Influence of forest decline on the abundance and diversity of Raphidioptera and Mecoptera species dwelling in oak canopies

Abstract

Trees in a state of decline exhibit a reduced foliage density and accumulate dead branches in their crowns. Consequently,
forest decline can markedly affect both the habitats and sources of food for canopy-dwelling insects. The decline-induced
increase in canopy openness may also modify the understory, shrub and ground layers, and have cascading effects on associated
species. Flight interception traps and green Lindgren traps were used to survey the canopy-dwelling insects in stands of
healthy and declining oak trees, in particular two insect orders: Raphidioptera, saproxylic insects associated with canopies, and
Mecoptera, necrophagous or opportunistic species associated with the herbaceous or shrub strata. Overall, green Lindgren traps
caught more of these insects than fl ight interception traps. The traps caught fi ve species of Raphidioptera. Three of them, Subilla
confi nis, Phaeostigma major and, to a lesser extent, Phaeostigma notata, were more abundant in stands or plots with declining
trees. However, the other two species of Raphidioptera, Atlantoraphidia maculicollis and Xanthostigma xanthostigma exhibited
a reverse trend. Two species of Mecoptera, Panorpa germanica and Panorpa communis, were particularly abundant, but unaffected
by the level of decline. Our results show that declining forests can either host more or fewer species of Raphidioptera with
saproxylic larvae, whereas Mecoptera with ground-living larvae were unaffected. Seasonal phenology and sex ratio of the species
are also discussed.

Full text

372
Final formatted article © Institute of Entomology, Biology Centre, Czech Academy of Sciences, České Budějovice.
An Open Access article distributed under the Creative Commons (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).
EUROPEAN JOURNAL OF ENTOMOLOG
Y
EUROPEAN JOURNAL OF ENTOMOLOGY
ISSN (online): 1802-8829
http://www.eje.cz
tent and frequency of declines are expected to increase
worldwide (Allen et al., 2010; IPCC, 2013). As decline
progresses, forest ecosystems undergo dramatic changes in
terms of composition, structure and functioning. In partic-
ular, there are conspicuous changes in the structure of the
canopy that are uncommon in those of healthy trees, such
as dead branches and cavities, which increase the structural
complexity of the canopy at scale levels of stand, tree and
branch (Ishii et al., 2004). However, it may also negatively
affect the amount and quality of other critical resources,
such as foliage, fruit and seed (Houston, 1981). Such
profound structural modifi cations can modulate micro-
climates, habitat opportunities, and trophic resources for
canopy-dwelling communities. For instance, oak decline
can promote saproxylic and generalist leaf-feeding beetles,
but have a negative effect on specialist leaf-feeders (Sallé
et al., 2020). Decline-induced environmental changes can
also shape habitat conditions at understory and ground lev-
els.
Raphidioptera and Mecoptera are two ancient taxonomic
orders of insects frequently encountered in moist temperate
forests (Byers & Thornhill, 1983; Aspöck, 2002). Raphidi-
Infl uence of forest decline on the abundance and diversity
of Raphidioptera and Mecoptera species dwelling in oak canopies
ALEXIS VINCENT 1, PIERRE TILLIER 2, CÉCILE VINCENT-BARBAROUX 1, CHRISTOPHE BOUGET 3 and AURÉLIEN SALLÉ 1
1 Laboratoire de Biologie des Ligneux et des Grandes Cultures, INRAE, Université d’Orléans, 45067 Orléans, France;
e-mails: alexis.vincent@etu.univ-orleans.fr, cecile.barbaroux@univ-orleans.fr, aurelien.salle@univ-orleans.fr
2 8 rue d’Aire, F-95660 Champagne-sur-Oise, France; e-mail: p.tillier.entomo@free.fr
3 INRAE, UR EFNO, Domaine des Barres, Nogent-sur-Vernisson, France; e-mail: christophe.bouget@inrae.fr
Key words. Raphidioptera, Mecoptera, forest decline, canopy, intermediate disturbance hypothesis, Quercus
Abstract. Trees in a state of decline exhibit a reduced foliage density and accumulate dead branches in their crowns. Conse-
quently, forest decline can markedly affect both the habitats and sources of food for canopy-dwelling insects. The decline-induced
increase in canopy openness may also modify the understory, shrub and ground layers, and have cascading effects on associ-
ated species. Flight interception traps and green Lindgren traps were used to survey the canopy-dwelling insects in stands of
healthy and declining oak trees, in particular two insect orders: Raphidioptera, saproxylic insects associated with canopies, and
Mecoptera, necrophagous or opportunistic species associated with the herbaceous or shrub strata. Overall, green Lindgren traps
caught more of these insects than fl ight interception traps. The traps caught fi ve species of Raphidioptera. Three of them, Subilla
confi nis, Phaeostigma major and, to a lesser extent, Phaeostigma notata, were more abundant in stands or plots with declining
trees. However, the other two species of Raphidioptera, Atlantoraphidia maculicollis and Xanthostigma xanthostigma exhibited
a reverse trend. Two species of Mecoptera, Panorpa germanica and Panorpa communis, were particularly abundant, but unaf-
fected by the level of decline. Our results show that declining forests can either host more or fewer species of Raphidioptera with
saproxylic larvae, whereas Mecoptera with ground-living larvae were unaffected. Seasonal phenology and sex ratio of the species
are also discussed.
