Several sampling methods for assessing infestation rate (number of wireworms per unit of surface) have been developed over time. The infestation thresholds above which treatments are recommended depend on the species, the proportion of voracious instars in the catches, the sampling method and the crops concerned [57]. Also, the limit of detection of traps must be smaller than the damaging population threshold [79]. Soil core sampling and counting of caught individuals lacks interest for growers and agronomists [79], because they are time-consuming [57,107] and subject to significant sampling errors, essentially due to the sampling effort, to the non-random aggregated distribution of the larvae [108,109] and particularly important in the case of low, yet damageable wireworm populations [110]. It is also important to consider that small larvae are more likely to be markedly aggregated, while medium-sized ones are more dispersed and large ones closer to random [109], a pattern involved in the interpretation of trap catches. Many bait-based sampling techniques have proven to be at least as effective as soil core for detecting wireworms [95,111,112]. Bait efficiency notably depends on the period of bait exposure (to be adapted according to case-specific parameters, such as trap handling or sowing period of the future crop) and the time of the year [28,112,113,114] and decreases when alternative food sources are present, which is often the case in freshly ploughed grass fields [24,79,115]. Germinating cereal seed baits, whose efficiency seems enhanced by daily irrigation [113], is the most efficient sampling technique (including cost and accuracy concerns) for determining wireworm populations in worldwide agricultural habitats and in grass fields intended for arable production [95,111,113,116,117].

Chabert and Blot (1992) used a modified version of traps designed originally against Melanotus spp. populations in the USA [118]. Briefly, they consisted of 650 ml plastic pots filled with vermiculite and a mix of 30 ml of maize and 30 ml of wheat as baits, entirely watered and buried in the ground, covered with soil and a plastic lid covered with a second soil layer. Traps were checked by hand after 10 to 15 days. For an equivalent number of samplings, baited traps allow the retrieval of two- to four-times more wireworms and save up to two thirds of the sampling time compared to simple soil core sampling [111,115]. On maize, the two techniques had comparable accuracies for high populations. When these decreased, accuracy remained acceptable with baited traps, while turning disastrous with soil cores. Such traps allowed linking the percentage of deceased plants to the wireworm infestation rate in several situations. For example, in maize, sowing before May 1^st in fields infested by five to 10 larvae per square meter caused the destruction of 30% of the plants [115]. Bait traps remain better presence/absence indicators of wireworm infestation than soil cores alone, especially where populations are below the limit of detection of soil cores (in potato – 62,500 wireworms/ha with 20 soil cores), even if the effective sampling area of a trap may be influenced by many factors [111,112]. In potato crops, there is a trend towards a higher level of damage in untreated plots with higher wireworm abundance, but even at densities of less than 100,000 larvae/ha, damage can range between 20% and 80% of tubers attacked [119]. Similar wide ranges were found in Romania: wireworm populations (mainly Agriotes) ranging from 5–10 larvae/sq.m to 70–100 larvae/sq.m will induce up to 31.6% of attacks on wheat, 22.5% on barley, 42.8% on corn seeds, 54.0% on corn at root neck, 64% on sunflower seed and 53.0% on sunflower at root neck [120]. This emphasizes the difficulty to define infestation thresholds above which treatment has to be applied and is partly attributable to sampling accuracy. A remarkable example of an efficient sampling plan to assess wireworm occurrence and control measure necessity was reported by Cherry et al. (2011). A sequential core-sampling plan at sugarcane planting was conducted on purpose to target areas needing treatment. If nine or more wireworms (Melanotus communis and Conoderus spp.) were retrieved from a total of 25 samples evenly spaced in a diagonal field transect, the latter were treated. By showing no significant differences in further yield and wireworm numbers between adjacent paired fields (identical agronomic conditions—one treated, the other untreated), they showed that both the sampling method and the used economic injury level allowed $100(US)/ha savings in unnecessary soil insecticide application [121]. In Italy, data collected over 15 years showed significant correlations between infestation rate or the average number of larvae per bait trap and damage in maize plants by A. brevis, sordidus and ustulatus [57]. Significant relationships between pre-planting catches of wireworms by baited traps and damage to maize plants were found for all three species, but symptoms of damage varied. When the population is greater than five larvae/trap, A. ustulatus larvae may significantly affect plant stands by damaging seeds. Agriotes brevis was found to be the most harmful, with catches above one per trap causing considerable plant damage that may result in a yield reduction. For the same damage level in maize fields, five-times more A. ustulatus larvae are needed. Agriotes sordidus has an intermediate damage potential (wireworm densities above two larvae/trap may cause a yield reduction). These thresholds are reliable for: (1) bare soil in which there are no alternative food sources, also after meadow (e.g., alfalfa and ryegrass) if this has previously been cultivated (in this case, the field has to have been ploughed at least three months before baited trap placement); (2) the average soil temperature at 10 cm depth is over 8 °C (for more than 10 days); and (3) the soil humidity is high (near to field water capacity) (Lorenzo Furlan—personal communication). These examples highlight the accuracy needed as to the conditions in which thresholds are described.

