This study aimed to disentangle the relative roles of climatic, landscape and local factors affecting fruit infestation rates of winter and spring host plants by the invasive fruit fly Drosophila suzukii. We assessed infestation in Aucuba japonica, Elaeagnus ×submacrophylla (syn. Elaeagnus ×ebbingei), Mahonia aquifolium, M. japonica and Viscum album fruit in the north of France, between January and July 2022.

Fruit infestation rates of four fleshy-fruited plant species by the invasive fly Drosophila suzukii were measured near Amiens city in northern France between January and July 2022. Each sample was accompanied by measurements of climatic and local and landscape variables in order to identify the environmental drivers of host fruit infestations.

Materials and Methods

Study area

The study was conducted between January and July 2022 in the region of Amiens (49°53′40″ N, 2°18′07″ E) in northern France. The region’s climate is oceanic with a mean annual temperature of 10.7°C and an average annual rainfall of 691.9 mm (data from meteorological station Dury-les-Amiens, StatIC network). The landscape of this region is generally characterised by agricultural production and consists of a mosaic of open fields cultivated for cereals, rapeseed and sugar beet, interspersed with orchards, grasslands, woodland patches and rivers.

Sampling design

Aucuba japonica var. variegata, Elaeagnus ×submacrophylla, Mahonia aquifolium, Mahonia japonica agg. (including a complex of related cultivars and hybrids) and V. album fruit were sampled every month, with individual plants of each species randomly sampled within a landscape area of 35 × 45 km. The number of individuals sampled per species and per month varied depending on fruit availability. When possible, a minimum of 100 fruit were randomly sampled from each plant individual, with some variation depending on seasonal availability of fruit.

Traits of host plant species and fruit

Collected fruit were separated into three subsets to monitor Drosophila emergence: undamaged (‘healthy’) and damaged fruit collected on the plant (‘damaged’) and on the ground (‘ground’; if present). The fruit were categorized as ‘damaged’ when they were opened / injured (incised skin) and/or rotten (brown spots). Several traits of the sampled plants were measured to characterise the local resources available for the flies and the local microhabitat. For each sampled plant, five berries were randomly selected to measure length and width and calculate volume (4/3 × π × mean radius³) and fruit skin area (4 × π × mean radius²). Five leaves were also taken, their length and width measured and the leaf surface index (length × width; Ulmer et al. 2022) calculated. Individual morphology of each plant species was characterised by measuring the minimum and maximum plant canopy diameter, the circumference of the largest trunk (except for V. album) and the total number of fruit present on the plant. Mistletoe being a parasitic shrub, we also recorded the host tree species and measured the height of the mistletoe individual on the tree (from the ground), the tree height, the crown diameter, the trunk circumference, and the number of mistletoe individuals present on the host tree and in a 20 m radius around it.

Environmental variables

Local, landscape and climatic variables were measured at each sampling site or extracted from online databases to examine the influence of regional and local environmental conditions on infestation rates.

Local environmental conditions were described as follows. First, within a 5 m radius plot centred on the sampled plant, the cover and height of the tree, shrub and herbaceous layers were estimated, as well as soil litter thickness . Second, within a 20 m radius, the percentage of local habitats surrounding the host was recorded (e.g., orchard, woodland, grassland, swamp, crop, garden, shrub, building, hedgerow, river, pond, poplar plantation, park, road), as well as the percentage of other plant species with maturing fruit.

The landscape composition around each sampled plant was then characterised. A geographic database was created using a Geographic Information System (GIS; ArcGIS Pro v.2.5, ESRI). The sampled plants were positioned in the GIS and buffers of 50, 100, 250, 500, 750, 1000, 1250, 1500, 1750, 2000, 2500 and 3000 m radii around each host tree were created for subsequent analyses of landscape composition. Landscape elements (crop, water, woodland, shrubland, grassland, road, urban area, orchard, industrial zone) were extracted from the OSO 2022 database (THEIA 2023) and updated using aerial photographs and, in buffers <100 m, field observations.

Macroclimatic conditions were characterised for each sampling site using regional measurements. Daily meteorological data were retrieved from the three meteorological stations closest to each site, from 1 January 2022 to each sampling date (https://www.historique-meteo.net/france/). Daily minimum, mean and maximum temperatures, rainfall and snowfall were calculated for all sites using inverse-distance weighting (IDW) interpolation (Willmott et al. 1985) from the data from the three nearest weather stations. Accumulated degree-days (“Growing Degree Days”, GDD) were calculated using a lower threshold of 0°C between 1 January 2022 and the sampling date (Baskerville and Emin 1969). The baseline value of 0°C is a standard threshold commonly used to calculate GDD in insect and plant studies (White et al. 2012; McNeil et al. 2020). It is particularly suitable to study the temporal synchrony between insects and plant resources (Iler et al. 2013; Ulmer et al. 2022). It was also chosen because (i) active D. suzukii can be observed even at very low positive temperature (< 5°C) during winter, including during periods of snowfall (Ulmer et al. 2024), (ii) flies are able to recover from chill coma after exposure to –1°C (Wallingford et al. 2016) and (iii) mistletoe fruit can undergo freeze-thaw cycles before ripening ends (Thomas et al. 2023). From daily precipitation values, we also calculated mean daily and cumulative precipitation between 1 January 2022 and each sampling date, and within the 7- or 14-days periods preceding each sampling date.

Microclimate temperatures were recorded at each sampling site using Hobo loggers (TIDBIT data logger V2 TEMP TBI-001, ONSET Company, Bourne MA, USA), recording every 60 min. In each sampling site a logger was suspended 1.5 m above the ground in the plant canopy, under the shade of a branch to avoid direct exposure to solar radiation and oriented northward. The minimum, mean and maximum air temperatures were extracted every day to compute the mean daily minimum, mean and maximum temperatures. As described for macroclimate data, GDD were calculated using Hobo logger data to characterise the local microclimate under the canopy of the sampled plants.

Emergence of Drosophila species

After collection, the fruit sampled from each plant were placed on wet cotton wool in cylindrical plastic transparent containers (diameter = 118 mm, height = 135 mm, volume = 1,476 cm³), covered with a nylon mesh, and maintained in a temperature-controlled room at 20°C under a 16:8 L:D regime. Adult flies emerging from the fruit were placed in 70% ethanol. They were identified to species level using Bächli et al. (2004) and specific criteria published for D. suzukii (Withers and Allemand 2012). Individuals of each species were sexed and counted using a Leica M205C stereomicroscope equipped with a Leica MC170 HD camera and the Leica Application Suite software.

Infestation variables

We examined the relationships between environmental variables and two common infestation variables that were either centred on the fruit (Fruit Infestation Rate for a given Drosophila species: FIR = 100 × number of emerged individuals from fruit collected from a given plant individual / total number of fruit collected from the same plant individual) or on the plant species (Plant Infestation Rate: PIR = 100 × number of infested plant individuals of a species in a month / total number of plant individuals of this same species sampled in a month). These variables can be interpreted as follows: FIR reflects the plant auto-contamination by the flies while PIR reflects fly dispersal between host plant individuals (e.g., when the FIR and PIR are both high, both auto-contamination within the plant individual and dispersal of the flies between plant individuals take place; when the FIR is high and the PIR is low, there is mostly plant auto-contamination; when the FIR is low and the PIR high, there is mostly fly dispersal; when both FIR and PIR are low, there is an absence of both auto-contamination and dispersal). These infestation variables were calculated for each fruit category (healthy or damaged fruit on the plant and fallen fruit on the ground) following Deconninck et al. (2024).