Identification of Agriotes wireworms – Are they always what they appear to be?

Identifikation von Agriotes-Drahtwürmern – Sind sie immer, was sie zu sein scheinen?

Material and methods

The field-collected wireworms (Tab. 1 and 2) originated mainly from a meadow near the Julius Kühn-Institut (Braunschweig, Messeweg 11/12), and were collected in 2011. The specimens from Tab. 3 were collected from potato fields in the vicinity of Uelzen (Lower Saxony), from potato fields in Deggendorf, Olching and Klotzau (Bavaria) and from a wheat field in the region of Tübingen (Baden-Württemberg) and were collected in 2013.

The specimens listed in Tab. 4 and 5 were not collected in the field. They were bred from known parents (adult beetles collected from the field) at the JKI wireworm breeding, which is described in more detail below, and from the wireworm breeding at Agroscope Reckenholz, Zürich, Switzerland (parents A. obscurus from Bärau, Berner Oberland and A. lineatus from Zürich, method after Kölliker et al., 2009).

The specimens from Tab. 6 were from six different sites (Neubörger, Dasselsbruch, Westenholz from Lower Saxony in 2011, Münzenberg from Hesse in 2011, Hildburghausen and Oldisleben from Thuringia in 2013) from the nationwide click beetle and wireworm monitoring (see Lehmhus, 2012, 2014).

Further tests were performed with about 40 morphologically unidentifiable Agriotes wireworms from field sites in the vicinity of Braunschweig.

To establish if there might be any indications of differences in feeding preferences from different species, wireworms from eight different potato fields were identified (Fig. 6) (Deggendorf, Olching and Klotzau in Bavaria, Natendorf, Oldendorf, Uelzen-Barnstedt, Uelzen-Borg and Zargleben in Lower Saxony).

Fig. 6. Results of wireworm identification from samples from potato. Identification by mor­phology, of problematic specimens additionally by PCR. Some damaged specimens could not be identified by PCR. Four sites close to Uelzen/Lower Saxony (upper row), three sites in Ba­varia (lower row), one site eastern Lower Sax­ony (Zargleben, lower row, right). n = number of wireworms.

After the morphological determination of the wireworms, parts of the worm were extracted and the DNA was extracted using the Qiagen DNA columns (DNeasy Plant mini Kit) according to the manufacturer’s instructions (Qiagen, Hilden). The DNA was collected and stored frozen.

A standard PCR was performed in a thermo cycler. The temperature settings were: 3´ at 96°C, 35 cycles: 30´´ at 96°C, 30´´ at 63°C, 30´´ at 72°C, and finally 5´ at 72°C. Taq-polymerase (0.5 units; …..) was used with an appropriate nucleotide mixture. The PCR-product was then separated on a 0.7% agarose gel and photographed.

The three primer pairs were used in a 10 μMol concentration and added to the reaction mix prior to the PCR-amplification step.

Results

Fig. 1. Granules between the legs (left) and at the base of abdominal segments (right) identify Agriotes sputator. Photographed specimen from the wi­reworm rearing at JKI.

Fig. 2. No granules between the legs (left) and no granules at the base of abdominal segments (right) indicate that this is either A. lineatus or A. obscurus. Photo­graphed specimen from the wireworm rearing at JKI, in this case A. obscurus.

When determining live wireworms compared to wireworms killed prior to their identification, a higher error rate occurred in the live, and therefore moving wireworms (Tab. 1). In contrast, morphological identification of dead larvae appeared straightforward following the two keys (Klausnitzer, 1994; Cocquempot et al., 1999). Using PCR it was difficult to assign wireworms to a single species for unknown reasons. Here, 65.6% (40 out of 61 wireworms) were identified as A. sputator by PCR and morphological identification. 14.8% (nine wireworms) were morphologically identified as A. sputator but no clear results were obtained by PCR. Seven specimens were morphologically identified as A. sputator when alive but as other species when dead.

A subsample of 12 of these morphologically identified larvae was analyzed by a further laboratory (LTZ Augustenberg) using the same PCR-method. For this subsample the results obtained for A. sputator were similar for all institutions and methods while the results for other species differed. Although the PCR-protocol used followed Staudacher et al. (2011) in both laboratories, cross-reactions with primers for two species (in most cases A. lineatus/A. obscurus) were again observed in the JKI labora­tory but not in the LTZ Augustenberg laboratory though some wireworms here also showed a very weak reaction with the primers for a 2^nd species.

Further PCR tests had the same result with cross reactions occurring in about 40 specimens from the vicinity of Braunschweig that were morphologically either A. lineatus or A. obscurus, but were not definitely identifiable due to abrasion. With the PCR following the method by Staudacher et al. (2011) each individual was assigned to both species. The reason remained unknown. Therefore, the primer development as the following step was undertaken for Agriotes lineatus and Agriotes obscurus.

Fig. 3. Multiplex-PCR with the three Agriotes species ob­scurus (1, 146 bp), lineatus (2, 289 bp) and sputa­tor (3, 289 bp) in two repeats (a, b). K represents the water control and S the size standard (100 bp ladder). There is no cross-reaction occurring bet­ween the single species, when applying the multip­lex-PCR. The size of the PCR-fragments obtained from the species can easily be differentiated.

Agriotes lineatus is lacking the smaller seta of two setae close to the stigma in abdominal segments 1–8 while A. obscurus (and also A. sputator) shows this seta (see Fig. 5) according to the key provided by Cocquempot et al. (1999). Additionally the angle of the small “substitute” tooth to the tip of the mandible is narrower in A. lineatus than in A. obscures (Fig. 4), but the difference appears often less marked than suggested in the identification keys using this trait (Schaerffenberg, 1940; Korschefsky, 1941; Klausnitzer, 1994).

Fig. 5. Setae above the stigma of A. lineatus (left) and A. obscurus (right). Apart from the large seta to the right of the stigma a fairly small, sometimes stubby seta is found above the stigma in A. obscurus and A. sputator. This seta is supposed to be absent in A. lineatus according to Cocquempot et al. (1999). Photographed specimens came from the wireworm breeding at JKI.

Fig. 4. “Substitute” tooth angle of A. lineatus (left) and A. obscurus (right). The angle is narrower in A. linea­tus than in A. obscurus, but the difference appears often less marked than suggested in the identifica­tion keys using this trait (Schaerffenberg, 1940; Korschefsky, 1941; Klausnitzer, 1994). Specimens from the wireworm rearing at JKI.

The 48 specimens from Tab. 3 were collected from potato fields in the vicinity of Uelzen (Lower Saxony), from potato fields in Deggendorf, Olching and Klotzau (Bavaria) and from a wheat field in the region of Tübingen (Baden-Württemberg).

In 21 of the 48 specimens (43.8%) no morphological identification was possible following the mandible, compared to only two specimens (4.2%), where no morphological identification was possible following the setae. However, in 9 cases (18.8%) differences between morphological identification and identification by PCR were observed. 10 of these wireworms (20.8%) did not react to any of the primers, although it was ensured that DNA was in the samples.

To test the accuracy of the results obtained, 31 larvae with known parentage (11 A. lineatus, 10 A. obscurus and 10 A. sputator) from the wireworm breeding at the JKI, established in 2013, were subjected to morphological and molecular species determination. Their parentage was known as they were offspring from adult click beetles caught at a single field site (Schandelaher Wohld) (methods according to Kölliker et al., 2009, modified). These click beetles were morphologically identified following Lohse (1979) and Laibner (2000).

The wireworm species identification via PCR confirmed all larvae to be the species expected from the parents used in the breeding. The morphological identification of the wireworms was less straightforward. Concerning the setae in the abdominal segments already two specimens of A. obscurus occurred in this sample where morphological identification was difficult. Furthermore, in one specimen of A. lineatus morphological identification following the setae would have resulted in identification as A. obscurus (Tab. 4).

For further investigation into the frequency of aberrant individuals in populations of the three species, a number of individuals of each species A. lineatus, A. obscurus and A. sputator from differend populations from wireworm breedings at the JKI and the Agroscope Reckenholz were examined morphologically. The wireworm specimens from JKI being alive, the angle of the substitute tooth could be examined only in those opening their mandibles. The results concerning the frequency of aberrant individuals from these populations of the three species are given below (Tab. 5). The average angle of the “substitute” tooth to the tip of the mandible differs between species, but it was not possible to safely determine the species from the mandible. The angle differed markedly even between individuals of the same species, while there was considerable overlap between species. The second dis­tinguishing trait used was the granulation between legs and at the base of each segment, which only 2% of the A. sputator did not show. The third trait was the small seta that should be present in A. obscurus (and A. sputator), but absent in A. lineatus. Contrary to the key, this seta was absent in some A. obscurus and present in some A. lineatus in all populations.

These data originated only from populations from 3 field sites for A. obscurus, from 2 field sites for A. lineatus and from 1 field site for A. sputator. In other populations of these three species the percentage of aberrant specimens may be different. Therefore, in a comparison of the PCR and morphological determination in different Agriotes populations 60 alcohol-conserved specimens from six different sites, collected in the German click beetle monitoring, were examined. In this assessment wireworms were identified morphologically with the key of Cocquempot et al. (1999). For the PCR the newly developed primers were used. In 50 cases congruent results were found between PCR and morphological identification, but in 10 cases results differed. Three (= 5% of the 60 specimens) morphologically not reliably identifiable specimens due to extreme abrasion (mandibles reduced to stumps and most setae lacking) were identified as A. obscurus via PCR. One wireworm (= 1.7% of the 60 specimens) appeared to be A. obscurus morphologically, showing no obvious granules between legs and at the base of abdominal segments 1–8, but was assigned to A. sputator by PCR. Six wireworms (= 10%) were identified as A. lineatus via PCR, but had been assigned morphologically to the species A. obscurus due to the occurrence of the small second seta above the stigma of the abdominal segments 1–8. Three of these aberrant A. lineatus specimens were from a single site (Westenholz, Lower Saxony) while the others came from different sites. Out of six different sites, four had morphologically aberrant wireworms even in this comparatively small number of samples.

Some further A. lineatus populations were already tested from three sites in Lower Saxony: Uelzen (PCR 41 A. lineatus, 14.6% with 2nd seta on more than two segments, 31.7% with 2^nd seta on at least one segment), Dasselsbruch (PCR 14 A. lineatus, 21.4% with 2^nd seta on all segments, 50% with 2^nd seta on at least one segment) and Vahlde (PCR 13 A. lineatus, one individual, 7.7% with 2^nd seta on all segments).

The importance of correct species identification is illustrated by the species spectrum of wireworms from eight different sites in potato. Substantial numbers of wireworms were collected in potato from eight sites (four sites close to Uelzen/Lower Saxony, three sites in Bavaria, one site in eastern Lower Saxony) to determine whether a species preference for a certain crop might occur. The identification was done mainly morphologically, but problematic specimens were additionally tested by PCR. These data suggest that there may indeed be preferences of certain species to certain crop situations, although some wireworms were too severely damaged during the potato harvest or during postal transport to be assigned to species. They were only identified as A. lineatus/A. obscurus. The data showed some similarities for six sites with the dominance of A. obscurus and occurrence of two to three further species, while the two other sites differed markedly (Fig. 6).

Since different Agriotes species can show differences in their biology and therefore can differ in their impact on a crop (Furlan, 2014), correct identification of species is important in practice. But if the commonly used keys do not result in correct identification of Agriotes species, the clarification of which species cause damage in which crop becomes problematic. In addition, in the majority of studies on Agriotes ecology and behaviour field-collected larvae are used. The outcome of experiments may be seriously affected by the difficulties in morphological identification of these cryptic pests. Therefore, morphological and molecular determination of Agriotes wireworms was compared here directly.

The morphological identification traits in Agriotes are vulnerable and may in part be lost. We consider the likely explanation an abrasion caused by the wireworm movement in the soil. With a range of 60°–120° for the mandible tooth angle for each of the three species, it seems that abrasion of the mandible can alter the angle in both directions, making it either narrower or broader. Abrasion makes this trait so variable that differences between species emerge only in the average angle (e.g. for Braunschweig 2013: A. sputator 94,8°, A. obscurus 100,6°, A. lineatus 78,3°). The differences between the angles were much less marked than the angles given in Klausnitzer (1994), i.e., A. sputator 90°, A. obscurus 120°, A. lineatus 60°. This trait is therefore only usable as additional information but not as a key feature for identification.