It is generally admitted that further improvements in bait traps or other sampling methods will be needed to accurately estimate the density of wireworms. Considering the factors involved, the outcomes in terms of treatment decision baseline may be as various as existing agronomic situations, spatial distribution (large-scale patterns) and crop-wireworm species combinations. The combined use of soil-core sampling and bait trapping may give a more accurate representation of potentially damaging species present, because it reflects more properly the wireworm species and distribution within the soil [122]. The closer to reality the estimation of the population is, the more precisely damaging thresholds can be defined, therefore reducing unnecessary treatment cases.

In-soil baits are useful in wireworm mass trapping, trap crops being the most efficient in that purpose, providing a very attractive alternative food source. By planting wheat rows eight days in advance of intercropped rows of strawberries, the seedling mortality due to Agriotes feeding can be reduced from 43% in unprotected plots to 5.3% in plots with an intercrop of wheat [5]. Maize protection can also be achieved and enhanced as the plant species diversity in the trap rows increases. Molecular analysis of gut content screening for plant DNA [86] outlines that wireworms are lured away from the target crop and also preferentially consume the intercrops [123], e.g., “trap peas” (Pisum sativum cv. Valverde) are more attractive to wireworms than potatoes [124]. These efficient techniques, difficult to set up at field scale, can only be developed with an economically interesting approach.

Three main elements account for the fascination of insect sex pheromones and their feasibility for insect management: (1) they are species-specific (even when synthetic and incomplete, and blends usually affect only the target, with the possible exception of taxonomically closely related species); (2) they are active in very small amounts; and (3) the vast majority are not known to be toxic to animals [125]. The identification of click beetles sex-pheromone main compounds in the late 1980s provided the qualitative bases of sampling techniques and mating disruption/mass trapping according to species.

Geranyl and (E,E)-farnesyl esters are major components of Agriotes natural sex pheromones [126]. Some species of click beetles may contain up to 24 substances in their pheromone glands [127]. Tóth (2012) summarized the status of identification of pheromones and attractants described in a variety of species among which 22 (73.3%) were Agriotes [128]. Click beetles show a variety of responses according to changes taking place in the structure of the pheromone component, especially in its acyl group or allylic fragment [129]. The sex attractant pheromone of one species may act as a sex inhibitor with respect to the other [129] through one of the compounds of the blend [130]. The same component can capture different target species, e.g., geranyl hexanoate actively attracts both A. sordidus and A. rufipalis Brullé [131]. Geranyl butyrate (GB) and (E,E)-farnesyl butyrate (FB) were identified in the pheromone gland extract of females A. brevis as the major sex pheromone components. Used as field baits (Italy, Hungary, Bulgaria) where both A.sputator and A. brevis were present, they selectively caught only A. brevis, whereas GB is also the main pheromone component of A. sputator, suggesting that FB has a role in reproductive isolation [132]. On the other hand, one sex-pheromone can attract individuals of species other than the target one, depending on geographic location [10,11,130,133], which shows that pheromone geographical “dialects” exist [9,127,129]. For A. lineatus and A. proximus, the comparative study between pheromone profile and the mitochondrial cytochrome c oxidase subunit I gene even led to the consideration of a necessary taxonomic revision [134].

The pheromone gland extract composition does not necessarily follow the proportions obtained by volatile collection when available, e.g., for A. lineatus and A. proximus [134], while the optimal ratio of concerned compounds in pheromone lures is also different from the two latter extracts [130]. In the future, if the pheromones of problematic species are to be analyzed by volatile collection (an unavailable method when most Agriotes gland extracts were studied in the 1980s), then more accurate pheromone elucidations may result. Besides, it has been noticed in some cases that females could be attracted towards blends of synthetic pheromonal compounds, revealing potential aggregation traits of the latter [128]. The same analytical improvements could provide information to understand these observations.

