Relationship between inoculum pressure of Fibularhizoctonia carotae in soil attached to the roots at harvest measured by quantitative PCR (pg DNA/g soil) and incidence of crater rot after storage. Data from naturally infected field trials with 10 replications at 5 farms over 3 years (2006 – 2008) 

Context in source publication

Context 1

... all tested isolates of the target organisms, but DNA from the other fungal species ( Table 1) did not give any signal. The detection limits from the two assays in singleplex reactions were deduced from the standard curves (Fig. 3a and b) to be 0.1 pg of DNA, and a linear range was observed over five orders of magnitude. The PCR efficiency of the F. carotae singleplex assay was lower than for the M. acerina assay, i.e. 79,8% compared to 101.9%. Both primer/probe sets were added in the same reaction mixture to facilitate the simultaneous detection and quantification of M. acerina and F. carotae in the samples to be tested. The detection limits were the same in the duplex reaction assay as in the single reactions (Fig. 3c and d). The PCR efficiency for F. carotae was however reduced to 75.8%, while the PCR efficiency of M. acerina was less affected (98.6%). For the singleplex assays as well as for the duplex qPCR assay, the correlations between the cycle threshold (Ct) values and the known DNA quantities in the dilution series were high (R 2 ≥ 0.99) (Fig. 3). Correspondence between PCR results and incidence of diseases Liquorice rot was found after storage (>0.1% diseased roots) in carrots from 14 of the 18 fields, and crater rot was only found after storage (>0.1% diseased roots) in carrots from 8 of the 18 fields. The average disease incidence within fields after storage ranged from 0% to 64% for liquorice rot and from 0% to 13% for crater rot. Primer pairs KLO15/KLO20 and FIB2/FIB 5 were used to test for the presence of the liquorice rot and crater rot pathogens in carrots and soils. For M. acerina , the variation in PCR results between fields was significant larger than the variation between the plots within the fields. For F. carotae the results were the opposite, which indicates that this pathogen has a more patchy distribution of the soil inoculum than M. acerina (Table 3). The best correlations were obtained between DNA from M. acerina and F. carotae in PCR data from carrot-soil samples and incidence of liquorice rot or crater rot (Table 3, sample types 6 and 7). The correlation between DNA from M. acerina and F. carotae in PCR data from soil samples (between rows) taken in the spring within 14 days after sowing (sample types 8 and 10) or in late fall after harvest (sample 9) and incidence of crater rot or liquorice rot was weak (Table 3). For the early carrot-soil samples in August (sample type 6), the model fit for liquorice rot was very good (R (adj) was 84.8%) (Table 3, Fig. 4), and also good for crater rot (R 2 (adj) 97.9%) (Table 3). The correspondence between the PCR results from samples containing carrot root tissue (sample types 1, 2, 4 and 5) and the incidence of the two diseases was generally low (R 2 (adj) <50%), except data on M. acerina — DNA from carrots in August (R 2 (adj) 77.4%) (Table 3). The calculation of the predictive ability of the different samples revealed that the carrot- soil samples from August was good for predicting liquorice rot (R 2 (pred) 55.5%). The predicting ability of the carrot root tissue samples from the same time point was less good (R 2 (pred) 37.4%) (Table 3). No sample types had predictive ability for crater rot (R 2 (pred) 0%) (Table 3). Liquorice rot was found in 9 of the 15 fields and 40% of the 69 stored carrot samples, and the disease incidence ranged up to 45%. Crater rot was found in 9 of the 15 fields and 25% of the 69 stored carrot samples, and the disease incidence ranged up to 57%. For crater rot there was large variation within fields in disease incidence. As in Experiment A, PCR data on liquorice rot generally correlated much better with disease data than was the case for data on crater rot. Data on M. acerina DNA from soil shaken from carrots generally correlated better with liquorice rot incidence in cold-stored roots than in DNA from ‘ bulk ’ soil taken from carrot fields in autumn or spring. For liquorice rot, data on PCR from soil samples taken from the roots at harvest correlated best and gave the best predictive ability (standard PCR: R 2 (adj) 75.5%, R 2 (pred) 74.9%, qPCR: R 2 (adj) 64.4%, R 2 (pred) 63.4%) (Table 4, Fig. 5), but PCR data from soil samples taken from carrots during the growing season also gave reasonable good predictive ability (R 2 (pred) ranging from 27.6% to 62.9% for the different samples and PCR methods). Data on PCR from samples taken in the autumn before sowing also showed some correlation with the disease data on liquorice rot and had a moderate predictive ability value for this disease (standard PCR: R (adj) 34.7%, R 2 (pred) 31.7%, qPCR: R 2 (adj) 40.7%, R 2 (pred) 38.6%) (Table 5). Correlations between detected DNA amounts of F. carotae and incidence of crater rot were generally very poor (Table 4); no PCR analyses of the different