INTRODUCTION
Global change is dramatically affecting forest ecosys-
tems, either by creating novel types or by increasing the
frequency and/or intensity of natural disturbances (Lieb-
hold et al., 2017; Seidl et al., 2017). According to the
intermediate disturbance hypothesis, while the outcome
may be detrimental at extreme levels, disturbance may
have positive outcomes on biodiversity at intermediate
levels of severity and frequency (Grime, 1973; Connell,
1978). For instance, wind throw gaps or fi res can promote
insect diversity at the landscape level (Bouget & Duelli,
2004; Moretti et al., 2004). Nonetheless, the outcome of
disturbance and whether the effects on insect diversity are
positive or negative, may depend on the taxonomic groups
and functional guilds considered (e.g. Moretti et al., 2004;
Sallé et al., 2020).
Forest decline consists of a progressive loss of vigour in
trees, over several years in response to multiple, succes-
sive or concomitant driving factors (Manion, 1981; Sallé
et al., 2014). In response to the predicted increase in the
frequency and severity of droughts and heat waves, which
are major inciting factors of forest decline, the spatial ex-
Eur. J. Entomol. 117: 372–379, 2020
doi: 10.14411/eje.2020.041
ORIGINAL ARTICLE

373
Vincent et al., Eur. J. Entomol. 117 : 372–379, 2020 doi: 10.14411/eje.2020.041
abundance of aphids and consequently have detrimental ef-
fects on the adults. Mecoptera are connected to the soil as
larvae and herbaceous or shrub layers as adults, and are
presumably less dependent on resources in the canopy.
However, increased canopy openness in declining stands
might favour lower strata of vegetation and their associ-
ated communities. Therefore, we hypothesized that decline
would have either no or slightly positive effects on Meco-
ptera, as suggested by Duelli et al. (2019).
To test these hypotheses, we sampled Raphidioptera
and Mecoptera communities in the canopy of mature oaks
stands in which the incidence of declining trees ranged
from zero to severe. This study was conducted within the
frame of an ongoing project investigating how oak decline
can affect the trophic guilds of canopy-dwelling insects.
For this project, we have been using a multi-taxa approach,
including several families of Coleoptera, Hemiptera and
Diptera belonging to different trophic guilds. Here we spe-
cifi cally report on the results concerning Raphidioptera and
Mecoptera. For these two groups, our objectives were to
(i) identify the canopy-dwelling species of Raphidioptera
and Mecoptera, (ii) evaluate the association between abun-
dance and diversity of these species and the local intensity
in forest decline, and (iii) provide additional information
on the ecology of these poorly studied groups.
MATERIAL AND METHODS
Study sites
This study was conducted in three state forests, located near
Orléans (Forêt domaniale d’Orléans, 47°98´97˝81N, 1°95´44˝E),
Vierzon (Forêt domaniale de Vierzon, 47°26´89˝N, 02°10´74˝E)
and Marcenat (Forêt domaniale de l’Abbaye, 46°21´12˝N,
3°36´13˝E). The forest at Orléans covers 35,000 ha and is domi-
nated by oaks (55%) and pines (39%). The forest at Vierzon, and
the adjacent forest at Vouzeron cover a surface area of 7,500 ha
and are dominated by oaks (61%) and pines (31%). Oaks in the
forest at Vierzon, in particular Q. robur, have regularly suffered
optera, or snakefl ies, have bark- or ground-dwelling larvae,
which generally prey on soft-bodied arthropods, eggs or
larvae. Adults are also predatory and mostly feed on aphids
and other Sternorrhyncha, but also occasionally consume
pollen (Aspöck, 2002). Several species preferentially live
in the forest mantle or canopy (Duelli et al., 2002). The
corticolous larvae of these species forage on the bark or
under loosened bark of trunks and branches (Wichmann,
1957). They can also colonize decayed branches and gal-
leries of bark and wood boring species, where they feed on
eggs, larvae, nymphs and immature adults of various spe-
cies, including saproxylic species (Wichmann, 1957; Kenis
et al., 2004; Weigelmeier & Gruppe, 2010). These snakefl y
larvae can therefore be considered as saproxylic. Mecop-
tera, or scorpion fl ies, have soil-dwelling saprophagous
larvae (Byers & Thornhill, 1983). Adults mostly feed on
dead soft-bodied arthropods and opportunistically on other
substances such as pollen or plant fl uids (Byers & Thorn-
hill, 1983). Although they can be found in the canopies of
trees (e.g. Barnard et al., 1983), adult Mecoptera are gener-
ally considered to be associated with herbaceous or shrub
strata in forest ecosystems (Byers & Thornhill, 1983; Du-
elli et al., 2002). Observations conducted in several forest
stands following either fi re or wind throw events indicate
that both Raphidioptera and Mecoptera may benefi t from a
certain level of disturbance (Duelli et al., 2019).
We hypothesized that oak decline could indirectly favour
species of Raphidioptera with saproxylic larvae, as the in-
creased structural complexity in the crown of declining
trees will provide a broader range of oviposition sites and
larval habitats, and the increased diversity and abundance
of saproxylic species more prey for the larvae. In addition,
the increased canopy openness can modify the microcli-
mates in tree crowns and the resulting warmer conditions
could promote larval development. On the other hand,
the reduced amount of foliage could negatively affect the
Table 1. Characteristics of the oak stands monitored.