Very recently, a binary combination of two synthetic floral compounds, (E)-anethol and (E)-cinnamaldehyde, was optimized as a female-targeted lure for female Agriotes ustulatus Schwartz click beetles, opening a new perspective for adult click beetle mass trapping [128,135], but most of the applied cases of adult monitoring have focused on male click beetles thanks to pheromone traps. Guidelines for the synthesis of sex-pheromone, highly effective for baits, are available [133,136]. Several sex pheromone traps have been developed for monitoring all the most important Agriotes species [9,137,138,139,140]. Among them, bottle traps (homemade funnel traps) and VARb funnel traps, an adaptation of CSALOMON Var funnel traps (Plant Protection Institute of Budapest, Hungary), were adapted for species privileging flying over crawling. TAL traps (adapted pitfall traps) and YATLOR traps, similar to the Estron trap described by Oleschenko (1987) and Kudryastev (1993) ([139] and [9]) were designed for species privileging crawling over flying. A combination of the bottom part of a YATLOR trap and the upper part resembling the bottle trap gave birth to a high capacity trap, the YATLORfunnel, or YATLORf trap, effective in monitoring all the target species during all the swarming season [137]. It allowed detecting wireworm population levels in different European countries [7] and establishing distribution maps of the predominant species, even at low population densities [10]. Trap design matters: during the early part of the trapping season, traps into which both crawling and flying beetles can enter may capture significantly more individuals than designs where insects can only fly in. This has been observed for A. brevis and disappears later in the season. This indicates the need for traps suitable for use throughout the whole season [132].

A ground-based pheromone trap highly efficient for monitoring A. lineatus and A. obscurus has also been developed in Canada (the “Vernon Beetle Trap”, Phero Tech Inc., Delta, British Columbia) [140], and attempts for their monitoring highlighted that pheromone trapping should be the choice method for their surveying in North America [141].

Blackshaw and Vernon (2006) demonstrated the usefulness of pheromone trapping as a sampling method in addressing ecological questions at a landscape scale, by studying the spatial stability in non-farmed habitats of male A. lineatus and A. obscurus [142]. Click beetle species have been monitored thanks to large-scale pheromone trapping in Croatia. The abundance and dominance of A. lineatus decreased compared to average data from 1961–1990, abundance being negatively correlated with average air temperature. This result was postulated to be a potential effect of climate change [12]. In Rhineland-Palatinate (Germany), pheromone trap catches performed from 2008 to 2010, with the aim to define click beetle communities, highlighted the unexpected presence of A. sputator, possibly related to the increased potato damage observed during the previous seven years [11]. Such studies perfectly illustrate large-scale direct uses of pheromone trapping. Interpreting adult male trap counts for quantitative predictions of population size is complex and limited, unless the observed differences between species can be explained by dissimilar movement rates of the species and interferences between traps [143], hence the necessity to acquire accurate knowledge of the range of attraction of the pheromone traps [144]. It is essential, although not necessarily always possible yet, to: (1) establish the biological significance of the pheromone trap catches; (2) determine the actual range of attractiveness (related to the movement rates—male A. obscurus are able to migrate at least 80 m and A. sputator up to 27 m, distances linked to site-specific factors [75,145]); (3) study the relationship between males captured and the level of the female population; and 4) establish a reliable correlation between adult trap catches and subsequent larval populations for all the species and varieties in different climatic and agronomic conditions (mainly rotation) [137]. In Italy, sex pheromone traps proved to be much more sensitive in detecting species than the tools used to monitor larval populations, and the trap catches were correlated with estimated larval populations [146]. A mark-release-recapture study performed in South Devon, U.K., in a topographically flat and wind obstacle-free area, helped in getting species-specific migration distance data. Inventories led to the definition of maximal sampling ranges and effective sampling areas of A. lineatus, A. obscurus and A. sputator traps. Recapture rates were significantly different between species and recapture distances (ranging from 4 to 32 m). With such results, an estimation of the minimum cost of mass trapping programs to prevent males from mating was possible: 165€/ha/year (A.lineatus), 247.5€/ha/year (A.obscurus) and 2343€/ha/year (A.sputator) [147], illustrating the importance of separating wireworms into species. Sufyan et al. (2011) worked on the same problem in Germany and obtained slight differences, mainly due to the employed regression method. They took account of the attractive potential, i.e., the maximum sampling area of pheromone traps in the equation. On the basis of the estimated probabilities of recapture, a maximum distance of 20 m between individual traps would be needed to ensure substantial mass trapping, i.e., 25 traps/ha to theoretically reduce the male click beetle populations by more than half [144].

