Abstract
Philaenus spumarius (Ps) is considered the main insect vector of the bacterium Xylella fastidiosa (Xf) in Europe. As such, it is a key actor of the Xf pathosystem on which surveillance and management strategies could be implemented. Although research effort has increased in the past years, the ecological factors shaping Ps abundance and distribution across landscapes are still poorly known in most regions of Europe. We selected 64 plots of 500m2 in Corsican semi-natural habitats in which we sampled nymphs and adults of Ps during three years. While local or surrounding vegetation structure (low or high scrubland) had little effect on Ps abundance, we highlighted a positive relationship between Ps abundance and the density of Cistus monspeliensis in the plots. We also found larger populations of Ps in cooler and moister plots. The pattern of host association highlighted here is unique, which calls for more studies on the ecology of Ps in Europe, to help designing surveillance and management strategy for Xf.
Similar content being viewed by others
Introduction
The meadow spittlebug Philaenus spumarius (Linnaeus, 1758) (Hemiptera: Aphrophoridae) (Ps) is considered the main vector of the plant pathogen Xylella fastidiosa (Xf) in Europe1,2. Research effort on its biology and ecology has recently increased, generally with the aim of proposing management strategies. However, effort is unequal among regions and habitats and it is still unknown whether environmental drivers of abundance and distribution of Ps (e.g., climate and host plants) are similar between different ecosystems and/or geographical regions.
Humidity and temperature are important drivers of the development and the geographic distribution of Ps2,3. Overall, studies have either addressed physiological or phenological responses of Ps to temperature and humidity at local spatial scales, or climate preferences at larger scales4, notably using species distribution models5,6,7. Regional studies investigating the effects of climatic conditions on populations of the meadow spittlebug are still scarce, with the exception of works by Drosopoulos and Asche8 in Greece, Godefroid and Durán9 in southern Spain, and Karban and Strauss10 in California. Authors agree that humidity and temperature affect the development time of immature stages of Ps3,11,12. While exact values vary, cool and moist conditions appear more favorable, with minimum temperatures for egg hatching and larval development ranging from 4 to 10 °C3,11,13. No data are available regarding maximum temperatures but nymphal development is still observed at 27 °C3,11. Humidity promotes hatching and survival of nymphs that need sufficient sap flow to feed. Nymphs produce a foam that covers them and to which they owe the name "spittlebugs"3. This foam likely protects them from predators and dehydration14. In this regard, extreme or prolonged summer droughts drastically reduce Ps populations10,15. Moreover, shift in host plants from dry vegetation to less water-stressed shrubs and trees or migration toward coolest places (e.g., near rivers) have been documented, showing that humidity and temperature play a major role in the ecology of this insect3,15,16,17,18,19.
Trophic interactions between insect vectors and their host plants can shape the epidemiology of plant diseases caused by Xf. In this regard, with more than 1000 host plant species in 75 families3,20,21, the highly polyphagous2,3,22 Ps is a key actor of the Xf-pathosystem. Several studies have explored local host associations of Ps in olive groves in Italy23,24, Portugal25, Spain19 and Greece26, where nymphs mostly feed on Asteraceae, Fabaceae and Apiaceae. However, no study has explored the link between the local abundance of host plants and the degree of polyphagy of Ps. Summer migration of adults from ground cover to olive foliage, linked with drying of herbaceous plants has been observed in southern Italy2,21. In particular, adults of Ps were observed in the vicinity of olive groves, from May to August, on Quercus ilex, Quercus x crenata, Pistacia lentiscus, P. terebinthus, Hedera helix and Myrtus communis17,21,24. In the Apulia region, Ps adults have also been found on Pinus halepensis from May to September and have been reported once on Ulmus minor in October17. On the other hand, in Belgium, Hasbroucq et al.27 reported adults of Ps on Ranunculaceae, Prunus sp., Rubus sp., Crataegus sp., Alnus sp., Picea sp., Quercus sp., Salix sp., Carex sp. and Urtica sp.. Recent studies performed in Corsica, France5,28,29 suggest that Ps has affinity with Cistus monspeliensis, at least locally. Provided this holds on a larger scale, Ps populations are expected to grow with C. monspeliensis cover30. However, it should be evaluated whether insect load (i.e., insect population size per plant unit) will increase, decrease, or be uncorrelated with host plant cover30,31. The degree of preference for a plant species over its neighbors seems to play a key role in this relationship and is often negative (i.e., insect load decreases with plant density) when the focal plant species is preferred over its neighbors31. This phenomenon is known as "resource dilution" and given that it is usually observed in specialist species32, it potentially does not contribute to determining the preference in a generalist insect such as Ps.
Another important aspect to consider within a vector-borne pathosystem is the dispersal ability of vectors. In flight mills, the average distance flown by Ps adults vary from one to five hundred meters33,34. In olive groves and meadow landscapes, the median distance covered during Ps adult life was assessed at 374 m and 507 m, respectively35. Although limited, this dispersal ability creates connection between habitats (e.g., grassland, meadow, forest, crops etc.) that are within a radius of some hundreds of meters. Consequently, Ps population size at a given site is likely to be influenced by the suitability of the surrounding habitat. Accordingly, population size of Ps has been shown to correlate positively with the surface of olive groves36,37 in a 125 m-radius buffer surrounding studied plots, while it negatively correlated with the surface of vineyards36. Ps density was also found to be lower in sites with high forest cover38. Considering landscape composition around study sites to investigate drivers of Ps population size seems therefore crucial.
