Silicon builds resilience in strawberry plants against both strawberry powdery mildew Podosphaera aphanis and two-spotted spider mites Tetranychus urticae

Silicon is found in all plants and the accumulation of silicon can improve plant tolerance to biotic stress. Strawberry powdery mildew (Podosphaera aphanis) and two-spotted spider mite (Tetranychus urticae) are both detrimental to strawberry production worldwide. Two field trials were done on a UK commercial strawberry farm in 2014 and 2015, to assess the effects of silicon nutrient applied via the fertigation system on P. aphanis and T. urticae. The silicon treatments decreased the severity of both P. aphanis and T. urticae in two consecutive years on different cultivars. The percentage leaf area infected with P. aphanis mycelium from silicon treated plants were 2.19 (in 2014) and 0.41 (in 2015) compared with 3.08 (in 2014) and 0.57 (in 2015) from the untreated plants. The etiology of the pathogen as measured by the Area Under the Disease Progress Curve from silicon (with and without fungicides) treatments was 152.7 compared with 217.5 from non-silicon (with and without fungicides) treatments for the overall period of 2014–2015. The average numbers of T. urticae recorded on strawberry leaves were 1.43 (in 2014) and 1.83 (in 2015) in plants treated with silicon compared with 8.82 (in 2014) and 6.69 (in 2015) in untreated plants. The silicon contents of the leaves from the silicon alone treatment were 26.8 μg mg-1 (in 2014) and 22.2 μg mg-1 (in 2015) compared with 19.7 μg mg-1 (in 2014) and 21.4 μg mg-1 (in 2015) from the untreated. The silicon nutrient root application contributed to improved plant resilience against P. aphanis and T. urticae. Silicon could play an important role in broad spectrum control of pests and diseases in commercial strawberry production.


Introduction
Silicon (Si) is the second most abundant mineral element in the soil and constitutes ca. 28% of the earth's crust [1,2]. Silicon is found in all plants but not considered an essential element for plant growth (International Plant Nutrition Institute, http://www.ipni.net/nutrifactsnorthamerican), as it is not directly involved in the plant metabolic process [2]. Nevertheless, threat to horticulture growers, partly because the large scale use of chemical insecticides has induced their insensitivity to chemical acaricides (e.g. cyhexatin, dicofol and azocyclotin) [24]. The intensive use of such insecticides has also reduced populations of their natural enemies, enabling the mite populations to increase exponentially [25]. As a result of EU directives to reduce the application of synthetic chemicals for the control of pests and diseases, alternative crop protection management strategies are of increasing importance to growers. Therefore, the aim of this investigation was to explore the effects of the silicon nutrient delivered via the fertigation system on strawberry powdery mildew P. aphanis and the mite T. urticae in a commercial cropping situation.

Description of the experimental site
Silicon fertigation field experiments were done in 2014 and 2015 on a commercial strawberry farm at Wisbech, Cambridgeshire, UK (PE14 0HS). The farm has a total cropping area of 113 ha, within which 14 ha was used for growing strawberries. The use of the commercial farm ensured that strawberries were being grown in optimum conditions to meet supermarket standards. Both field trials were set up in polyethylene tunnels commercially managed and harvested; such a commercial design enabled the results to be applicable for commercial growers. Each tunnel had five raised soil beds (180m long) running in parallel (1m spacing between beds) (S1A Fig). Six strawberry plants were grown in each coir bag (1m long) placed on raised soil beds (S1B Fig). There were approximately 5,000-5,300 strawberry plants in one tunnel. Water (from the mains) and nutrients were delivered to the plants through irrigation drippers (four per coir bag) connected to the fertigation system (an irrigation system that delivers both water and fertilizers) five times per day, and commercial fungicide applications were done based on the normal farm spray schedule. The grower also used biocontrol agents (e.g. Beauveria bassiana) to control strawberry pests such as two-spotted spider mites. The silicon product used was Sirius 1 (a nutrient to increase the strength and health of the plants, main active ingredient: 70-80% tetraethyl silicate) provided by Orion Future Technology (Kent, UK). The silicon nutrient was added in the irrigation water at a concentration of 0.017% (by volume, 0.003 mg ml -1 in the irrigation water) and applied once per week via the fertigation system. The adoption of the farm fertigation system for silicon treatments necessitated the use of large sample sizes (e.g. 15 leaves x five replicates per treatment), because feeding through the irrigation pipes could be controlled only on a tunnel basis. The leaf assessments of severity of powdery mildew and numbers of spider mites were done every two weeks.

