Evaluation of 31 Potential Biofumigant Brassicaceous Plants as Hosts for Three Meloiodogyne Species

Abstract

Brassicaceous cover crops can be used for biofumigation after soil incorporation of the mowed crop. This strategy can be used to manage root-knot nematodes (Meloidogyne spp.), but the fact that many of these crops are host to root-knot nematodes can result in an undesired nematode population increase during the cultivation of the cover crop. To avoid this, cover crop cultivars that are poor or nonhosts should be selected. In this study, the host status of 31 plants in the family Brassicaceae for the three root-knot nematode species M. incognita, M. javanica, and M. hapla were evaluated, and compared with a susceptible tomato host in repeated greenhouse pot trials. The results showed that M. incognita and M. javanica responded in a similar fashion to the different cover cultivars. Indian mustard (Brassica juncea) and turnip (B. rapa) were generally good hosts, whereas most oil radish cultivars (Raphanus. sativus ssp. oleiferus) were poor hosts. However, some oil radish cultivars were among the best hosts for M. hapla. The arugula (Eruca sativa) cultivar Nemat was a poor host for all three nematode species tested. This study provides important information for chosing a cover crop with the purpose of managing root-knot nematodes.

Root-knot nematodes (RKN, Meloidogyne spp.) are economically the most damaging plant-parasitic nematodes in California vegetable production (Koenning et al., 1999). Nematode control with fumigant pesticides, although still used extensively in some crops, is highly regulated in California because of negative effects on human health and the environment. Fumigants have been identified as contributing to the emission of volatile organic compounds (VOCs), leading to poor air quality in several major crop growing areas of California. Under the 2007 Ozone State Implementation Plan, the California Department of Pesticide Regulation is required to reduce emission of smog forming VOCs from soil fumigants (Wang et al., 2009). Currently, restrictions on fumigant use during the May-October period are in place in five “non-attainment areas” in California. These restrictions may involve covering treated fields with tarpaulin, several postfumigation water treatments, application through drip tubing, reducing rates, increasing injection depths, and requirements for soil compaction (EPA, 2012). These restrictions generally result in higher costs to the grower, and therefore there is a need for economically viable alternatives for soil disinfestation, without the negative side effects. Biofumigation and anaerobic soil disinfestation (ASD) have been proposed as methods that can meet these requirements. With the ASD method, large amounts of fresh organic matter are incorporated in the soil followed by irrigation and sealing with plastic tarp for several weeks. During this period anaerobic conditions and fermentation products with pesticidal activity develop (Lamers et al., 2010). Biofumigation is a similar method and occurs when volatile compounds with pesticidal properties are released into the soil during decomposition of plant material or animal by-products (Halbrendt, 1996; Kirkegaard and Sarwar, 1998; Bello et al., 2000a, 2000b). Most research on biofumigation, however, has focused on using brassicaceous crops (Kirkegaard and Matthiessen, 2004). Upon tissue disruption, glucosinolate compounds in brassicas produce biocidal isothiocyanates that are released in the soil when the crop is shredded and incorporated (Chew, 1988; Brown et al., 1991). The suppressive effect of brassicaceous biofumigants on soilborne pathogens, weeds, and plant-parasitic nematodes has been demonstrated in numerous laboratory, greenhouse, and field studies (Ploeg and Stapleton, 2001; Ploeg, 2008; Zasada et al., 2010). To qualify as a good cover crop for the management of plant-parasitic nematodes, the crop should be a poor host for the nematodes and lower the population after incorporation of the crop into the soil (Viaene and Abawi, 1998). A complication with growing brassicas or other crops as cover crops for biofumigation or for ASD to control RKN, is that they may multiply the target nematode population (Ploeg, 2008; Zasada et al., 2010). To avoid this, cover crops can be grown when soil temperatures are sufficiently low to prevent nematode activity (Roberts, 1987), or resistant or nonhost cover crop cultivars can be grown (Stirling and Stirling, 2003; Pattison et al., 2006). The objective of this greenhouse study was to compare the host status of a range of brassicaceous crop cultivars for three species of RKN (M. incognita, M. javanica, M. hapla) that are common in California, and identify cultivars that have a low risk of increasing the nematode population when grown as a biofumigant cover crop.

Materials and Methods

Nematode origin:

Populations of three locally occurring Meloidogyne spp. were used in the experiments: M. incognita race 3, originally isolated from cotton in the San Joaquin Valley, CA; M. javanica from cowpea, Chino, CA; and M. hapla from alfalfa in San Bernardino, CA. Species and race identifications were done with isozyme electrophoresis and on differential host tests (Eisenback and Triantaphyllou, 1991). Populations were increased and maintained on tomato ‘UC82’ grown in coarse sand in a greenhouse. To prepare inoculum, Meloidogyne eggs were extracted from heavily infested tomato by shaking the roots in a 0.525% NaOCl solution (Radewald et al., 2003). Eggs released from the roots were collected on a 25-μm pore-size sieve and were counted in two 0.025-ml subsamples. Egg suspensions were then adjusted to contain 20,000 eggs in 5 ml.

