Evaluation of root-knot nematode resistance assays for sugarcane accession lines in Australia

A total of 802 sugarcane accession lines were tested against M. javanica from 2012 to 2019 in eight trials at SRA Woodford Pathology Farm, Woodford, Queensland (26.9550° S, 152.7780° E) (Table 1). Among these, 648 lines (year 2012-2016) were from introgression populations, and 154 lines (year 2017-2019) were from the core breeding germplasm populations of SRA. In total, 22 accession lines, mainly commercial varieties/parents (including two susceptible controls), were incorporated in multiple trials to obtain data on their resistance to M. javanica (Table 2).

Details of trial procedures and design of experiments were described elsewhere (Bhuiyan et al., 2014, 2016; Croft et al., 2015). In short, all accession lines were collected from the SRA germplasm collection, Meringa, Queensland. Stalks of accession lines were stripped of leaves and leaf sheaths, cut into one-budded setts using an electric saw, hot water treated at 50^oC for 30 min to eliminate systemic fungal and bacterial diseases and placed in a germination chamber in trays with moist vermiculite. Germinated young plants were planted to 700 ml pots filled with sterilized fine white washed sands (Caboolture Gravel & Landscapes, Morayfield, Qld), maintained in a glasshouse for inoculation. Each trial was set out on air conditioned benches using a randomized complete block design in five replicate pots, and one plant per replication. Two data loggers (Thermodata Pty Ltd, Brisbane, Australia) were included in each trial by burying them approximately 5 cm deep in potting media to record the temperature of the potting media during the trial period. Two commercial varieties Q208 and Q135 that were highly susceptible to M. javanica were included in each trial as controls for comparison. The M. javanica were reared in the roots of tomato plants in a glasshouse as described by Van den Bergh et al. Eggs of M. javanica were extracted from the tomato plants using the bleach method (Stirling et al., 2011). Inoculation was done by applying approximately 5,000 eggs per pot. Eggs were added in 5 ml deionized water at the base of the plant in two holes using a pipette. Plants were maintained for approximately 12 weeks before harvesting and assessing for nematode population and resistance based on number of eggs per plant, number of eggs per g of fresh roots, and visual rating of gall severity. Root and shoot biomass were measured for each plant (Bhuiyan et al., 2016, 2019). For visual ratings, the roots were washed free of sands and visually rated for galling using a previously described method (Shepherd, 1979) with slight modification on a 0 to 5 scale based on the percentage of roots with galls, as shown in Figure 1, where 0 = no galls, 1 = ≤1 to 2%, 2 = 2 to 25%, 3 = 25 to 50%, 4 = 51 to 75%, 5 ≥ 75%.

Figure 1:

For the extraction of eggs, roots were washed free of sands, submerged in 1% NaOCl (bleach) and agitated for 5 min. The bleach solution was then poured through two sieves (a 150 μm sieve over a 38 μm sieve), with eggs being collected on the lower sieve, and finally washed into a 30-ml plastic vial. Extracted egg samples were stored at 6°C until counting. Enumeration of a subsample of nematode eggs were performed under a compound microscope (10×- 40×) using a Hawksley slide counting chamber of 1 ml capacity.

The number of eggs per plant was recorded after the enumeration, and the number of eggs per g of root was estimated from the total number of eggs present in the pot divided by the fresh weight of the root system. Data were analyzed by fitting a linear mixed model to all datasets using Proc Mixed in SAS version 9.4 (SAS Institute, Cary, NC). For each trial, accession lines were treated as a fixed effect. Block (replication) and the error term (residual) were treated as a random effect. The interactions of the fixed effects were included in the model. Degrees of freedom were adjusted using the Kenward–Roger method (Kenward and Roger, 1997) and normality of residuals was tested using Proc Univariate of SAS. The number of eggs per plant, and the number of eggs per g root data were log-transformed (ln (x+1)) before analysis. Proc Corr of SAS was used to calculate correlations between mean of nematode resistance parameters and root or shoot biomass. Corr plot = matrix in SAS was used to calculate the correlation matrix to determine the linear relationship of root and shoot biomass with nematode parameters. On the basis of the linearity, Proc Reg of SAS was used to determine the relationship between root biomass and eggs per g of roots. The percent of coefficient of variation (CV) within each trial for root and shoot biomass, and visual ratings for gall were calculated as CV = 100%(σ/µ), where σ and µ are the standard deviation and mean, respectively. As for the log-transformed data a formula: CV = 100 % e ln ( 10 ) 2 σ 2 − 1 , where σ is the standard deviation of the log-transformed data was used (Canchola et al., 2017). In total, 22 accession lines, as described earlier, were selected to determine the repeatability among the trials in relation to nematode resistance parameters (Table 2). Each accession line was included in at least four of the screening trials. SAS Corr in Fisher command was used to transform the data to normalize the distribution and stabilize the variance (Steel and Torrie, 1960).