The potential of using pheromone trapping for male click beetles with subsequent wireworm reduction remains unclear, but will definitely require a dense network of traps [144]. The spatial relationship between aboveground adults and belowground larval distributions may not always be straightforward, unlike for other pests, which is important in cases where the monitoring and/or management is carried out on a different life stage to that causing damage [122,148]. Unfortunately, the spatial relationships between adult male catches and the distribution of their larvae is such that pheromone traps alone will not reliably indicate where wireworms occur in a field or an area [143].

The use of sex-pheromones to monitor and trap Agriotes click beetles illustrates the issues separating the understanding of an ecological phenomenon at the individual scale and its application in IPM strategies. Behavioral and ecological differences between species need to be acknowledged when researching the pest complex at the field and larger scale [106,148,149]. Separating wireworms to species instead of grouping as a complex reveals more as to the relationship between adult and wireworm distributions [122]. Also, the usefulness of pheromone traps in assessing Agriotes spp. populations could be restricted to larger scales than those at which pest management is normally undertaken. Distribution maps should be defined using sex pheromone traps within zones defined according to prevalent crop rotation organic matter content, soil type and precipitation [57], as illustrated above. In fragmented rural landscapes, the monitoring of Elateridae should take into account the heterogeneous pattern of the adult population (stratified sampling) [150]. Combining a geostatistic analysis of agronomic characteristics of soil and climatic conditions of the investigated area with adult distribution pattern, it should be possible to draw risk-areas on a predetermined scale of investigation (such as a province) and to provide more precise monitoring and management, provided that further studies show more details of how key variables influence the distribution [150].

Efficient active substances should be available in soil during critical cropping phases. Wireworms temporarily repelled or deterred from feeding on germinating seeds and seedlings will subsequently feed on older and less vulnerable tissue, which could be tolerated [21]. In any case, the timing of sowing in relation to upward wireworm movements in the root layer is a key factor to consider [158]. When corn is protected from damage during the first three weeks of growth, economic impacts can be minimized [40]. With seed coatings, the area treated with chemicals is considerably reduced and they appear just as effective as full-soil treatments. Moreover, they may well be combined with fungicide seed treatments, therefore providing easy handling by farmers together with adequate wireworm control, notably on wheat and barley [120,159]. Good candidates for seed coatings should unite four characteristics. They must (1) be efficient through wireworm feeding, i.e., below integumental penetration rate insecticides, usually inefficient on wireworms [160,161], (2) be integrated in a solid state in the seed-coating mixture in order to avoid atmospheric dispersion during mechanical sowing, (3) have a low water solubility and a high affinity for soil particles and (4) remain active against wireworms after sowing (approximately one month for corn), during the whole seedling establishment/strong wireworm feeding phase [161]. Neonicotinoid insecticides as a treatment applied to potato seed at planting can enhance a significant reduction of attacked tubers between sowing and harvest (imidacloprid-70 g/ton seed, thiamethoxam-50 g/ton seed). Compared with broadcast organophosphorous soil treatments, the amount of needed insecticide can be reduced by over 90% depending on the amount of potato seeds per hectare. Such numbers, altogether with the lack of an adverse effect on potato yield, stand for an adequate alternative to soil treatments with organophosphorous compounds [159].

Seed treatments cannot automatically be linked with wireworm population mortality. On wheat, neonicotinoid (imidacloprid, clothianidin and thiamethoxam), pyrethroid (tefluthrin) and a combination of the latter and thiamethoxam applied as seed treatments can provide excellent stand protection, likely through prolonged wireworm intoxication, but the reductions of neonate and in place wireworm populations are not significant. To the contrary, fipronil seed treatment provides an excellent protection with population reduction, which makes it as efficient as lindane [162]. If blended with thiamethoxam, wheat stand and yield are improved over the individual chemicals applied alone. They lead to high mortality of resident and neonate wireworms, which implies this treatment should only be applied once every three years during which time susceptible crops could be grown, with much lower rates (50×) than the formerly used Vitavax Dual containing lindane [163]. Furthermore, if using seed treatment on only a fraction of the seeds, the proportion of treated seeds could even be adjusted to the infestation level: decreased proportions of treated seeds are needed when higher levels of larval infestation are observed, reducing even more the used quantity of active ingredients (0.095 g/ha). Such mixtures of untreated and blend-treated seeds could be used in push-pull strategies as trap crops [163]. This rate of application is as far as we know the lowest ever envisaged for wireworm control.