Experimental design and treatments:

Thirty-one cruciferous cultivars comprising eight species (Table 1) were seeded in 1-gal (3.8-liter) plastic pots filled with steam-sterilized sandy soil (93% sand, 4% silt, 3% clay; pH 7.1). Each pot received 10 seeds, and five pots were seeded for each cultivar. In addition, five pots were planted with one 3-wk-old tomato ‘UC82’ seedling. Pots were placed in a completely randomized order on a greenhouse bench (17°C to 29°C, natural light), 10-g slow release fertilizer (Osmocote®) was added to each pot, and pots were watered daily through an automated drip system. Two weeks after emergence, enough seedlings were removed to give three plants per pot. One week later, three 3-cm-deep small holes were made in a triangular pattern over the surface of each pot, 20,000 RKN eggs in 15-ml water were inoculated by adding 5 ml into each hole, and the holes were then covered with soil. Eight weeks after inoculation, three soil cores including root material, were removed from each pot from top to bottom (17 cm) with a 2-cm diam. sampling rod. The soil removed from each pot was thoroughly mixed, and 100g was used for nematode extraction over a 5-d period with a modified Baermann funnel-technique (Rodriguez-Kabana and Pope, 1981). Second-stage Meloidogyne juveniles (J2) were counted using a dissecting microscope at 40× magnification. Plants were carefully removed from the pots, the root systems were washed free of soil, and were thoroughly examined for the presence of galls. The severity of root-galling was indexed on a scale from 0 (no galls) to 10 (100% of roots galled) (Bridge and Page, 1980). Separate experiments were done with M. incognita, M. javanica, and M. hapla as inoculum, and the entire experiment was repeated once for each of the three Meloidogyne spp.

Table 1.

Thirty-one brassicaceous crop cultivars used in Meloiodgyne greenhouse pot experiments.

Statistical analysis:

Raw nematode data (number of RKN J2 recovered from soil) was log₁₀(X+1)-transformed before analysis. Effects of the plants on RKN J2 levels and root galling were analyzed in an analysis of variance (ANOVA) procedure, and means were compared using Fisher’s protected least significant difference (LSD) test (P ≤ 0.05) using SAS statistical software (SAS Institute, Cary, NC). All data presented are nontransformed means ± standard errors. Spearman rank correlation coefficients between J2 levels and galling, and between data from the repeated experiments were also calculated using SAS software.

Results

Host status for M. incognita:

Average J2 counts were significantly different between the two replicated experiments (P = 0.0001) and between the 32 plant cultivars tested (P = 0.0001). There was no significant interaction between the experiment and cultivar (P = 0.097) indicating that the 32 varieties gave similar results in the two experiments. This was confirmed by ranking the cultivars according to J2 levels at harvest and subjecting the ranking to a Spearman rank correlation test. Results show that there was a significant positive correlation of the 32 cultivars in J2 numbers between Experiments 1 and 2 (Spearmann rank correlation coefficient 0.73; P = 0.0001). In both experiments, B. rapa ‘Br02206’ was as good a host as tomato (P ≤ 0.05; Table 2). All three B. juncea cultivars also were relatively good hosts ranking among the best seven plants in the two experiments. Ratios of J2 populations of brassicaceous plants relative to tomato (P_b/P_t) for B. juncea ranged from 0.75 (‘Pacific Gold’, Experiment 1) to 0.12 (‘ISCI99’, Experiment 2). The poorest hosts for M. incognita were broccoli (B. oleracea) ‘Liberty’ and R. sativus ‘Boss’, both reducing J2 numbers compared with tomato on average by more than 99%, and resulting in significantly lower J2 levels than any of the other cultivars. Other cultivars that were among the 10 poorest hosts in both experiments included the five R. sativus cultivars ‘Adagio’, ‘Colonel’, ‘Comet’, ‘Defender’, and ‘TerraNova’, E. sativa ‘Nemat’, and S. alba ‘Abraham’.

Table 2.

Root galling and Meloidogyne incognita second-stage root-knot nematode juvenile numbers 8 wk after inoculating brassicaceous cultivars and tomato with 20,000 eggs (N = 5) per 3.8-liter pot. Ranking from 1 (highest) to 32 (lowest).

