Modeling the potential global distribution of suitable habitat for the biological control agent Heterorhabditis indica

Abstract Entomopathogenic nematode (EPN) Heterorhabditis indica is a promising biocontrol candidate. Despite the acknowledged importance of EPN in pest control, no extensive data sets or maps have been developed on their distribution at global level. This study is the first attempt to generate Ecological Niche Models (ENM) for H. indica and its global Habitat Suitability Map (HSM) for H. indica to generate biogeographical information and predicts its global geographical range and help identify of prospective areas for its exploration and to help identify the suitable release areas for biocontrol purpose. The aim of the modeling exercise was to access the influence of temperature and soil moisture on the biogeographical patterns of H. indica at the global level. Global Heterorhabditis indica ecosystems. CLIMEX software was used to model the distribution of H. indica and assess the influence of environmental variable on its global distribution. In total, 162 records of H. indica occurrence from 27 countries over 25 years were combined to generate the known distribution data. The model was further fine‐tuned using the direct experimental observations of the H. indica's growth response to temperature and soil moisture. Model predicts that much of the tropics and subtropics have suitable climatic conditions for H. indica. It further predicts that H. indica distribution can extend into warmer temperate climates. Examination of the model output, predictions maps at a global level indicate that H. indica distribution may be limited by cold stress, heat stress, and dry stresses in different areas. However, cold stress appears to be the major limiting factor. This study highlighted an efficient way to construct HSM for EPN potentially useful in the search/release of target species in new locations. The study showed that H. indica which is known as warm adapted EPN generally found in tropics and subtropics can potentially establish itself in warmer temperate climates as well. The model can also be used to decide the release timing of EPN by adjusting with season for maximum growth. The model developed in this study clearly identified the value and potential of Habitat Suitability Map (HSM) in planning of future surveys and application of H. indica.


| INTRODUC TI ON
Entomopathogenic nematodes (EPN) have proved useful in controlling insect pests belonging to order Coleoptera, Lepidoptera, and Diptera and have the potential to replace costly and hazardous chemical pesticides to an extent or be an effective option in Integrated Pest Management (IPM) programs (Georgis et al., 2006).
Extensive surveys have been conducted across the globe to isolate EPN species and exploit them as biocontrol agent of soil-dwelling and foliar insect pests in agricultural fields. Till date, around 121 valid species of entomopathogenic nematode (EPN) belonging to genera Steinernema (100 species) and Heterorhabditis (21 species) have been identified from different countries of the world (Bhat et al., 2020). For economic and successful use, it is important that the naturally occurring species of biocontrol agents are discovered and tested against high priority local crop pests. Also, it is widely accepted that locally collected EPN should be released for biocontrol as this would exclude many risks associated with introducing exotic species into new environments (Abate et al., 2017;Askary et al., 2018;Noosidum et al., 2010). Hence, search for EPN in their natural habitat is the first step toward their use in pest control programs (Slininger et al., 2003). Even though EPN are ubiquitous in nature (Hominick, 2002), the recovery frequency during surveys conducted worldwide is usually low. EPN recovery from the field requires considerable resources, expertise, and time. The search for suitable EPN to employ as biocontrol agent may become more efficient if geographical distribution of target EPN species is known. Despite the acknowledged importance of EPN in pest control (Askary & Abd-Elgawad, 2017Lacey & Georgis, 2012), no extensive data sets or maps have been developed on their distribution at country, regional, or global level. For many of the described EPN species, only a few exact localities with geographic coordinates have been published and their occurrence is generally presented in the form of point maps. Such point maps show the occurrence data but convey no information on the likelihood of their occurrence in nonsurveyed areas. It is, therefore, difficult to prioritize search areas for successful isolation of EPN and to effectively plan their use in biocontrol programs. In such cases, Habitat Suitability Maps (HSM) presenting potential habitats where the species are likely to be found, can help in prioritizing areas for searching and increase chances of successful isolation of EPNs. HSM can also help in identifying potential habitats where species can establish itself post release. Ecological Niche Models (ENM) are frequently used to predict geographic distribution of suitable habitats of modeled species and to generate their HSM (Franklin, 2010;Mukherjee et al., 2011). ENM is a GISbased method that commonly utilizes the information on climatic tolerances and ecology of an organism in their native habitat, to identify habitats in other geographical regions where populations could potentially occur (Pearson, 2010). This approach has proven valuable for generating biogeographical information and predicts the geographical range of a species without conducting extensive surveys (Steinbauer et al., 2002). CLIMEX software has proved useful in biocontrol research, in identification of prospective areas for the exploration and to study post release establishment patterns of biological control agents (Firehun et al., 2013;Goolsby et al., 2005;Japoshvili et al., 2015;Lawson et al., 2010;Senaratne et al., 2006;Shrestha et al., 2015;Zalucki & Van Klinken, 2006). This software has found to be ideal for modeling the species with patchy distribution and point distribution record (Hill & Thomson, 2015) and can be used to model EPN species distribution. CLIMEX models have also been widely used for many pests including a few nematode species and have been useful in projecting their distributions (Boag et al., 1997;Singh et al., 2021;Yeates et al., 1998).
CLIMEX model makes use of both species' distribution and phenology data (temperature, moisture preferences, wet, cold, heat, and dry stresses) and correlates with metrological datasets to make predictions. Thus, the HSM generated provides a scientific basis for identifying areas where there is high likelihood of species being found or has the potential to successfully establish itself.
However, CLIMEX or any other simulation software has never been used for modeling the distribution of any EPN species.
Despite 155 published records of H. indica's occurrence from 27 countries, there is no species distribution map at global level. Therefore, this study develops a predictive model of the likely global distribution of H. indica and its suitable habitat. It is believed that this approach can be both economic and time efficient. The aim of the modeling exercise of this study was to develop an ecological niche model and to generate HSM for H. indica using the CLIMEX software to know H. indica's possible global geographic range to better design the future surveys and to help identify the suitable release areas. The study also aimed to investigate the importance of environmental variables when projecting the range and spatial pattern of the modeled EPN species. These methods may also be used for modeling the distribution of other EPN species and to make better informed decisions while using EPN as biocontrol agents.