Forest Stand
ID Overstorey Understorey Stand
height (m)
Mean ± SE
oak DBH (cm)
Tree density
(n/ha)
Total basal
area (m2/ha)
Vierzon 19 Oak with scattered beech and Scots pine Hornbeam 26 57 ± 3 89 18
Vierzon 35 Oak with scattered beech and Scots pine 26 54 ± 3 105 21
Vierzon 70 Oak with scattered beech and Scots pine 24 49 ± 6 58 9
Vierzon 71 Oak with scattered beech and Scots pine Beech 25 40 ± 4 105 15
Vierzon 81 Oak Hornbeam 24 43 ± 2 197 15
Vierzon 179 Oak 29 64 ± 6 81 21
Vierzon 236 Oak with scattered beech 24 73 ± 4 88 23
Vierzon 249 Oak with scattered beech and hornbeam 25 52 ± 2 127 22
Vierzon 290 Oak Scattered wild service trees 23 51 ± 2 160 22
Orléans 351 Oak with scattered Scots pine Chestnut and wild service trees 24 63 ± 5 136 18
Orléans 751 Mixed stand of oak and beech,
with scattered Scots pine Hornbeam 27 64 ± 4 107 24
Orléans 1140 Oak Scattered glossy buckthorn 25 72 ± 6 90 26
Orléans 1343 Oak 28 75 ± 4 100 28
Orléans 1344 Oak 24 78 ± 5 74 27
Orléans 1427 Oak 26 NA 140 15
Orléans 1490 Oak and scattered aspen Hornbeam 23 70 ± 9 73 17
Orléans 1491 Oak and scattered aspen Hornbeam 23 65 ± 3 90 17
Orléans 1502 Oak and scattered aspen Hornbeam 21 48 ± 3 222 16
Marcenat 12 Oak Hornbeam 29 56 ± 6 154 19
Marcenat 37 Oak Hornbeam and linden 37 79 ± 4 76 18
Marcenat 58 Oak Scattered hornbeam 23 40 ± 4 389 33

374
Vincent et al., Eur. J. Entomol. 117 : 372–379, 2020 doi: 10.14411/eje.2020.041
from decline (Sallé et al., 2020). The forest at Marcenat covers
2,100 ha, and is largely dominated by oaks (71%) with some
pines (11%) and Douglas fi r (8%). Several stands are currently
experiencing decline, triggered by recurrent summer droughts
(Baubet O., Département de la Santé des Forêts, pers. comm.).
We selected 21 mature oak-dominated stands (minimal surface
area: 3 ha): nine in the forest of Orléans, nine in the forest of Vi-
erzon and three in the forest of Marcenat. Stands were located at
least 500 m from one another. Within each stand, two trees were
selected, one healthy and one declining (if both tree types were
present) at least 50 m apart. Traps were placed in these trees (see
sampling protocols).
In each stand we recorded average tree height, diameter and
density, as well as stand composition (Table 1). We also evalu-
ated the level of decline at three embedded spatial scales: (i) trees
with a trap, hereafter referred to as the tree scale, (ii) the ten clos-
est oaks surrounding the trees with traps, hereafter referred as
a plot, and (iii) 30 trees in the stand (i.e. the 20 trees in the two
monitored plots and 10 trees in an additional plot located between
the two monitored plots), hereafter referred as a stand (Table 2).
We evaluated the decline level of each tree using the DEPERIS
protocol developed by the French Forest Health Department (Dé-
partement de la Santé des Forêts) (Goudet & Nageleisen, 2019).
In summary, percentage of dead branches and ramifi cation loss
in the canopy were evaluated. Based on these criteria, each tree
was assigned to a class of decline ranging from A (no decline) to
F (severe decline). Trees belonging to A–C classes were consid-
ered healthy. Trees belonging to D–F classes were considered in
decline. At plot and stand scales, the percentage of declining trees
(falling within D–F classes) was calculated (Table 2).
Sampling
Insects were collected using green multi-funnel traps (GMFT)
(Lindgren traps, Chemtica Internacional, San Jose, Costa Rica),
each with 12 fl uon-coated funnels, and cross-vane fl ight intercep-
tion traps (FIT) (PolytrapsTM). Both types of traps were suspend-
ed approximately 15 m above the ground (i.e. among the lower
branches in the canopy), in the same tree. The collectors were
fi lled with a solution of 50% (v/v) monopropylene glycol and
water plus a drop of detergent. The traps were set from the end of
March to the beginning of September 2019. The collectors were
emptied every month and the species caught recorded.
Identifi cation
Most of the specimens were identifi ed by AV and validated by
PT. We used the Mecoptera identifi cation key of Tillier (2006)
and unpublished key of B. Bal and P. Tillier for the Raphidio-
ptera, which is a modifi ed version of the key of Semeria and Ber-
land (1988). The sex ratio (males / females) was calculated for
each species. For both Raphidioptera and Mecoptera, we record-
ed the number of individuals (abundance) and for Raphidioptera
the number of species (richness) at the tree scale.
Data analysis
All analyses were performed in R, version 3.5.1 (R Core Team
2018). To assess the effect of trap type on the insects sampled,
we performed Wilcoxon signed-ranked tests. Spearman correla-
tion tests were used to study the abundance of species at the tree
level. To rank the effect of the level of decline at tree, plot or stand
scales on variations in average univariate metrics (mean values of
abundance of each species per trap, taxonomical order abundance,
group species richness), we used the differences in the Akaike
information criterion (AICc) scores to compare the fi t between
the generalized linear mixed models including each of the three
explanatory variables separately. To assess the signifi cance of the
estimates of the best decline feature for each response variable,
the error structure of the generalized linear mixed-effects models
was adjusted to better fi t the data. In order to do so, glmm were
fi tted for negative binomial, Gaussian, log-normal and Poisson
family (functions lmer, glmer.nb and glmer, lme4 R-package).
Since both types of traps were present in the same tree, two trees
within the same stand had traps and some stands were in the same
forest, we added forest, stand and tree as nested random factors
to our mixed model. We also added trap type as a fi xed effect. To
plot the predictions of generalized linear mixed-effects models
based on abundance per tree (i.e. with catches of GMFT and FIT
pooled), we used the ggpredict() function in the ggeffects pack-
age.
RESULTS
Totals of 5374 Mecoptera and 572 Raphidioptera were
caught. All Mecoptera were either Panorpa germanica L.
(3072 ind.) or Panorpa communis L. (2302 ind.). Panorpa
germanica was more abundant than P. communis at Marce-
nat and Orléans, while P. communis dominated at Vierzon.
Five species of Raphidioptera were collected of which
the most abundant in the three forests was Subilla confi nis
(Stephens) (310 ind.), followed by Phaeostigma major
(Burmeister) (95 ind.), Phaeostigma notata (Fabricius) (66
ind.), Atlantoraphidia maculicollis (Stephens) (54 ind.) and
Xanthostigma xanthostigma (Schummel) (47 ind.). Neither
P. notata nor A. maculicollis were caught at Marcenat.
Sex ratio
For both species of Mecoptera their sex ratio was slight-
ly skewed towards female (0.79 and 0.68 for P. communis
and P. germanica, respectively). In contrast, the sex ratio of
the species of Raphidioptera caught was consistently and
sometimes markedly skewed towards male (X. xanthostig-
ma: 1.19, A. maculicollis: 1.32, P. major: 2.13, P. notata:
2.73, S. confi nis: 3.03).
Table 2. Decline indices of trees bearing traps and percentage of
surrounding declining trees in the monitored oak plots and stands.
Forest Stand
ID
Decline index
of trees with
traps
Percentage of
declining trees
at the plot scale
Percentage
of declining
trees at the
stand scale
Tree #1Tree #2 Plot #1 Plot #2
Vierzon 19 C D 100 80 93
Vierzon 35 B C 20 80 43
Vierzon 70 C D 50 100 73
Vierzon 71 B E 40 90 63
Vierzon 81 C D 5 50 18
Vierzon 179 C D 55 50 63
Vierzon 236 D E 80 35 45
Vierzon 249 E F 80 30 35
Vierzon 290 C D 15 50 28
Orléans 351 C D 0 10 3
Orléans 751 C D 50 20 33
Orléans 1140 D E 60 60 63
Orléans 1343 B B 0 10 3
Orléans 1344 D E 30 70 47
Orléans 1427 C E 20 60 37
Orléans 1490 B C 20 10 30
Orléans 1491 B C 0 20 30
Orléans 1502 B D 10 40 23
Marcenat 12 D D 60 60 55
Marcenat 37 C C 10 0 30
Marcenat 58 C C 0 60 47

375
Vincent et al., Eur. J. Entomol. 117 : 372–379, 2020 doi: 10.14411/eje.2020.041
Effect of trap type
For all species, many more individuals were caught in
green multi-funnel traps than in fl ight interception traps (P
< 0.01 for all species Fig. 1).
Seasonal activity
For the analysis of seasonal activity, catches of both
types of trap were pooled. The two species of Mecoptera
differed markedly in their seasonal activity (Fig. 2). While
most individuals of P. communis were caught in June, two
peaks of activity were recorded for P. germanica, one in
May and the other in August. For Raphidioptera, P. major,
P. notata and S. confi nis were mostly caught in May and
A. maculicollis and X. xanthostigma mostly in April. For
most species of Mecoptera and Raphidioptera the seasonal
activity of the males and females was similar except for
A. maculicollis, where males were more abundant in April
(80% of individuals) and females more abundant in May
(64% of individuals).
Site-related variations
Both species of Mecoptera were caught in most of the
trees and stands monitored. The abundances of the two
species were signifi cantly and positively correlated (rho =
0.81, P < 0.001).
Species of Raphidioptera were more unevenly distrib-
uted (Fig. 1, Table S1). Half of the GMFT caught less than
5 individuals and fi ve no specimens. Conversely, several
traps caught large numbers of individuals (e.g. 146, 94, 32,
28 and 28). Species frequently co-occurred, and eight traps
collected four different species. Abundances of S. confi nis,
P. major and P. notata were positively correlated (rho >
0.33, P < 0.002).
Effects of decline on diversity metrics
The abundance of Mecoptera was not associated with the
level of decline at any of the three spatial scales considered
(Table 3). On the other hand, the abundance of Raphidi-
optera in general was positively associated with the level
of decline, even though abundances recorded at high lev-
Fig. 2. Seasonal variations in mean (± SE) number of species of Mecoptera (left) and Raphidioptera (right) caught both green multi-funnel
traps and fl ight interception traps (pooled data) suspended in oak canopies.
Fig. 1. Boxplots of the number of Mecoptera (left) and Raphidioptera (right) caught by green multi-funnel traps (GMFT) and fl ight intercep-
tion traps (FIT) in oak canopies.

376
Vincent et al., Eur. J. Entomol. 117 : 372–379, 2020 doi: 10.14411/eje.2020.041
els of decline were very variable (Table 3, Fig. 3). Abun-
dance increased approximately nine-fold over the gradient
in level of decline at the plot scale (Fig. 3). Although not
very diverse, the mean local species richness of Raphidi-
optera also increased with the level of decline at the stand
scale (Table 2). At the species level, signifi cant increases
in abundance were detected at the plot scale for both S.
confi nis and P. major, and a similar trend was recorded at
the stand scale for P. notata. Nonetheless, A. maculicollis
and X. xanthostigma tended to be more abundantly caught
on healthy trees than on those in decline.
DISCUSSION
Large numbers of Mecoptera were caught by the GMFTs
placed in the canopies of trees. Panorpa germanica and
P. communis are very common species and frequently en-
countered in temperate forests, where they are generally
associated with the herbaceous layer (Byers & Thornhill,
1983; Duelli, 2002). The fact that we caught so many in
the upper layers could indicate a shift in their spatial niche
within closed stands where the herbaceous and shrub lay-
ers are poorly developed, adults may turn to the canopy in
order to forage. The high catches might also be a result of a
strong attractiveness of the GMFTs. Mecoptera are known
to be attracted to yellow traps, which are frequently used
to catch them (e.g. Duelli et al., 2002; Mignon, 2002). The
green colour of the GMFT mimics the colour of foliage
and attracts several guilds of phyllophagous insects (Sallé
et al., 2020), and might evoke a favourable hunting ground
for Mecoptera.
All the species of Raphidioptera caught have bark-dwell-
ing larvae, which can develop under the bark of oaks (As-
pöck et al., 1980 in Hiermann et al., 2018; Aspöck, 2002;
Alexander, 2004). They are recorded occurring in the
canopy of pure or mixed oak stands in Europe and on oak
branches (e.g. Barnard et al., 1986; Czechowska, 1997;
Gruppe & Schubert, 2001; Alexander, 2004; Gruppe et al.,
2004; Gruppe, 2008; Weigelmeier & Gruppe, 2010). Sev-
eral species reported in previous surveys, such as Puncha
ratzeburgi (Brauer), Venustoraphidia nigricollis (Albarda)
and Dichrostigma fl avipes (Stein), were not recorded dur-
ing this study as they are restricted to Central Europe (As-
pöck & Aspöck, 2007). Conversely, surveys in Central Eu-
rope have never reported the presence of A. maculicollis,
which is restricted to Western Europe (Aspöck & Aspöck,
2007). The species recorded in this study are in accord with
both the ecological preferences and known geographical
distribution of the species. GMFTs were just as effective
at catching adult Raphidioptera as Mecoptera. The green
colour might be attractive for adults looking for aphids on
Fig. 3. Relationship between the number of Raphidioptera caught by both green multi-funnel traps and fl ight interception traps in the same
trees, and the percentage of declining trees in the plot (10 trees) surrounding the trap-bearing tree. Left: actual data, right: model prediction
and confi dence interval (see material and methods).
Table 3. Association between abundance and species richness of the different species of Mecoptera and Raphidioptera caught in oak
canopies and the incidence of declining trees.
Variable †
(mean value per tree) Best ecological model Effect of decline
Estimate SE P-value
Mecoptera
Abundance 1Stand –0.87 1.00 0.381
Panorpa communis 1Stand –0.42 1.17 0.721
Panorpa germanica 1Stand –1.14 1.21 0.347
Raphidioptera
Abundance 1Plot 2.07 0.63 0.001
Species richness 1Stand 1.29 0.51 0.011
Subilla confi nis 1Plot 3.51 0.91 < 0.001
Phaeostigma major 2Plot 4.02 0.69 < 0.001
Phaeostigma notata 2Stand 3.26 1.73 0.061
Atlantoraphidia maculicollis 2Tree –1.20 0.63 0.057
Xanthostigma xanthostigma 2Tree –0.77 0.45 0.092
Generalized linear mixed-effects models fi tted for a negative binomial 1 or Poisson distribution 2 with forest, stand, tree and trap type as
random effects; † – the variable considered for species is abundance. Lines in bold indicate signifi cant relationship between species abun-
dance and decline level.

377
Vincent et al., Eur. J. Entomol. 117 : 372–379, 2020 doi: 10.14411/eje.2020.041
foliage. However, the reason why we caught more males
than females is unknown and apparently not previously
reported.
Our results support the hypothesis that forest decline
provides opportunities for colonization by species of both
xylophagous and non-xylophagous saproxylic beetles
(Sallé et al., 2020). Although there were not many species
of Raphidioptera and they were not very abundant, mark-
edly more were caught in declining trees, plots and stands.
Duelli et al. (2019) report that most species of Neuroptera
and Raphidioptera, including Phaeostigma notata, are
more abundant in the years following a fi re or wind throws.
The abundance of larvae of Raphidioptera is positively
correlated with arthropod abundance in their larval habi-
tat (Weigelmeier & Gruppe, 2010). The accumulation of
dead branches in the canopy, with galleries and loose bark,
probably provided a wider range of oviposition sites, larval
habitats and suitable microclimates for S. confi nis, P. major
and P. notata. In addition, it could also have provided more
prey for the larvae of these species, since declining trees or
plots host more individuals, more biomass and more spe-
cies of saproxylic insects (e.g. Sallé et al., 2020). However,
the reduction in the density of foliage in the crowns of de-
clining trees may result in a lower abundance of aphids
and other Sternorrhyncha, which in turn may affect the
abundance of adult Raphidioptera. This may explain why
more A. maculicollis and X. xanthostigma tended to be
caught by traps in healthy than in trees in decline. Their
abundance, however, was quite low, and further sampling
would be necessary to confi rm this trend. Nevertheless, it
indicates that these two species may have different eco-
logical preferences compared to the species of Subilla and
Phaeostigma and is in accord with the fact that they are ac-
tive at different times each year (see below, Aspöck, 2002).
The abundance of Mecoptera was not associated with
either tree or forest health. Mecoptera are presumably
loosely dependent on trees for habitat and food since they
have ground dwelling larvae and the adults are opportunis-
tic feeders (Byers & Thornill, 1983). Nonetheless, decline-
induced defoliation can affect the microclimates in forests,
within and below the canopy and so affect the understory
and composition of the litter. This may in turn indirectly af-
fect the communities dwelling in the herbaceous and shrub
layers, as is reported for Mecoptera (Duelli et al., 2002),
but not supported by our data.
