Temporal Interactions between Root-Lesion Nematodes and the Fungus Rhizoctonia Solani Lead to Reduced Potato Yield

Soil microorganisms and soil fauna may have a large impact on the tuber yield of potato crops. The interaction between root-lesion nematodes and the pathogenic fungus Rhizoctonia solani Kühn was studied on potato plants grown in pots under controlled conditions. In two similar experiments, different combinations of nematodes and fungal mycelium were added to the pots at three occasions; at planting, after 14 days, and after 28 days. The nematodes reduced root biomass and the combination of nematodes and R. solani resulted in reduced tuber yield in both experiments, but the interaction was not synergistic. In contrast, the number of stem canker lesions decreased in the presence of nematodes compared to treatments with R. solani only. The time of inoculation influenced the severity of both fungal and nematode damage. The nematode damage on tubers was less severe if the nematodes were added at 28 days, while the number of severe stem canker lesions increased if the fungus was added at 28 days. However, the time of nematode inoculation did not affect the incidence of fungal damage, hence the nematodes did not assist R. solani to infect the plant. Our results highlight the underestimated importance of root-lesion nematodes, not resulting in obvious above ground symptoms or misshaped tubers yet affecting the performance of other pathogens.


Introduction
Producing a high-yielding and high-quality potato crop (Solanum tuberosum L.) is a complex task with many aspects to consider, including cultivar, soil parameters, rainfall, and agricultural measures [1,2]. In addition, the organisms in the soil, such as fungal species, insects, and bacteria, may very well influence the development of the potato crop depending on their food preferences [3]. The complexity of the interaction between multiple species present in the soil and the cultivated crop is difficult to entangle [4]. Here, we investigated the temporal interactions between the potato plant root-lesion nematodes and the pathogenic fungus Rhizoctonia solani Kühn.
Free-living and sedentary plant-parasitic nematodes, representing 15% of the total number of nematode species described, are present in soils worldwide and are significant pathogens in agriculture [5]. Free-living plant-parasitic nematodes have a vast host range, including potato, and are

Fungus
An isolate of R. solani, AG2-1, originating from Vara, Sweden, was kindly provided by Dr. S. Ahlström (Dept. of Forest Mycology and Plant Pathology SLU, Uppsala, Sweden), which was the same isolate as that used in the experiments previously published [12,21]. The methodology was also the same as in those experiments, that is, the fungal mycelium was mixed to small pieces and diluted in tap water to correspond to 0.01 g mycelium pot −1 within the same volume as for the inoculation of the nematodes.

Nematodes
Two types of nematode inocula were used for the experiments; in experiment 1, a full nematode community dominated by root-lesion nematodes, in particular P. crenatus, and in experiment 2, a pure culture of the root-lesion nematode P. penetrans. The full nematode community was derived from field soil taken from a potato field in the county of Östergötland in south-central Sweden, whereas P. penetrans were originally bought from Plant Research International Wageningen, the Netherlands, and kept in culture on maize. The maize plants were grown in sterilized sand and new seeds were sown after the senescence of the plants, which were cut, and the roots were left in the pots as feed for the nematodes. The nematodes were extracted from soil and root tissues, respectively, using Whitehead and Hemming trays (Whitehead and Hemming 1965), kept in a cold storage room 4 • C, and aerated with an aquarium air pump (superfish Air-Flow mini). The concentration of plant-parasitic nematodes in the solution was determined by counting the nematode numbers in subsamples under the microscope.