2014-2015 silicon fertigation experiments
The experiment in 2014 was set up in Blackberry Field in April (S2A Fig). Plants of strawberry cultivar 'Driscoll Jubilee TM ' (June bearer-one harvest per year in June/July) [26] were planted in coir bags on 20 March 2014. Plants were then covered by fleece to protect them from frost until late April. Application of the silicon nutrient started on 09 May and four treatments were used in two tunnels between 09 May and 12 August 2014.
In one tunnel, no silicon was applied. The first 15m of five growing beds received no fungicide sprays (untreated control) and the remaining parts of beds received fungicide sprays in accordance with commercial spraying practice (commercial fungicide only) ( Table 1). In another nearby (< 10m distance) tunnel, all five beds received 0.017% silicon nutrient through the fertigation system once per week. The first 15m of each bed received no fungicide (0.017% Si alone) and the remaining parts of beds received commercial fungicide according to the normal farm practice (0.017% Si plus commercial fungicide).

Assessment of Podosphaera aphanis development on the leaf surface
The assessment of P. aphanis development was based on 75 leaves per treatment (five replicates of 15 leaves per strawberry bed). Pre-assessments before the silicon treatment started on 08 and 22 April in 2014 and on 21 April in 2015, respectively. Samples were collected at twoweekly intervals commencing 08 April 2014 and 21 April 2015. Each leaflet of the sampled leaf was placed under a dissecting microscope (GX microscopes, 1x and 3x objectives; 10x eyepieces, GT Vision Ltd, Suffolk, UK) to be assessed for P. aphanis development using the assessment key developed by Jin [15]. The disease severity was expressed as % leaf area covered by P. aphanis colonies (amount of mycelium).

Assessment of Tetranychus urticae presence on the leaf surface
The same batches of leaf samples collected for P. aphanis assessment (20 May-12 August in 2014 and 21 April-11 August in 2015) were also used for the assessment of T. urticae infestation. Each sampled leaf was placed under a dissecting microscope (as described above for P. aphanis assessment) at ×10 magnification, and the number of T. urticae (at all life-stages apart from eggs) that were present on the leaf surface was recorded. The observation of the leaf was done by moving the leaf from the right to the left side in a regular pattern and observations started from the top part of the leaf surface then moved systematically to the bottom part of the leaf ensuring that every part of the leaf surface was observed without repetition, thereby avoiding miscounting or double counting. In addition, if the movement of T. urticae followed the same route as the observation, its movement direction was specially noted, and it was excluded from the count if it appeared in the next observation field. Since the strawberry leaf was observed under x10 magnification, one observation field could cover a relatively large percentage of the leaf area (25% of the whole leaf in most cases), thus avoiding multiple counting due to spending a long time on the same leaf. Also, since the size of the T. urticae was small compared with the observation field, its movement usually stayed within the same field. Even if it was moving, its movement direction could easily be traced; therefore the risk of repeated counting could be minimized.

