Conservation In Practice |
Corresponding author: Emanuela Maurizi ( emanuela.maurizi@uniroma3.it ) Academic editor: Franco Mason
© 2017 Emanuela Maurizi, Alessandro Campanaro, Stefano Chiari, Michela Maura, Fabio Mosconi, Simone Sabatelli, Agnese Zauli, Paolo Audisio, Giuseppe Maria Carpaneto.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Citation:
Maurizi E, Campanaro A, Chiari S, Maura M, Mosconi F, Sabatelli S, Zauli A, Audisio P, Carpaneto GM (2017) Guidelines for the monitoring of Osmoderma eremita and closely related species. In: Carpaneto GM, Audisio P, Bologna MA, Roversi PF, Mason F (Eds) Guidelines for the Monitoring of the Saproxylic Beetles protected in Europe. Nature Conservation 20: 79-128. https://doi.org/10.3897/natureconservation.20.12658
|
Osmoderma eremita (Scopoli, 1763) is a saproxylic scarab beetle protected by the Habitats Directive in the European Union. The present paper is part of a special issue on monitoring of saproxylic beetles protected in Europe and starts with a revision of the current knowledge on systematics, ecology, ethology and conservation of O. eremita and its allied species, followed by experimental tests of different methods for monitoring its populations. Two methods were compared in several localities of central Italy: (1) the widely used pitfall traps into tree cavities and (2) black cross windows traps baited with a specific pheromone produced by male beetles. The first method, often used in northern and central Europe, did not give acceptable results in Italy probably because of the scarcity of veteran trees with large hollows. It could only be used successfully in areas where: 1) tree hollows were abundant, large enough and with sufficient amounts of wood mould for planting pitfall traps and 2) the team is composed of several operators in order to ensure the checking of at least 150 traps every two days during the whole period of mating activities (15 July–25 August). The second method, consisting of hanging 30 black cross window traps during the mating period and checking them every two days, turned out to be better for capturing a significant number of individuals but cannot be used every year because of the possible disturbance on mating activities of the species.
Saproxylic beetles, old-growth forests, hollow trees, dead wood, trap sampling, Habitats Directive
The hermit beetle, Osmoderma eremita (Scopoli, 1763), is a large saproxylic chafer (Coleoptera: Scarabaeidae: Cetoniinae) associated with hollow veteran trees of the European broadleaf woodlands. It is included in Annexes II and IV of the Habitats Directive (Council Directive 92/43/EEC of 21 May 1992 on the conservation of natural habitats and of wild fauna and flora). The Habitats Directive requires that Member States prepare, every six years, a report on the conservation status of the species listed in the Annexes. In order to facilitate the answer to this obligation, the Life Project “Monitoring of insects with public participation” (LIFE11 NAT/IT/000252: hereafter, MIPP) conducted experimental fieldwork to develop a standardised method for monitoring the saproxylic beetle species for the project: Osmoderma eremita (hermit beetle, Scarabaeidae), Lucanus cervus (European stag beetle, Lucanidae), Cerambyx cerdo (great capricorn beetle, Cerambycidae), Rosalia alpina (rosalia longicorn, Cerambycidae) and Morimus asper/funereus (morimus longicorn, Cerambycidae) (
The present paper is part of a special issue on the results for monitoring saproxylic beetles which are protected in Europe and it is dedicated to Osmoderma eremita. The paper starts with a significant revision of the current knowledge on systematics, distribution, ecology, ethology and conservation of Osmoderma eremita and allied species, mostly derived from a subsequent split of the species and hence worthy of the same protection level. Such a review is followed by the experimental test of methods for monitoring its populations.
The genus Osmoderma, established in 1828 by Le Peletier de Saint-Fargeau and Audinet-Serville, is currently assigned to the superfamily Scarabaeoidea, family Scarabaeidae (scarab beetles) and subfamily Cetoniinae (fruit or flower chafers). Nevertheless, in many European scientific papers and databases produced in the second half of the last century, the subfamily was raised to family rank and Osmoderma was reported as a genus of the family Cetoniidae. This change in systematic arrangement caused some confusion in the documents of the Habitats Directive and is explained by the fact that some specialists, (i.e. authors of the most important works used for identification of the European scarab beetles (
Within the subfamily Cetoniinae, traditional taxonomy assigned the genus Osmoderma to the subtribe Osmodermina (two genera) of the tribe Trichiini, a large clade of Cetoniinae that includes the widespread genera Trichius, Gnorimus and others (Howden 1964, Krikken 1978,
At global level, the genus includes at least 12 species throughout the Holarctic Region (Howden 1964,
The North American sector includes mainly the potential range of the US eastern broadleaf deciduous forests and extends southwards up to the 35th parallel, where two species occur sympatrically: O. eremicola (Knoch, 1801) and O. scabrum (Palisot de Beauvois, 1805). A third species, O. subplanatum Casey, 1915, unexpectedly lives in the Great Plains, probably restricted to fragmented woodlands along lake shores and the river basins of the Mississipi and Missouri. All three of them overrun the southern border of Canada to a limited extent.
The three species living in the Asian Far East have recently been revised by
In order to address the great confusion in taxonomy and nomenclature of the European hermit beetles, a DNA-approach of species delimitation was attempted by
Osmoderma eremita (Scopoli, 1763) in most of western Europe;
Osmoderma barnabita Motschulsky, 1845 (sensu
Osmoderma lassallei Baraud and Tauzin, 1991, endemic to northern Greece and European Turkey;
Osmoderma cristinae Sparacio, 1994, endemic to Sicily.
Genetic distance and parsimony analysis well supported the delimitation of two clusters, each formed by two species: the first cluster is restricted to western Europe (O. eremita and O. cristinae) and the second one to eastern Europe (O. barnabita and O. lassallei) (Figure
Distribution map of the genus Osmoderma in Europe (from
The morphological traits used for delimiting the above mentioned species and subspecies are poorly marked and subject to a significant variation. They mainly concern the reliefs on head and pronotum, as well as the shape of male genitalia (Figures
Habitus of Osmoderma in dorsal view. A O. eremita male B O. eremita female C O. cristinae male D O. cristinae female (photographs by A. Ballerio and M. Uliana, from
Lateral side of parameres, in O. eremita (A) and O. barnabita (B). The arrows indicate the lack or the presence of concavity under the protuberance (From
An integrated approach to the Italian populations, based on morphological and molecular analyses (
Additional poorly investigated taxa occur from the Balkan peninsula to the Caucasus mountains (apart from O. lassallei and O. barnabita): at least one undescribed species or semispecies, related to O. lassallei, in southern Greece (P. Audisio, P. Petrakis and G.M. Carpaneto, unpublished data); the rare O. brevipenne Pic, 1904, described from the Taurus Mountains, southern Turkey; O. richteri Medvedev, 1953, from Georgia, Caucasus (only two females captured until now).