Experimental Setup
The experiments were designed as pot experiments with eight and sixteen treatments, respectively, ( Table 1) with eight replicates. Washed and sterilized (200 • C for six hours) sand (0-3 mm Ø) was used as potting medium. In each pot (12 × 12 × 25 cm), one filter paper was used to cover the bottom to avoid leakage of the 1600 g dry sand that was first added. The potting medium was then wetted with 200 ml of tap water. One pre-germinated mini tuber of cultivar King Edward VII (tubers produced from meristem cultures) (Agrico Nordic, cultivated at The Finnish Seed Potato Centre Ltd (SPK), Tyrnävä, Finland) was placed in each pot with the most germinated sprouts facing upwards. Another 900 g of sand was added on top of the tuber and 100 ml tap water was applied.
For both experiments, the first inoculation of organisms occurred the day after planting. Nematodes, fungus, and water were added in different combinations to the pots: control (tap water); fungus (0.1 g mycelium pot −1 ); nematodes (2 plant-parasitic nematodes g −1 sand) at start, after 14 days, and after 28 days. During the inoculation of nematodes in experiment 1, three subsamples were taken out to determine the complete composition of the nematode community added to the pots. The actual inoculum volume differed between the inoculation events as a result of variations in nematode density, but the volumes added for each treatment were always the same within each inoculation time. For instance, at one inoculation event, the volume of 90 ml of liquid solution contained either 45 ml of nematode solution, 45 ml fungal solution, and/or tap water, according to the treatments in Table 1. The remaining pots received the corresponding amount of water at each inoculation event.
The experiments were carried out in a climate chamber set to 12 • C and with a day/night cycle of 16/8 h of artificial light (152 LUX on average). The pots were placed on trolleys in a randomized complete block design and the trolleys were moved once a week to provide a uniform exposure to light. The plants were initially watered (200 mL) twice a week to ensure normal moisture and during the last weeks of the experiments, depending on the growth level, the plants were watered thrice a week. In addition, all pots were fertilized at three occasions in experiment 1 and four occasions in experiment 2 with a complete fertilizer (Blomstra NPK: 100:18:86 and micronutrients, Orkla Care, Solna, Sweden). In total, each plant received 180 mg nitrogen and 200 mg nitrogen, respectively.

Harvest
The potato plants were harvested ten weeks after the first inoculation. This was performed as previously described [12,21]. In short, the potato plant was taken out of the pot; washed carefully with tap water; dried with a paper towel; and divided into stems, roots, stolons, and tubers. The tubers were divided into small (0.5-2 cm in diameter) and large (>2 cm in diameter) tubers. Tubers smaller than 0.5 cm were regarded as stolons. In experiment 1, the plant parts were graded regarding nematode damage, stem canker, sclerotia, and "elephant hide" (Table A1). The severity of stem canker and "elephant hide" were graded using the same method as in experiment 2, while the actual number of sclerotia and nematode damage was instead counted on each plant part.
For the treatments with nematodes, the submerged part of the main stem (from the mother tuber up to the surface level, 5-8 cm), the largest tuber, approximately ten roots, and 20 g of the potting medium were weighed separately and put in plastic bags for cold storage until later extraction of nematodes. The remaining plant parts and 100 g of the potting medium were weighed and dried for dry weight measurements. The mother tuber was measured in length and diameter and graded regarding nematode damage and sclerotia (black scurf).

Nematode Extraction
The below ground part of the main stem and the roots were cut into 1 cm pieces and each put in one vlieseline covered mesh net sieve. The tuber was divided into four pieces and put skin side down in each mesh net sieve or, if too large, two sieves per tuber were used. The entire 20 g sample of potting sand was placed in a mesh net sieve. The sieves were placed in Baermann funnels, and the nematodes were extracted for 24 h, heat-killed, and fixated in formalin [22]. The number of nematodes was estimated in the suspensions from each extraction under low magnification (50×) and expressed as the number of nematodes per gram dry weight of each plant part or per gram dry potting sand.

Statistical Analyses
The two experiments were analysed separately. In experiment 1, the effects of additions of fungus and nematodes on dry weight of stems, tubers, roots, and stolons were analysed with a randomized complete block design analysis of variance (ANOVA) using R version 3.1.1 [23]. To account for extreme values and heteroscedasticity, we used robust standard errors, packages multcomp [1], and sandwich [2,3]. The number of tubers and stems was analysed with Poisson regression using R, with the control treatment used as baseline.
The proportion of small tubers was analysed with binomial regression using SAS for Windows 9.3 (SAS Institute Inc., Cary, NC, USA). To analyse the effect of fungal and nematode damage, we used a dichotomous variable-damage or no damage. The data were analysed with binomial regression using R. The number of nematodes in the different plant parts and in the sand was analysed with randomized complete block design ANOVA in R with robust standard errors.
For experiment 2, stem, root, and stolon biomass; number of tubers; and abundance of root-lesion nematodes in roots were analyzed with linear mixed models with block as a random factor using R version 3.5.2 [23] and packages lme4 [24] and nlme [25]. The log-transformed biomass was used for stems and stolons as a result of better results in tests for normality and equal variance for the residuals (diagnostics test performed using the car package [26]). Abundance of root-lesion nematodes in roots was modelled using log-transformed abundance and a treatment-dependent variance, thanks to differences in variance.
The probability of elephant hide was modelled with a logistic model. The factor of nematode addition was dropped thanks to non-significance in the model with interaction. Block and treatment interacting with block were used as random factors. Stem canker was modelled with a negative-binomial model on number of lesions. Separate models were estimated for lesions of severity 1 and for severity 2 and 3 combined. Block and treatment nested within block were used as random factors.
Post-hoc comparisons between factor levels were done using the Tukey honestly significant difference (HSD) method in the emmeans package [27]. For all models in experiment 2, model selection was performed through backward elimination based on significance in Type II Wald-tests. The interaction term was tested by comparing the full-factorial model to the additive model. Non-significant factors were first reduced to binary factors (added or not added nematodes or fungi respectively) and, if that binary factor was non-significant, removed from the model.

