A predator-prey system: Phytoseiulus persimilis (Acari: Phytoseiidae) and Tetranychus urticae (Acari: Tetranychidae): worldwide occurrence datasets

The predator-prey system Phytoseiulus persimilis Athias-Henriot (Acari: Phytoseiidae) and Tetranychus urticae (Koch) (Acari: Tetranychidae) represents the most studied and the most known system among the Acari. Because of the agronomical importance of the two mite species, a wealth of modelling studies investigated the interactions of the two species at the individual level. However, regional or global level works on this system are lacking, which is impeding the investigation of climate change effects on biological control effectiveness. Here we compile and geo-locate worldwide occurrences for the two species considered, based on literature, collection and field survey data. The datasets presented in this document gather most of the literature records of both species for which locality data were available for geo-referencing (1,037 for T. urticae and 126 for P. persimilis). Geolocated data from collections and field surveys including host-plants are also presented for the first time (322 for T. urticae and 65 for P. persimilis). Phytoseiulus persimilis is also reported for the first time from Kenya and La Martinique.


Introduction
The development of effective biocontrol strategies relies on an in-depth understanding of predator-prey relationships (McMurtry and Croft, 1997).However, while many modelling studies focus at fine scale interactions (Takabayashi et al., 2000;Ellner et al., 2001), few studies address interactions at coarser levels.Furthermore, the investigation of climate change effects on biological control requires studies that expand beyond the local to the regional and global level.Such studies depend however, on the availability of detailed geo-located data on the distribution of both prey and predators.
Among the Acari, due to their economic importance, Tetranychus urticae (Koch, 1836) (Acari: Tetranychidae) and its predator Phytoseiulus persimilis Athias-Henriot, 1957 (Acari: Phytoseiidae) are the most studied, the most well-known and the most widely distributed preypredator pair globally.For both species, databases recording host plants (for phytophagous mites) or plant support (for predatory mites) and country level distribution are available (Demite et al., 2018;Migeon and Dorkeld, 2018) but geo-located occurrences are not included in these databases and remain scarce.This lack of information combined with a country scale mapping leads to an over-representation on a global scale as exemplified figures 1 and 2. Furthermore, some reports are made from glasshouses in countries with too harsh winters and unsuitable for outdoor surviving and have no ecological significance.
Tetranychus urticae is the species that attracts the most interest and has generated the greatest number of works among the family Tetranychidae (Vacante, 2015); the Google Scholar query "Tetranychus urticae" returns 50,600 records (Google Scholar October 2018).It has been recorded from 1,140 host plants in 124 countries (Migeon and Dorkeld, 2018).The annual value of pesticides used against the pest represents 400 million dollars (Van Leeuwen et al., 2015).
Meanwhile P. persimilis is the species that attracts the most interest and has generated the greatest number of works among predatory mites; the Google Scholar query "Phytoseiulus persimilis" returns 10,200 records (Google Scholar October 2018).It has been recorded in 36 countries from all the continents (Demite et al., 2018), but mainly around the Mediterranean Basin especially before its large introduction all over the world for biological control purposes (Kanouh et al., 2010).This predatory mite is mass-produced by biological control companies throughout the world.Phytoseiulus persimilis feeds mainly on T. urticae and belongs to the specialist class of Phytoseiidae predators, as it is closely associated with its "unique" prey T. urticae (McMurtry & Croft, 1997;McMurtry et al., 2015).
Few datasets regarding Phytoseiidae or Tetranychidae have been published.GBIF (2018) gathered 10,572 occurrences of Phytoseiidae, of which only 69% are geo-located.In the GBIF (Global Biological Information Facility, https://www.gbif.org)dataset we found 121 P. persimilis records but only 74 with suitable coordinates (match of location and coordinates and 0.1 or more degree precision) and 65 unique locations.Regarding the Tetranychidae, two datasets exist.The first one deals with the genus Mononychellus (Vásquez-Ordóñez and Parsa 2014); the second deals with the whole family (Migeon 2015) and provides 271 T. urticae records but only 248 with suitable coordinates and 69 unique locations.To fill the gap in the availability of detailed geo-located data for both species, we publish four datasets.Two are issued from literature records and contain the occurrences retrieved for the two mite species considered.The two others contain new unpublished data from our collections and surveys and report occurrences and host plants or plant support for the two species.(Migeon and Dorkeld, 2018).Migeon (2015) data (green dots) are also added.