Control with naturally occurring active substances is possible. Both mortality and morbidity depend on the target species, and the specific effects should be taken into account [164]. Besides, if a biopesticide, intended as a plant-derived component, has a beneficial effect on the control of a pest species, potential phytotoxic effects on specific crops must be tested before any field trial [40].

The bioactive hydrolysis products of glucosinolates (GLs - contained in several Brassica species), particularly isothiocyanates (ITC), can be used to control soil pests and weeds through in-soil incorporation of GLs-containing plant material, a practice known as biofumigation. The relatively rapid sorption and degradation of ITC in the days after incorporation minimizes its persistence and leaching risks, which makes it a promising technique in the aim to reduce reliance on synthetic pesticides [165]. Allyl-isothiocyanate (AITC) has been tested on several important wireworm species. To our knowledge, Tattersfield and Roberts (1920) were the first to show its high respiratory-toxic effect on click beetle larvae, among a wide range of other substances [166]. Laboratory assays led to the assessment of the AITC LC₅₀ for Limonius californicus and showed its lethal and sublethal effects, which in the meantime highlighted the potential of that product for other wireworm species control [21]. The key factor of efficiency is the GL richness of the plant tissue or derived meal used for control [167]. Biocidal plant root systems (Brassica juncea var. ISCI 99) do not necessarily cause significant mortality, while the incorporation of aboveground material at a dosage of 55 t/ha of fresh matter, corresponding to about 290 µmoles of GLs/l of soil, can significantly reduce wireworm populations [78]. In both pot assays and field trials for Agriotes brevis, A. sordidus and A. ustulatus management, the AITC tested sources were chopped Brassica juncea (L.) Czern fresh plants and biofumigant meals derived from defatted seeds of Brassica carinata A. Braun. In pot assays, a clear rate effect was demonstrated, with sufficient seed meal to supply approximately 160 µmoles of GLs/l of soil, resulting in significant wireworm mortality. The effect of the chopped B. juncea plants was less consistent. Accurate incorporation of defatted seed meals resulted in an efficacy level sufficient to protect susceptible crops from damage. Globally, the level of crop protection was comparable to that of conventional insecticide treatments and no evidence of phytotoxicity was noticed [168]. Spotting the appropriate timing for broadcast application is the critical factor here, because the active volatile substances remain available for only 24–36 hours [169]. Successful practical results depend on simultaneously united conditions: (1) suitable GL dosage (the target dose is related to infestation rate and degradation parameters); (2) homogeneous broadcast of biofumigant defatted seed meals, (3) effective and prompt soil incorporation, (4) suitable soil temperature (10.5–16 °C, available in spring and autumn in most European countries) and sufficient soil water content (enzymatic hydrolysis of glucosinolates) and (5) the occurrence of wireworms in the upper soil layers [168,169]. To ensure this, 10–15 days with a soil temperature over 18–20 °C after cold or dry conditions should pass so that the larvae in the little mobile stage in the deeper soil layers can turn into the feeding stage and move upwards looking for food [169]. It may also be possible to improve population management by applying biofumigant defatted seed meals when the first instar larvae are present, which would induce significant reduction in the following year’s population [168,169].

A few other plant-derived active substances affect wireworms. Cinnamaldehyde can significantly reduce damage to mother tubers when applied as a drench at 150 g a.i./ton tubers [170]. The most effective active ingredient in the neem tree Azadirachta indica A. Juss insecticides is azadirachtin. Its repellent effect on the sugarcane wireworm Melanotus communis decreases overtime. Future research determining whether repellency at different rates may be useful in reducing wireworm damage in both sugarcane and other agroecosystems and on other wireworm species [20]. Naturally occurring monoterpenoid essential oil constituents thymol, citronellal, eugenol and rosemary essential oil have been tested for their acute toxicity (LD₅₀ and LC₅₀) against late instars of A. obscurus. Their phytotoxicity was evaluated on corn germination and seedling establishment. Thymol had the greatest contact toxicity, while rosemary oil did not show any. Citronellal had the greatest volatile toxicity, before rosemary oil, thymol and eugenol. Thymol, eugenol and citronellal had significant phytotoxic effects (inhibition of corn seed germination and seedling establishment) and rosemary had only minimal ones. Acceptable qualities of a biopesticide therefore include direct toxicity and repellency effects on the pest. Also, for economic reasons, the best way to use these compounds would be according to their repellency and not their toxicity, therefore, in push-pull strategies [40].