The genus Osmoderma includes the largest fruit chafers (Cetoniinae) of the European fauna (body length up to 36mm, head included; greatest width of elytra up to 19mm). The head is small in comparison with the body, with a clypeus broader than long. Antenna is 10-segmented with a small terminal club formed by three short lamellae. The body is wide, suboval and more or less flattened dorsally. The prothorax is significantly narrower than the bases of the elytra, more or less convex, with lateral sides rounded or sharp-cornered just before the middle. The legs are strong, with moderately long tibiae and short tarsi. Fore tibiae have three distinct teeth on their external side. The wings are well developed in both sexes. Colour is black to piceus or chestnut, often shiny with cupreous or green metallic lustre.
In the European species, males have a clypeus concave, more or less reflexed at apex, a slightly convex clypeo-frontal transition area and two lateral elevations above the antennal insertion that look like two horns. Females have a clypeus flat and not reflexed at apex, a tumid clypeo-frontal transition area and lack elevations at sides. Head surface is punctured in males and rugose in females. The pronotum of males has distinct basal angles, with lateral sides obtuse-cornered just before the middle; it is about one-third wider than long, strongly convex, with two distinct longitudinal carinae, parallel and separated by a more or less deep groove. The pronotum of females has basal angles rounded and an overall rounded outline, it is less convex than males and shows only vestigial ridges, almost vanished or reduced to two tubercles. Pygidium is large and strongly convex in males, smaller and slightly convex in female.
As already mentioned in the previous paragraph, the morphological traits used for delimiting the European species and subspecies are poorly marked and subject to significant variations. The morphological identification of Osmoderma is a task for specialists who have examined many specimens from different locations (in museums or other collections) and progressively acquired a deep knowledge of the intraspecific variation within single populations and throughout the geographic range of the species. Another problem is that the specimens occurring in many collections and museums cannot be used as a tool for making comparisons as they have often been improperly identified, owing to the extensive taxonomic confusion in the past decades. Luckily, the European species have different ranges and, in most cases, the populations can be simply assigned to a species by the locality. The problem arises when a locality is situated in the overlapping area between the ranges of two parapatric species, as in the case of O. eremita and O. barnabita in central Europe and the Balkans, where the possibility cannot be ruled out that they mated with each other. The less variable trait for separation of these two species is perhaps the shape of the parameres of the male genital organ (
1. | Parameres, on lateral side, show a strong upward protuberance like a triangular bump rounded at apex, followed by an almost straight or very slightly convex slope continuing up to their base. | O. eremita |
2. | Parameres, on lateral side, show a very strong upward protuberance like a square bump rectangular at apex, followed by a strong concavity and then a straight slope continuing up to their base. | O. barnabita |
The male genitalia also allow the separation of O. eremita and O. barnabita from O. cristinae, as the latter does not show a real upward extension on parameres but only a moderate swelling followed by a straight or very slightly convex slope continuing up to the base (Figure
Lateral side of parameres, in O. eremita (A–C) and O. cristinae (D–F), showing intraspecific variation and difference on the parameres between the two species (photo by Federico Romiti).
All morphological characters used to separate the European species of Osmoderma are quantitative and change between them without a sharp difference; furthermore, they are subjected to significant variation and cannot give certainty in identification. The sexual differences in the O. barnabita/lassallei-group are less prominent than in the O. eremita/cristinae-group.
Non-expert operators engaged in searching the remains of Osmoderma in wood mould inside tree cavities, may find it hard to distinguish them from those of Gnorimus variabilis (Linnaeus, 1758). However, the remains of the latter are clearly smaller and slender, with proportionately longer and fine tarsi and middle tibiae are strongly recurved in males. Moreover, the outer side of the fore tibiae have two well developed teeth (instead of three teeth in Osmoderma), while the scutellum is shorter and half-circle shaped (instead of the almost isosceles triangle-shaped scutellum of Osmoderma). Elytra of G. variabilis can have whitish spots on the dorsal surface.
The eggs of Osmoderma are globe-like and white, with a diameter of 4-5mm. Larvae are classified in the scarabaeiform type and usually named ‘white grubs’ or ‘curl grubs’ by applied entomologists. They are characterised by a stout and downward projecting (hypognathous) head and a large fat, white or blueish, C-shaped body (Hayes 1929,
Larva of Osmoderma eremita. A Habitus and B Raster (ventral spiny area of the last abdominal segment). Note the lack of the palidia (the two longitudinal series of spinules occurring in the raster of most fruit chafers) and the end of the legs provided with a short and stout claw (see text for detailed explanation). (A photo by Sonia Dourlot, from
Larval characters are only used for the identification of wholly developed larvae (mainly belonging to the third stage, i.e. the last phase of their life cycle). A good combination of characters for separating hermit beetle larvae from those of all other European Cetoniinae is the following: 1) the absence of palidia across the raster, 2) the short, blunted and subconical terminal claw of tarsus and 3) the body size between 65 and 75mm, considering that the large larvae of Protaetia speciosissima are never longer than 65mm (Korschefsky 1940). In fact, almost all European saproxylophagous fruit chafers which live in dead wood (i.e. Cetonia, Protaetia and Gnorimus) have distinct palidia and legs which end without a claw or with a normally elongate and slender claw (never thick, subconical and blunted) (Figure
Shape of the terminal claw in European fruits chafer larvae. A leg of Osmoderma eremita B last two segments and claw of O. eremita C idem of Gnorimus variabilis D idem in Trichius fasciatus E last three segments in Cetonia aurata F last three segments in Protaetia cuprea (outlined by Federico Romiti).
The following key, mainly valid for third instar larvae was constructed by one of us (G. Carpaneto) and was based on information drawn from several authors (Hayes 1929, Böving and Craighead 1931, Korschefsky1940,
1. | Tarsi ending with a claw (long or short, sharpened or blunted) | 2 |
– | Tarsi rounded apically, without claws | 4 |
2. | Tarsal claws slender, more or less sharpened and recurvate. Ninth and tenth abdominal terga separated in two single dorsal units. Raster with or without palidia. Anterior margin of labrum not distinctly trilobed, almost straight or slightly emarginate in the middle. Larvae in dead wood, up to 60mm | 3 |
– | Tarsal claws neither sharpened nor recurvate, but like a black, short and thick conus with a large base and a blunted apex. Ninth and tenth abdominal terga fused into a single dorsal unit. Raster without palidia. Anterior margin of labrum distinctly trilobed. Larvae in dead wood, up to 75mm long | Osmoderma |
3. | Raster with two distinct palidia forming an oval Figure. Larvae in dead wood, up to 50 or 60mm long, according to the species | Gnorimus |
_ | Raster without palidia. Last abdominal segment normally shaped. Larvae in dead wood, up to 40mm long | Trichius |
_ | Raster without palidia. Last abdominal segment ends with a pair of rounded areas delimited by a slight wrinkle. Larvae in dead wood, up to 25mm | Valgus |
4. | Larvae small (up to 25mm), amongst the roots of herbaceous plants | Oxythyrea and Tropinota |
– | Larvae large (25-65mm), amongst plant roots, humus and dead wood, according to species | 5 |
5. | Raster with two well distinct palidia, each formed by18-28 spinules (mostly 22-24). Larvae up to 45mm long | Cetonia |
– | Raster with two well distinct palidia, usually formed by less than 22 spinules (usually 14-20, exceptionally up to 26 in the rare P. fieberi). Larvae, up to 65mm long | Protaetia and Aethiessa |
Non-expert operators may have difficulty in recognising the larvae of hermit beetles from those of the other two large and widespread scarab beetles occurring in deadwood i.e.Oryctes (Scarabaeidae Dynastinae) and Lucanus (Lucanidae Lucaninae). Apart from microscopy characters involving antennae and mouth parts, often very hard to detect during the fieldwork, the larva of Oryctes is easily distinguished because its last abdominal segment is divided into two halves by a transverse narrowing, so that, in dorsal view, the abdomen seems to be formed by ten clearly observed segments. On the other hand, the larva of Lucanus is immediately recognisable by the abdominal terga not plicate, the longitudinal anal slit (instead of the transverse anal slit of Osmoderma and other Cetoniinae), the large but much narrower and elongate abdominal spiracles (see Bardiani et al. 2017) and the stridulating organs consisting of a series of teeth on second and third pairs of legs (instead of being present on the ventral side of mandibles and dorsal side of maxillae as in Cetoniinae).
The freshly laid egg of Osmoderma is globe-like and opaque white but, after 2-3 weeks, it becomes yellowish and doubles its size, reaching a diameter of 4-5mm. Each female lays 20-80 eggs deeply in a tree cavity, between the inner wall and the wood mould, protected by a flexible coating of wood pulp and a variable number of larvae (from 12 to 18) hatch after 14-20 days. The newly hatched larvae measure just 6mm but they grow quickly, feeding on decaying wood and its mixture of rot-causing fungi. Larvae go through three different instars (L1, L2 and L3) across a developing time of 2-4 years and overwinter at their first or second instar, depending on the oviposition date and local microclimatic conditions (Tauzin 1994,
At the end of the second or third summer, usually in September, when the third instar larva has reached its full development (75mm long), it builds an oval-shaped cocoon by clumping fine wood debris with its salivary and faecal secretions. Within this protecting case, the larva spends its last winter and then transforms into a pupa in the spring (normally in May) and into an adult (June) (Tauzin 1994). The emergence of the adults occurs in July and in early August, when many individuals show a shiny integument suggesting they had just emerged from the cocoon.
The seasonal activity of the adults is correlated with climatic factors, such as altitude, latitude and habitat type. In Sweden, the adults are active from July to September, while in central and southern Europe, there have been several observations in May-June and even a few at the end of April (
One or two days after their emergence, the males of the genus Osmoderma start to emit a strong odour which can be perceived by humans at a distance of several metres, reported by French entomologists of the 19th century as “cuir de Russie”, namely Russian leather (
Many authors have indicated this scent to be similar to fermented fruit (peaches, plums or apricots) on which the adults were reported to feed. The French vernacular name “pique-prune” (= who picks the plum) (
The primary habitat of hermit beetles consisted of broadleaf old-growth forests with an abundance of hollow trees and dead wood. In particular, the species seems to prefer areas where the canopy is not dense because of the age of the plants and the occurrence of open space left by the fall of veteran trees. In fact, according to many literature data and the authors’ long personal experience, hermit beetles are saproxylophagous species whose larvae meet their habitat and food requirements in tree cavities rich in wood mould. Nevertheless, as it is more than a strictly forest dwelling species, the hermit beetle seems to be an ecotonal species of the forest edge and clearings which can be found in any context where there are old hollow trees, not only in mature forests but also in the agricultural and urban landscapes. In contrast to logs and snags, live hollow trees provide long-lasting resources (
The altitudinal range of the hermit beetle varies from sea level (evergreen forest dominated by holm oak, Quercus ilex or cork oak, Q. suber) up to almost 1400m (mountain forests of beech tree, Fagus sylvatica), through all intermediate altitudinal belts (covered by mixed deciduous forests and woodlands of Quercus spp., Acer spp., Fraxinus spp., Tilia spp., Ulmus spp., Carpinus spp., Salix spp., Populus spp., Betula spp., Alnus spp., Castanea sativa etc.) (
Hermit beetle larvae usually occur in hollow but still living, standing trees but they have also been found in dead standing trees as long as the dead wood is able to retain a suitable degree of moisture (Rainus et al. 2005). More rarely, the species has been observed in felled tree trunks or in stubs but it mainly occurs after a recent cut. The age of habitat trees suitable for hermit beetles depends on the features of wood. Fast-growing species such as poplars (Populus spp.) and willows (Salix spp.) may harbour hermit beetles when they are only a few decades old (