In the present study, we aimed at comparing Ps abundance in plots distributed along a gradient of climate conditions and characterized by contrasted vegetation structure, and various landscape contexts in mesomediterranean areas of Corsica. We addressed five research questions stemming from our review of the literature: (1) Are populations of Ps larger in areas densely covered with C. monspeliensis?; (2) Does Ps load on C. monspeliensis plant increase, decrease or remain steady with C. monspeliensis cover?; (3) Are populations of Ps larger in cooler and moister plots, especially during the hot and dry seasons?; (4) Does vegetation structure (tree cover versus open areas) impact the abundance of Ps, especially during the hot season? and (5) Does vegetation structure of the surrounding environment impact the abundance of Ps?.
Material and methods
Sampling plots
Sixty-four plots of 500 m2 were distributed in different regions of Corsica (Fig. 1; Supplementary map) along two gradients: one gradient of C. monspeliensis cover (from 0 to 95%) and one elevation gradient (from 6 to 595 m above sea level). These two uncorrelated gradients (Spearman’s rank correlation test p-value: 0.7586) allowed us to independently assess the effects of vegetation and climate on Ps abundance.
Sampling period and frequency
First, a phenological survey of populations of Ps was performed bi-weekly on two plots (CAN-01 and CAN-02, Table S1) to identify periods during which densities of nymphs and adults were the highest. Based on this phenological survey, field work was conducted on all sampling plots in April (highest density of nymphs); June (highest density of young adults) and October (breeding season). Conducting a field survey for a fortnight at peak abundance indeed contributed to minimize differences in abundance between plots. These three sampling sessions were repeated during three consecutive years (2018, 2019 and 2020).
Within-plot vegetation structure
All studied plots were located in the scrubland vegetation, a common vegetation type at the mesomediterranean level39,40 (see botanic inventory in Table S1 for details). The percent cover of live C. monspeliensis was visually estimated on each plot at each sampling session. The median values over all sessions were used in subsequent analyses for all but two plots (VENT-03 and VENT-04) on which vegetation was cut by municipal services to prevent bushfires. For these two plots, C. monspeliensis covers measured at each session were used instead.
Vegetation structure within plot was retrieved from the OCS GE database (© IGN–2022, https://geoservices.ign.fr/ocsge) provided by the French National Institute of Geographic and Forestry Information. This database includes land cover and land use, photo-interpreted from 50 cm-pixels orthophotographs, in the form of a polygon database whose contours are positioned with a precision of ca. 1 m. Vegetation structure within plot was coded as follows: "high scrubland" (when tree cover exceeded 25%); number of plots n = 33, and "low scrubland" (shrub and herbaceous formations with tree cover below 25%; n = 31; Fig. 1; Table S1; Supplementary map).
Vegetation structure around plots
Vegetation structure around plots was also characterized through the OCS GE database. Based on Ps dispersal ability reported by Bodino et al.35 in natural conditions, we calculated the proportion of each type of vegetation structure within a radius buffer zone of 374 m, which was drawn around the plot but excluding the plot itself (see Supplementary map). It is noteworthy that results were unchanged with a radius buffer zone of 500 m (data not shown). Four vegetation structure were identified: "high scrubland", "low scrubland" (see previous section), "unvegetated" (artificialized areas, bare ground, water and sea surfaces), and "orchards" (tree or woody vines cover exceeding 25% with agricultural use, i.e., mainly olive groves, Citrus groves and vineyards).
Vegetation structure around plots was then analyzed by a Principal Component Analysis (PCA). The proportion of each of the four vegetation structures within each buffer was used as input for the PCA (Table S1). The PCA was performed using the R package ade441. Scores of each plot on the first (PC1landscape) axis and second (PC2landscape) axis of the PCA were used as input variables in all generalized linear mixed models (GLMMs).
Abundance of P. spumarius
In a preliminary experiment, we counted the number of nymphs in 20 randomly selected foams on each plot. An average of 1.24 nymphs per foam was found (± 0.57 standard deviation, Fig. S1) and this number did not differ significantly among plots (minimum p-value of Dunn test for multiple comparisons of means = 0.09, Fig. S1). Thus, we considered that the number of foams could be used as a good proxy for the number of nymphs and nymph's abundance was defined as the number of foams counted in 10 min in the vegetation of the plot (herbaceous, shrub, tree foliage of lowest branches, undergrowth) by the same operator.
Adult abundance was defined as the number of adults collected by two operators sweep netting the vegetation for 10 min, a time window sufficient to sample almost all the vegetation present in the plot (i.e., herbaceous, shrub, lowest branches of trees, and undergrowth vegetation). Sweep netting consisted in periods of 30 s of alternate backhand and forehand strokes with the net followed by 10–30 s of collection of adults with mouth aspirators. Sweeping the vegetation over a longer period would have increased risks of missing the target species in the net among plant debris and non-target arthropods. At any rate, after 10-min, adults collected in the mouth aspirators were counted and released afterward to preserve local insect populations.
To ascertain that foams/adults were correctly identified on sight, nymphs used to estimate the average number of nymphs per foam in the preliminary experiment and about 30 adults collected in the vicinity of each plot every June and October were killed using ethyl acetate, stored in 75° ethanol, and brought to the laboratory. We collected a total of 11,018 adults and 1911 nymphs. These specimens were identified to species level in the lab using a binocular microscope, the identification key of Biedermann and Niedringhaus42; illustrations from Stöckmann et al.43 and the Arthemis database (https://arthemisdb.supagro.inrae.fr44). All nymphs (1911) and 11,003 adults (99.9%) were confirmed as Ps. The remaining 0.1% of adults (N = 15), were re-identified as Neophilaenus campestris. Given this negligible error rate, counts based on identification on sight in the field were considered as reliable.