Results
The potato plants developed well and the shoots were approximately two cm above the surface of the potting medium at the time of the second inoculation and all, except two tubers in experiment 2, had germinated. The plants were approximately 10 cm at the inoculation at day 28 after planting. If the potato plant was inoculated with R. solani, it was clearly affected and had symptoms like stem canker, black scurf, and elephant hide. Plant-parasitic nematodes were present in all tested plant parts as well as in the potting medium from the pots where nematodes had been added. In the first experiment with a full nematode community, plant-parasitic nematodes dominated the inoculated community together with bacterivorous nematodes (47% each) and consisted of the taxa Pratylenchus (41% of total nematode abundance), Tylenchorhynchus (5%), and Trichodoridae (1.4%). As expected, only the endoparasitic Pratylenchus spp. were found in the different plant parts at harvest. The nematode community also consisted of fungal-feeding nematodes from the genera Aphelenchus (1.4%) and Aphelenchoides (2.9%), and some omnivors (1.4%) were also present.

Impact on Stems
The dry weight and number of stems were not significantly affected by treatment in either of the experiments (E1: p = 0.31 and p = 0.34, respectively; E2: p = 0.52 and p = 0.39, respectively; Table 2). The majority of the stems got stem canker in both experiments regardless of the time that R. solani was added (Tables 3 and 4). Nematode addition reduced the severity of stem canker in experiment 2, as the number of stem canker lesions of severity 1 (small lesions) was reduced to 2.81 lesions stem −1 in the treatments that also contained nematodes, compared with 4.15 lesions stem −1 without nematode addition (p = 0.0024). Likewise, the number of lesions with severity 2 (large lesion) or 3 (completely girdled) increased with 82% in absence of nematodes (p = 0.0096). The occurrence of the fungus did not affect the number of nematodes in the stems in either of the experiments (p = 0.54 and p = 0.12, respectively; Table 5).
The time of fungal inoculation influenced the number of stem canker lesions, as the likelihood of severe stem canker was lower in the treatment with nematode addition at planting and fungus at 28 days in experiment 1 (p = 0.029). In addition, in experiment 2, the number of stem canker lesions of severity 2 and 3 was 1.94 times higher when the fungus was inoculated at 28 days compared with inoculation at 14 days (p = 0.027). The nematode damage in experiment 2, as well as the number of nematodes per gram stem, were significantly higher if the nematodes were added at the start and at 14 days compared with at 28 days (p < 0.001; Table 6). Table 2. Dry weights (g; mean (SE)) and numbers of different parts of potato plants subjected to different combinations of the plant pathogenic fungus Rhizoctonia solani (F) and nematodes (N)-full community in experiment 1 and the root-lesion nematode Pratylenchus penetrans in experiment 2. Treatments with the same letter within a column and experiment are not significantly different at level p < 0.05.

Treatment
No     In experiment 1, the number of fungal-feeding nematodes differed among the treatments (p = 0.013), with more fungal-feeding nematodes in the stem in the treatment with nematodes at planting and fungal addition after 14 days (p < 0.001) and fungus at planting and nematode addition after 28 days (p = 0.009) compared with the treatments with only nematodes (Table 5). There were also more fungal-feeding nematodes in these two combination treatments compared with the treatment with nematodes at planting and fungus added after 28 days (p = 0.002 and p = 0.044, respectively).

Impact on Stolons
In experiment 1, the dry weight of stolons was significantly affected by treatment with the fungus reducing the biomass regardless of time of inoculation (p < 0.001; Table 2), apart from when nematodes were added at start and the fungus at 28 days, which was not significantly different from the control and nematode only treatment. In the second experiment, the dry weight of the stolons was reduced by fungal addition at the start and at 14 days (p < 0.001; Table 2). The fungus produced sclerotia on the stolons, which were more numerous when added late and the number of necrotic lesions was also affected by time of fungal inoculation (Table 7).

