Entomopathogenic Bombs – Sweet Potato Weevils Be Gone

Sweet potato weevil causes serious yield loss to growers wherever sweet potato is grown, including Hawaii. Current grower practices rely upon regular pesticide applications which have environmental, economic, and societal drawbacks. We have completed several field experiments and demonstrations to show how Entomopathogenic Nematodes (EPN) can be incorporated into the sweet potato cropping cycle to manage sweet potato weevils and reduce insecticide usage.

One field experiment focused on demonstrating the comparative effectiveness of Heterorhabditis indica and Steinernema feltiae, two EPNs, in aqueous and cadaver applications. EPN-infected cadavers or living bombs provide the EPN a more protected environment and result in applications of where the EPN are more effective. The EPN-infected cadavers tested were larvae of Galleria mellonella. This field experiment consisted of cadavers infected with H. indica, cadavers infected with S. feltiae, aqueous application of H. indica, aqueous application of S. feltiae; and an untreated water control. Plots were planted with 20-cm long slips of Ipomea batatas Okinawan and EPN treatments commenced once sweet potato roots began to swell, around 3 months after planting. The EPN treatments were applied monthly until harvest or 4 to 6 months after planting. The EPN-infected cadaver treatment rate was based on average EPN yield recorded for a cadaver in the laboratory such that EPN-infected cadaver and aqueous applications were at 2.0 x 10⁹ IJs/ha. At harvest, sweet potato yield was recorded, swollen roots were graded, and a subsample of sweet potato was assayed for sweet potato weevil infection.

The sweet potato weevil pressure was low in the test field. Less than 10 males were captured in a pheromone trap during the entire sweet potato growth period. Consequently, few sweet potato roots were damaged by the weevil even in the untreated controls. Nevertheless, plots treated with S. feltia-infected cadavers and plots receiving aqueous sprays of H. indica had the highest yields which was greater than the untreated control (P<0.05). We attribute the increased yield to the effects that the EPNs had on other aspects of soil-surface and soil biological community.

A second field demonstration was conducted to optimize application rates for the EPN-infected cadavers for effective sweet potato weevil control. Plots were established similar to the ones in the previous demonstration. When sweet potato root swelling commenced, the EPN-infected cadavers were applied on the soil under the sweet potato canopy. Larvae of G. mellonella infected with H. indica were used in this test. Four treatments were evaluated consisting of 0 IJ/ha, 1 x 10⁹ IJs/ha (0.5X), 2 x 10⁹ IJs/ha (1X), or 4 x 10⁹ IJs/ha (2X). Sweet potato harvest data similar to the previous demonstration was collected.

Similar to the previous field, sweet potato weevil pressure in the field was low with little loss even in the untreated plots. However, the 0.5X application rate or 1 x 10⁹ IJs/ha resulted in the greatest sweet potato yield of all treatments (P<0.05). It may have been that the higher rates of cadaver application resulted in greater predation of the cadavers by geckos, birds, or ants in the field, thus reducing the actual amount to EPN application.
Another series of experiments have been undertake to facilitate the application of EPN-infected cadavers. The infected cadavers can be challenging to transport and apply in the field. A gentle hand is required so as not to break the cadavers and allow the EPN to be exposed prematurely. Desiccation of the cadavers, or removal of 15% of their moisture, resulting in EPN-infected cadavers that are leather like. These leather-like cadavers are easier to transport and less likely to rupture when applied in the field.

Laboratory experiments were conducted with larvae of G. mellonella infected with H. indica that were desiccated to 85% relative humidity. The desiccated larvae are leathery and not as prone to rupture during handling. IJ emergence did not differ between non-desiccated and desiccated larvae. The average IJ emergence of both desiccated and non-desiccated larvae was 107,564 and 116,049 IJs/cadave, respectively (P<0.05). No difference was observed in IJs during subsequent infection of G. mellonella larvae. Both the desiccated and non desiccated IJs resulted in 100% mortality of G. mellonella. Desiccation does not appear to adversely IJ. Future research should investigate holding desiccated cadavers for longer periods as a means to store cadavers for future use.

Steinernema feltiae was less desirable as a candidate for desiccation due to IJ emergence from the host insect cadaver during desiccation. Galleria mellonella infected with S. feltiae were easily ruptured during handling and application. Cadavers of G. mellonella deteriorated rapidly when infected with S. feltiae. When Tenebrio molitar was used as the host, IJ of S. feltiae would emerge during the desiccation process. This early emergence of IJ lessens the potential effectiveness of the EPN application. These two reasons-fragility of cadavers and early IJ emergence, made S. feltiae a less than ideal candidate for EPN-infected cadaver application.

Heterorhabditis indica is a good candidate for desiccation and storage when reared on G. mellonella. H. indica is not adversely affected by desiccation and the desiccated cadavers are easier to apply in the field.

A large plot field experiment was conducted to elucidate the effectiveness of desiccated EPN-infected cadavers. Cadavers were applied at a rate equivalent to2.0 x 10⁹ IJs/ha commencing when sweet potato roots began to swell, about 4 months after planting. EPN were applied monthly until harvest. Other treatment applications included a Metarhizium-amended mulch and post-plant application of Metarhizium. The EPN-infected cadaver applications reduced sweet potato weevil damage compared to the untreated controls but not as much as the Metarhizium-amended mulch plots. Plants in the Metarhizium-amend mulch plots tended to be larger than the other plots regardless of postplant treatments. A noticeable yield gradient was evident in total yields from replication 1 to replication 4. It is likely that EPN application should have commenced sooner to offer earlier protection from the sweet potato weevils.