Ecological drivers of P. spumarius abundance using bioclimatic data (GLMM1 models)
A first set of GLMMs (hereafter referred to as GLMM1 models) was used to assess the effects of percent cover of C. monspeliensis, climate, plot vegetation structure and surrounding vegetation structure on Ps abundance. Three GLMMs were built, one for each sampling month: GLMM1-April (nymphs), GLMM1-June (young adults) and GLMM1-October (breeding adults).
In this set of GLMMs, climate was described by a PCA performed on 19 bioclimatic variables (Table S1)45 retrieved from the SAFRAN model from Météo France46, which interpolates data measured several times a day by a network of over 1000 meteorological stations spread over the French territory. SAFRAN provides daily data of temperature (2 m above ground) and precipitation interpolated at a resolution of 8 km. Bioclimatic variables were derived from the monthly minimum, maximum, mean temperature and mean precipitation values, and were computed over the whole sampling period (2018–2020). They represented annual trends, seasonality, and extreme values of annual climate conditions. As no strong message emerged from the literature about the bioclimatic factors driving Ps abundance, we used all 19 bioclimatic variables. Scores of each sampling plot on the first axis (PC1bioclim) and second axis (PC2bioclim) of the PCA (Table S1) were used as input variables for all three GLMM1 models. The 19 bioclimatic variables were computed for all plots using the function biovars of the R package dismo47 to the three-year mean values of monthly temperatures (minimum and maximum) and precipitations (sum).
Structure of GLMM1 models, following glmmTMB notation was:
where Cm, (1|id) and (1|observation) respectively stand for C. monspeliensis, random effect on the identifier of the plot and observation-level random effect, and \(m\in \left\{April, June, October\right\}\).
P. spumarius load per host plant unit C. monspeliensis cover (GLMM2 models)
A second set of models (hereafter referred to as GLMM2 models; one model per sampling month) was built with the same structure as GLMM1 models but the response analyzed was the Ps load per unit cover of C. monspeliensis. To do that we used rate models48, adding an offset—\(offset(\mathrm{log}\left(\%cover \,\,of \,\,Cm\right))\)—in the glmmTMB formula reported above. It should be noted that the response analyzed was only a proxy of Ps density on C. monspeliensis: we did not record the plant host of every Ps sampled. This proxy is thus slightly overestimating of the actual density, as some (but not much) individuals were collected on other host plants. Only significant variables of each GLMM1 models were included in corresponding GLMM2 models and only the 61 plots where C. monspeliensis was present were analyzed.
Ecological drivers of P. spumarius abundance using seasonal climate data (GLMM1bis models)
A last set of GLMMs (hereafter referred to as GLMM1bis models) was used to check the robustness of the trends observed with GLMM1 models. They only differed from GLMM1 models in the way the effect of climate was tested. In GLMM1bis models (one for each sampling month), climate was described by a set of seasonal variables: average day temperatures (minimum, mean and maximum), sum and maximum day precipitations over the 1 or 2-month period preceding field surveys, all averaged over the 3 years of the study (10 climate descriptors, Table S1). A PCA was performed for each sampling month: April, June and October and scores of each plot on the first (PC1seasonal) and second (PC2seasonal) axis of the PCA were used for each corresponding GLMMs (GLMM1bis-April, GLMM1bis-June, GLMM1bis-October). This analysis is complementary to that of GLMM1 models because it relies on raw climate data (instead of elaborated bioclimatic variables) retrieved on a relevant period with respect to Ps phenology. The structure of GLMM1bis models was the same as for GLMM1 models. No significant differences between the results of GLMM1 and GLMM1bis models were found. Therefore, we did not use seasonal climate data to analyze Ps load per unit C. monspeliensis cover.
Modelling framework
All GLMMs were built with the R package glmmTMB49,50 using linearly parameterized negative binomial distributions ("nbinom1"51). To account for repeated measures on the same plot, we added a random effect on the identifier of the plot48. An observation-level random effect was also added to correct overdispersion52. Year was included as an experimental design fixed effect because the number of factor levels was below 553. As preliminary data analyses suggested a nonlinear relationship between Ps abundance and percent cover of C. monspeliensis, quadratic and cubic components of this variable were added to GLMMs. We used the DHARMa54 and performance55 packages to detect possible significant deviations from model assumptions (normality of residuals, homoscedasticity, the absence of collinearity between factors and the absence of highly influential data points by calculating Cook’s distance). To explore the results of the fitted GLMMs, type II analyses of deviance (R package car56) and post-hoc pairwise comparisons of factor levels were performed (R package emmeans57 and multcomp58).
Results
For all correlations mentioned in the text, p-values are below 0.005. Details on χ2, df and exact p-values are available in Table S2.
Spatiotemporal variability of the abundance of P. spumarius
A total of 19,808 foams (used as a proxy for nymphs, see methods) and 18,645 adults of Ps were counted during our three-year survey. Spatial and temporal variability of Ps abundance is illustrated in Fig. 2 (raw data are provided in Table S1). No data was collected in April 2020 because of restrictions due to the Covid-19 pandemic.