Average galling was not significantly different between the two experiments (P = 0.81), nor was the interaction between “experiment” and “cultivar” (P = 0.37). Galling on R. sativus ‘Boss’ was very minor in both experiments, but none of the cultivars remained free of root-galling. Cultivars of B. juncea and B. rapa all showed very obvious root-galling. The severity of galling was positively correlated with the J2 levels in both experiments (Spearmann rank correlation coefficient 0.59, P = 0.0003; and 0.59, P = 0.0004 for Experiments 1 and 2, respectively) (Table 2).

Host status for M. javanica:

There was a significant effect of the “experiment” (P = 0.0001), of the “cultivar” (P = 0.0001), and of the interaction between “experiment” and “cultivar” (P = 0.0001) on average J2 counts. Still, the ranking of the cultivars according to J2 numbers was very similar between the two experiments (Spearmann rank correlation coefficient 0.73; P = 0.0001). Good hosts for M javanica in both experiments included B. juncea ‘ISCI99’ and ‘Nemfix’ and B. rapa ‘Rondo’. These cultivars resulted in J2 numbers that were not significantly lower than after tomato. In the first experiment, R. sativus ‘Boss’, ‘TerraNova’, and ‘Defender’, and E. sativa ‘Nemat’ yielded significantly fewer J2 than the other cultivars. These four cultivars were also among the five poorest hosts in Experiment 2. Broccoli ‘Liberty’ was a very poor host in Experiment 2 (ranking 30 of 32; P_b/P_t = 0.003), but a moderately good host in Experiment 1 (ranking 18 of 32; P_b/P_t = 0.26).

“Experiment,” “cultivar,” and “experiment × cultivar” effects on galling were also significant (P = 0.0001). Like with M. incognita, cultivars of B. juncea and B. rapa generally showed obvious root galling, where as galling on R. sativus ‘Boss’ was very minor. None of the cultivars remained free of galling, and there was a significant positive correlation between root-galling and J2 levels (Spearmann rank correlation coefficient 0.67, P = 0.0001; and 0.68, P = 0.0001 for Experiments 1 and 2, respectively) (Table 3).

Table 3.

Root galling and Meloidogyne javanica second-stage root-knot nematode juvenile numbers 8 wk after inoculating brassicaceous cultivars and tomato with 20,000 eggs (N = 5) per 3.8-liter pot. Ranking from 1 (highest) to 32 (lowest).

Host status for M. hapla:

J2 levels were quite different between the two replicated experiments (P = 0.0001). For example, in the first experiment on average only 804 J2/100 g soil were obtained after 8 wk under tomato, whereas in the second experiment 27,272 J2/100 g soil were found. Effects of “cultivar” and “experiment × cultivar” were also highly significant (P = 0.0001). In the first experiment, J2 numbers after B. rapa ‘Br02005’ and ‘Br02006’, and R. sativus ‘Colonel’, ‘Doublet’, ‘Final’, and ‘TerraNova’ were not significantly different from those after tomato. In the second experiment, none of the brassicaceous cultivars were as good a host as tomato. The two B. rapa cultivars ‘Rondo’ and ‘Samson’, and the two R. sativus cultivars ‘Colonel’ and ‘TerraNova’ were among the 10 best brassicaceous hosts for M. hapla in both experiments. Conversely, ‘Adagio’, ‘Condor’, and ‘Nemat’ were among the five poorest hosts in both experiments. Reductions in J2 numbers by the latter three cultivars relative to tomato ranged from 90% (P_b/P_t = 0.10) to 99% (P_b/P_t = 0.01) for ‘Condor’ in Experiments 1 and 2, respectively. Despite the variability in J2 numbers between the two experiments, the ranking of the cultivars regarding J2 numbers was significantly positively correlated (Spearmann rank correlation coefficient 0.52, P = 0.002).

The overall average galling was similar between the two experiments (“experiment”; P = 0.416), although there was some variability in galling on the individual cultivars between the two experiments (“experiment × cultivar”: P = 0.015). Galling occurred on all brassicaceous cultivars, ranging from a galling index of 1.8 on B. napus ‘Winfred’ and B. oleracea ‘Liberty’ in Experiment 1, to 6.6 on R. sativus ‘TerraNova’ in Experiment 2. In both experiments, cultivars that ranked high in galling generally also ranked high in J2 numbers and vice versa (Spearmann rank correlation coefficient 0.51, P = 0.0028; and 0.74, P = 0.0001 for Experiments 1 and 2, respectively) (Table 4).

Table 4.

Root galling and Meloidogyne hapla second-stage root-knot nematode juvenile numbers eight weeks after inoculating brassicaceous cultivars and tomato with 20,000 eggs (N = 5) per 3.8-liter pot. Ranking from 1 (highest) to 32 (lowest).