The analysis of 2014 and 2015 experimental results
The analysis was done for the 2014 and 2015 seasons individually, as well as for the combined seasons to assess the consistency of the effects of silicon on P. aphanis development and T. urticae presence. Area Under the Disease Progress Curve (AUDPC) (The American Phytopathological Society, https://www.apsnet.org) was used for the analysis of P. aphanis development. The calculation was done using the equation AUDPC = P nÀ 1 i¼1 ððx iþ1 þ x i Þ=2Þðt iþ1 À t i Þ, where x i is a measure of disease severity (% area of leaf coverage by P. aphanis) at the i th sampling date, t is a measure of time (i.e. days), and n is the total number of samplings. The calculation of the AUDPC for each treatment was based on fortnightly disease assessment results from five strawberry beds. Each bed was a replicate, which had an AUDPC value based on the average % area of leaf coverage by P. aphanis mycelium of 15 leaves collected randomly from this bed. Comparisons of five AUDPC values for each of those four treatments were made by the analysis of variance (ANOVA) test using R software (The R Foundation for Statistical Computing, 3.3.0 GUI 1.68). In addition, apparent infection rate (r) was calculated to assess the rate of epidemic development for each treatment using the equation r = (lnq 2 -lnq 1 )/(t 2 -t 1 ), where q 2 and q 1 are the quantities of disease present at times t 1 and t 2 , respectively [27].
Similar to the calculation of the AUDPC, the overall sum of the number of T. urticae for the entire experimental period could be represented by the overall Area Under the 'Pest' Progress Curve (AUPPC). The AUPPC value for each sampling date was calculated based on the mean value of the total number of T. urticae on 15 leaf samples from each strawberry bed of all five beds (each bed was a replicate, five replicates in total). The overall analysis for the combined two seasons was based on the overall number of spider mites per treatment in 2014 and 2015 and five sum values from each treatment were compared using ANOVA.

Silicon extraction
Three mature strawberry leaves were collected (one leaf per plant) per strawberry bed for silicon extraction. Three beds from each treatment were sampled (i.e. three replicates of 3 leaves per treatment). The Autoclave-induced digestion (AID) [28] method was used to extract silicon from strawberry leaves (oven dried (at 60˚C) leaf powder (0.1g) per three leaf samples x three replicates per treatment). The leaf silicon extraction was done monthly between 08 April and 23 September in 2014 and between 21 April and 29 September in 2015. A silicon standard curve was made to calculate the concentration of silicon in the strawberry plant material [15]. The silicon concentration of the sample material was then calculated by using the equation y = 1.0957x, where y is the absorbance of silicon at 650nm (CECIL 1021 Spectrophotometer, 1000 series, Cambridge, UK) and x is the concentration of silicon (mg ml -1 ).

Results
Results from both 2014 and 2015 experiments were consistent; strawberry plants that received weekly silicon application (with or without fungicide) had reduced severities of both P. aphanis and T. urticae compared with the untreated control plants.

Assessment of Podosphaera aphanis development on the leaf surface
In the 2014 experiment, strawberry plants from the 0.017% Si plus commercial fungicide treatment had the smallest disease scores (AUDPC = 63) (P < 0.001) and infection rate (r = 0.0012) among all treatments ( Table 2; Fig 1A). There was a significant difference (P < 0.05) in disease severity between the untreated control (AUDPC = 662) and 0.017% Si alone (AUDPC = 475) treatments from 17 June 2014 onwards. In addition, it was shown that the onset of epidemic development was delayed by approximately 14 days for two silicon treatments (17 June 2014) compared with the untreated control (03 June 2014) ( Fig 1A). Furthermore, the 0.017% Si plus commercial fungicide treatment had a smaller disease severity compared with commercial fungicide only treatment (AUDPC = 106, r = 0.0017) ( Table 2), which indicated that the use of silicon and fungicide together may enhance the effectiveness of fungicide treatments.
In the 2015 experiment, plants from 0.017% Si plus commercial fungicide treatment had the smallest disease severity (AUDPC = 53, r = 0.0004) throughout the experimental period ( Table 2; Fig 1B). A significant difference (P < 0.001) in disease severity was found between this treatment and the untreated control (AUDPC = 281, r = 0.0011). Plants from this treatment developed less P. aphanis than those from the commercial fungicide only treatment (AUDPC = 69, r = 0.0005) ( Table 2). Moreover, plants from the treatment 0.017% Si alone also The two-weekly leaf assessment results are presented separately for P. aphanis (Fig 1) and T. urticae (Fig 2).

PLOS ONE
Silicon builds resilience in strawberries against P. aphanis and T. urticae developed less disease (AUDPC = 267, r = 0.001) than the untreated control (Fig 1B), which was consistent with the 2014 results ( Fig 1A).

