Abstract
Euschistus servus (Say) can develop a generation on wheat, Triticum aestivum L., before moving to corn, Zea mays L., where it can be a pest. Because effective management methods are unknown, this study sought to describe the spatial distribution and dispersal of E. servus in the wheat and corn interface. In addition, Oebalus pugnax (F.) densities were documented in both crops. Wheat fields adjacent to the corn were sampled before harvest and stink bugs were marked using a product containing egg whites. Dispersal into the adjacent corn was measured using grid sampling, and dispersion was measured over time using an immunoassay targeting egg albumin on E. servus collected in corn. Dispersion was measured using Anselin Local Moran's I for unmarked stink bugs only. O. pugnax was prominent in wheat but was rarely recovered from corn. In contrast, E. servus was common in wheat during both years and dispersed into the adjacent corn. E. servus nymph and adult densities increased quadratically over time in corn in 2011. In contrast, E. servus nymph densities decreased over time in corn in 2012, while adult densities remained static. Most aggregations of E. servus nymphs and adults were located on the edge of the corn, directly adjacent to the harvested wheat. This is likely the first study to directly document the movement of E. servus nymphs to the adjacent crop. Movement from wheat to corn was not consistent between the years and may have been influenced by factors such as variations in weather, timing of wheat harvest, or other available alternative hosts.
In the southern United States, stink bugs are polyphagous hemipteran pests that feed on many crops and uncultivated hosts (McPherson and McPherson 2000). They are also important pests of both soybean, Glycine max (L.), and cotton, Gossypium hirsutum (L.) (Musser et al. 2012, William 2012). In addition, stink bugs are commonly found in wheat, Triticum aestivum (L.) but do not cause economic injury to grain yield (Blinka 2008). Finally, stink bugs can be pests of corn, Zea mays (L.), injuring it during both the vegetative and reproductive stages (Apriyanto et al. 1989, Ni et al. 2010).
In North Carolina, Euschistus servus (Say) can undergo their first generation completely on wheat (Blinka 2008). Soft red winter wheat is a common crop that is planted annually in the fall and harvested in the spring in the southeastern United States. North Carolina growers annually planted between 255,000 and 348,000 ha of wheat from 2007 to 2011 (U.S. Department of Agriculture–National Agricultural Statistics [USDA–NASS] 2008, 2009, 2010, 2011, 2012). Growers in North Carolina also annually planted ≈350,000 ha of corn during this time period (USDA–NASS 2007, 2008, 2009, 2010, 2011, 2012), often adjacent to or near wheat fields that were beginning to mature. Because of this proximity, wheat serves as a source for E. servus infestation in corn during the late spring and early summer (Blinka 2008, Reisig 2011). It is thought that adult E. servus disperses from wheat during harvest, moving into the adjacent corn. In other areas of the southeast, such as Georgia, E. servus is not found in corn until June, likely dispersing into this crop from wild hosts as well as wheat (Herbert and Toews 2011). Although other alternative hosts, such as cotton, peanut, Arachis hypogaea L., and soybean are present during this time, E. servus is most abundant in corn until the summer solstice. Before the summer solstice, E. servus dispersal into and abundance in corn seems to be most associated with food quality of the host (Herbert and Toews 2011).
Some growers apply an insecticide for E. servus by air after tasseling in corn generally aerially and tank mixed with a fungicide. It is not necessarily targeted at E. servus but is applied with the hope of increasing yield by eliminating some disease and insect pressure. However, even when aerial insecticide treatments are targeted for E. servus in VT-stage (Hanway 1963) corn, they are not an effective management method (Reisig 2011).