Despite widespread opinion supporting the extreme polyphagy of hermit beetles, detailed studies by larval sampling in wood mould showed that these insects have preferred tree species at the local level. The northern-most populations of Osmoderma (O. eremita in Sweden and O. barnabita in Finland) exploit mainly the huge trees of pedunculated oaks (Quercus robur) which form small stands in the agricultural landscape (
Despite being able to fly, hermit beetles show a low dispersion distance. In Sweden, O. eremita showed a limited dispersal rate and mean dispersal distances of less than 200m, with only 15% of the adults leaving their natal hollow tree (
Tethered flight experiments were conducted in the laboratory by
As the daily activity of hermit beetles varies according to light, temperature and humidity, they have been recorded as active both day and night by different studies. In Sweden, during a long term research project, the adults were only found active at daytime, with flying individuals mainly seen in the early afternoon on warm, sunny days (Ranius 2005); on the contrary, in central-southern European countries (Germany and France), they have been observed to be active at dawn (
Hermit beetles represent one of the most important “flagship” species and “umbrella” species for the protection of the local saproxylic communities (
The conservation status and threats of the European hermit beetle species have been assessed for the IUCN Red List: O. eremita (
Decreasing number of hollow veteran trees (habitat trees) in forest ecosystems due to commercial management which was not suitable for the conservation of biological diversity;
Fragmentation of old-growth forests and lack of connection amongst the habitat trees;
Felling of chestnut orchards or their transformation to coppices;
Transformation or felling of ancient orchards with old fruit trees of the rose family (especially Prunus);
Abandonment of pollarding practices on willows, poplars, mulberry trees and other species across the rural landscape;
Removal of dead wood and old trees from forests or hedgerows for civic use such as domestic fuel;
Increasing use of old trees and dead wood in general as biomass for industrial fuel production;
The lack of entomologists involved in risk management assessment of dangerous old trees in urban parks and country roads for ensuring both human safety and biodiversity conservation;
Risk of pheromone use by insect dealers for international trade of rare species.
The long-term research conducted in Sweden, where hermit beetles live aggregated into large hollow oaks within small tree stands, concluded that there is a considerable extinction risk for many populations as they mainly rely on only one or a few trees with large amounts of wood mould and 10% of the hollow trees harbour two thirds of the individuals (
Until the present time, several methods have been used to study and monitor the O. eremita populations, mainly based on: (1) searching or capturing the adults by either pitfall and pheromone traps or visual survey and (2) wood mould sampling, searching for the remains of adults and frass or larvae in hollow trees. These methods are however often used in combination to achieve a greater sampling efficacy. In the following paragraphs, an overview is presented of the monitoring methods used for the different European countries on the genus Osmoderma.
Sweden
In this country, the standard method for the monitoring of O. eremita populations was developed by using traditional methods, like pitfall trapping and/or wood mould sampling (
The identification of the peach-like, fruity odour of the hermit beetles as (R)-(+)-γ-decalactone and the identification of its function as a sex pheromone released by males to attract females, allowed the development of different methods useful for monitoring the species.
Svensson and Larsson (2008) for the first time tested the “custom built trap”, also called the pheromone trap, baited with the pheromone of the species. They positioned 12-20 replicates of traps for three years. The traps were suspended from tree branches at 2-4m height and at least 10m apart. The traps would be checked every second day. They suggested undertaking the monitoring from the end of June until the end of August (Svensson and Larsson 2008). Pheromone traps outside trees and pitfall traps inside tree hollows, facilitate the mark-recapture of beetles (
Furthermore, in 2016, a national monitoring programme was carried out in 73 sites, from 19 July until 1 August using 198 pheromone traps and 114 pitfall traps. The traps were checked every three days. During the sampling, 56 hermit beetles from 43 pheromone traps and 29 beetles from 12 pitfall traps were detected respectively (N. Janssons, pers. com.).
Finland
In 2012, in order to achieve a comprehensive assessment of the national distribution of O. barnabita, a pheromone-based plan was designed and implemented by volunteers in several urban parks and in all major oak woodlands in Finland (Landvik et al. 2015). The volunteers were amateur and professional entomologists, contacted by advertisement; the protocol and the sampling set (trap, bait etc.) were supplied by the project. The sampling was carried out from 2012 until 2014 in 52 sites, during the flight activity of the species from mid July to mid August, using 2-4 traps per site. Despite the massive sampling effort of more than 3,500 trap-days, O. barnabita was only detected in the Ruissalo Island (SW Finland). The presence of O. barnabita at the level of host trees was evaluated by a standardised method: the site was divided into squares of 250 x 250m and, in each square, the hollow trees were detected and analysed by sieving the wood mould from any remains of adults, frass and larvae of the species (Landvik et al. 2015). Signs of hermit beetles were detected in 62 trees out of 192 total examined (32%) and the most preferred host was pedunculate oak (Quercus robur), with additional occurrences in lime (Tilia) and alder (Alnus). Large trees were more frequently occupied than smaller ones, with an incidence of 90% on the largest oaks.
Norway
During the last 100 years, live specimens of the O. eremita had not been found in Norway and the species was presumed regionally extinct (category “Extinct?”) (
Poland
Despite the large number of occurence sites of hermit beetles in Poland (see
Germany
In southern Germany, where both O. eremita and O. barnabita are known to occur, the monitoring method proposed is the wood mould sampling by using the “vacuum cleaner method”, used for the first time by Bußler and Müller (2008). A suction tube of a vacuum cleaner (Nilfisk Backuum battery) was introduced into the cavity, extracting each time 3.5 litres of wood mould. The sample was checked for frass, pellets, adults, larvae and remains of the hermit beetles. Sampling was undertaken on 127 tree cavities during 2006 and 2007, in eight Nature 2000 areas in southern and northern Bavaria. Hermit beetles were recorded in 39 tree cavities. In particular, larvae were recorded in 11 trees and remains in 30 samples. This method allows good results to be obtained but has a significant impact on the microhabitat structure and negatively affects the saproxylic communities.
France
In France, several monitoring programmes for hermit beetle populations were undertaken by using different methods separately and in combination: wood mould sampling (
Romania
In Romania, the standard method for the monitoring of Osmoderma (locally represented only by O. barnabita sensu
Austria
In Austria, the monitoring protocol on hermit beetle populations has still not been proposed, although O. eremita was considered a highly endangered species from past knowledge and the species was thought to have become extinct due to habitat loss (
Switzerland
In 2004, a management plan for old trees was developed to protect saproxylic beetles in urban and periurban areas (
Slovenia
In this country, the monitoring method for O. eremita (s.l.) populations was published for the first time by
Italy
In Italy,
Regarding the setting of the traps in sampling areas,
Three methods were selected at the beginning of the project (for details on the project, see
A preparatory phase was conducted to select the sites and the period for monitoring of hermit beetle populations. Firstly, an analysis of previous data from literature and the study of collections of museums and private owners were carried out for an overview of the presence of the species and its phenology. Moreover, preparatory surveys were made during autumn-winter 2013-2014 in order to select the best suited sites in the two macroareas (FC and PA) for setting traps in the coming years (considering accessibility, presence of the species, forest structure etc.). During these surveys all potential suitable trees were mapped by GPS and the main features were recorded to create an inventory.
Black cross window traps (hereafter: BCWT) were used, functioning as interception air traps baited with the pheromone produced by males of O. eremita (Svensson and Larsson 2008,
Dispensers for pheromone were 1.5ml plastic eppendorf vials loaded with 1200μl of the racemic mixture of γ-decalactone (catalogue no. W236004, Sigma-Aldrich, USA). Cut strings of cotton dental rolls (Celluron, Paul Hartmann, S.A., France) were inserted as wicks into the dispenser. The dispensers were attached to the traps with a metal hook (Figure
BCWT for capturing O. eremita. A Outline of a trap with an eppendorf vial containing the pheromone B positioning on tree branch C insertion of the funnel into the bottle, with details of hermit beetle captured by the trap (note the four small holes in the base to avoid water filling) D O. eremita with double marking procedure on elytra: dimples made by a small drill and a numbered sticker (photographs by E. Capogna: A–B, F. Bernardini: C, S. Chiari: D).
Pitfall traps (hereafter: PT) were used to catch beetles walking on the wood mould surface inside the cavities. PT consisted of empty jars, with an opening diameter of around 7cm, placed inside the wood mould with the opening at the surface of the mould (for details see Campanaro et al. 2014,
At the first capture, the sex of each hermit beetle was determined by examining head and pronotum morphology. Afterwards, each beetle was marked on the elytra by a numeric code following a double procedure (Figure
Within the macroarea FC, trap sampling was undertaken in two sites: Foresta della Lama (FL) and Camaldoli (CA) which were dominated by beech and chestnut respectively. Within the macroarea PA, traps were set in three study sites: Difesa di Pescasseroli (DP), Val Fondillo (VF) and the Riserva Naturale Feudo Intramonti e Colle di Licco (FI). The first two areas were dominated by beech forests and the third area was occupied by oak forests (for other information on the study areas, see
In 2014, trapping was only undertaken in the macroarea PA (in two sites, DP and VF), during July (Table
Sampling design used during the MIPP project. Summary of the sampling design followed during the study of O. eremita in three years. (PA= Parco Nazionale d’Abruzzo Lazio e Molise; FC= Foreste Casentinesi; VF = Val Fondillo; DP = Difesa di Pescasseroli; FI= Feudo Intramonti e Colle di Licco; CA= Camaldoli; FL = Foresta della Lama).
Macroarea | Site | Method | Study period | # Survey | |
PA | VF | 45 BCWT | – | 21/07-31/07/2014 | 5 |
DP | 45 BCWT | – | 21/07-31/07/2014 | 5 | |
PA | VF | 10 BCWT | 10 PT | 15/07-28/08/2015 | 21 |
DP | 10 BCWT | 10 PT | 15/07-28/08/2015 | 21 | |
FI | 10 BCWT | 10 PT | 15/07-28/08/2015 | 21 | |
FC | CA | 10 BCWT | 10 PT | 15/07-18/08/2015 | 14 |
FL | 10 BCWT | 10 PT | 15/07-18/08/2015 | 14 | |
PA | DP | 10 BCWT | 10 PT | 18/07-12/08/2016 | 12 |
FC | FL | 10 BCWT | 10 PT | 18/07-12/08/2016 | 12 |
These field experiments addressed the following questions: (i) are the sites suitable for the species? (ii) what is the most efficient marking technique for hermit beetle populations? (iii) does the use of 1200μl in one or two dispensers affect the efficacy of the method? (iv) is the method suitable for detecting the species and estimating relative abundance?
In 2015, trapping was conducted in both macroareas: PA (in three sites DP, VF and FI) and FC (in two sites, FL and CA). For each site, 10 BCWT and 10 PT were used. BCWT were positioned on tree branches at 2-4m height and baited with one eppendorf vial containing 1200μl of racemic mixture. PT were positioned inside the tree hollows amongst suitable trees which had previously been investigated in 2014. The distance between BCWT and PT was at least 100m. Traps were checked for the presence of beetles every second day, three times a week (Table
These field experiments addressed the following questions: (i) between BCWT and PT, what is the better method to detect hermit beetles? (ii) is the colour numbered sticker a reliable method to mark hermit beetles? (iii) what is the lower limit of population abundance for using the capture-release method? (iv) what is the minimum sampling effort (in terms of trap and survey number) for detecting the species? (v) do environmental variables of the plot have a relationship with the species occurrence?
In 2016, trapping was again conducted in both macroareas: PA (in two sites DP and VF) and in FC (in one site FL). For each site, 10 BCWT and 10 PT were used. The BCWT were positioned on tree branches at 2-4m height and baited with one eppendorf vial containing 1200μl of racemic mixture. PT were positioned inside the tree hollows, previously selected amongst suitable trees investigated in 2014. The distance between BCWT and PT was at least 100m. Traps were checked for the presence of beetles every second day, for three times a week for a total of 12 surveys (Table
For the dataset of 2014, the Yates corrected χ2 test was applied to investigate differences between male and female captures. Analysis was performed with the programme R version 3.2.5 (R Core Team, 2016), using a significance level of 0.05 to reject the null hypothesis.
The large-scale occupancy probability (ψ) was firstly explored by condensing detections from each capture method and fitting single species, single season and multi-method models. Then, the detection probability estimate (p^) was calculated applying the single species, single season and multi-method models separately to each capture method (
For the dataset of 2015, the Yates corrected χ2 test was used to investigate differences between sampling methods and sex. Analysis was performed as for the 2014 dataset.
Occupancy models for single species, single season were used to calculate site occupancy (ψ), i.e. an estimate of the probability that a randomly selected site of the study area is occupied and the detection probability (p), i.e. the probability of detecting the species by the applied monitoring method (
Craig’s model for closed populations (Craig 1953) was used to estimate the population size for each study area. The closed population model was used due to the low number of recaptures. The Coefficient Variation (CV) was calculated as the standard error divided by the number of individuals estimated and indicates the precision of the population size estimate. Population size estimates for each study area were generated with male and female data pooled together.
The species habitat relationship was analysed pooling data from all the study areas. A set of a priori models were defined with the covariates varying in a way that could explain patterns of trap occupancy. Models with single, additive and multiplicative effects of site covariates (HN, MC, TN, MA) were hypothesised.
Occupancy models, for both 2014 and 2015 data analysis, were fitted and maximum-likelihood estimates were obtained using the programme PRESENCE (
In 2014, a total of 17 captures of 15 individuals and 8 captures of 8 individuals of O. eremita were made in VF and DP respectively (Table
Dataset of mark recapture during 2014. Summary of capture-mark-recapture data of O. eremita obtained during 2014 in PA (PA = Parco Nazionale d’Abruzzo Lazio e Molise; VF = Val Fondillo; DP = Difesa di Pescasseroli).
Site | Sex | Marked | Captures | ||
control | one dispenser | two dispensers | |||
VF | male | 4 | 0 | 3 | 2 |
female | 11 | 0 | 11 | 1 | |
DP | male | 1 | 0 | 1 | 0 |
female | 7 | 0 | 4 | 3 |
In both sites, the estimated large-scale probability of O. eremita occupancy was high, but considerably uncertain (VF ψ = 0.55, SE = 0.18; DP ψ = 0.80, SE = 0.57). Model selection statistics provided strong evidence that occupancy probabilities, in both sites, were firstly influenced by the sampling method (s) (VF ws = 0.98; DP ws = 0.96) (Table
Occupancy probabilities model selection. Summary of the model selection statistics for the study of O. eremita during 2014 in PA (PA = Parco Nazionale d’Abruzzo Lazio e Molise; VF = Val Fondillo; DP = Difesa di Pescasseroli).
Site | Model | K * | -2Log (L) | ΔAIC | w |
VF | ψ, θt, ps+t | 9 | 61.54 | 0.00 | 0.55 |
ψ, θt, ps | 9 | 62.05 | 0.51 | 0.43 | |
DP | ψ, θt, ps+t | 9 | 42.99 | 0.00 | 0.52 |
ψ, θt, ps | 9 | 43.48 | 0.49 | 0.41 |
Detection probabilities estimates by monitoring during 2014. O. eremita detection probability estimates (p^) and associated standard errors (in parenthesis) are given for the plausible candidate models (ΔAIC < 2) (w = Akaike’s weight for each model), during 2014 in PA (PA = Parco Nazionale d’Abruzzo Lazio e Molise; VF = Val Fondillo; DP = Difesa di Pescasseroli).
Site | Model | w | p^contro l | (SE) | p^one disp. | (SE) | p^two disp. | (SE) |
VF | ψ, θt, ps+t | 0.55 | 0.00* | (0.00) | 0.34* | (0.16) | 0.10* | (0.08) |
ψ, θt, ps | 0.43 | 0.00 | (0.00) | 0.67 | (0.27) | 0.20 | (0.13) | |
DP | ψ, θt, ps+t | 0.52 | 0.00* | (0.00) | 0.21* | (0.14) | 0.07* | (0.07) |
ψ, θt, ps | 0.41 | 0.00 | (0.00) | 0.50 | (0.35) | 0.17 | (0.15) |
In 2015, in total 116 capture events of 105 individuals were registered by BCWT (Table
Time interval of capture events during 2015. Capture events of O. eremita during the MIPP project monitoring in 2015 in four Italian study sites (grey line = Foresta della Lama; black line = Feudo Intramonti e Colle di Licco; black dashed line = Val Fondillo; black dotted line = Difesa di Pescasseroli).
Dataset of mark recapture during 2015. Summary of mark-recapture data obtained during the study of O. eremita in Italy with different capture methods during 2015 in the two macroareas PA and FC (BCWT = Black Cross Windows Traps; PT = Pitfall Traps; PA= Parco Nazionale d’Abruzzo Lazio e Molise; FC = Foreste Casentinesi; VF = Val Fondillo; DP = Difesa di Pescasseroli; FI = Feudo Intramonti e Colle di Licco; CA = Camaldoli; FL = Foresta della Lama).
Site | BCWT | PT | ||
Marked beetles (♀/♂) | Recapture events (♀/♂) | Marked beetles (♀/♂) | Recapture events (♀/♂) | |
VF | 31/3 | 2/0 | 0/0 | 0/0 |
DP | 37/5 | 6/0 | 0/0 | 0/0 |
FI | 12/1 | 4/0 | 0/0 | 0/0 |
CA | 2/0 | 0/0 | 0/0 | 0/0 |
FL | 8/2 | 3/0 | 0/0 | 0/0 |
The mean value amongst macroareas for ψ was 0.98 (SE = 0.03) and for p it was 0.10 (SE = 0.04) (Table
Occupancy and detection probability estimates by monitoring during 2015. O. eremita occupancy (ψ) and detection (p) probability estimates and associated standard errors (in parenthesis), are given for plausible candidate models (within 2 ΔAIC of the top model). The data were obtained by monitoring during 2015 in the two macroareas PA and FC (PA= Parco Nazionale d’Abruzzo Lazio e Molise; FC = Foreste Casentinesi; VF = Val Fondillo; DP = Difesa di Pescasseroli; FI = Feudo Intramonti e Colle di Licco; FL = Foresta della Lama).
Site | Model | K * | -2Log (L) | ΔAIC | w | ψ | SE | p | SE |
VF | ψ(.), p (.) | 2 | 167.93 | 0.00 | 1.00 | 0.93 | 0.10 | 0.15 | 0.03 |
DP | ψ(.), p (.) | 2 | 150.84 | 0.00 | 1.00 | 1.00 | 0.00 | 0.09 | 0.07 |
FI | ψ(.), p (.) | 2 | 108.07 | 0.00 | 0.99 | 1.00 | 0.00 | 0.07 | 0.02 |
FL | ψ(.), p (.) | 2 | 94.95 | 0.00 | 1.00 | 1.00 | 0.00 | 0.09 | 0.02 |
For selection of a standard error for the estimated level of occupancy of 0.10, the most efficient allocation of resources is to use 30 BCWT, each checked 23 times (690 total number of surveys).
Amongst the marked beetles, not one lost the colour-numbered sticker.
The most abundant population was the one for VF 304±104 (C.V. 0.34), the less abundant being that of FL 22±5 (C.V. 0.23) (Table
Estimation of population size. O. eremita population size estimates obtained by monitoring during 2015 in the two macroareas PA and FC (PA= Parco Nazionale d’Abruzzo Lazio e Molise; FC = Foreste Casentinesi; VF = Val Fondillo; DP = Difesa di Pescasseroli; FI = Feudo Intramonti e Colle di Licco; FL = Foresta della Lama).
Site | Population size estimate | ±95 % confidence interval | Coefficient of variation |
VF | 304 | 104 | 0.34 |
DP | 173 | 45 | 0.26 |
FI | 22 | 5 | 0.23 |
FL | 19 | 2 | 0.10 |
Trap occupancy was favoured by the additive effect of the hole number (wHN = 0.72) and microhabitat amount (wMA = 0.38) (Table
BCWT occupancy data. BCWT occupancy of O. eremita during 2015 in relation to the number of holes (grey empty circles) and microhabitat amount (black dots). Vertical bars indicate the standard error.
Dataset of environmental variables. Environmental characteristics (within 2 ΔAIC from the top model) affecting the occurrence of O. eremita in BCWT. The data were obtained during 2015 in the two macroareas PA and FC (PA= Parco Nazionale d’Abruzzo Lazio e Molise; FC = Parco Nazionale delle Foreste Casentinesi).
Model | K | -2Log (L) | ΔAIC | w |
ψ(HN), p (t) | 18 | 492.66 | 0.00 | 0.50 |
ψ(HN+MA), p (t) | 19 | 492.29 | 1.63 | 0.22 |
In 2016, in total 32 captures of 31 individuals and 11 captures of 10 individuals of O. eremita were made, all by BCWT in PA and FC respectively (Table
Dataset of mark-recapture during 2016. Summary of mark-recapture data obtained during the study of O. eremita in Italy with different capture methods during 2016 in the two macroareas, PA and FC (BCWT = Black Cross Windows Traps; PT = Pitfall Traps; PA= Parco Nazionale d’Abruzzo Lazio e Molise; FC = Parco Nazionale delle Foreste Casentinesi; DP = Difesa di Pescasseroli; FL = Foresta della Lama).
Site | BCWT | PT | ||
Marked beetles (♀/♂) | Recapture events (♀/♂) | Marked beetles (♀/♂) | Recapture events (♀/♂) | |
DP | 22/9 | 1/0 | 0/0 | 0/0 |
FL | 5/5 | 1/0 | 0/0 | 0/0 |
Comparison of the results obtained from all methods allowed the best standard method for the monitoring of O. eremita to be selected. The analysis of the total number of hermit beetles captured by the two trap types during the sampling of 2015 and 2016 in the two macroareas (FC and PA) showed that only BCWT captured the beetles. The use of PT provided zero captures in this research, notwithstanding this method was successfully used in different environmental contexts, such as in Sweden and France (
These results suggested that, to carry out monitoring of a southern population of the hermit beetle (at least for the Apennine range), a number of 30 BCWT checked every second day for eight weeks (from early July to the end of August) is the most efficient method; traps can also be placed on trees without hollows due to the high attraction of the pheromone on the species. The activity period can vary according to altitude and latitude, other than seasonal weather conditions and then the sampling period for monitoring could be shorter where there is a greater knowledge of the life cycle of the species at local level (e.g. Figure
A high number of PT (at least 150) placed in the same number of tree cavities (preferably one trap per tree) may be used as an alternative method. In fact, this method is cheaper and less invasive (due to the breeding activities by pheromones not being disturbed), but may be used only in areas where: 1) tree hollows are abundant, large enough and with sufficient amounts of wood mould to set pitfall traps and 2) the team is composed of several people in order to ensure the checking of so many traps every two days.
According to N. Jansson (pers. com.), the bottom of the traps (both BCWT and PT) should be lined with wet moss, in order to reduce temperature, drought, stress and energy consumption of the beetles captured.
The first step in carrying out the monitoring plan is to choose a suitable area inhabited by O. eremita. Ideally, an area should be investigated in detail, in order to map the hollow tree by GPS, to take photographs and to record the characteristics of the trees (e.g. species, size, DBH, height, number of hollows etc.). This information, gathered in the field, will be useful for setting the traps. Afterwards, the BCWT would be built during the spring, before starting the monitoring. (For details and materials about the BCWT, see Methods and supplementary file 1.) The traps are set in the field at least a week before starting the sampling, without both pheromone and plastic collecting jars. The BCWT could be positioned randomly in a woodland or along a transect (e.g. along tree-lined streets, rows of pollarded trees bordering rivers and channels etc.) or in a grid (e.g. 5 rows with 6 traps in each one; the size of the grid is about 500m long and 600m wide), setting the traps at least 100m from each other. Every trap should be mapped by GPS and marked by a label with its number.
The monitoring can be carried out by two operators checking the traps every two days for 8 weeks, from July to September. The surveys can be conducted in two different ways:
Three times a week (Monday, Wednesday and Friday), the traps can be activated the day before starting the check (Sunday) and deactivated after the third weekly check, on Friday. The BCWT is activated when the caps of the plastic collection jar are removed and the eppendorf vial with pheromone is baited; in the opposite case, deactivation means that the collection jar is closed and the eppendorf vial has been removed. The activation and deactivation of the traps is crucial to avoid the death of individuals.