Regarding seasonal activity, our observations indicate
that P. germanica is bivoltine, as previously reported in
Belgium and the Czech Republic (Meurisse & Magis, 1989;
Mignon, 2002; Vidlicka & Holusa, 2007). Some authors
also suggest that P. communis could have a facultative sec-
ond generation in August (Meurisse & Magis, 1989; Mi-
gnon, 2002), which was not detected in our study. We also
did not record an earlier emergence of males, although it is
apparently common in other species of Mecoptera (Byers
& Thornill, 1983). As previously reported, there was a tem-
poral shift in the main activity period of the two species
of Mecoptera (Sauer 1970 in Meurisse & Magis, 1989;
Mignon, 2002; Vidlicka & Holusa, 2007). The positively
correlated abundances of these two species indicate they
have similar habitat preferences, although they may exploit
the vegetation in different layers (Sauer, 1973 in Byers &
Thornhill, 1983). As is suggested by Byers and Thornhill
(1983), the temporal shift in the periods when they are ac-
tive might result from a temporal segregation in order to
avoid competition as species of Panorpa actively compete
for food (Byers & Thornhill, 1983; Thornhill, 1987). The
species of Raphidioptera had two distinct seasonal patterns
of activity, with an early peak for A. maculicollis and X.
xanthostigma, and a slightly later peak for S. confi nis, P.
major and P. notata. This is in accord with previous obser-
vations for P. notata and S. confi nis (Barnard et al., 1986;
Duelli et al., 2002), but not those for X. xanthostigma and
A. maculicollis (Alexander, 2004). Atlantoraphidia macu-
licollis overwinter as pupae, with adults emerging early in
spring (Aspöck, 2002), as recorded in this study. The four
other species overwinter as larvae, a short pupation occurs
in spring and then the adults emerge in mid or late spring
(Aspöck, 2002). This is also similar to our results for S.
confi nis, P. major and P. notata, but not in accord with
the earlier activity of X. xanthostigma. As already men-
tioned, the abundance of the latter species was quite low
and, therefore, seasonal activity recorded for this species
needs to be confi rmed. Since the abundances of S. confi nis,
P. major and P. notata are positively correlated with each
other there is likely to be both a strong spatial and temporal
overlap in their ecological niches.
CONCLUSIONS
Our study provides new data on canopy dwelling
Mecoptera and Raphidioptera in temperate broadleaved
forests in western Europe, two taxonomic groups that are
largely understudied in that context. Our sampling proved
to be effi cient and is recommended for use in future stud-
ies on these insects. We demonstrate that some species of
Raphidioptera prefer weakened and decaying plots and are
promoted by forest decline.
ACKNOWLEDGEMENTS. This work was supported by Région
Centre-Val de Loire Project no. 2018-00124136 (CANOPEE)
coordinated by A. Sallé. We thank C. Moliard and G. Parmain
(INRAE) and X. Pineau (University of Orléans) for their techni-
cal assistance. We are also grateful to the National Forestry Offi ce
(Offi ce National des Forêts) and Forest Health Service (Départe-
ment de la Santé des Forêts), with special thanks to S. Chevalier,
A. Hachette, Y. Baugin, F. Mouy, H. Dézélut and Y. Deboisse for
their fi eld assistance. D. Fourcin edited the manuscript. Finally,
we thank the two reviewers for their valuable comments.
REFERENCES
ALEXANDER K.N.A. 2004: The larval habits of snakefl ies (Raphid-
ioptera: Raphidiidae). — Entomol. Record J. Variation 11 6:
113–114.
ALLEN C.D., MACALADY A.K., CHENCHOUNI H., BACHELET D.,
MCDOWELL N., VENNETIER M., KITZBERGER T., RIGLING A.,
BRESHEARS D.D., HOGG E.H. ET AL. 2010: A global overview
of drought and heat-induced tree mortality reveals emerging
climate change risks for forests. — For. Ecol. Manage. 259:
660–684.

378
Vincent et al., Eur. J. Entomol. 117 : 372–379, 2020 doi: 10.14411/eje.2020.041
ASPÖCK H. 2002: The biology of Raphidioptera: a review of pre-
sent knowledge. — Acta Zool. Acad. Sci. Hung. 48: 35–50.
ASPÖCK H. & ASPÖCK U. 2007: The Raphidioptera of the Apen-
nines Peninsula: a biogeographical analysis. — Ann. Mus. Civ.
Stor. Nat. Ferrara 8: 95–106.
ASPÖ CK H., ASPÖ CK U. & HÖ LZEL H. 1980: Neuropteren Europas.
Goecke & Evers, Krefeld, 495 pp.
BARNARD P.C., BROOKS S.J. & STORK N.E. 1986: The seasonality
and distribution of Neuroptera, Raphidioptera and Mecoptera
on oaks in Richmond Park, Surrey, as revealed by insecticide
knock-down sampling. — J. Nat. Hist. 20: 1321–1331.
BOUGET C. & DUELLI P. 2004: The effects of windthrow on forest
insect communities: a literature review. — Biol. Conserv. 11 8:
281–299.
BYERS G.W. & THORNHILL R. 1983: Biology of the Mecoptera.
Annu. Rev. Entomol. 28: 203–228.
CONNELL J.H. 1978: Diversity in tropical rain forests and coral
reefs. — Science 199: 1302–1310.
CZECHOWSKA W. 1997: A comparative analysis of the structure of
Neuropteroidea communities of tree canopies in linden-oak-
hornbeam forests, light oak forests, mixed coniferous forests
and pine forests. — Fragm. Faun. 40: 127–168.
DUELLI P., OBRIST M.K. & FLUCKIGER P.F. 2002: Forest edges are
biodiversity hotspots-also for Neuroptera. — Acta Zool. Acad.
Sci. Hung. 48: 75–87.
DUELLI P., WERMELINGER B., MORETTI M. & OBRIST M.K. 2019:
Fire and windthrow in forests: winners and losers in Neuro-
pterida and Mecoptera. — Alp. Entomol. 3: 39–50
GOUDET M. & NAGELEISEN L.M. 2019: Protocole Dépéris: Mé-
thode de notation simplifi ée de l’aspect du houppier des arbres
forestiers dans un contexte de dépérissement. — For. Entrepr.
