were significantly positively correlated to mite density. In addition, the T_SUM of the rest seasons and the RAH of all seasons did not show a significant correlation to mite density dynamic (

Table 4. Pearson’s correlation analyses among the mite density and weather elements.

Seasons	2018–2019	2019–2020	2020–2021	2021–2022	2022–2023
Weather Lements	Average Mite Density/Leaf
TSUM	0.5715 *	−0.1489	0.1663	−0.2919	0.6100 *
PSUM	−0.8186 **	−0.5941 *	−0.5733 *	−0.6765 **	−0.7499 **
RAH	−0.0499	0.1493	0.2297	0.2079	0.2047

Note: RAH, Tsum, and Psum denote relative air humidity, sum of temperature, and sum of precipitation, respectively. Asterisks denote significant difference in the correlation (* p < 0.05, ** p < 0.01).

).

3.6. Impact of Climate Elements on MRSM-Resistance Performance

We used the MDI and local climate data to set up MLR equations to analyze the influence of climate elements (i.e., P_SUM, T_SUM, and RAH) on the resistance level. According to the selected equation (Y (MDI) = β0 + β1 × P_SUM + β2 × T_SUM + β3 × RAH) (see all the equations in Table S14), high associations between the MDI and climate elements were found for the highly susceptible cultivars but not for the resistant ones. During these periods, adjusted R² values for the highly susceptible cultivars like Dashehari, Keitt, Yuexi, and Zihua ranged from 0.8875 to 1.0000 (Figure 6A). In addition, the corresponding p values for those cultivars were accordingly low (p < 0.05) (Figure 6B). In contrast, the adjusted R² values for the resistant cultivars such as Sanya, Sunrise, and Tainong No. 1 were usually lower than 0.5376, and the p values were higher than 0.05 (Figure 6A,B).

3.7. The Yield Losses of Different Mango Cultivars Caused by MRSM

To evaluate the capacity to reduce the losses of yield, the 12 mango cultivars were subjected to yield determination under either acaricide-free or acaricide-applied conditions. The yields for acaricide application ranged from 7.5 to 27.2 tons/ha (Figure 7A–L), depending on the tested cultivars. Conversely, without acaricide application, the yields decreased significantly; specifically, without acaricide application, the highly MRSM-resistant cultivars (Tainong No. 1) could maintain about 67.9% of the yield (Figure 7L), and the resistant cultivar (Sunrise) could maintain about 58% of the yield (Figure 7K). However, the MRSM-susceptible cultivars suffered the most remarkable drop in production, with an approximately 53–61% reduction in yield (Figure 7A–H). In addition, the GLMM analysis showed that, for each cultivar, the acaricide application always presented a significantly higher yield (all p-values lower than 0.05), while the yield in different survey season also presented a statistical difference, depending on the MRSM-resistance level. These results indicated that the higher the MRSM-resistance level, the better the performance in maintaining the yield.

4. Discussion

Identifying crop resistance to mite requires a scientific, accurate, and easy-to-handle method. Several studies are defining the leaf damage phenotypes by introducing concepts such as damage level, damage scale, or damage rating. For instance, a damage rating from 1 to 6 was applied to define the resistance level of cassava cultivars against green mites by the International Center for Tropical Agriculture (CIAT), and 72 cultivars among the total tested 300 obtained damage ratings lower than 3.0, demonstrating low-to-moderate resistance [19]. Similarly, the leaf damage index, which ranges from 0 to 5 based on the visual evaluation of leaf damage caused by the two-spotted spider mite, was also developed to evaluate mite damage [20]. However, these methods either rely on experience or are time- and labor-consuming. Moreover, resistance level judgment depends only on damage scale rather than accurate quantitative calculation, which seems empirical and not precise. In this study, an accurate and convenient measurement of leaf damage rate was developed, and the derived resistance level of mango cultivars against the MRSM was first identified based on the MDI. The degree of inhibition on mite reproduction and development can also reflect the resistance levels of the plant. As studies have pointed out, assessing host plant resistance must take into consideration mite population development and plant damage [21]. Here, when the MRSM was fed on resistant mango cultivars, the survival, egg production, and hatchability were all significantly inhibited, while the developmental durations were prolonged. These consequences will greatly reduce the mite density and decrease leaf damage. Moreover, those adverse effects on the MRSM were differentiated by the twelve mango cultivars with distinct resistant levels against MRSM. These results can explain the disparate MRSM damage phenotypes of different cultivars in both greenhouse and field tests, wherein cultivars with higher resistance levels showed stronger inhibition to MRSM reproduction. Similar results also can be seen in several mite–plant interaction studies, i.e., rubber tree germplasms and six-spotted mite Eotetranychus sexmaculatus [22], cassava varieties and two-spotted spider mite [6], and cotton germplasms and two-spotted spider mite [23]. It is noteworthy that some cultivars presented with fluctuations in resistance performance in the field and were not consistent to the results in the greenhouse. This result is in agreement with several previous studies. For instance, nine aphid-resistant soybean genotypes were identified in greenhouse, but only two of them were identified as resistant in the field test [24]. Zhu et al. [25] conducted identification of resistance to Bemisia tabaci using 550 cotton genotypes in greenhouse and field experiments. Although the greenhouse test identified 100 resistant and susceptible genotypes, there were only 42 genotypes that showed identical resistance performance in the field test. We deduce that this inconsistency is probably due to the environmental variability in the field, compared with the stable and normalized culture condition in the greenhouse; therefore, we explored how the climate elements would alter the resistance performance.