The efficacy of tefluthrin in wheat seed treatments is likely due to a combination of repulsion and short-term morbidity events [162], but although repellency to an insecticide may provide temporary protection to plants and allow stand establishment, wireworms initially repelled may damage crops later in the season. Thus, it is important to observe the direct effects of novel insecticides on them. Chemicals identified as eliciting repellency should be tested to determine whether the attractive cues of host plants dominate repellent stimuli [171]. Moreover, the ability of wireworms to recover from an active substance-induced morbidity may seriously limit the efficacy of the substance, as it has been shown for L. canus and tefluthrin-treated seeds [172]. Their sensitivity to repellent compounds may decrease when repeatedly made moribund. They may even be capable of associative learning [23]. Whenever they are conducted, toxicity trials should include observations on morbidity, and the trials should last until morbidity symptoms cease. Long-term morbidity (lethal, sublethal and behavioral effects) and potential recovery or death of wireworms exposed to certain insecticides have implications on how laboratory and field studies should be designed and interpreted [30,162]. The on-going work on alternative active substances that affect wireworms can benefit lessons from previous research on synthetic chemicals. Table 1 shows insecticides tested against Agriotes wireworms since 1990. Some studies in which identification to the genus wasn’t stated were cited for their impact on research against wireworms. We distinguish between studies with behavioral observations on the larvae in regard to the insecticidal effect and those with field efficiency and crop protection considerations alone. Most of the behavioral effects (repellency, morbidity, sublethal effect, etc.,) concern A. obscurus and should be conducted on other species. Wireworm mobility categories proposed by van Herk and Vernon (2011) should allow comparing active substance efficiencies and species resistance [173]. Side effects on non-target organisms and acute or chronic effects on human health will always have to be investigated [40,169]. For any substance, field-realistic doses should be tested on potentially armed organisms, such as studied on bees for imidacloprid and clothianidin [174]. Globally, considering the overall apparent recovery and avoidance abilities of wireworms, push-pull strategies could be an interesting approach, with the advantage of limiting natural selection [40,175].

The efforts to develop the use of entomopathogenic fungi (EPF) against wireworms are very recent, although these enemies were known since early in the 20th century. To our knowledge, the first report of Metarhizium anisopliae Sorokin as able to infect wireworms in nature was made in 1932 [192]. It can be applied in the field and infect and kill wireworms [193]. Laboratory assays allowed identification of virulent strains of EPF against larvae of A. lineatus, namely V1002 and LRC181A, respectively able to cause 90 and 100% mortality three weeks post-inoculation [194]. Applied before planting in an area with wireworms (95% A. obscurus) as a corn seed treatment alone or in combination with clothianidin or spinosad, the strain F52 can induce significant increases in stand density and stock and foliage area fresh weight yield. Not only does M. anisopliae strain F52 growth observed on retrieved wireworm cadavers link the positive effects on corn with the wireworm control, but it also suggests a synergistic interaction between EPF and insecticides [195], like when the strain is combined with spinosad: A. obscurus and A. lineatus wireworm mortality is higher when treatments are combined than for each treatment applied alone [190]. The efficacy of M. anisopliae can also be enhanced by incorporation of neem seed cakes, as shown for the control of the black vine weevil Otiorhynchus sulcatus (Coleoptera: Curculionidae) [196]. Moreover, because M. anisopliae gradually degrades in the soil, the inclusion of spinosad or other insecticide treatments could extend the total control period provided by one application [190].

Beauveria bassiana (Bals.-Criv.) Vuill. is also efficient, even more with low doses of imidacloprid [170]. Presumably, the low rate (bio)insecticides stress the target or alter their behavior and immune mechanisms in such a way as to make them more susceptible to infection [190,197,198]. Some insecticides may increase pest mobility and increase acquisition of EPF conidia and, since mortality is dose-related, increase mortality [198]. Low levels of reduced-risk pesticides can be combined with biological agents without environmentally harmful traditional pesticide strategies [194,198], which clearly represents potential savings for growers [198].