246: 36–40.
GRIME J.P. 1973: Competitive exclusion in herbaceous vegetation.
Nature 242: 344–347.
GRUPPE A. & SCHUBERT H. 2001: The spatial distribution and plant
specifi city of Neuropterida in different forest sites in Southern
Germany (Raphidioptera and Neuroptera). — Beitr. Entomol.
51: 517–527.
GRUPPE A. & SOBEK S. 2011: Effect of tree species diversity on the
neuropterid community in a deciduous forest. — Acta Entomol.
Sloven. 19: 17–28.
GRUPPE A., GOSSNER M. & SIMON U. 2004: Neuropteren in Wäl-
dern Schwabens und Oberbayerns. — Beitr. Bayer. Entomo-
faunistik 6: 115–126.
GRUPPE A., GOSSNER M., ENGEL K. & SIMON U. 2008: Vertical and
horizontal distribution of arthropods in temperate forests. In
Floren A. & Schmidl J. (eds): Canopy Arthropod Research in
Central Europe-Basic and Applied Studies from the High Fron-
tier. Bioform, Nürnberg, pp. 383–405.
HIERMANN U., KOPF T. & GRUPPE A. 2018: Nachweise von Kamel-
halsfl iegen (Insecta: Neuropterida: Raphidioptera) aus Vorarl-
berg und Nordtirol (Austria occ.), sowie Streudaten aus Euro-
pa. — Inatura – Fors. Online 55: 5 S, 5 pp.
HOUSTON D.R. 1981: Stress triggered tree diseases: the diebacks
and declines. — USDA For. Serv. Rep. 41: 81, 36 pp.
IPCC 2013: IPCC, 2013: summary for policymakers. In Stocker
T.F., Qin D., Plattner G.K. et al. (eds): Climate Change 2013:
The Physical Science Basis. Cambridge University Press,
Cambridge (UK) and New York (USA), pp. 3–32.
ISHII H.T., TANABE S.I. & HIURA T. 2004: Exploring the relation-
ships among canopy structure, stand productivity, and biodiver-
sity of temperate forest ecosystems. — For. Sci. 50: 342–355.
KENIS M., WERMELINGER B. & GRÉGOIRE J.C. 2007: Research on
parasitoids and predators of Scolytidae – a review. In Lieutier
F., Day K.R., Battisti A. & Grégoire J.C. (eds): Bark and Wood
Boring Insects in Living Trees in Europe, a Synthesis. Springer,
Dordrecht, pp. 237–290.
LIEBHOLD A.M., BROCKERHOFF E.G., KALISZ S., NUÑEZ M.A., WAR-
DLE D.A. & WINGFIELD M.J. 2017: Biological invasions in for-
est ecosystems. — Biol. Invasions 19: 3437–3458.
MAGIS N. & MEURISSE X. 1989: Précisions sur la phénologie des
Panorpa Linné, 1758 de la faune de Belgique (Insecta: Meco-
ptera). — Bull. Ann. Soc. Roy. Belg. Entomol. 125: 265–272.
MANION P.D. 1981: Tree Disease Concepts. Prentice-Hall, Engle-
wood Cliffs, NJ, 399 pp.
MIGNON J. 2002: Evaluation de la diversité des Mécoptères d’un
site de la région liégeoise (ru de Targnon, Theux, Belgique). —
Notes Fauniques Gembloux 46: 84–86.
MORETTI M., OBRIST M.K. & DUELLI P. 2004: Arthropod biodi-
versity after forest fi res: winners and losers in the winter fi re
regime of the southern Alps. — Ecography 27: 173–186.
R CORE TEAM 2018: R: A Language and Environment for Statisti-
cal Computing. R Foundation for Statistical Computing, Vi-
enna. URL: https://www.R-project.org
SALLÉ A., NAGELEISEN L.M. & LIEUTIER F. 2014: Bark and wood
boring insects involved in oak declines in Europe: current
knowledge and future prospects in a context of climate change.
— For. Ecol. Manage. 328: 79–93.
SALLÉ A., PARMAIN G., NUSILLARD B., PINEAU X., BROUSSE R., FON-
TAINE-GUENEL T., LEDET R., VINCENT-BARBAROUX C. & BOUGET
C. 2020: Forest decline differentially affects trophic guilds of
canopy-dwelling beetles. — Ann. For. Sci. 77: 86, 21 pp.
SAUER K.P. 1970: Zur Monotopbindung einheimischer Arten der
Gattung Panorpa (Mecoptera) nach Untersuchungen im Frei-
land und im Laboratorium. — Zool. Jb. Abt. Syst. Geog. Biol.
Tiere 97: 201–284.
SAUER K.P. 1973: Untersuchungen zur Habitatselektion bei Pa-
norpa communis L. mit einem Beitrag zur Theorie des Begriffs
Monotop und seiner Beziehung zur okologischen Nische. —
Zool. Jb. Abt. Syst. Geog. Biol. Tiere 100: 477–496.
SEIDL R., THOM D., KAUTZ M., MARTIN-BENITO D., PELTONIEMI M.,
VACCHIANO G., WILD J., ASCOLI D., PETR M., HONKANIEMI J. ET
AL. 2017: Forest disturbances under climate change. — Nat.
Clim. Change 7: 395–402.
SÉMÉRIA Y. & BERLAND L. 1988: Atlas des Névroptères de France
et d’Europe: Mégaloptères, Raphidioptères, Névroptères Pla-
nipennes, Mécoptères. Boubée, Paris, 192 pp.
THORNHILL R. 1987: The relative importance of intra-and interspe-
cifi c competition in scorpionfl y mating systems. — Am. Nat.
130: 711–729.
TILLIER P. 2008: Contribution à l’étude des Mécoptères de France.