To better focus management efforts for E. servus, it has been suggested that insecticide sprays should be concentrated on stink bugs when they are aggregated in a single crop, rather than dispersed throughout the farmscape (Blinka 2008, Herbert and Toews 2011). Growers who are concerned about stink bugs dispersing from wheat to corn can apply an insecticide to wheat that is adjacent to corn. At this point in the season in North Carolina, E. servus is concentrated in wheat and nearly undetectable in other agronomic field crops (Blinka 2008). However, the most effective registered insecticides (pyrethroid-class) have a 30-d preharvest interval restriction. If applied by ground, growers are disinclined to drive over wheat after growth stage (GS) 31 (Zadoks et al. 1974) because of the associated yield loss (Lee et al. 2009). Finally, the F1 generation of E. servus is primarily produced from GS 80 to maturity, GS 92 (Blinka 2008). The time period for this to occur in North Carolina is generally less than a month (Blinka 2008), negating the effects of an insecticide spray 30 d before harvest.
The dispersion of E. servus has been described in wheat. Populations of E. servus are generally aggregated in wheat, with more aggregations occurring within the first 10 m of a field border (Reay-Jones 2010). Similarly, kernels injured by stink bugs and populations of E. servus are aggregated near corn field edges (Ni et al. 2011; Tillman 2010, 2011). However, the ecological explanation for this phenomenon remains unexplained.
In contrast, little is known about the dispersal of this insect in wheat and corn. E. servus resides in and feeds on corn during vegetative and reproductive phenological stages, but adults are more common after the VT stage (Herbert and Toews 2011, Tillman 2010). In Georgia farmscapes, E. servus moves among corn, peanut, and cotton (Tillman 2011), presumably during the F1 and F2 generations. However, description of the dispersal and oviposition of E. servus as they emerge from overwintering coupled with the development of the F1 generation remains unknown in this system.
Presumably, the wheat and corn interface is not as prevalent in the Georgia system compared with North Carolina. This study sought to describe the spatial distribution and dispersal of E. servus in a North Carolina wheat and corn interface, with attention to when stink bugs would move and how far they might move. It was hoped that by better describing the ecology of this system, management efforts could be shifted from poorly timed insecticide sprays applied on a whole-field basis.
Materials and Methods
In late-May 2011, a 29-ha commercial wheat field (35.781, −76.357) was selected near Phelps Lake, NC, with relatively high abundances of E. servus (one per pendulum sweep using a 38-cm diameter net). This was located directly adjacent to a 29-ha corn field (35.781, −76.356) separated by a 1 m-wide drainage ditch. In 2012, the corn field from the previous year was planted with wheat and was selected during early June for experimentation (35.779, −76.353). As in the previous year, this was located directly adjacent to a 29-ha corn field (35.779, −76.352) separated by a 1-m drainage ditch. The fields each consisted of five 5.7-ha cuts and were located in an area of intense agricultural cultivation. They were surrounded on three sides by other wheat and corn fields as well as a dirt road with full-season soybean on one side in 2011 and corn in 2012. Typical for the area, in both years, the growers applied a pyrethroid insecticide with a fungicide when nitrogen was applied to the wheat field directly before GS 31 (during March of both years). No other foliar insecticide was applied to either the corn or wheat field.
A sampling grid (Fig. 1) was placed in both fields 100 m from the road. Each point was identified using a 1.8-m fiberglass pole and was georeferenced using a Trimble GeoXT (Trimble, Sunnyvale, CA). There were 30 sampling points in wheat and 30 sampling points in corn. Stink bugs were marked in the single cut of wheat adjacent to the corn field containing the sampling arena. Total length of the cut was 800 m, while the width was 83 m. A John Deere 6,000 Hi-Boy (Deere & Company, Moline, IL) was used to spray a 20% vol:vol egg beaters (ConAgra Foods, Omaha, NE) solution on 3 June 2011 at 188 liters/ha and on 11 June 2012 at 168 liters/ha. In both years, wheat was at GS 92 and the corn was GS V7. In 2011, the grower harvested approximately half of the wheat field (the portion most distant to the corn) on 3 June, 4 h postmarking, resuming, and completing the harvest on 6 June. In 2012, the grower harvested the entire wheat field 3 h postmarking.