Impact on Tubers
The dry weight of the tubers was not affected by the fungus and nematodes in experiment 1 (p = 0.057, Figure 1). In experiment 2, the dry weight of tubers was reduced by both fungus and nematodes, regardless of the time of inoculation (Figure 2). The combination of fungus and nematodes reduced on average the yield by 11.7 g (p < 0.001) compared with the control, 6.6 g compared with the average of the fungus only treatments (p < 0.001), and 2.8 g compared with the nematode only treatments (p = 0.023). Addition of fungus at planting and nematodes at 28 days resulted in a higher dry weight than addition of both nematodes and fungus at planting (p = 0.022) and addition of fungus at planting and nematodes after 14 days (p = 0.023). Addition of fungus and nematodes had no significant effect on the number of tubers in either of the experiments.  The time of fungal inoculation influenced the amount of black scurf and elephant hide on the tubers. The probability of both black scurf and elephant hide was higher when the fungus was added at 28 days compared with when the fungus was added at planting in both experiments (Table 3, Table 4, and Table 7). In the second experiment, the odds increased for elephant hide and black scurf if the fungus was added at 14 days as well. Nematode damage was visible on the tubers in experiment 2, while symptoms of nematode feeding were not observed on the tubers in experiment 1. Addition of fungus did not interact with nematodes regarding symptoms, but the time of nematode inoculation was important for the amount of nematode damage and number of nematodes in the tubers ( Table 6). The amount of nematode damage was significantly lower when nematodes were added at 28 days and the probability of nematode damage on tubers was highest when nematodes were added at 14 days.
Although there was no visible damage of nematodes in experiment 1, the tubers did contain plant-parasitic nematodes in both experiments ( Table 5). The number of plant-parasitic nematodes in the tubers was higher when the nematodes were added at the start and at 14 days in experiment 2 ( Table 6). Fungal addition at 28 days resulted in lower numbers of plant-parasitic nematodes in all nematode treatments, compared with addition at the start in experiment 2 (p = 0.014, Table 7). The number of nematodes in tubers was not different between the treatments in experiment 1 (p = 0.09) ( Table 5).

Impact on Roots
The dry weight of roots was not significantly affected by treatment in experiment 1 (p = 0.10). Addition of nematodes reduced the root dry weight (p < 0.001) in experiment 2 and there was an interaction between addition of nematode and fungus (p = 0.047) ( Table 2). The roots were always covered with sclerotia if the fungus had been added and had brown lesions or sections regardless of treatment and experiment. Plant-parasitic nematodes were abundant in the roots, especially in the second experiment, where the plants that were inoculated at 14 days had the highest abundance (Tables 5 and 7). There were, however, no significant differences among the treatments in experiment 1. On the other hand, the number of fungal-feeding nematodes differed among the treatments in experiment 1 (p = 0.013) with more fungal-feeding nematodes when the fungus was added at planting and nematodes after 14 days compared with three other nematode treatments (Table 5).

Potting Medium
Both plant-parasitic and fungal-feeding nematodes were found in the potting medium in experiment 1, but neither group was significantly affected by the different treatments (Table 5). In experiment 2, more nematodes were found in the potting medium when inoculated after 14 days and the lowest amount was found after the latest addition (Table 6).

Discussion
Addition of root-lesion nematodes and R. solani affected the potato plants in some way for almost all treatments. Two of our hypotheses were confirmed as we found that the tuber yield was affected by the combination of nematodes and R. solani and that the time of inoculation influenced the severity of both fungal and nematode damage. However, contrary to our hypotheses, we found that the quality of the tubers and the fungal damage on the plant were not significantly dependent on the presence of nematodes prior to fungal inoculation (hypothesis no. 5 not confirmed). The severity of stem canker did not increase in the presence of nematodes and the stem canker was even less severe when nematodes were added in combination with the fungus, whereof hypothesis no. 1 was not confirmed.
In both experiments, the interaction between the two pathogenic organisms was most substantial regarding the tubers. The combination of plant-parasitic nematodes and R. solani resulted in lower tuber yield in both experiments, which was particularly observed in the second experiment, but the interaction was not synergistic. The root-lesion nematode P. penetrans may alone cause considerable yield loss, by 30% to 70%, in the potato crop, mainly through impact on the roots [28][29][30]. The reduction in root biomass restricts the optimal uptake of water and nutrients, which is needed for adequate potato yield. Therefore, the yield loss found in our experiments most likely depended on the observed reduction of root biomass in the presence of nematodes, regardless of the time of inoculation (Table 2, Figure 1, and Figure 2). In addition, in some treatments, there was an interaction effect of nematodes and fungus on the root biomass, and similar results were also observed in a previous study, where the potato cultivar Kuras was inoculated with a full nematode community dominated by root-lesion nematodes in combination with R. solani [12].
Contrary to our first hypothesis, plant-parasitic nematodes reduced the number of stem canker lesions on the stems instead of making them more severe, and the nematodes did not affect fungal damage on stolons or tubers either, as hypothesized. One possible explanation may be that the nematodes activated resistance mechanisms in the potato plants in the same way as root-knot nematodes may induce defence mechanisms in tomato [31].
In accordance with our fourth hypothesis, the time of inoculation influenced fungal and nematode damage. Fungal skin tuber damage increased if the fungus was added at 14 and 28 days, which may be because of infiltration of the added mycelium directly on the developing tubers. The fungal damage on the stolons was also affected by the time of fungal addition with increased necrosis if the fungus was added early and late, which may be because of the prolonged time of exposure when the mycelium was present during growth and that the mycelium could infiltrate directly on the stolons when added late, respectively.
Regarding nematode damage, the earlier the nematodes had the opportunity to colonise the plantlet, the more severe damage they achieved. The nematode damage and the number of nematodes in the plant parts were higher if the nematodes were added early, denoting that the nematodes thrived on the potato plants, even though not shown as visual damage in the first experiment. The necrotic lesions on tubers and stems, as well as the number of nematodes in the plant parts, were fewer when the nematodes were added late (at 28 days). The number of nematodes also increased in the tubers when the fungus was added early in experiment 2. Surprisingly, root-lesion nematodes were extracted from the stems. These nematodes generally do not occur in stems, but possibly did here because of the high competition for food in the limited soil volume. The stems may, however, be more difficult to feed from when they grow older, hence the lower abundance when nematodes were added after 28 days. Nematodes may first attack the roots and then later attack the stem when the fungus has affected the stem tissue [32].
The time of inoculation did also influence the number of plant-parasitic nematodes in the potting material, because the highest abundance was found when added after two weeks and the lowest abundance was found after addition at day 28 (experiment 2). This lower abundance in the potting material coincides with the higher abundances in the plant parts and indicates that the nematodes must have entered and left the plant parts regularly. Pratylenchus penetrans have a short life cycle of 4-8 weeks and the females can produce thousands of nematodes in the roots at once, which may explain the high amount of nematodes found at harvest [33].
Although no damage of nematodes was visible in the first experiment, the tubers did contain both plant-parasitic and fungal-feeding nematodes. The number of plant-parasitic nematodes in the tubers differed among the treatments, but there were differences between the two experiments, whereof no general conclusions about the time of inoculation and order of appearance could be drawn. Increased numbers of plant-parasitic nematodes in the tubers were observed for the combination treatment in five of six cultivars in a previous experiment [12].
The anastomosis group of the isolate used was not the one that usually affects potato (AG3), but AG2-1 is also pathogenic on potato and can be more aggressive on shoots and stems, and produces many small canker lesions [34,35]. The anastomosis group occurs in Swedish and Finnish agricultural soils, but was reported as less aggressive, and to at least form less sclerotia on the tubers [36,37]. The fungal strain used in these experiments did, however, demonstrate good efficacy in producing both stem canker and sclerotia on the tubers.

Conclusions
It is essential to understand and appreciate the importance of relationships between pathogens in order to control disease through appropriate management methods; thereof, more research is needed to unravel these questions. Our main conclusions are that the yield is affected by co-occurrence of root-lesion nematodes and R. solani, and that the nematodes may interact with the potato plant, leading to less stem canker. Our results highlight the importance of analysing the presence of nematodes in the field to be able to create protective strategies for an efficient potato production. Crop rotation with at least four potato free years and other agricultural measurements may be useful to reduce the population of R. solani in the field [1,38], and the effect of the crops grown in the potato free period on root-lesion nematodes needs to be taken into consideration.