| MATERIAL S AND ME THODS
CLIMEX software version 4.0 (Hearne Scientific Software Company) for windows was used to model the distribution of potential biocontrol agent H. indica under current climatic conditions. "Compare Locations model" for single species was utilized for the purposes of this study. Annual growth index (GI) was calculated as a function of soil moisture and temperature, and stress indices (cold, wet, hot The model was parameterized using global distribution of H. indica and published laboratory derived ecological study results. The parameters were iteratively adjusted depending on satisfactory agreement between the potential and known worldwide distribution of H. indica. The parameters were subsequently verified by comparing to ecological and physiological data of H. indica to ensure that they were biologically reasonable. ArcMap version 10.4.1 (Minami, 2000) was used to map CLIMEX output.

| Heterorhabditis indica occurrence data
In total, 155 records of H. indica occurrence from published research papers and 7 records from online database Global Biodiversity Information Facility database (GBIF) (http://www.gbif.org) were combined to generate the known distribution data. Not all published research papers provided exact geographic coordinates for positive sites so in those cases the Google Earth (https://www.google.com/ earth/) and the Google Maps (https://www.google.com/maps/preview) were used to best determine coordinates for the occurrence site. Duplicate records were removed.

| Model fitting and verifications
Since H. indica occurs primarily in tropical and sub-tropical climatic conditions, the Wet Tropical template provided with CLIMEX software was selected to create the "species parameter file." The model was calibrated using species-specific physiological tolerance thresholds. The climatic requirements of H. indica are inferred from its current known occurrence. Developmental threshold temperature data of H. indica and moisture requirement was inferred from laboratory data given by Kour et al. (2021)

| Moisture index
A value of SM = 0 indicates no soil moisture; SM = 0.5 indicates soil moisture content is 50% of Held capacity; SM = 1 indicates that the soil moisture content is 100% of capacity (Sutherst et al., 2007). To allow reasonable species growth, SM0 was set at 0.08. Soil moisture values for optimum growth (SM1 and SM2) were set at 0.1 and 0.7, respectively, to fit observed occurrence of species in United Arab Emeritus, Palestine, and Egypt. An upper soil moisture level (SM3) of 1.7 was adapted following a similar approach.