Assessment of Tetranychus urticae presence on the leaf surface
Results from the 2014 experiment showed that 0.017% Si alone and 0.017% Si plus commercial fungicide treatments had average smaller numbers of T. urticae per strawberry leaf (< 2) than the untreated control and commercial fungicide only treatment (Fig 2A). The fungicide only treatment had an average of ten T. urticae per leaf, which was the highest among all treatments. The treatment 0.017% Si plus commercial fungicide had the smallest AUPPC value  (Table 2). It can be seen that both treatments with silicon had smaller values of AUPPC than treatments without silicon.
In addition, all four treatments showed a similar trend, with the number of T. urticae per leaf increasing from < 2 on 20 May 2014 to a maximum (> 40 in the commercial fungicide only treatment) on 01 July and then gradually decreasing to nearly 0 by late July/early August 2014 (Fig 2A). There was a significant difference between commercial fungicide only and

PLOS ONE
Silicon builds resilience in strawberries against P. aphanis and T. urticae 0.017% Si alone (P < 0.001) treatments, and between commercial fungicide only and 0.017% Si plus commercial fungicide (P < 0.001) treatments on 01 July 2014. Results from the 2015 experiment showed that there were fewer T. urticae present on leaves from the silicon treatments than treatments without silicon throughout the experimental period ( Fig 2B). Leaves from 0.017% Si alone and 0.017% Si plus commercial fungicide treatments had less T. urticae present. These two treatments had an average of two T. urticae per leaf compared to nine in the untreated control. Similarly, the 0.017% Si alone treatment had the smallest AUPPC value (AUPPC = 2,265), followed by the 0.017% Si plus commercial fungicide treatment (AUPPC = 2,681) ( Table 2). The untreated control had a greater AUPPC value (AUPPC = 13,149) than the commercial fungicide only treatment (AUPPC = 8,149), which was slightly different from the 2014 results. The AUPPC results from 2014/2015 showed that treatments with silicon had less T. urticae infestation than the untreated control and commercial fungicide only treatments. Interestingly, for treatments with the same application rate of silicon, those with added fungicide had greater T. urticae infestation numbers than those without fungicide (i.e. commercial fungicide only vs. untreated control in 2014; 0.017% silicon nutrient plus commercial fungicide vs. 0.017% silicon nutrient in 2015) ( Table 2).

Overall analysis of combined 2014/2015 Podosphaera aphanis severity and Tetranychus urticae infestation results
Apart from the individual analyses of the 2014 and 2015 results, an overall analysis of all data from these two years was also done (Table 3). Strawberry plants that received silicon developed significantly less P. aphanis (AUDPC = 152.7) (P < 0.05) and were less infested by T. urticae (AUPPC = 170) (P < 0.001) compared with plants from non-silicon treatments (AUDPC = 217.5, AUPPC = 876) in both 2014 and 2015. It was shown that both strawberry   powdery mildew severity (Fig 1) and numbers of spider mites present (Fig 2) were significantly less in the two silicon treatments (0.017% Si alone; 0.017% Si plus commercial fungicide) compared with the two non-silicon treatments (untreated control; commercial fungicide only) and this significant difference was consistent throughout 2014 and 2015 (Table 3). There was a clear difference (P < 0.001) in powdery mildew severity between the 2014 and 2015 experiments. The 2014 experiment had much more severe disease (AUDPC = 662, r = 0.0042 in untreated control) than the 2015 experiment (AUDPC = 281, r = 0.0011 in untreated control) ( Table 2; Fig 1). Nevertheless, when analysing the 2014 and 2015 disease results separately, even though disease severity differed between these two years, plants in the silicon treatments still had less powdery mildew than those in the non-silicon treatments (P < 0.05) in both years ( Table 3). The disease AUDPC from the untreated control treatment was significantly higher (P < 0.05) in 2015 but only slightly higher (P < 0.1) in 2014 than that from the 0.017% Si alone treatment in the same year ( Table 2). No significant difference was found between commercial fungicide only and 0.017% Si plus commercial fungicide treatments.
The  (Table 2). Moreover, in the 2014 experiment, strawberry plants in the 0.017% Si alone treatment were less infested by T. urticae (AUPPC = 2,222) (P < 0.05) than those in the untreated control. Similarly, strawberry plants in the 0.017% Si plus commercial fungicide treatment were less infested by T. urticae (AUPPC = 1,977) (P < 0.001) than those which received only commercial fungicide. In the 2015 experiment, even though the overall numbers of T. urticae present were different from the previous year, strawberry leaves in the 0.017% Si alone treatment were observed to have fewer T. urticae (AUPPC = 2,265) than those in the untreated control (P < 0.05) (Fig 2B).