Structure of vegetation surrounding sampling plots
PC1 and PC2 of the PCA performed on surrounding vegetation structure accounted for 50.8% and 30.7% of the variability, respectively (Fig. 3). PC1landscape represented a gradient of vegetation density and opposed plots located in high scrubland landscapes (tree cover above 25%; correlation to PC1landscape: 0.997) to plots located in low scrubland landscapes (correlation to PC1landscape: −0.945). The area covered by high scrubland in the landscape was therefore strongly negatively correlated with the area covered by low scrubland. PC2landscape opposed plots located in landscapes with large unvegetated areas (correlation to PC2landscape: 0.794) to plots located in the vicinity of orchards (correlation to PC2landscape: −0.756). Plots were first separated according to the scrubland structure (high or low) in their vicinity. Then, rarer elements of the landscape, unvegetated soils and orchards, were captured on PC2landscape which distinguished plots with more rocky soil and distributed on the coast (e.g., LUM-07, CALV-04, BONI-02) from plots located in the Balagne region where the landscape is historically rich in olive orchards (e.g., LAG-01, URT-13).
Ecological drivers of P. spumarius abundance using bioclimatic data (GLMM1 models)
The transformation of the 19 bioclimatic variables through the PCA yielded two synthetic variables, PC1 and PC2, which supported 61.9% and 17.8% of the climate variability, respectively (Fig. 4). PC1bioclim opposed plots with high temperatures and low precipitations to plots with high precipitations and low temperatures. The variables exhibiting the largest scores upon PC1bioclim were the temperature of the warmest month and of the wettest quarter (right side of the axis, Fig. 4) and the temperature seasonality and the precipitation of the driest month (left side of the axis, Fig. 4). PC2bioclim opposed plots with high seasonal climate variations to plots with low seasonal climate variations. The bioclim variables that displayed the largest scores on PC2bioclim were the diurnal and annual temperature ranges (the "high seasonal variation" end of the gradient), and many equally projected precipitation-related variables in the positive direction (the "low seasonal variation" end of the gradient). Examples of raw climate values of the most contrasted plots according to the PCA are given in Table S3.
Models showed that Ps abundance was positively correlated with the percent cover of C. monspeliensis for each sampling month (Table S2). This correlation was characterized by a strong rise in Ps abundance for C. monspeliensis cover ranging from 0 to 25%, then abundance levelled off and finally rose again for C. monspeliensis cover exceeding 80% (Fig. 5A). Climate had a significant effect only on the abundance of adults (June and October). The abundance of adults decreased with sample plot scores on PC1bioclim (Fig. 5B). Therefore, Ps abundance increased with decreasing temperatures and increasing precipitations. Abundance of adults also increased with the scores of sample plots on PC2bioclim, that is with decreasing seasonal contrast in climatic conditions (Fig. 5C). Local (i.e., in plot) and surrounding vegetation structure correlated with Ps abundance only in June. Plots with high scrubland vegetation hosted higher abundances of young adults than plots with low scrubland vegetation (Fig. 5D), but vegetation structure of the plot had no effect on the abundance of nymphs or adults during the breeding season (Table S2). Abundance of young adults significantly increased with sample plot scores on PC1landscape (Fig. 5E), which means that abundances of Ps was higher when tree cover in surrounding vegetation was higher. Contrastingly, structure of landscape vegetation had no effect on the abundance of nymphs or adults during the breeding season (Table S2). In all years, adult abundance was higher in October than in June (Figs. 4, S2).
P. spumarius load per unit C. monspeliensis cover (GLMM2 models)
The effects of climate, local and landscape vegetation structures observed in GLMM1 models were identical to those observed with GLMM2 models (Table S2). In addition, for all sampling months, Ps load per unit C. monspeliensis cover was negatively correlated with C. monspeliensis percent cover (Fig. 5F).
Ecological drivers of P. spumarius abundance using seasonal climate data (GLMM1bis models)
The PCA performed on the datasets associated to each date yielded two principal axes (PC1 and PC2) that accounted for 72.8% and 14.9% (April), 71.2% and 15.4% (June), and 62.3% and 31.8% (October) of the climate variability, respectively (Fig. S3). For the three PCA, PC1 opposed plots with high temperatures to plots with high precipitations and PC2 opposed plots with high temperatures and precipitations to plots with low temperatures and precipitations.
As in GLMM1 models, there was a positive correlation between Ps abundance and C. monspeliensis cover, with similar sigmoid pattern (Table S2; Fig. S4A). The abundance of adults increased with plot scores on PC1seasonal in June and October (Fig. S4B), that is with increasing precipitations and decreasing temperatures. Only in June, abundance of Ps decreased with plot scores on PC2seasonal (Fig. S4C), that is with decreasing precipitations and temperatures. Finally, neither local nor landscape vegetation structure had a significant effect on Ps abundance in GLMM1bis models.
Discussion
As for our first question, we found that Ps population size increased (although not linearly) with C. monspeliensis cover in mesomediterranean semi-natural habitats of Corsica. In addition, this relationship holds at both nymphal and adult stages, demonstrating that the affinity of Ps for C. monspeliensis goes beyond previously described local trends28,29. To answer the second question, we did not use the number of Ps found on each C. monspeliensis plant but added instead an offset to the model, which was equivalent to using the number of Ps divided by the area covered by C. monspeliensis in each plot as the output variable. Using this proxy, we found that Ps load on C. monspeliensis decreased with C. monspeliensis cover, especially from 0 to 50% plot cover. This pattern is expected when an insect has preference for one plant species over all its neighbors31,32. Insect load per C. monspeliensis plant was maximal when C. monspeliensis plants were scarce and disseminated in a matrix of other plant species. Then, when C. monspeliensis cover increases, Ps populations tend to dilute over the resource.