Climate elements significantly influence the mite population dynamic [12]. For example, heat and drought are known to induce T. urticae dispersal into trees. Hardman et al. [26] found that greater heat was sufficient to produce more generation of T. urticae, which would have promoted a greater increase in mite populations in the apple orchards. Bostanian et al. [27] reported that abundant rainfall could strongly suppress the population growth of Panonychus ulmi on the apple trees. In addition, the abundance of mite Polyphagotarsonemus latus in the chili varieties showed a significant positive correlation with temperature, and the relationship between relative humidity and rainfall was negative [28]. In the present study, we found that during the five seasons, the precipitation data showed low concordance, while temperature showed moderate concordance and humidity showed high concordance within the five survey seasons. To go along with this, MRSM density was significantly negatively correlated to precipitation, while temperature presented a moderate correlation, and humidity did not exhibit a great impact on mite density. This phenomenon indicated that the rainfall would be the primary factor influencing MRSM occurrence and the derived plant damage. Furthermore, conducting plant resistance identification within the period of pest peak occurrence (December in the previous year to April in the current year) rather than a random time can better represent the genuine resistance performance. This viewpoint is supported by the studies regarding cassava resistance to T. urticae [6], in which the pest dynamic survey was first performed to schedule the resistance identification time and give reliable results.

Different environments can result in a variation in pest-resistance performance. For example, when planted in different environments, faba bean (Vicia faba) resistance to fungi exhibited distinct resistance levels [29]. In another study identifying Botrytis fabae resistance, a field assay also revealed the instability of resistance performance across different environments [30]. Similarly, although 300 cultivars with low to moderate resistance to cassava green mite were identified by the CIAT in Columbia (in the tropical lowland that possessed prolonged dry season and endured high CGM populations) [31], only 72 cultivars consistently demonstrated the same resistance level in Brazil (primarily in the northeastern semiarid regions) [19]. Due to the environmental variability in different fields, abiotic stresses in the field, like drought, chilling, loss of applied fertilizers, and waterlogging, normal physiological growth may be hindered, which may cause the deterioration of pest resistance [32]. For a certain field environment, the climate condition is the major uncontrollable factor during the planting season; thus, figuring out how the climate elements influence plant resistance against pests is quite necessary, especially for a perennial crop such as the mango tree. A recent study developed a data-driven simulation platform to predict wheat cultivars’ yield performances under uncertain climate conditions [33]. Nevertheless, the studies on the impact of climate elements on resistance performance have rarely been conducted. In the present study, the MK mutation test illustrated that the examined time point that did not fit the overall trend (mutation time) of precipitation and temperature collectively appeared in the year of 2020 (less rainfall and a higher temperature compared with the rest seasons), and several cultivars showed enhanced resistance against the MRSM. In addition, for certain cultivars, the resistance fluctuation was not only limited to 2020 but could also be seen throughout the five planting years if abnormal precipitation or temperature occurred in a specific month for resistance evaluation. Furthermore, it is noteworthy that cultivars with higher resistance were less affected by changing climate elements. For instance, the resistant cultivar Sunrise and highly resistant cultivar Tainong No. 1 demonstrated stable resistance performance throughout all the seasons. These results suggested that these resistant cultivars can be considered ideal materials for MRSM control or for a future breeding program of resistant plants.

The yield of the twelve mango tree cultivars can reflect their resistance levels in the field. Although Sunrise and Tainong No. 1 were identified as MRSM-resistant cultivars, they still suffered about 20–30% yield loss without acaricide application. Comparatively, cultivars that were susceptible to the MRSM suffered about 50–60% of yield loss, and cultivars with moderate resistance levels possessed a medium level of maintaining the yield. In addition, without acaricide application, cultivars with higher resistance usually presented a more stable yield among different seasons. Several studies released insect pest-resistant cultivars to the field for pest control and rendered a good profit [29,32]. A select number of moderate mite-resistant cultivars were introduced to growers by breeders and entomologists of cassava [31]. Moreover, a cultivar named Nataima-31 was cultivated in Tolima, Columbia. This cultivar can attain a high yield of 33 t/ha (34% higher than the regional farmers’ variety) without pesticide applications, and now this cultivar is being grown commercially in different regions of Colombia, Ecuador, and Brazil [34]. The chain effect of the climate on MRSM-resistance occurrence and, therefore, performance yield can provide a better understanding of pest management and breeding of novel mango germplasms. Although the cultivars Tainong No. 1 and Sunrise exhibit excellent resistant traits, there is still considerable work to be performed in the future. For example, is the resistance performance consistent in different producing areas? Will other environmental factors, such as soil microorganisms and the landscape, affect the resistance and their interactions? Nevertheless, the present study provides promising materials for effective mite control or as good materials for deciphering the mite resistance mechanism, as well as benefiting future breeding program of mite resistance.

5. Conclusions

Based on three rounds of greenhouse tests and five seasons of field tests, cultivars with distinct resistance levels were identified. In addition, some cultivars presented with fluctuations of resistance performance in the field and were not consistent to the greenhouse results. When exploring the climate impact, we found that precipitation, temperature, and air humidity presented primary, secondary, and no impact on resistance performance of the susceptible cultivars, respectively. By contrast, those changing climate elements do not influence the resistance performance of the resistant cultivars. Furthermore, yield estimation speculated that the resistant mango cultivars can maintain about 70% of the yield without acaricide application. The present study can provide insight into insect pest management and future breeding programs under changing climate conditions.