Silicon content in strawberry plants in the 2014 and 2015 experiments
The leaf silicon content from the untreated control was significantly less (19.7 μg mg -1 ) than that in the 0.017% Si alone (26.8 μg mg -1 ) and 0.017% Si plus commercial fungicide (26.8 μg mg -1 ) treatments in 2014 (P < 0.001); however in 2015, there was a difference (P < 0.05) between the commercial fungicide only (21.1 μg mg -1 ) and the 0.017% Si plus commercial fungicide (25.3 μg mg -1 ) treatments (Table 4). In both years, there was a significant difference (P < 0.001) in amounts of leaf silicon content between the first assessment (19 April in 2014; 21 April in 2015) and following assessments in the same year.

Treatments
Mean monthly leaf silicon content tested a (μg mg -1 ) Data are the mean of three replicates ± standard error (mean ± SE). b Data followed by same letter within each column indicate no significant difference (P > 0.05) between treatments using ANOVA followed by TukeyHSD test.

Discussion
The work reported here has shown that strawberry plants treated with silicon were more resilient and had significantly less severe P. aphanis infection and less severe T. urticae infestation compared with those from the untreated control. This effect was found to be consistent over the two-year research and on different cultivars. Silicon has been used for disease control in many economic crops including barley, cucumber, rice and strawberry [29]. Studies showed that plants that received silicon were more resilient under the stresses of fungal diseases such as powdery mildew, rust, and leaf spot [30]. One explanation for this is that the accumulation of silicon in the form of amorphous silica, forms a barrier to prevent penetration by the pathogen [31]. Work has shown that silicon treatment of strawberry plants increased leaf cuticle thickness, density of leaf wax and numbers of leaf trichomes [15], and silicon treated coffee seedlings developed a thicker epicuticular wax layer [29], suggesting that silicon induces the formation of physical defence barriers against attack by pathogens and pests [32]. It was found that soluble silicon was effective as a preventive measure for increasing plant resistance at an early stage of pathogen colonisation (e.g. conidial germination, germ tube elongation etc.), and the effect was greater in very susceptible cultivars (up to 86%) compared to less susceptible cultivars (up to 58%) [33].
Silicon plays a role in inducing plant natural defence responses, by interacting with a number of key components of plant stress signalling systems [30]. For example, soluble silicon treated cucumber and rice demonstrated increased accumulation of phenolics and phytoalexins when infected by powdery mildew and blast [4,30,34]. Research on wheat also found that silicon reduces the production of reactive oxygen species (ROS), which causes oxidative damage to plant cells, therefore reducing the cellular damage caused by wheat blast (Pyricularia oryzae) [31]. Results from the present study showed that the rate of P. aphanis epidemic development in the 0.017% Si alone treatment was ca.14% less than that in the untreated control in the 2014 experiment; whereas it was only ca.9% less when compared with the untreated control in 2015, when there was a less disease severity than in 2014 (P < 0.05). This further indicated that plants may benefit more from the supply of silicon when disease severity is greater.
The field experiment results showed that treatment 0.017% Si plus commercial fungicide had the smallest disease severity in both 2014 and 2015. Recent study also found that strawberry plants received silicon with potassium carbonate were more resilient against p. aphanis (AUDPC = 6) compared with potassium carbonate treatment alone (AUDPC = 24) [11,15]. The combined treatment also had a delay in onset of epidemic development of P. aphanis. Similar findings were also shown between treatments silicon with fungicides and fungicides alone in the same study [15]. The above evidence indicated that plants more benefit more from the combination of silicon and fungicides than the use of fungicides alone. There was a different infestation level of T. urticae between 2014 and 2015 for the same treatment. The average number of T. urticae per leaf was two to three times greater in the untreated control than in the silicon treatments in 2014 and more