To our knowledge, Corsica is the only region where Ps population size shows such a pattern of strong association with C. monspeliensis density, although this plant is widely distributed across the Mediterranean region59. The availability of C. monspeliensis in mesomediterranean habitats of Corsica might primarily explain this unusual relationship given that Ps is highly polyphagous. The observed preference may be driven by retained chemical signal in generations of P. spumarius succeeding on C. monspeliensis as already observed in other taxa60, and/or by specific communities of endosymbionts61. The question remains opened and more investigations are required to understand the determinisms of the association observed in Corsica (e.g., tests of host preferences in controlled conditions). In the light of this study and a previous one28, assessing the consequences of pulling away C. monspeliensis from the direct vicinity of agricultural areas as a risk mitigation strategy against Xf appears relevant and so seems the release of biocontrol agents62 near C. monspeliensis bushes.
C. monspeliensis-dense areas favoring larger populations of Ps and dilution over the resource may have contributed to the spread of Xf over Corsica5,7. Indeed, C. monspeliensis can be infected by Xf63,64 and is very abundant in early stages of mesomediterranean vegetation series in Corsica39. It is noteworthy that C. monspeliensis falls well within the description of the "hidden reservoir" suggested by epidemiological modelling65. C. monspeliensis being a pyrophyte66 and an early colonizer39, fire breaks and land clearings may have contributed to the spread of Xf in Corsica by increasing cover of C. monspeliensis. The strong association between Ps and C. monspeliensis in mesomediterranean habitats might contribute to explain why we did not observe any major outbreaks on plants sensitive to the strains of Xf ssp. multiplex present in Corsica so far63.
Populations of adults of Ps were larger in cooler and moister habitats with lower seasonal fluctuations, whatever the sampling month, and this pattern was similar whatever the type of climate data used (i.e., bioclimatic variables or seasonal raw data). This result was expected from the physiological requirements of Ps2,11,12 and from observations made in Italian grasslands, where larger populations of Ps are found in cooler sites37. This also fits with species modelling approaches, which predict low suitability for hot and dry areas in Mediterranean regions5,6 and is consistent with results of Avosani et al.38, who highlighted that the probability of observing Ps decreased when the average number of sunlight hours increased. It is noteworthy that the correlation between nymph abundance and either PC1bioclim or PC1seasonal was not significant, which could be the result of non-restrictive climatic conditions at sampling sites in April. The number of overwintering eggs is not well predicted by the number of gravid females, but depends primarily on weather conditions during the oviposition period67. Oviposition and winter survival of eggs add variability to population dynamics and may blur the climate effect found at adult stages.
Interestingly, the slope of the correlation with PC1bioclim and PC1seasonal was steeper in June than in October. Furthermore, local or surrounding vegetation structure had effect only on the abundance of young adults when bioclimatic data were used, with larger populations observed in plots with high scrubland vegetation and surrounded by high scrubland landscapes (tree cover > 25%). This pattern could be explained by the migration to summer hosts17,29 outside of the sampling plot in hotter and drier low scrubland places and a microclimatic effect benefiting Ps in higher scrubland vegetation9. Apart from this situation, local and landscape vegetation structures had no effect on Ps populations and C. monspeliensis cover was much more important to explain Ps abundance than broad vegetation structure.
We believe that our results can help designing surveillance plans for Xf that are adapted to the specificities of Corsica. Surveillance may indeed target C. monspeliensis-rich habitats and use Ps sampled there as sentinel insects to track Xf in the environment5,7. As they highlight an unexpected and unique association pattern between Ps and C. monspeliensis, our results call for more studies on the ecology of Xf vectors in Europe, including common ones, to better understand, anticipate and control the spread of Xf in a climatically changing world. So far, only a few studies have investigated spatial distribution of vectors across European semi-natural landscapes9,37,68. In Mediterranean regions, studies principally focused on agroecosystems and targeted only a few species of potential vectors of Xf. They were primarily conducted in Italian olive groves, which is explained by the dramatic economic and socio-cultural consequences of Xf in this area69. A few were also performed in almond orchards, citrus orchards or vineyards in other Mediterranean areas16,28,70. Yet, only studies that attempt to decipher plant-vector trophic networks or investigate the main drivers of local abundance of vectors will help to better anticipate future outbreaks and provide clues on how to manage this bacterium with sustainable practices.
Data availability
The data sets generated and analyzed during the current study are available in Table S1.
References
Saponari, M. et al. Infectivity and transmission of Xylella fastidiosa by Philaenus spumarius (Hemiptera: Aphrophoridae) in Apulia, Italy. J. Econ. Entomol. 107, 1316–1319. https://doi.org/10.1603/EC14142 (2014).
Cornara, D., Bosco, D. & Fereres, A. Philaenus spumarius: When an old acquaintance becomes a new threat to European agriculture. J. Pest Sci. 91, 957–972. https://doi.org/10.1007/s10340-018-0966-0 (2018).
Weaver, C. R. & King, D. Meadow spittlebug, Philaenus leucophthalmus (L.). Ohio Agric. Exp. Stn. Res. Bull. 741, 258 (1954).
Halkka, A., Halkka, L., Halkka, O., Roukka, K. & Pokki, J. Lagged effects of North Atlantic Oscillation on spittlebug Philaenus spumarius (Homoptera) abundance and survival. Glob. Change Biol. 12, 2250–2262. https://doi.org/10.1111/j.1365-2486.2006.01266.x (2006).