| Cold stress
The cold stress temperature threshold (TTCS) mechanism was used to describe the response of H. indica to cold. The TTCS procedure assumes that stress will start below a given temperature and accumulate at a given weekly rate. The cold stress was calculated as per the method described by Pivonia and Yang (2004). The following equation was used to calculate weekly cold stress: weekly stress = (THCS) × (TTCS) average weekly min temperature) × number of weeks, where THCS is the rate at which cold stress accumulates when minimum temperature is less than TTCS and number of weeks stands for the successive number of weeks with stress.
THCS was calculated using the maximum time permitted between two infections to maintain a population. In soil, H. indica can survive for 7-8 months before they find a new host (Mauleon et al., 2006

| Heat stress
Temperature is one of the most important factors affecting the infectivity and survival of infective juveniles (IJ). Temperatures above 35°C are sub lethal and above 40°C is lethal for IJ (Sandouka et al., 2003). To estimate heat stress, the method described by Pivonia and Yang (2004)

| Wet stress
The wet stress parameter was fitted through iteration where parameter was adjusted to match the known occurrence of H. indica. The threshold value for wet stress (SMWS) of 1.7 with an accumulation rate (HWS) of 0.003 was used.

| Model verification
In the final development phase, CLIMEX model was verified qualitatively by evaluating its ability to predict currently known occurrence of H indica. Occurrence with EI values > 0 were interpreted as correctly predicted presence. Following verification, the parameter file was run using the "Compare Locations" function to examine predicted global distributions.

| Model fitting and verification
The current recorded global occurrence of H. indica includes 27 countries and island groups between 20°S and 40°N (Figure 1).
Under a current climatic scenario, the predicted global distribution of H. indica shows a very close agreement with the known geographic range of H. indica (Figure 1). Occurrence records accord well with the modeled climate suitability for the area, and the present distribution is consistent with the EI values except for one occurrence record in Saudi Arabia which falls in the region predicted to be unsuitable for its occurrence.

| Model projection at global level
In  is limited by cold stress, heat stress, and dry stresses in different area ( Figure 8). However, cold stress appears to be the major limiting factor, whereas wet stress is the least limiting factor.

Projected land area that is climatically suitable for H. indica
under the current climate was quantified for each region ( programs, it is best to select and obtain species or strains from areas with climatic conditions like that of the release area. Maladaptation of EPN species to the climate of the release area can be a major limiting factor to the success of a biological control programs (Haye et al., 2013;Hoddle et al., 2014). Since many of the commercially available EPN species are isolated from either North America or Europe (Grewal & Peters, 2005), it is important to know the release areas most suitable for their introduction. Therefore, it is import- is unable to detect microhabitats which are very important as well (Unwin & Corbet, 1991).
While the cold and dry stress conditions are the primary characteristics defining H. indica range frontiers, it is the rainfall and the soil moisture that characterize its population growth.  tors that also determine habitat suitability for a particular species (Kriticos et al., 2003). There are many biotic and abiotic factors that affect EPN geographical distribution (Hominick et al., 1996). Other factors such as soil properties and biotic interactions may prevent species from colonizing sites that are otherwise climatically suitable.
Inclusion of soil type and host distribution into the climate projections will further improve the predictive ability of CLIMEX output.

| LI M ITATI O N
In pest management, the benefits and dangers of using modeling will not be fully appreciated until limitations of specific models are discussed openly (Worner, 1991). Modeling the potential distribution of a species using climatic mapping has received some criticism because the potential distributions cannot be predicted based on climate alone (Japoshvili et al., 2015;Soberon & Peterson, 2005).
Biotic interactions, such as host availability, soil type, geographical barriers to dispersal, are also limiting factors in species distribution (Legaspi & Legaspi, 2010;Ni et al., 2012;Zalucki & Van Klinken, 2006). However, we consider that the method described herein is a very useful predictive tool and should be seen as "First approximations" of species distribution (Pearson & Dawson, 2003).

CO N FLI C T O F I NTE R E S T
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.