Discussion

Our results show that there are significant differences within and between brassicaceous species with regard to host status for RKNs, and furthermore that the host suitability of a particular brassicaceous cultivar for RKNs can differ depending on the nematode species. Therefore, identifcation of the target RKN to species level is important to optimize this management strategy. Results for M. incognita and M. javanica were similar, with B. juncea and B. rapa cultivars generally being good hosts, and E. sativa Nemat and the R. sativus cultivars Boss and TerraNova consistently ranking among the poorest hosts. Whether inoculated with M. incognita or M. javanica, reproduction on the latter two cultivars was reduced by at least 98% compared with reproduction on tomato. These results are in agreement with findings by Curto et al. (2005) who also reported that B. juncea cultivars were among the better hosts for M. incognita, and that R. sativus Boss and E. sativa Nemat were poor or nonhost. Broccoli, in our study a poor host for M. incognita and to a lesser degree for M. javanica, was also identified by McSorley and Frederick (1995) as a crop that showed potential to reduce Meloidogyne populations, although they used a different cultivar. Brassica juncea Nemfix, in our study among the best hosts for M. incognita and M. javanica, was also found a good host for M. javanica by Stirling and Stirling (2003), which led them to conclude that “it is imperative that in warm climates only poor hosts are grown” and that “a screening program could identify such material.” Monfort et al. (2007) included B. juncea Pacific Gold in a cover crop field study on a M. incognita–infested site and showed that the nematodes reproduced on the cover crop. They also concluded that information on the level of susceptibility of Brassica species to RKN is needed for biofumigation to become a succesful nematode management strategy. Johnson et al. (1992) and McLeod et al. (2002) reported that M. javanica and/or M. incognita caused galls on B. napus Humus and that populations were maintained on this cultivar. In our study, both these Meloidogyne species also caused galls on Humus roots, and the cultivar was not a particularly good or bad host, despite containing high levels of glucosinolates (Johnson et al., 1992). The R. sativus cultivar Adagio, in our study a poor host for both M. incognita and M. javanica, was also a poor host for M. javanica in a study by McLeod et al. (2001). Gardner and Caswell-Chen (1994), however, reported that both M. incognita and M. javanica produced numerous females in the roots of this cultivar, and recommended that it should not be grown on M. incognita– or M. javanica–infested sites for risk of nematode population increase. The response of M. hapla to the different brassica cultivars was sometimes very different from results with M. incognita or M. javanica. For example, R. sativus Colonel and TerraNova were among the best hosts for M. hapla, but poor hosts for M. incognita and M. javanica. Also, the variability in host status for M. hapla within the same brassica species was greater. For example, within R. sativus there were good hosts (Colonel, TerraNova) as well as poor hosts (Adagio, Condor). The fact that plants resistant to M. incognita and M. javanica are often susceptible to M. hapla (Roberts, 1992) is thought to be the result of basic differences in the inheritance of resistance (Bünte et al., 1997). Within R. sativus lines, it is also thought that differences in reponse to M. incognita or M. javanica, and M. hapla may be because of a single gene resistance mechanism for the former two species, and a polygenic resistance mechanism for M. hapla (Bünte et al., 1997). Ongoing breeding efforts may lead to more cultivars with resistance to the range of economically inportant Meloidogyne species. In our study, poor hosts for all three Meloidogyne species include R. sativus Adagio and E. sativa Nemat. The latter cultivar was also reported a poor host for M. hapla by Melakeberhan et al. (2006, 2010). Others (Curto et al., 2005; Melakerberhan et al., 2006) studied the mechanism responsible for the differences in host status of brassica crops, and found that in poor or nonhosts, the nematodes do invade the root systems but fail to develop into females and/or develop very slowly, rendering the crops a trap crop.

Although this study did not include the biofumigant effect that occurs during cover crop decomposition after soil incorporation of the crop, it would be unproductive to grow a host crop as a biofumigant. This was also concluded by Monfort et al. (2007) who observed a large net reduction in M. incognita population levels from brassica cover crop incorporation to planting of the next vegetable crop, but also noted that the increase in nematode levels during cover crop growth was a major obstacle. This study clearly demonstrates that large differences occur between different brassica cultivars with respect to their host status for three Meloidogyne species. Based on this study, the R. sativus cultivars Boss, Terranova, or E. sativa Nemat would be good choices for cover crops on M. incognita– or M. javanica–infested sites. Conversely, B. rapa or B. juncea cover crops would carry a risk of substantial nematode multiplication. On M. hapla–infested sites, E. sativa Nemat would carry little risk of nematode multplication, whereas B. juncea cultivars should be avoided. Evaluation of the validity of these recommendations under field conditions is an important next step.