Further investigations concerning virulent strains of EPF should be undertaken taking into account factors, including temperature, duration of exposure to the studied strain, soil concentration of conidia and food availability. These will affect the levels of mortality and the rate of emigration of wireworms repelled by the fungi [199]. The use of fungus against click beetles is less developed, though Zoophthora elateridiphaga has been found to be an important mortality factor of the click beetle A. sputator in Swiss meadows [200] and strains efficient against larvae have been suggested as potentially equally efficient against click beetles [194].

Entomopathogenic nematodes (EPN) are also efficient against wireworms. Steinernema feltiae and S. glaseri caused a high mortality on the sugar beet wireworm Limonius californicus, however, field experimentations did not show very efficient and economically interesting results. The timing of application in this regard is a very important factor [201]. On Agriotes wireworms, S. feltiae furrow treatments have been tested, but were not efficient [170]. The strain UWS1 of Heterorhabditis bacteriophora Poinar has proven very efficient against A. lineatus (67% mortality three weeks post-inoculation) [194]. Hexamermis spp. is also a known parasite of A. obscurus [202]. Field damage to maize by wireworm can significantly be reduced with applications of H. bacteriophora and Steinernema species [203]. The most efficient strains among those screened by Ansari et al. (2009) were local and originated directly from wireworm cadavers [194]. Where IPM is considered, native strains that co-evoluted with a target pest, and therefore adapted to local conditions, should be screened for their efficiency. More than identifying perfectly adapted strains less risky as to their impact on non-target organisms, this helps outline their optimal development conditions. This was undertaken for Spanish strains of S. feltiae, even if, in this case, the most efficient one only led to a 7% larval mortality on A. sordidus [204]. Any potential remains interesting because of possible synergies [198], such as observed with M. anisopliae in the control of the pest Hoplia philanthus Fuessly (Coleoptera: Scarabaeidae) and the black vine weevil O. sulcatus [205,206]. Synergies could equally work against wireworms and offer an organic or chemical-free approach in the control of A. lineatus [194] and potentially other Agriotes species.

The mechanisms underpinning synergy between agents remain unclear. Furthermore, not all interactions are synergistic. Antagonistic or barely additive effects have been observed in a wide range of situations involving other pests, strains and pesticides. The outcomes appear to be dependent on factors, including EPF strain (a virulent strain is vitally important), dose and the synergist type or efficacy enhancing agent used, which overall underlines the need to identify and optimize synergies [198]. Differences in wireworm species susceptibility to EPN exist and may be due to a range of physical barriers, including morphological barriers or behavioral responses to attacks [207]. Once more, this highlights the necessity to consider the overall pest complex case by case.

Comparisons between existing management options is necessary, whether they concern different treatments or consideration of potential control agents in place. The outcomes of such comparisons may differ drastically between agronomic situations. For example, natural infestations of Metarhizium spp., Beauveria spp. and nematodes should be evaluated when studying any treatment, as was done for biocidal seed meals in a field trial with rearing cages (A. sordidus and A. ustulatus) in different crop rotation systems. In this case, their efficiency was overall low and inferior to that of GL-containing plant material and did not differ between rotations [78], but one can imagine the biases caused by omitting their roles in the results interpretation.

Searching among dead wireworms collected in Germany in the aim to find new entomopathogenic taxa, an intracellular bacterium was recently identified: a pathogen belonging to the genus Rickettsiella [208,209]. Its suitability for IPM strategies hasn’t been tested yet. The insecticidal potential of the bacterial flora of A. lineatus and other hosts was also studied. Of the isolated strains, five induced 100% mortality on third instar A. lineatus larvae 10 days after treatment, three of them belonging to the wireworm flora, highlighting their microbial control agents’ potential [210]. Among potentially interesting predators of wireworms, some Carabidae and Staphylinidae (Coleoptera) are known to feed on A. sputator [211], but their use as control agents is inexistent. Other predators or parasitic pathogens should be researched, for example, in systems where everything points to wireworm infestations without occurrence. Then again, such natural control is possible in very diversified agricultural systems, where there is room for species other than crops and their pests. As stated above, grassy landscape margins would be key factors explaining wireworm potato damage. Their role for the maintenance of natural pest control should be investigated [101], especially since weed spots are key landscape elements for oviposition [103].