Geographic coverage
The spatial coverage differs among the two species.
The known distribution of T. urticae is shown in the figure 1 along with the geo-located records.Among the 411 references on T. urticae distribution in the Spider Mites Web database (Migeon and Dorkeld, 2018) representing 124 countries, we were able to find 246 references with localities (see Supplementary materials).These references correspond to 96 countries.For 29 countries, the information was not accurate enough to be included in the dataset (Table 1).
The known distribution of P. persimilis is reported on figure 2 along with the geo-located records.Among the 86 references dealing with the P. persimilis distribution in the Phytoseiidae Database (Demite et al. 2018) representing 36 countries, we were able to find 54 references with localities (see Supplementary materials).These references correspond to 29 countries.For 17 countries, the information was not accurate enough to be included in the dataset (Table 2).Kenya and Martinique are new records.For two countries (Finland and Hungary), only indoors records were reported and not included in the dataset.

Method step description
The datasets pool two different data types.(i) All available literature for the two species was analysed.Geographical data (coordinates) were compiled when available.When only textual descriptions of locations were available, occurrence data were assigned using several geolocation tools like GoogleMaps, GeoNames and other gazetteers   (Demite et al., 2018).GBIF (GBIF, 2018) (green dots) are also added.were added for the two species.These latter records are also completed by host (or support) plant when available.To avoid confusions and provide solely "natural records", only records not reported, according to the literature, from glasshouses or indoors supports were used.The coordinates are in decimal degrees (DD) based on the WGS84 geodetic system.Geographic coordinates were visually verified using the Check Coordinates tool in Diva-GIS (Hijmans et al., 2012) and manual verification (e.g. points in the sea).
Host plant nomenclature is in accordance to current taxonomy reference (The Plant List, 2013).

Uncertainty issues
Collection and survey records -Location precision was set to 0.001°DD for collection and surveys where recent locations had coordinates records Literature records -Location precision has been set to 0.001°DD when coordinates or precise locality was indicated in the publication to 0.01 DD when the place was found to correspond to a bigger area (10 to 50 km²).Less precise locations were not recorded.

Figure 1
Figure 1 Tetranychus urticae.World map representing all the locations included in the dataset.Global literature survey (yellow dots), collections and fields surveys (blue dots).Brown countries indicate literature country records(Migeon and Dorkeld, 2018).Migeon (2015) data (green dots) are also added.
, completed when necessary by textual search.Each record contains an indication of uncertainty.(ii) Unpublished data from INRA-CBGP collection, SupAgro collection and Cyprus University of Technology, Department of Agricultural Sciences, Biotechnology and Food Science (CUT-ABF) survey
A total of 1,037 T. urticae occurrences were retrieved from literature; 941 of them are unique location records.Unpublished data from INRA-CBGP collection gather 178 occurrences / host plant records.Unpublished data from CUT-ABF gather 144 occurrences / host plant records.Altogether unpublished data represent 212 unique location records.A total of 126 P. persimilis occurrences were retrieved from literature; 103 of them are unique location records.Unpublished data from Montpellier SupAgro-CBGP collection gather 65 occurrences / support plant records.These unpublished data represent 35 unique location records.

Table 1
Countries for which Tetranychus urticae occurrence was not accurate enough to determine the corresponding geographical coordinates and then including them in the dataset.

Table 2
Countries for which Phytoseiulus persimilis occurrence was not accurate enough to determine the corresponding geographical coordinates and then including them in the dataset.