Every two days without any breaks, continuously; the traps can be activated the day before to start the check and never deactivated until the end of sampling.
The plastic eppendorf vials (capacity of 1.5ml) loaded with 1200μl of the racemic mixture of γ-decalactone can be prepared before starting the sampling and preserved in a freezer. It is suggested that the pheromone vials be changed once a week.
The protocol requires the presence of two operators who should fill the field sheet of each survey with the date, the weather conditions (rain and cloud) and the starting time (see supplementary file 2). After that, the operators check the traps and, once an adult is found, it is temporarily placed into the plastic jar with the cap to prevent escape. Once the check of BCWT has been completed, the number of individuals collected is counted, specifying the number of males and females for each trap in the field sheet (see supplementary file 2). Afterwards, the beetles were released in the same hollow tree from where they were captured. At the end of each survey, the operators recorded the weather conditions (rain and cloud) and the finishing time (see supplementary file 2).
The equipment required is: GPS, clipbord, field sheet, pencil, clock, plastic eppendorf vial with pheromone, cotton dental rolls, tweezers and camera.
O. eremita is a species with a low dispersal rate. In fact, the majority of the individuals can move only a few hundred metres (from 100m to 250m), while some have a dispersal rate exceeding 2000m, as reported in several studies summarised by
The major risk associated with the use of the BCWT is related to adverse weather conditions; in fact, in case of heavy rains, the plastic collection bottle, despite the presence of drainage holes, may be filled by water, resulting in an obvious danger to the specimens trapped inside. In case of heavy rainfall, it is recommended to increase the rate of checks, verifying daily if there are specimens inside the traps and emptying water from the collection chambers and from any material that can clog the holes (e.g. leaves).
Finally, O. eremita is a rare and elusive species, very sought after by collectors and the specimens trapped inside BCWT may represent easy prey for them. Thus, to avoid risks related to unauthorised collections, it is recommended to set the traps on higher branches with the help of an extension, so that the traps are not easily accessible.
In order to identify the population trend for O. eremita at local level during the years, a reference value can be calculated as follows:
For each survey, sum the total number of individuals (males + females) captured by all 30 BCWT.
Calculate the mean value of the number of individuals from all surveys, excluding the survey with the lowest number. Removing the lowest number, as proposed for other insect species (FLA 2014, Hardersen and Toni 2015), allows the elimination of eventual outlier values due to adverse climatic conditions (e.g. low temperature and/or rainfall) or other factors, that may affect the activity of the individuals and increase data heterogeneity.
Table
Example of calculation of mean value of captured beetles. Subset of traps and surveys used for calculation of the reference value of the monitoring trend of O. eremita population in one site (BCWT: Black Cross Window Trap).
Survey | BCWT 1 | BCWT 2 | BCWT 3 | BCWT 4 | BCWT 5 | BCWT 6 | BCWT 7 | Total |
1 | 1 | 2 | 4 | 3 | 2 | 3 | 1 | 16 |
2 | 3 | 6 | 5 | 5 | 6 | 3 | 3 | 31 |
3 | 3 | 5 | 7 | 7 | 7 | 4 | 5 | 38 |
4 | 2 | 4 | 3 | 2 | 2 | 0 | 0 | 13 |
5 | 4 | 7 | 8 | 5 | 3 | 5 | 4 | 36 |
Average value for the four counts with the highest average total | 30.25 |
The present work was developed within the EU project LIFE11 NAT/IT/000252, with the contribution of the LIFE financial instrument of the European Union. We are grateful to all the staff of the National Forestry service, local biodiversity office of Castel di Sangro, in particular to Tiziana Altea, Mario Romano, Lucia Eusepi, Federica Desprini, Ilaria Filippone; to all the staff of the National Forestry service, local biodiversity office of Pratovecchio, in particular to Giovanni Quilghini, Antonio Zoccola, Barbara Rossi, Sandro Marsella, Silvia Bertinelli, Valerio Mazzoli and Angelo Lamberti; to all the staff of Parco Nazionale d’Abruzzo, Lazio e Molise, in particular to Dario Febbo, Cinzia Sulli and Paola Tollis and all the park rangers; to the staff of the Foreste Casentinesi, Monte Falterona e Campigna National Park and Cooperativa “In Quiete” and their volunteers. We would like to thank our colleagues and field assistants who helped during the surveys: Emiliano Mancini, Lara Redolfi De Zan, Ilaria Toni, Alessandro Cini, Sonke Hardersen, Sarah Rossi De Gasparis, Giulio Nigro, Marco Molfini, Francesca Bernardini, Emma Pellegrini, Randi Rollins, Sara Amendolia, Giulia Albani Rocchetti, Fabio Garzuglia, Rosaria Santoro, Sara Benelli, Valentina Stagno, Marco Boscaro, Silvia De Michelis, Cristina Mantoni, Nino Loreto, Davide Caprioli, Alessandro Cortini, Matteo De Santis, Daniele Di Giacomo, Giulia Caruso, Eleonora Cammarano, Ilaria Zappitelli, Kevin Mancini, Flavio Cennamo, Giacomo Grosso, Gioia Burini, Silvio Kroha, Federico Grant, Luca Gallitelli, Livia Benedini, Emilia Capogna, Elisa Tamilia, Laura Petruccelli, Michela Perrone, Adriano Sanno, Gabriele Miserendino, Alessandro Morelli, Luca Bisegna, Cristiana Baldi, Luciano Caporale and Emanuela Rosi. We would like to thank everyone for allowing us to use their photos (Alberto Ballerio, Benjamin Calmont, Sonia Dourlot, Federico Romiti, Marco Uliana, Francesca Bernardini, Emilia Capogna, Stefano Chiari). We thank Meike Liu (Sapienza - Università di Roma, Dipartimento “Charles Darwin”, Italia and Northwest A&F University of Yangling, China) for preparing the new version of the European distribution map of the genus Osmoderma.
A special permit was obtained from the Italian Ministry of Environment for handling and capturing individuals of the target species (collection permit: Ministero dell’Ambiente e della Tutela del Territorio e del Mare - DG Protezione della Natura e del Mare, U. prot PNM 2012-0010890 del 28/05/2012).
With the contribution of the LIFE financial instrument of the European Union.