Cruaud, A. et al. Using insects to detect, monitor and predict the distribution of Xylella fastidiosa: A case study in Corsica. Sci. Rep. 8, 15628. https://doi.org/10.1038/s41598-018-33957-z (2018).
Godefroid, M. et al. Climate tolerances of Philaenus spumarius should be considered in risk assessment of disease outbreaks related to Xylella fastidiosa. J. Pest Sci. 2021, 1–14. https://doi.org/10.1007/s10340-021-01413-z (2021).
Farigoule, P. et al. Vectors as sentinels: Rising temperatures increase the risk of Xylella fastidiosa outbreaks. Biology 11, 1299. https://doi.org/10.3390/biology11091299 (2022).
Drosopoulos, S. & Asche, M. Biosystematic studies on the spittlebug genus Philaenus with the description of a new species. Zool. J. Linn. Soc. 101, 169–177. https://doi.org/10.1111/j.1096-3642.1991.tb00891.x (1991).
Godefroid, M. & Durán, J. M. Composition of landscape impacts the distribution of the main vectors of Xylella fastidiosa in southern Spain. J. Appl. Entomol. 146, 666–675. https://doi.org/10.1111/jen.13003 (2022).
Karban, R. & Strauss, S. Y. Physiological tolerance, climate change, and a northward range shift in the spittlebug, Philaenus spumarius. Ecol. Entomol. 29, 251–254. https://doi.org/10.1111/j.1365-2311.2004.00576.x (2004).
Chmiel, S. M. & Wilson, M. C. Estimation of the lower and upper developmental threshold temperatures and duration of the nymphal stages of the meadow Spittlebug, Philaenus spumarius. Environ. Entomol. 8, 682–685. https://doi.org/10.1093/ee/8.4.682 (1979).
Yurtsever, S. On the polymorphic meadow spittlebug, Philaenus spumarius (L.) (Homoptera: Cercopidae). Turk. J. Zool. 24, 447–460 (2000).
Ahmed, D. D. & Davidson, R. H. Life history of the meadow spittlebug in Ohio. J. Econ. Entomol. 43, 905–908. https://doi.org/10.1093/jee/43.6.905 (1950).
Whittaker, J. B. Cercopid spittle as a microhabitat. Oikos 21, 59–64. https://doi.org/10.2307/3543839 (1970).
Drosopoulos, S. New data on the nature and origin of colour polymorphism in the spittlebug genus Philaenus (Hemiptera: Aphorophoridae). Ann. Soc. Entomol. Fr. NS 39, 31–42. https://doi.org/10.1080/00379271.2003.10697360 (2003).
Bodino, N. et al. Phenology, seasonal abundance, and host-plant association of spittlebugs (Hemiptera: Aphrophoridae) in vineyards of Northwestern Italy. Insects 12, 1012. https://doi.org/10.3390/insects12111012 (2021).
Cornara, D. et al. Natural areas as reservoir of candidate vectors of Xylella fastidiosa. Bull. Insectol. 74, 173–180 (2021).
Gargani, E. et al. A five-year survey in Tuscany (Italy) and detection of Xylella fastidiosa subspecies multiplex in potential insect vectors, collected in Monte Argentario. Redia 104, 75–88. https://doi.org/10.19263/REDIA-104.21.09 (2021).
Morente, M. et al. Distribution and relative abundance of insect vectors of Xylella fastidiosa in olive groves of the iberian peninsula. Insects 9, 175. https://doi.org/10.3390/insects9040175 (2018).
Delong, D. et al. Spittle-insect vectors of Pierce’s disease virus. I. Characters, distribution, and food plants. Hilgardia 19, 339–356 (1950).
Bodino, N. et al. Phenology, seasonal abundance and stage-structure of spittlebug (Hemiptera: Aphrophoridae) populations in olive groves in Italy. Sci. Rep. 9, 1–17. https://doi.org/10.1038/s41598-019-54279-8 (2019).
Wiegert, R. G. Population energetics of meadow spittlebugs (Philaenus spumarius L.) as affected by migration and habitat. Ecol. Monogr. 34, 217–241. https://doi.org/10.2307/1948501 (1964).
Dongiovanni, C. et al. Plant selection and population trend of spittlebug immatures (Hemiptera: Aphrophoridae) in olive groves of the Apulia region of Italy. J. Econ. Entomol. 112, 67–74. https://doi.org/10.1093/jee/toy289 (2019).
Bodino, N. et al. Spittlebugs of mediterranean olive groves: Host-plant exploitation throughout the year. Insects 11, 130. https://doi.org/10.3390/insects11020130 (2020).
Villa, M., Rodrigues, I., Baptista, P., Fereres, A. & Pereira, J. A. Populations and host/non-host plants of spittlebugs nymphs in olive orchards from northeastern Portugal. Insects 11, 720. https://doi.org/10.3390/insects11100720 (2020).
Antonatos, S. et al. Seasonal appearance, abundance, and host preference of Philaenus spumarius and Neophilaenus campestris (Hemiptera: Aphrophoridae) in olive groves in Greece. Environ. Entomol. 50, 1474–1482. https://doi.org/10.1093/ee/nvab093 (2021).
Hasbroucq, S., Casarin, N., Ewelina, C., Bragard, C. & Grégoire, J.-C. Distribution, adult phenology and life history traits of potential insect vectors of Xylella fastidiosa in Belgium. Belg. J. Entomol. 92, 2569 (2020).
