2.1.2. Decision-Support Systems

Building on the knowledge of risk factors, a range of models have been able to predict wireworm occurrence based on soil and meteorological data coupled with a hydrologic model [22]; click beetle abundance based on climatic and edaphic factors [24]; wireworm activity based on soil characteristics [21]; their abundance and community structure [53]; correlation between the damage caused in potato fields and landscape structure [56]; and to determine the key climate and agro-environmental factors impacting wireworm damage [23,25,52].

The hypothesis of the vertical distribution of wireworms depending on soil moisture, soil temperature, and soil type was verified by Jung et al. [21], who developed the prognosis model SIMAGRIO-W used as a decision-support system to forecast the (*Agriotes* spp.) wireworm activity based on edaphic properties. Albeit successfully applied in field tests in western Germany, the model performed poorly when it was evaluated in eastern Austria, and research effort is still needed to improve the current model.

Analyzing long-term survey data from maize fields in northern Italy, in which *Agriotes brevis*, *A. sordidus*, and *A. ustulatus* were identified as the predominant pest species, Furlan et al. [23] calculated risk level based on the different weights of the studied risk factors (defined by relative risk values). A simple decision tree was suggested for practical IPM of wireworms [57,58].

The decision-support system VFF-QC (web application: https://cerom.qc.ca/vffqc/, accessed on 9th of May 2021) was originally developed in Quebec (Canada) from a huge database that included more than 800 fields (maize, soybean, cereals, and grasslands), which were characterized by a set of factors (e.g., agricultural practices, soil type, humidity, and organic matter content) and wireworm trapping between 2011 and 2016 [59]. A predictive model based on boosted regression trees assessed the risk level (low, moderate, or high) of finding wireworms in abundance and determined if the field had reached a threshold that would justify treatment. To the best of our knowledge, VFF-QC is the most-used decision-support system for wireworm risk assessment, partly due to rules adopted in 2018 by the Government of Québec that force agronomists to justify the need for seed treatment before prescribing or recommending them to growers.

Using a similar statistical approach, Poggi et al. [25] examined the relative influence of putative key explanatory variables on wireworm damage in maize fields and derived a model for the prediction of the damage risk; they also assessed their model's relevance in providing the cornerstone of a decision support system for the managemen<sup>t</sup> of damage caused by wireworms in maize crops.

As a whole, these decision-support systems rely on correlative approaches that unravel the potential of a dynamic landscape to shape wireworm populations and eventually crop damages. The development of models that describe the mechanisms driving wireworm colonization, and subsequently elucidate the ecological processes that operate at the landscape scale, remains an avenue for future research.

#### *2.2. Monitoring and Thresholds*
