Date Published: March 31, 2017
Publisher: Public Library of Science
Author(s): Noel G. Hahn, Cesar Rodriguez-Saona, George C. Hamilton, Patrizia Falabella.
Geospatial analyses were used to investigate the spatial distribution of populations of Halyomorpha halys, an important invasive agricultural pest in mid-Atlantic peach orchards. This spatial analysis will improve efficiency by allowing growers and farm managers to predict insect arrangement and target management strategies. Data on the presence of H. halys were collected from five peach orchards at four farms in New Jersey from 2012–2014 located in different land-use contexts. A point pattern analysis, using Ripley’s K function, was used to describe clustering of H. halys. In addition, the clustering of damage indicative of H. halys feeding was described. With low populations early in the growing season, H. halys did not exhibit signs of clustering in the orchards at most distances. At sites with low populations throughout the season, clustering was not apparent. However, later in the season, high infestation levels led to more evident clustering of H. halys. Damage, although present throughout the entire orchard, was found at low levels. When looking at trees with greater than 10% fruit damage, damage was shown to cluster in orchards. The Moran’s I statistic showed that spatial autocorrelation of H. halys was present within the orchards on the August sample dates, in relation to both populations density and levels of damage. Kriging the abundance of H. halys and the severity of damage to peaches revealed that the estimations of these are generally found in the same region of the orchards. This information on the clustering of H. halys populations will be useful to help predict presence of insects for use in management or scouting programs.
The spatial arrangement of insect populations in an area could influence actions taken by pest managers, and therefore must be understood. Since insects are mobile organisms governed by the need to feed and reproduce, they often disperse themselves in space in predictable ways. In stores containing food products, Indian meal moth (Plodia interpunctella Hübner) has been found concentrated around areas with concentrations of birdseed and pet products . Broadly speaking, insect populations can exhibit either a homogeneous (uniform), random, or clustered arrangement in space . These patterns are affected by the spatial scale of investigation. A uniform pattern is one that is evenly distributed across a landscape. A random pattern of distribution means that individuals are equally likely to occur at any location in the area. A clustered distribution is one in which many individuals are concentrated closely together with large areas containing few or no individuals. This spatially heterogeneous distribution could result from environmental constraints, such as a landscape feature like availability of resources, climate, or soil type. For example, the area of gypsy moth and spruce budworm defoliation has been found to increase or decrease depending on fluctuations in temperature and precipitation . The input of chemicals and physical manipulation on farms results in changes to the landscape. Understanding these changes is important to our understanding of how agricultural pests are spatially distributed.
The landscape within 1 km around each orchard varied among the sites. Agriculture was predominant around CR1, CR2, and R1, while forest surrounded the majority of Farm N (S1 Fig). The sites CR and R1 are within highly productive agricultural regions in New Jersey, whereas Farm N is closer to more forested, mountainous areas. The agriculture surrounding these orchards consisted of a mixture of tree fruit orchards, ornamental farms, soybeans, and Christmas tree farms. Plantings at Cream Ridge were comprised of predominantly apricots, apples, and peaches with small amounts of raspberry, strawberry and squash. Farm S was surrounded by a relatively equal mix of agriculture, forest, and urban areas. Within 5 km of each orchard, the patterns of land use remained similar to that within 1 km. However, around CR1 and CR2, agricultural land use dropped from 72% to 46% of the total land use (S1 Fig).
This is the first study to document the distribution of H. halys in peach orchards. Our results provide strong evidence for clustering that varied over time. There was greater clustering in August at high population levels, and low clustering in July at low population levels. Clustering of instances of damage occurred when levels of damage were high later in the season. Investigating the location and levels of H. halys populations in orchards enables us to learn about the distribution of the pest in relation to time and space. Since populations of H. halys varied throughout the sampling season at different sites, we believed that they would exhibit different levels of clustering and spatial autocorrelation throughout this period. Feeding in aggregations has been linked to the tendency of insect populations to outbreak . Our results show that clustering of H. halys populations occurred mostly on the later sampling dates, when population levels were higher in the orchards. Since this is most prevalent during the August sample dates when fruit is ripe or ripening, it suggests that populations are clustering at times and in areas with suitable resources. In addition, later sampling dates could have had H. halys immigrants from nearby areas attracted by aggregating individuals already in the orchard. Our kriging analysis of R1 estimated large populations of H. halys at the northeastern part of the orchard in addition to large levels of damage throughout the area, especially towards the southern section. There is a forested edge past the northeastern part of the orchard, leadings us to believe that many H. halys migrate from there into the northeastern row. Although there was no indication of high populations of H. halys in the southern part of the orchard on the day of sampling, the high levels of damage there could be a result of migration to and from an adjacent field of soybeans south of the orchard. A helpful follow-up study investigating the clustering on field edges in relation to the phenology of surrounding crops within the farmscape. A recent study has found that injury to apple due to H. halys is not uniformly distributed in commercial apple orchards, and injury was generally higher along borders at forested edges .