Mesmin, X. et al. Interaction networks between spittlebugs and vegetation types in and around olive and clementine groves of Corsica; implications for the spread of Xylella fastidiosa. Agric. Ecosyst. Environ. 334, 107979. https://doi.org/10.1016/j.agee.2022.107979 (2022).
Albre, J., García-Carrasco, J. M. & Gibernau, M. Ecology of the meadow spittlebug Philaenus spumarius in the Ajaccio region (Corsica)—I: Spring. Bull. Entomol. Res. 111, 246–256. https://doi.org/10.1017/S0007485320000711 (2021).
Andersson, P., Löfstedt, C. & Hambäck, P. A. Insect density–plant density relationships: A modified view of insect responses to resource concentrations. Oecologia 173, 1333–1344. https://doi.org/10.1007/s00442-013-2737-1 (2013).
Hambäck, P. A., Inouye, B. D., Andersson, P. & Underwood, N. Effects of plant neighborhoods on plant–herbivore interactions: Resource dilution and associational effects. Ecology 95, 1370–1383. https://doi.org/10.1890/13-0793.1 (2014).
Otway, S. J., Hector, A. & Lawton, J. H. Resource dilution effects on specialist insect herbivores in a grassland biodiversity experiment. J. Anim. Ecol. 74, 234–240 (2005).
Lago, C. et al. Flight performance and the factors affecting the flight behaviour of Philaenus spumarius the main vector of Xylella fastidiosa in Europe. Sci. Rep. 11, 17608. https://doi.org/10.1038/s41598-021-96904-5 (2021).
Casarin, N. et al. Investigating dispersal abilities of Aphrophoridae in European temperate regions to assess the threat of potential Xylella fastidiosa-based pathosystems. J. Pest Sci. https://doi.org/10.1007/s10340-022-01562-9 (2022).
Bodino, N. et al. Dispersal of Philaenus spumarius (Hemiptera: Aphrophoridae), a vector of Xylella fastidiosa, in olive grove and meadow agroecosystems. Environ. Entomol. 50, 267–279. https://doi.org/10.1093/ee/nvaa140 (2020).
Santoiemma, G., Tamburini, G., Sanna, F., Mori, N. & Marini, L. Landscape composition predicts the distribution of Philaenus spumarius, vector of Xylella fastidiosa, in olive groves. J. Pest. Sci. 92, 1101–1109. https://doi.org/10.1007/s10340-019-01095-8 (2019).
Cappellari, A. et al. Spatio-temporal dynamics of vectors of Xylella fastidiosa subsp. pauca across heterogeneous landscapes. Entomol. Gen. 42, 515–521. https://doi.org/10.1127/entomologia/2022/1427 (2022).
Avosani, S., Tattoni, C., Mazzoni, V. & Ciolli, M. Occupancy and detection of agricultural threats: The case of Philaenus spumarius, European vector of Xylella fastidiosa. Agric. Ecosyst. Environ. 324, 107707. https://doi.org/10.1016/j.agee.2021.107707 (2022).
Allier, C. & Lacoste, A. Processus dynamiques de reconstitution dans la série du Quercus ilex en Corse. In Vegetation Dynamics in Grasslans, Healthlands and Mediterranean Ligneous Formations 83–91 (Springer, 1981).
Delbosc, P., Bioret, F. & Panaïotis, C. Plant landscape of Corsica: Typology and mapping plant landscape of Cap Corse region and Biguglia Pond (Springer Nature, 2020).
Chessel, D., Dufour, A.-B. & Thioulouse, J. The ade4 package—I: One-table methods. R. News 4, 5–10 (2004).
Biedermann, R. & Niedringhaus, R. The Plant-and Leafhoppers of Germany: Identification Key to All Species (Wabv Fründ, 2009).
Stöckmann, M., Biedermann, R., Nickel, H. & Niedringhaus, R. The Nymphs of the Planthoppers and Leafhoppers of Germany (WABV, 2013).
INRAE-CBGP. Arthemis DB@se - ARTHropod Ecology, Molecular Identification and Systematics. https://arthemisdb.supagro.inrae.frhttps://doi.org/10.15454/TBGRIB. Accessed 2021.
Xu, T. & Hutchinson, M. ANUCLIM version 6.1 user guide. Aust. Natl. Univ. Fenner Sch. Environ. Soc. Canberra 2011, 256 (2011).
Quintana-Seguí, P. et al. Analysis of near-surface atmospheric variables: Validation of the SAFRAN analysis over France. J. Appl. Meteorol. Climatol. 47, 92–107. https://doi.org/10.1175/2007JAMC1636.1 (2008).
Hijmans, R. J., Phillips, S., Leathwick, J. & Elith, J. Dismo: Species Distribution Modeling https://CRAN.R-project.org/package=dismo (2017).
Faraway, J. J. Extending the Linear Model with R: Generalized Linear, Mixed Effects and Nonparametric Regression Models (Chapman and Hall/CRC, 2006).
Brooks, M. E. et al. glmmTMB balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling. R. J. 9, 378–400. https://doi.org/10.3929/ethz-b-000240890 (2017).
R Core Team. R: A Language and Environment for Statistical Computing https://www.R-project.org/ (2019).
Hardin, J. W. & Hilbe, J. M. Generalized Linear Models and Extensions 4th edn. (Stata Press, 2018).
Harrison, X. A. Using observation-level random effects to model overdispersion in count data in ecology and evolution. PeerJ 2, e616. https://doi.org/10.7717/peerj.616 (2014).