Schematic of sampling grid. Fields were separated by a 1-m ditch.
Before the spraying and on the same day, E. servus nymphs and adults were collected from each sampling location in wheat using a sweep net (38 cm in diameter) and from each location in corn by hand collection. These were to serve as a negative control for determining marked individuals (dispersal) as well as for dispersion and abundance information. At each location in the wheat, 25 sweeps of 180° were taken surrounding the sampling flag and at each location in the corn, 40 consecutive plants were searched down the row, centered on the flag. Field-collected data included stink bug species, number, and whether it was an adult or nymph. To avoid cross-contamination of the protein, sweep net bags were changed between every sampling location. Furthermore, in both the wheat and corn, stink bugs were collected from either the sweep net or from the corn plant using a single Chem-Wipe (Chem-Wipe Industries, Ltd., Edmonton, Alberta, Canada) and quickly transferred to a 1.5-ml microfuge tube. The tubes were transported to the laboratory on ice and stored at −10°C for later testing. Stink bugs were sampled from both wheat and corn at each location in a similar fashion 3 h after spraying. Further samples were taken from just the corn on 5, 7, 9, 11, and 13 June 2011 and on 13, 15, 18, and 20 June 2012. Corn growth stages ranged from V7 to V8 during the time observations were taken. In both years, all adult stink bugs were collected to test for the mark. In 2012, all nymphs were collected in corn (n=9), but in 2011, not all nymphs were collected in corn (n=46 out of 363 nymphs counted) to reduce the cost of testing for the mark.
E. servus adults and nymphs collected in the vials were tested for the presence of egg albumin, using a sandwich enzyme-linked immunosorbent assay (ELISA) kit (Egg Protein ELISA kit, Catalog No. 141OA, Morinaga Institute of Biological Science, Inc., Kanazawa, Yokohama, Japan). Rather than grinding the sample, as suggested in the kit instructions, stink bugs were simply suspended in the sample extraction solution that was provided. E. servus collected from wheat before the marking were used as negative controls, while those collected from the wheat 2 h after the marking served as positive controls. Optical density of the samples in the ELISA plates was estimated using a SpectraMax Plus 384 Absorbance Microplate Reader (Molecular Devices, LLC, Sunnyvale, CA). A stink bug was considered marked using the conservative threshold recommendation from Sivakoff et al. (2011). No Oebalus pugnax (F.) stink bugs were tested because so few were collected in corn. The possible importance of E. servus nymphs in stink bug dispersal was not considered until part-way through the experiment in 2011. Hence, E. servus nymphs were not collected for analysis on 5 and 7 June 2011 for testing. Nymphs were not collected for testing on 13 June 2011 because of the high cost to perform the enzyme-linked immunosorbent assay (ELISA) analysis.
Both nymph and adult E. servus densities in corn were compared over time using a separate generalized linear mixed model for each year (PROC GLIMMIX; SAS Institute 2008). The single fixed factor was sample date, while sampling point (location) was modeled as a random factor. A Poisson distribution was found to be the best fit for all the models, but the 2011 data were over-dispersed (Littell et al. 2006). Therefore, a random statement was added in the 2011 models to include an overdispersion parameter (RANDOM_RESIDUAL_). If the model was significant, contrast statements were used to measure trends using orthogonal polynomial coefficients. Because few O. pugnax were recovered in corn, data were not analyzed for this species. Spatial dispersion of E. servus was analyzed using Anselin Local Moran's I (Anselin 1995) in the geographical information systems software ArcGIS 10.0 (Environmental Systems Research Institute [ESRI] 2011). Values were categorized by the program into the following categories: not-significant, high values surrounded primarily by high values, low values surrounded primarily by low values, and outliers, either a high value surrounded primarily by low values or a low value surrounded primarily by high values. Significance was determined based on the z-score, with a P < 0.05 level. Z-scores are not reported in the manuscript because each location on each date has a unique score.
Results
Dispersal to Corn in 2011.
O. pugnax adults and nymphs were detectable in wheat but were very rare in corn (Table 1). E. servus adults and nymphs were present in both wheat and corn. In corn, nymph E. servus densities increased over time to a peak of 0.138 per plant averaged across all sampling dates on 11 June from a low of 0.001 per plant on 3 June (F=26.85; df=5, 145; P < 0.0001; Fig. 2). The increase was quadratic (F=3.93; df=1, 145; P=0.0493). Similarly, adults increased 31.5-fold from the beginning (3 June, 0.008 adults per plant) to the termination of the experiment (13 June, 0.252 adults per plant; F=55.77; df=5, 145; P < 0.0001; Fig. 3). The increase was also quadratic (F=13.52; df=1, 145; P=0.0003). Across all sampling dates in corn, 23 E. servus adults (6% of those tested) and 22 E. servus nymphs (85% of those tested) were classified as marked. Adults classified as marked were detected from 3 h after marking until 6 d after marking but not at 8 and 10 d after marking (Table 1). Nymphs classified as marked were detected from 3 h after marking until 8 d after marking, although they were not tested at 10 d after marking. Moreover, a nymph classified as marked was recovered in the corn 15.5 m from the adjacent wheat 3 h after marking. Finally, a marked adult was found at the furthest extent of the sampling grid, 59 m from the nearest marked wheat at 3 h after marking.
Number of E. servus nymphs in wheat (25 sweeps per location) and in corn (total in 40 plants at each location) at each sampling point over time in (A) 2011 and (B) 2012.
Number of E. servus adults in wheat (25 sweeps per location) and in corn (total in 40 plants at each location) at each sampling point over time in (A) 2011 and (B) 2012.
Means (± SEM) per sweep (wheat) or per plant (corn) averaged across 30 sampling points
Marked represents the percentage of sampled E. servus that tested positive using the immunomarking technique. Although commonly found in wheat, O. pugnax was not tested using the marking technique because it was rare in corn.
Means (± SEM) per sweep (wheat) or per plant (corn) averaged across 30 sampling points
Marked represents the percentage of sampled E. servus that tested positive using the immunomarking technique. Although commonly found in wheat, O. pugnax was not tested using the marking technique because it was rare in corn.
Dispersal to Corn in 2012.
As in 2011, O. pugnax adults and nymphs were detectable in wheat but were very rare in corn (Table 1). In contrast, E. servus adults and nymphs were present in both wheat and corn. In corn, nymph E. servus densities decreased over the course of the experiment (F=3.09; df=4, 116; P=0.0185). This decrease was linear (F=10.46; df=1, 116; P=0.0016). Adult densities did not vary, fluctuating between 0.017 and 0.009 over the course of the experiment (F=0.80; df=4, 116; P=0.5256). Six E. servus adults (8% of those tested) and no E. servus nymphs were classified as marked. Adults classified as marked were detected from 3 h after marking until 9 d after marking (Table 1). Moreover, all nymphs (unmarked) were recovered at the edge of the corn adjacent to wheat. Furthermore, two marked adults were found at the furthest extent of the sampling grid, 59 m from the nearest marked wheat, at 3 h after marking.
Dispersion in Wheat and Corn in 2011.
E. servus adults and nymphs were significantly clustered near the edge of wheat adjacent to corn on 3 June, but only adults were clustered in the corn row adjacent to wheat (Figs. 4 and 5). On 5 June, when the adjacent wheat was only partially harvested, adults were clumped in a single area of the corn field, 15–30 m from the wheat edge, while nymphs were rare in the corn and not clumped. On 7 June, when the adjacent wheat was harvested, adults were clumped at the farthest extent from the sampling grid, 60 m from the wheat stubble. Nymphs were clustered on the row of corn directly adjacent to the wheat stubble on 7, 9, 11, and 13 June as were adults on 11 and 13 June. Similarly, adults were clustered on the row of corn directly adjacent to the wheat stubble on 9, 11, and 13 June.
Anselin Local Moran's I classifications for E. servus nymphs in wheat (25 sweeps per location) and in corn (total in 40 plants at each location) at each sampling point over time in (A) 2011 and (B) 2012. NS, not significant; HL, high value surrounded primarily by low values; LL, low value surrounded primarily by low values; HH, high values surrounded primarily by high values; LH, low values surrounded primarily by high values; *, no nymphs found at any location on 11–18 June in corn.
Anselin Local Moran's I classifications for E. servus adults in wheat (25 sweeps per location) and in corn (total in 40 plants at each location) at each sampling point over time in (A) 2011 and (B) 2012. NS, not significant; HL, high value surrounded primarily by low values; LL, low value surrounded primarily by low values; HH, high values surrounded primarily by high values; LH, low values surrounded primarily by high values.
Dispersion in Wheat and Corn in 2012.
E. servus adults and nymphs were clustered near the edge of wheat adjacent to corn on 11 June but were not clustered in corn (Figs. 4 and 5). Nymphs were clustered on the edge of corn relative to wheat on 15 June and in one isolated location (high Moran's I value surrounded by low values) on 20 June. Adults were clustered on the edge of corn relative to wheat on 15, 18, and 20 June. In one part of the field on 20 June, a cluster of adults extended 15 m into the field from the edge.
Discussion
E. servus nymphs and adults move directly from wheat to corn before the wheat harvest and directly after harvest. Marked individuals were found in corn 3 h after application of the marking material in the adjacent wheat and throughout the observation period, which extended 9 and 10 d postmarking in 2011 and 2012, respectively. This study is the first to demonstrate that E. servus nymphs can disperse among adjacent crops and that their contribution to the overall population density as dispersers may be significant. E. servus emigration from wheat and immigration to corn may be correlated with the timing of wheat harvest because stink adult densities in corn increased >30-fold after the harvest of adjacent wheat during the course of the 2011 sampling. However, other ecological factors are likely involved because E. servus densities remained static in corn during 2012, even though adjacent wheat was harvested.
Dispersal.
This study confirms previous studies in eastern North Carolina (Blinka 2008, Reisig 2011) that wheat is a major source of E. servus in corn. In contrast to previous studies, this study demonstrates direct movement of marked individuals from wheat that were recovered in the adjacent corn. Differences in dispersal and abundance in corn between the 2 yr of the study were striking and are difficult to explain given the number of factors known to influence dispersal, individually and in combination. For example, dispersal of insects can be influenced by density-dependent factors (Sinclair 1989). Overall population density of E. servus in wheat in the 2011 experiment was comparable (≈25% higher) with that in 2012. Therefore, it is unlikely that density-dependent effects affected dispersal. In addition, the nymph to adult ratio in wheat was numerically similar in 2011 (61% adults) compared with 2012 (73% adults). Two noticeable differences between the 2 yr of study were weather and timing of the study. Temperatures in May and June 2011 were above average and there was no precipitation during the course of the experiment (State Climate Office of NC 2011a, b). In contrast, temperatures in May 2012 were above average, but precipitation was above average. Temperatures in June 2012 were below average during the course of the experiment and precipitation was near normal to just below normal (State Climate Office of NC 2012a, b).
Because of these differences in weather, wheat was harvested earlier in 2011 than 2012. Conditions were also conducive to weed growth during 2012, with more and lusher weedy hosts for stink bugs available than those in 2011. As a result, dispersal could have been influenced by the availability of suitable hosts. Pentatomids that switch feeding hosts can increase performance and host switching is common during the preovipositional adult period (Panizzi 1997). For example, performance of Euschistus heros (F.) can be influenced by changing the food source of both nymphs and adults (Pinto and Panizzi 1994). Furthermore, stink bugs within the genus can be found on a variety of plants, including those that are not good hosts. For example, Euschistus paranticus (Grazia) can be found on two wild plants that have been shown to be poor hosts (Smaniotto and Panizzi 2013). It is speculated that these plants are used as temporary hosts until more preferred hosts are available or that these plants are used as iterant shelters, while other plants are used as feeding hosts.