than three times greater in 2015. This suggested that there could be an interaction between abiotic effects (e.g. temperature and humidity) and cultivars. Nevertheless, strawberry plants from 0.017% Si alone treatment still had fewer T. urticae compared with those from the untreated control over two years. Silicon has been reported to improve plants resistance against many pests such as spittlebug in sugarcane, rice green leafhopper, stem borer and brown planthopper [4,35,36]. Silicon can affect insect biology such as food intake, nymph development and adult longevity [37]. High concentration of silicon on sugarcane has been observed to extend the nymphal stage of spittlebug Mahanarva fimbriolata Stal and shorten the adult life [35]. Soluble silicon is particularly effective in enhancing plant physical defence against piercing-sucking insects. For example, a study has shown that K 2 SiO 2 reduced the population fitness of green peach aphid Myzus persicae on Zinnia elegans [37]. This could be partly attributed to the deposition of silicon beneath the cuticle, forming a physical barrier in the cell wall to prevent the penetration by insects [4,38].
The work reported here also showed that commercial fungicide treatments (with or without silicon) increased T. urticae infestation more than the application of silicon alone. It therefore suggested that there might be interactions between the use of fungicide and the presence of T. urticae. Research demonstrated that the fungicides fenhexamid and cyprodinil+fludioxinil reduced the mortality of T. urticae inoculated with a fungal pathogen Neozygites floridana, a natural enemy of T. urticae [39]. The fungicide cyprodinil+fludioxinil was also found to reduce sporulation of N. floridana, thus subsequently inhibiting establishment of N. floridana in T. urticae populations. The grower at Maltmas farm uses both the fungicides fenhexamid and cyprodinil+fludioxinil, as well as biocontrol agents such as B. bassiana and predatory mites such as N. cucumeris for disease and pest control. Although it is not clear whether these two fungicides would have similar inhibition effects on B. bassiana as they had on N. floridana, the greater number of T. urticae in the fungicide only treatment indicated that frequent use of some fungicide applications may inhibit the sporulation and establishment of some biocontrol agents, thereby reducing their efficacy against target pests.
It is considered that the location where silicon is deposited in the plant and how it is deposited significantly influence plant resilience. Work in 2012 and 2013 demonstrated that the plants from the silicon root treatment had less severe P. aphanis severity, and also a delayed epidemic development by up to two weeks, than those from the foliar treatment [11,15]. Another study showed that strawberry plants were less susceptible to P. aphanis when grown in the silicon-saturated soil prior to planting than those treated with silicon after the first infection [33]. Therefore, root application of silicon was used in both 2014 and 2015 experiments. When silicon was applied through the fertigation system, it was then absorbed by plant roots, and transported through the xylem and finally deposited in the leaf epidermal cells and xylem vessels [8,29]. Many other studies also reached similar conclusions; root silicon application stimulated plants both to form physical barriers and to interact with stress response metabolism, whereas foliar applied silicon may only create a chemical-physical barrier (e.g. a change in surface pH) through its deposition on the leaf surface [8,29,40]. Thus, good pest and disease management can be achieved by a continuous supply of silicon via the roots, resulting in improved plant resilience, and potentially reduce usage of pesticides [18], which subsequently contribute to sustainable management of strawberry production.

Conclusions
Results from the two seasons of silicon fertigation experiments suggested that silicon improved strawberry plants' resilience against strawberry powdery mildew and the infestation of twospotted spider mites. Silicon was applied in a bioavailable form via the fertigation system and was taken up by the plants through the roots. It was suggested that the use of silicon could play an important role in the combined integration of pest and disease management in sustainable strawberry production.