Though the setae are less vulnerable to abrasion, they also may break due to movements through the soil. If this happens it may be possible to see the socket where the seta is inserted, but only under good magnification. Abrasion is therefore one reason for specimens being not reliably identifiable.

Yet not all difficulties are explained by abrasion or loss of setae. The regular occurrence of individuals showing mixed traits, e.g. a mandible like A. lineatus and the setae pattern of A. obscurus (with the additional small seta) or vice versa, complicates the morphological differentiation. The occurrence of A. lineatus specimens showing the additional small seta clearly demonstrates that not all identification errors occur due to abrasion of important key features (Tab. 3–6). The PCR to support the morphological identification results was used to differentiate between the Agriotes specimens via genetic differences. The PCR results (following Staudacher et al., 2011, and using self-developed primers) clearly show that morphological identification of wireworms of the genus Agriotes is not always possible.

In general, the morphological key provided by Cocquempot et al. (1999) for Agriotes wireworms from France is also correct for most individuals of the same species in Germany. But the results showed that there are aberrant individuals that have morphological characters contradicting the results produced by PCR. When used on wireworms from the JKI rearing, the PCR corresponded with the morphological species identification of the parents (= adult click beetles). Thus it is assumed that the PCR with the newly developed primers produces the correct species identification. The populations of the three most common wireworm species (A. sputator, A. obscurus and A. lineatus from the wireworm breeding) originating from sites in Northern Germany and in Switzerland all held aberrant specimens. In these populations a range from 2.0% (A. sputator) up to 15.9% (A. obscurus) of the individuals, depending on species, would have been misidentified using the key by Cocquempot et al. (1999). In smaller samples from other sites this percentage varied further and was sometimes either higher or lower. Morphological identification of A. sputator may be easier than of A. lineatus and A. obscurus, because the granules at the base of the segments appear to be less vulnerable to abrasion than setae or mandible teeth. The latter trait appears to be the most sensitive and therefore the least reliable trait.

The data from further wireworm monitoring sites indicate that the percentage of aberrant specimens may vary between populations from different sites, but also that the majority of sites holds aberrant specimens. The frequency of misidentifications may therefore strongly vary between different sites. Others, e.g. Jossi et al. (2008) and Staudacher et al. (2011) mention similar difficulties with morphological determination of these three important species in their larval stage. Therefore, at the moment there is no morphological key to reliably differentiate between all specimens of these Agriotes larvae. A variable portion of the larvae, depending on the location, will be misidentified just relying on morphological keys. Nevertheless the majority of larvae are correctly identified with morphological keys.

However, when exact identification of each Agriotes specimen is needed, it is necessary to combine the morphological differentiation with the now developed PCR-detection. Our newly synthesized primers are suitable to distinguish all three Agriotes species in a combined (multiplex) PCR, where all three primer pairs have been applied in one reaction mix. With the new primers, the difficulties in differentiating between A. lineatus and A. obscurus encountered before with the primers given by Staudacher et al. (2011) did not occur.

Precise identification is crucial when trying to determine whether there are ecological differences, e.g. in food preferences of these three widespread species. Ellis et al. (2009) and Lole (2010) conclude similarly that stating ecological differences between these three species is hampered by the lack of a method to distinguish between the wireworms merely by morphological (physical) features, and they emphasize the necessity of PCR-methods for precise identification.

Frequent determination errors, together with species differences in their effects on a given crop, would make a prognosis of damage difficult. For example, Parker and Howard (2001) state that the effort to predict the likely level of damage to potatoes caused by a given wireworm population has not been consistently successful. This may in part also be due to incorrect wireworm species identification. The wireworm species spectrum from eight different potato field sites indicates a possible preference of A. obscurus for the potato crop even though only part of the wireworms was identified via PCR. Similarly, Kempkens et al. (2004) also found a dominance of A. obscurus (above 90%) and 9% Hemicrepidus niger in potato. But in their work only morphological identification after Klausnitzer (1994) was performed, indicating that part of the Agriotes obscurus might have belonged to other Agriotes species.

In view of the results presented here, this work will be continued with further wireworm populations from the German click beetle and wireworm monitoring, including further wireworm samples from additional crops.