E. servus is known to use a variety of wild and cultivated species as reproductive and feeding hosts, with wheat identified as a particularly important host among others (Jones and Sullivan 1982). O. pugnax has been reported to feed on rice (spp.) and wild grasses (Panizzi 1997) and can be an important pest of wheat (Viator et al. 1983). Corn is not known to be a major reproductive host of this species, although all developmental stages can occasionally be found in this crop (Tillman 2010). In the present experiment, it is likely that wheat was a better developmental host than corn for both O. pugnax and E. servus because many adults and nymphs were present in wheat but not in the adjacent corn in the beginning of the experiment. Wheat harvest may have provided a no-choice situation, where the preferred host was removed from the system, forcing individuals to move to adjacent hosts. Noticeably, nymph O. pugnax densities did not increase in corn, while E. servus nymphs dispersed to the adjacent corn in 2012. Tillman (2010) found E. servus nymphs on corn during mid-May, July, and August (Tillman 2011) in Georgia, presuming that these developed entirely in corn. However, the current study demonstrates that nymphs can disperse into corn, suggesting that the presence of nymphs in a single crop does not imply the development of nymphs in a single crop. Finally, fewer nymphs were found in the 2012 corn arena (9 nymphs over 5 sampling dates) than in 2011 (365 nymphs over 6 sampling dates) and no nymphs recovered in 2012 corn were classified as marked, whereas 85% of the nymphs tested in 2011 corn were classified as marked.
Wheat adjacent to the corn was the only area where the marking material was applied in both years. Therefore, it is possible that nymphs crawled from the weeds in the ditchbank to the corn or that they were already present in the corn that was not marked. Tillman et al. (2009) marked Nezara viridula L. nymphs in peanuts and recovered them in the adjacent cotton. However, in this study, it is also possible that nymphs crawled into the sampling arena from other unmarked areas, given that a nymph had dispersed at least 15.5 m in 3 h in 2011. Although data are not presented, nymphs were visually observed crawling from wheat stubble into the corn in 2011 and many were found feeding at the base of corn plants on the row directly adjacent to the wheat. In contrast to 2011, when the ditchbank was dry with senesced weeds, in 2012, the ditch was full of water and was full of large actively growing weeds. When the weeds were swept with a 38-cm diameter sweep net, an average of six adult and nymph E. servus were recovered in 25 sweeps. Hence, it is likely that nymph stink bugs were prevented from dispersing into the adjacent corn by the water in the ditch and that they had dispersed into the weeds separating the two fields.
Similar to nymphs, adults dispersed into corn from the wheat. However, it is difficult to ascertain where the majority of the adults immigrated from in the sample area. Unlike nymphs in 2012, a much smaller proportion of adults captured in the corn were classified as marked. Adults could have immigrated to the corn sampling arena from an unmarked area, whether that was wheat or an alternative host (such as the weeds in the ditchbank), they could have immigrated from the marked area in the wheat after molting from the juvenile stages, presumably leaving most of the marking material with the shed cuticle, or they could have immigrated as nymphs into the corn, eclosing into the adult stage before they were sampled. Stink bug adults are more common in corn after the VT stage (Herbert and Toews 2011, Tillman 2010). However, in this study, both adults and nymphs immigrated to corn during the V7 stage, well before the VT stage. Hence, in this system, the presence of stink bugs in corn can be influenced by the presence of adjacent wheat.
Mark-Recapture.
The mark-recapture technique used in this study is a relatively recent method to document movement. Another method that has been used to investigate stink bug movement is to capture the insects, mark them with paint (Jones and Sullivan 1982), release them, and recapture them. The efficiency of this technique can be low, with ≈12% of marked E. servus recaptured in one study (Jones and Sullivan 1982). In addition, it was not an effective method to document movement of Chinavia (Acrosternum) hilare Say or Euschistus tristigmus Say in this study. However, it has been successfully used to document movement of N. viridula between sorghum and cotton (Tillman 2006) and peanuts and cotton (Tillman et al. 2009). In the 2009 study, the paint marking method was used along with a permanent marker. It is unclear how efficient recovery of marked individuals was in these studies, but the marker was very durable, lasting at least 54 d (Tillman 2006). In contrast to the paint or marker method, the marking method used in this study can mark many individuals at one time, because insects do not have to be captured, marked, and release. This presumably increases the recapture rates of marked individuals because more can be marked at one time. In addition, the efficiency of acquiring the marker is high. In this study, 75–100% of adults and 100% of the nymphs recovered in the adjacent unmarked corn were marked. In another study, 78.9% of Cacopsylla pyricola Foerster acquired the marking protein simply from walking across apple leaves treated with egg whites (Jones et al. 2006). However, the marker is likely less durable. Although egg markers are detectable at least 19 d after marking, there are indications that this type of marker is water soluble and can be washed off during rain (Jones et al. 2006). As noted above, precipitation was low in 2011 but average in 2012, with precipitation present during the 2012 experiment but not in 2011.
Dispersion.
Both E. servus adults and nymphs were aggregated in wheat directly before harvest in 2011 and 2012, but the degree of aggregation was stronger in 2011. These aggregations tended to be along the field border, agreeing with previous studies in wheat (Reay-Jones 2010) as well as with other stink bug species in other crops (Panizzi et al. 1980) but were also along the border with other available hosts, which included weeds in the ditchbank and corn. E. servus nymphs were not aggregated in corn until after the adjacent wheat was harvested during both years. There was a single aggregated area of adults in the corn sampling arena during 2011 before the wheat harvest. After the wheat harvest, all adult and nymph aggregations were located almost entirely along the border with wheat, which mirrors findings from previous studies that have found E. servus densities and associated injury near corn field edges (Ni et al. 2011, Tillman 2010, 2011). The current study offers one possible explanation for this phenomenon, which is the previous presence of an available developmental and feeding host (wheat) that is removed from the system and forces emigration.
These results have implications on management efforts for E. servus in corn. After the 2011 results, it was hoped that an insecticide spray targeted against nymphs in wheat stubble might be effective in presenting nymph immigration to the adjacent corn. However, based on 2012 results, it is clear that nymphs do not always move directly from wheat stubble into the adjacent corn. Border sprays in corn might be effective in reducing E. servus abundances and could be timed with harvest of the adjacent wheat. Several factors may reduce the efficacy of this tactic and the residual efficacy for such a spray will likely be short. For example, efficacy could be reduced by the prolonged movement of nymphs from the adjacent wheat stubble. In addition, many of the adults in this study were not classified as marked. It is possible that these individuals could have moved from other wheat fields that were harvested later that the adjacent focal field or from other alternative hosts. If these sources prove important, border sprays would need to be repeated to be effective.
It is clear that wheat is an important source of E. servus in corn and that wheat harvest can be an important trigger of dispersal of these stink bugs into corn. However, wheat harvest is likely not the only cause of dispersal in this system as results between the 2 yr were so different. One strategy to prevent E. servus injury in corn would be to co-locate wheat of similar maturity near next season's corn fields. Wheat could then be harvested at a single time, forcing stink bugs to emigrate at a single time. One well-timed spray in corn might have efficacy, if properly timed and with proper canopy penetration. More research on the contribution of hosts other than wheat to E. servus abundance in corn might also benefit the development of management tactics and strategies.
This work was supported by the North Carolina Corn Growers Association. Special thanks to Frances Sivakoff for her advice on mark-recapture, as well as a threshold value for determining marked individuals. We would also like to acknowledge Michael Toews for the PROC IML language for determining the appropriate orthogonal contrasts. Finally, we thank Francis Reay-Jones for editing the first draft of this manuscript.
References