Bolker, B. M. et al. Generalized linear mixed models: A practical guide for ecology and evolution. Trends Ecol. Evol. 24, 127–135. https://doi.org/10.1016/j.tree.2008.10.008 (2009).
Hartig, F. DHARMa: Residual Diagnostics for Hierarchical (Multi-Level/Mixed) Regression Models https://CRAN.R-project.org/package=DHARMa (2020).
Lüdecke, D., Ben-Shachar, M. S., Patil, I., Waggoner, P. & Makowski, D. performance: An R package for assessment, comparison and testing of statistical models. J. Open Sourc. Softw. 6, 3139. https://doi.org/10.21105/joss.03139 (2021).
Fox, J. & Weisberg, S. An {R} Companion to Applied Regression Third edn, https://socialsciences.mcmaster.ca/jfox/Books/Companion/ (Sage, Thousand Oaks CA, 2019).
Lenth, R. V. Emmeans: Estimated Marginal Means, Aka Least-Squares Means https://CRAN.R-project.org/package=emmeans (2021).
Hothorn, T., Bretz, F. & Westfall, P. Simultaneous inference in general parametric models. Biom. J. 50, 346–363. https://doi.org/10.1002/bimj.200810425 (2008).
Fernández-Mazuecos, M. & Vargas, P. Ecological rather than geographical isolation dominates Quaternary formation of Mediterranean Cistus species. Mol. Ecol. 19, 1381–1395. https://doi.org/10.1111/j.1365-294X.2010.04549.x (2010).
Berenbaum, M. R. & Feeny, P. P. 1. Chemical mediation of host-plant specialization: The papilionid paradigm. In Specialization, Speciation, and Radiation (ed. Tilmon, K.) 3–19 (University of California Press, 2008). https://doi.org/10.1525/california/9780520251328.003.0001.
Kapantaidaki, D. E., Antonatos, S., Evangelou, V., Papachristos, D. P. & Milonas, P. Genetic and endosymbiotic diversity of Greek populations of Philaenus spumarius, Philaenus signatus and Neophilaenus campestris, vectors of Xylella fastidiosa. Sci. Rep. 11, 3752. https://doi.org/10.1038/s41598-021-83109-z (2021).
Mesmin, X. et al. Ooctonus vulgatus (Hymenoptera, Mymaridae), a potential biocontrol agent to reduce populations of Philaenus spumarius (Hemiptera, Aphrophoridae) the main vector of Xylella fastidiosa in Europe. PeerJ 8, e8591. https://doi.org/10.7717/peerj.8591 (2020).
Denancé, N. et al. Several subspecies and sequence types are associated with the emergence of Xylella fastidiosa in natural settings in France. Plant Pathol. 66, 1054–1064. https://doi.org/10.1111/ppa.12695 (2017).
EFSA, Delbianco, A., Gibin, D., Pasinato, L. & Morelli, M. Update of the Xylella spp host plant database—systematic literature search up to 31 December 2020. EFSA J. 19, 6. https://doi.org/10.2903/j.efsa.2021.6674 (2021).
Soubeyrand, S. et al. Inferring pathogen dynamics from temporal count data: The emergence of Xylella fastidiosa in France is probably not recent. New Phytol. 219, 824–836. https://doi.org/10.1111/nph.15177 (2018).
Roy, J. & Sonié, L. Germination and population dynamics of Cistus species in relation to fire. J. Appl. Ecol. 29, 647–655. https://doi.org/10.2307/2404472 (1992).
Whittaker, J. B. Density regulation in a population of Philaenus spumarius (L.) (Homoptera: Cercopidae). J. Anim. Ecol. 42, 163–172. https://doi.org/10.2307/3410 (1973).
Chapman, D. et al. Improving knowledge of Xylella fastidiosa vector ecology: modelling vector occurrence and abundance in the wider landscape in Scotland. Project Final Report. PHC2020/04, Scotland's Centre of Expertise for Plant Health (PHC) https://doi.org/10.5281/zenodo.6523478 (2022).
Saponari, M., Giampetruzzi, A., Loconsole, G., Boscia, D. & Saldarelli, P. Xylella fastidiosa in olive in Apulia: Where we stand. Phytopathology 109, 175–186. https://doi.org/10.1094/PHYTO-08-18-0319-FI (2019).
López-Mercadal, J. et al. Collection of data and information in Balearic Islands on biology of vectors and potential vectors of Xylella fastidiosa (GP/EFSA/ALPHA/017/01). EFSA Supp. Publ. 18, 10. https://doi.org/10.2903/sp.efsa.2021.EN-6925 (2021).
Acknowledgements
This work was funded by the Collectivité Territoriale de Corse and the European Union Horizon 2020 research and innovation program under Grant Agreement No. 727987 XF-ACTORS. We are grateful to three anonymous reviewers for their feedback on this work.
Author information
Authors and Affiliations
Contributions
A.C. and J.Y.R. designed research. A.C., J.Y.R., L.H. and F.C. provided funding. M.C., X.M., I.Q., S.B., P.F., E.P., J.M.T., J.S.C., J.Y.R. and A.C. conducted experiments. M.C., X.M. and J.P.R. analyzed data. M.C., X.M., J.P.R. and A.C. drafted the manuscript. All authors reviewed the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Chartois, M., Mesmin, X., Quiquerez, I. et al. Environmental factors driving the abundance of Philaenus spumarius in mesomediterranean habitats of Corsica (France). Sci Rep 13, 1901 (2023). https://doi.org/10.1038/s41598-023-28601-4
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-023-28601-4
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.