Deuxième partie: clé de détermination des Panorpa de France
(Mecoptera: Panorpidae). — Entomologiste 64: 21–30.
VIDLIČKA Ľ. & HOLUŠA J. 2007: Neuroptera, Raphidioptera and
Mecoptera assemblages inhabiting young spruce (Picea abies)
forests: dominance structure and seasonal activity patterns. —
J. For. Sci. 53: 74–81.
WEIGELMEIER S. & GRUPPE A. 2010: Occurence of Raphidiop-
tera larvae in dead wood of Quercus petraea (Matt.) Liebl. In
Devetak D., Lipovšek S. & Arnett A.E. (eds): Proceedings of
the Tenth International Symposium on Neuropterology. Piran,
Slovenia, June 22–25, 2008. Faculty of Natural Sciences and
Mathematics, University of Maribor, Maribor, pp. 301–306.
WICHMANN H.E. 1957: Untersuchungen an Ips typographus L.
und seiner Umwelt: Die Kamelhalsfl iegen. — Z. Angew. En-
tomol. 40: 433–440.
Received June 26, 2020; revised and accepted September 7, 2020
Published online October 6, 2020
Supplementary material follows (Table S1).

379
Vincent et al., Eur. J. Entomol. 117 : 372–379, 2020 doi: 10.14411/eje.2020.041
Table S1. Number of individuals of each species of Raphidioptera and Mecoptera caught per tree and type of trap.
Forest Stand
ID Tree # P. communis P. germanica P. major P. notata S. confi nis X. xanthostigma A. maculicollis
GMFT FIT GMFT FIT GMFT FIT GMFT FIT GMFT FIT GMFT FIT GMFT FIT
Marcenat 12 1 62 1 331 2 1 1 0 0 12 0 0 0 0 0
Marcenat 12 2 95 0 145 6 1 1 0 0 6 1 3 0 0 0
Marcenat 37 1 129 0 382 1 0 0 0 0 0 0 1 0 0 0
Marcenat 37 2 9 0 29 1 1 0 0 0 0 1 1 0 0 0
Marcenat 58 1 39 0 77 0 0 0 0 0 5 0 7 0 0 0
Marcenat 58 2 46 0 45 0 0 0 0 0 26 2 2 0 0 0
Orléans 351 1 45 3 23 1 1 0 2 0 0 0 0 0 1 0
Orléans 351 2 179 8 57 0 1 0 0 0 0 0 0 0 0 0
Orléans 751 1 11 1 22 0 0 0 0 0 0 0 1 0 3 0
Orléans 751 2 0 0 17 0 0 0 0 0 0 0 1 0 11 0
Orléans 1140 1 0 0 7 0 2 0 2 1 7 0 0 0 0 0
Orléans 1140 2 0 0 1 0 0 0 0 1 6 0 0 0 0 1
Orléans 1343 1 31 0 138 0 0 0 2 0 1 0 0 0 0 0
Orléans 1343 2 18 5 56 11 1 0 0 0 0 0 0 0 4 0
Orléans 1346 1 13 1 92 3 0 0 1 0 2 0 1 0 0 0
Orléans 1346 2 15 1 29 1 0 0 0 0 0 0 0 0 0 0
Orléans 1427 1 16 2 31 2 1 0 5 0 15 1 1 0 6 0
Orléans 1427 2 25 0 60 3 0 0 1 0 2 0 1 0 3 1
Orléans 1490 1 99 12 172 9 0 0 0 0 0 0 0 0 0 0
Orléans 1490 2 10 0 23 0 0 0 0 0 0 0 0 0 0 0
Orléans 1491 1 20 0 100 2 0 0 0 0 0 0 0 0 1 0
Orléans 1491 2 60 0 319 1 0 0 0 0 0 0 2 0 0 0
Orléans 1502 1 140 4 96 2 1 0 0 0 4 0 0 0 1 0
Orléans 1502 2 97 2 139 0 0 0 0 0 0 0 0 0 0 0
Vierzon 19 1 36 1 50 0 0 0 0 0 0 0 0 0 2 0
Vierzon 19 2 28 1 14 0 0 0 1 0 0 0 3 0 2 0
Vierzon 35 1 13 0 19 0 0 0 1 0 1 0 0 0 6 0
Vierzon 35 2 54 3 65 0 0 0 0 0 0 1 1 0 2 0
Vierzon 70 1 57 1 40 0 5 2 9 0 8 1 5 0 1 1
Vierzon 70 2 32 0 14 0 32 1 11 0 102 0 0 0 1 0
Vierzon 71 1 137 2 99 3 2 2 1 0 2 0 2 0 1 0
Vierzon 71 2 309 5 118 0 31 1 2 0 56 3 5 0 0 0
Vierzon 81 1 14 1 15 0 0 1 1 0 5 0 0 0 0 0
Vierzon 81 2 26 1 63 0 2 0 0 0 4 0 1 0 1 0
Vierzon 179 1 34 1 2 0 0 0 19 0 4 4 1 1 4 0
Vierzon 179 2 52 2 9 0 2 1 5 0 25 0 0 0 0 0
Vierzon 236 1 2 0 2 0 0 0 0 0 0 0 0 0 0 0
Vierzon 236 2 44 4 14 0 1 0 1 0 0 0 1 0 0 0
Vierzon 249 1 115 1 54 0 0 0 0 0 0 0 1 0 0 0
Vierzon 249 2 86 1 39 1 0 0 0 0 1 0 0 0 1 0
Vierzon 290 1 17 0 4 0 0 0 0 0 1 0 2 0 0 0
Vierzon 290 2 23 0 11 0 0 0 0 0 1 0 3 0 0 0
Total 2238 64 3023 49 85 10 64 2 296 14 46 1 51 3