samples gave R 2 (pred) value >1% for crater rot. The lack of relationship between PCR data for F. carotae and crater rot incidence in soil samples taken at harvest is shown in Fig. 6. Both false positive and false negative PCR data was obtained. This was also found in some cases for M. acerina (Fig. 5). Standard PCR and quantitative PCR gave similar results with slight variations, and data on both methods are presented in Tables 4 and 5. Liquorice was found in stored carrots from 9 of 14 fields, and crater rot was found in 7 of the 14 fields. The densest sampling strategy (25 subsamples) generally gave better correlation between PCR data and disease data than the 9 subsample strategy (Table 5). For crater rot only the samples from the denser strategy were predictive, although the values were low (R 2 (pred) 11.0% in spring, and 7.4% for early August samples for standard PCR). For liquorice rot, increasing the number of early August samples from 9 to 25 increased the predictive ability value from R 2 (pred) 55.6% to 67.3% using standard PCR and from R 2 (pred) 42.5% to 48.3% using qPCR (Table 5). With the advent of quantitative molecular methods, it is attractive to make use of this type of data to assess whether fields have suitably low levels of specific pathogens to prevent disease development in the field or, in the case of vegetables and fruit that are stored after harvest, during long-term storage. We have developed specific PCR methods for identifying the pathogens M. acerina and F. carotae , which are serious causes of storage rot in carrots during prolonged low temperature storage. The method is rapid and sensitive, allowing routine detection of less than 0.3 pg of M. acerina DNA, and less than 0.03 pg of F. carotae DNA, even in the presence of large excesses of plant or soil DNA. Further large increases in the sensitivity of a PCR assay could be attained using a primer pair that amplified a shorter product, but at the cost of some loss of specificity (data not shown). Use of quantitative PCR did not provide improvement in the results, and either of the PCR methods could be used in a practical situation. qPCR requires more equipment but is less laborious than standard PCR, where products are displayed and quantified on gels. The PCR primers developed were species-specific for the target fungal pathogens M. acerina and F. carotae . Specificity was tested on a range of different soilborne fungal species including other pathogens that might attack carrots. We did not include close relatives of M. acerina in the test of specificity, but such species are not known to be common in agricultural soils (see references on . speciesfungorum.org/Names/Names.asp?strGenus= Mycocentrospora). Issues that can affect the accuracy of DNA-based soil testing include sampling strategies, sample size, DNA extraction from soil and assignation of disease risk categories. These have been discussed in detail by Ophel-Keller et al. (2008), and we present here our results on these issues. In small plot experiments (Experiments A and B), soil sampling was relatively dense, down to less than 5 m between the sampling points. Such dense sampling is not feasible in a practical field situation. Thus in Experiment C we tested a sampling strategy with 20 versus 40 m between sampling points (9 versus 25 cores/sampling points per ha). PCR data from samples taken using a dense sampling strategy gave higher correlation coefficients, and higher predictive ability values than using the less extensive sampling strategy. This is not surprising if the fungal propagules have a patchy distribution. Therefore the denser sampling strategy is most cost-effective with regard to the potential for a more reliable prediction. The sampling density of 25 cores/sampling points per ha is similar to sampling densities used for detection and quantification of other soil-borne pathogens, like Colletotrichum coccodes and Rhizoctonia solani in potatoes in the UK (Lees et al. 2010; Peters et al. 2011). The distribution of the propagules of our examined pathogens in field soils is not very well documented. In the absence of a host, M. acerina chlamydospores usually remain quiescent in soil and growth of this pathogen in soil is negligible (Wall and Lewis 1980). M. acerina that establishes in the rhizosphere or in the crown of the root is probably the main source of inoculum of this pathogen. Splash dispersal of conidia of M. acerina is possible within fields, but aerial spread by wind of these conidia appears to be insignificant (Evenhuis et al. 1997; Hermansen and Amundsen 2000). M. acerina is a polyphageous fungus including several weeds as hosts (Hermansen 1992). F. carotae is considered to be a soilborne pathogen (Rader 1948; Davis and Raid 2002). F. carotae has relatively few reported hosts, but the pathogen has the ability to live as a saprophyte, including growth on wooden crates (Davis and Raid 2002). Basidiospores from the anamorph Athelia arachnoidea growing on dead deciduous leaves might be important in the epidemiology of this pathogen (Adams and Kropp 1996). If basidiospores infect senescent carrot leaves, this ...