The CVs for sugarcane biomass and nematode parameters varied among trials. The highest variations were in root biomass, ranging from 30 to 49% (Table 3). The CVs in the nematode parameters, eggs/plant, and eggs/g roots were higher in the early trials in years 2012 (Nem12-1) and 2013 (Nem13-1 and Nem13-2), ranging from 25 to 47%. In all other trials, the nematode parameter, eggs/plant had the lowest variations ranged from 2.3 to 9.2%, followed by eggs/g root (3.2-12%).

In 22 accessions, eggs per plant were more repeatable than eggs per g of root and visual ratings across the trials (Table 4). In 18 out of 28 possible trial combinations that include the same accessions, correlation of eggs per plant among trials were significant (p < 0.05). In 14 out of 28 possible trial combinations eggs per g of root for the same accession lines were correlated among the trials (p < 0.05). Only 11 out of 28 combinations for the same accession lines, was there a correlation between visual gall ratings among trials were significant (p < 0.05) with r values ranged from 0.72 to 0.92. Root biomass was variable in 22 accessions, varied from just over 3 g in commercial variety SRA2 to 25.3 g in IJ76-333 (Erianthus sp) (Table 2). Two wild accessions (IJ76-333 and IJ76-370) and only introgression accession QBYN05-20563 were highly resistant, supporting the lowest number of eggs per plant (3.7-12.4%) in relation to susceptible controls. Entries IJ76-370 and QBYN05-20563 supported only 7.8 and 3.7% eggs per g root compared to susceptible control, respectively. All other accessions had close to 20% or more eggs per plant or g root compared to susceptible controls. Among commercial varieties, Q231, Q241, and Q249 had less than 40% relative eggs per plant compared to the controls.

There were no significant correlations of nematode eggs per plant with root and shoot biomass in any of the eight trials (Table 5). Nematode eggs per g of root were negatively correlated (r = −0.24 to −0.47, p≤0.001) with shoot biomass in four out of seven trials. Low (r≤−0.19, p≤0.05) to strong (r = −0.65 and −0.74, p < 0.0001) negative correlations of eggs per g of root were observed with root biomass in seven out of eight trials. In two out of seven trials, visual ratings had moderate positive correlations with shoot biomass (r = 0.24 and 0.30, p < 0.05) and the rest of the trials had no significant correlations. Moderate negative correlations were observed between visual rating and root biomass in two out of eight trials, and one trial showed moderate positive correlations between these two parameters.

The relationships between eggs per g of root and root biomass varied among the trials (Fig. 2). In the trials Nem19-1 and Nem17-1 eggs per g of root explained 58% of variations in root biomass, and the regression and correlation coefficients were highly significant (p < 0.0001) (Fig. 2). In the trials Nem18-1 and Nem16-2, eggs per g of roots were able to explained 20 and 15% of variation in root biomass, and the correlation and regression coefficients were significant (p < 0.001). In three trials, Nem15-1, Nem13-2, and Nem12-1 eggs per g of roots were able to explain only 7% of variation in root biomass; although, correlations and regression coefficients were significant (p < 0.001). In all trials 5 to 8% reductions of root biomass were estimated in each unit increase of log-egg numbers per g of root. Slopes for the regression lines were significant in all trials (p < 0.05 to p < 0.0001), and varied from −0.11 in Nem15-1 to −3.02 in Nem12-1.

Figure 2: