Proposal for a time-based standard sampling method for the monitoring of Gomphus flavipes (Charpentier, 1825) and Ophiogomphus cecilia (Fourcroy, 1785) (Odonata: Gomphidae)

Abstract

Monitoring of conservation status is an obligation arising from Article 11 of the Habitats Directive for all species of community interest. However, the development of monitoring methods for invertebrate species has received relatively little attention. Gomphus flavipes (Charpentier, 1825) and Ophiogomphus cecilia (Fourcroy, 1785) are two dragonfly species, listed in the annexes of the Habitats Directive, which suffered severe declines in the last century and have since recovered. Methods for the monitoring of these two gomphids have been proposed, but these have not been extensively tested and no abundance classes have been proposed for the evaluation of the conservation status of these species. A time-based standard sampling method is proposed for both species and results from numerous sites in Lombardy, northern Italy, are presented. Applying the standard method revealed that it is common for rivers that high water levels preclude sampling of exuviae through the summer and it is better to allow for two seasons when planning the monitoring. A further result is the fact that it was not always possible to sample the same stretches as the dynamic nature of the rivers and fluctuations in water level lead to some river banks becoming unsuitable for sampling during some visits. In these cases the time-based approach was advantageous, as the method did not need to be modified in response to the original bank section becoming unsuitable.

Full text

55
Fragmenta entomologica, 51 (1): 55-62 (2019)
Research article
Submitted: November 20th, 2018 - Accepted: May 2nd, 2019 - Published: May 31st, 2019
Proposal for a time-based standard sampling method for the monitoring
of Gomphus avipes (Charpentier, 1825) and Ophiogomphus cecilia
(Fourcroy, 1785)
(Odonata: Gomphidae)
Sönke HARDERSEN 1, Ilaria TONI 1,*
1 Centro Nazionale per lo Studio e la Conservazione della Biodiversità Forestale “Bosco Fontana” Carabinieri - Strada Mantova 29,
I-46045 Marmirolo (MN), Italy - s.hardersen@gmail.com; ilariatoni2010@gmail.com
* Corresponding author
Abstract
Monitoring of conservation status is an obligation arising from Article 11 of the Habitats Directive for all species of community inter-
est. However, the development of monitoring methods for invertebrate species has received relatively little attention. Gomphus avipes
(Charpentier, 1825) and Ophiogomphus cecilia (Fourcroy, 1785) are two dragony species, listed in the annexes of the Habitats Direc-
tive, which suffered severe declines in the last century and have since recovered. Methods for the monitoring of these two gomphids
have been proposed, but these have not been extensively tested and no abundance classes have been proposed for the evaluation of the
conservation status of these species. A time-based standard sampling method is proposed for both species and results from numerous
sites in Lombardy, northern Italy, are presented. Applying the standard method revealed that it is common for rivers that high water lev-
els preclude sampling of exuviae through the summer and it is better to allow for two seasons when planning the monitoring. A further
result is the fact that it was not always possible to sample the same stretches as the dynamic nature of the rivers and uctuations in water
level lead to some river banks becoming unsuitable for sampling during some visits. In these cases the time-based approach was advan-
tageous, as the method did not need to be modied in response to the original bank section becoming unsuitable.
Key words: Dragony, monitoring protocol, exuviae, Habitats Directive, Gomphidae, Italy.
Introduction
Halting the loss of biodiversity is probably the most im-
portant challenge in nature conservation. The recognition
that the threat to species and ecosystems has never been so
great as today has led to the convention on biological di-
versity (Rio de Janeiro, 5 June 1992). As a response many
national authorities developed biodiversity action plans
and improved the monitoring of biodiversity. However,
invertebrates are mostly under-represented in internation-
al conservation measures and have largely been neglect-
ed in the literature on conservation (Clark & May 2002;
Zamin et al. 2010; Cardoso 2012), even if they dominate
among organisms in terms of richness, abundance and of-
ten biomass (Cardoso et al. 2011). This bias is also reect-
ed in the Habitats Directive (European Union 1992), as its
lists of the protected species contain few arthropod species
(Haslett 2007; Cardoso 2012).
Monitoring of conservation status is an obligation aris-
ing from Article 11 of the Habitats Directive for all species
of community interest and this provision is not restricted to
Natura 2000 sites. Even though monitoring is compulsory,
the development of methods for invertebrate species has
received little attention and currently well tested methods
are available only for few species (e.g. Svensson & Lars-
son 2008; Campanaro et al. 2016; Hardersen et al. 2017).
Monitoring is critically essential to determine the extent
to which protected areas are effective in conserving biodi-
versity (Rao & Ginsberg 2010). As the overall objective of
the Habitats Directive is to achieve and maintain favorable
conservation status for all habitats and species of commu-
nity interest, monitoring must lead to a clear picture of the
actual conservation status and its trends on various levels
and the information on biodiversity, collected locally, is
needed to report trends at local, national or international
scales and to support policies at all these scales (Mace et
al. 2005).
Gomphus avipes (Charpentier, 1825) and Ophiogom-
phus cecilia (Fourcroy, 1785), both listed in the annexes of
the Habitats Directive, suffered a severe decline in the last
century and have both recovered since the 1990s (Bou-
dot & Kalkman 2015) and this also applies to Italy, where
these species are currently categorized as “least concern”
in terms of extinction risk (Riservato et al. 2014). The
eISSN: 2284-4880 (online version)
pISSN: 0429-288X (print version)

56
Hardersen & Toni
knowledge on recovery and expansion of these two spe-
cies is mainly based on distributional data, but has not been
documented by monitoring single sites and it is often not
clear if newly discovered populations have been found-
ed recently or had been previously overlooked (Schiel &
Hunger 2006; Ketelaar 2010; Westermann 2011; Boudot
& Kalkman 2015). A number of methods for the monitor-
ing of these two gomphids have been proposed (Schnit-
ter et al. 2006; Trizzino et al. 2013; Janák et al. 2015) and
all of these are based on the collection of exuviae, as it is
well known that for gomphids exuviae are more observ-
able than adults (Hardersen 2008; Hunt et al. 2010). Ad-
ditionally, only exuviae in situ constitute completely de-
pendable evidence that larval development has been com-
pleted successfully (Corbet 1993). However, none of the
above methods proposed for the monitoring of G. avipes
and O. cecilia has been extensively tested. Such testing
seems particularly important for river systems, where it
is common that high water levels make sampling of exu-
viae during some parts of the summer impossible (Hunt et
al. 2010; Farkas et al. 2012; Janák et al. 2015). Addition-
ally, our experience has shown that the dynamic nature
and large uctuations in water level of the Italian river
systems often leads to river sections, which were initially
suitable for sampling of exuviae, can become unsuitable
only some weeks later. A further problem is that there are
currently no accepted criteria to class populations in the
three levels of conservation status recognized by the Habi-
tats Directive.
The aims of our study was to contribute to the devel-
opment of a standard method for the monitoring of the
two dragonies G. avipes and O. cecilia; in particular the
aims of this study were: 1. to test the practicability of a
time-based standard method in numerous sites. 2. to pro-
pose thresholds for three classes of abundance based on
the data gathered.
Material and Methods
Study sites
The monitoring of the two species (G. avipes and O. ce-
cilia) by means of collecting exuviae was carried out at
13 sites located in the Lombardy region. For all locations
we report identication number (id), site name, province,
geographic coordinates and dates of monitoring (Tables 1
& 2).
Seven sites were located on the river Po, three sites
on the river Oglio, and one site each on the river Terdop-
pio, the river Adda, and on the articial channel “Naviglio
Langosco” (Fig. 1).
The monitoring method for G. avipes was applied in
11 sites (located on rivers: Po, Oglio, Adda, Terdoppio,
Naviglio Langosco) and the monitoring method for Ophi-
ogomphus cecilia was applied in 6 sites (located on rivers:
Oglio, Adda, Terdoppio, Naviglio Langosco). The sites
were chosen on the basis of prior knowledge on the pres-
ence of the two species.
Sampling
In this study we applied a time-based sampling meth-
od, as the dynamic nature of Italian river systems can lead
to river sections, initially suitable for sampling of exuviae,
becoming unsuitable only some weeks later. Therefore the
methods standardize the time to be employed in river sec-
tions suitable for sampling, rather than the length of the
section (cf. Schnitter et al. 2006; Trizzino et al. 2013; Janák
et al. 2015). A section suitable for sampling was dened as
a river bank with an angle between 10° and 45° and with
the bank substrate exposed (not covered by leaves, etc.).
Shallow banks and sandbanks (< 10°) were excluded be-
cause exuviae tent not to accumulate here as any rise or
fall in water level leads to large changes in the position
of the shoreline (pers. obs.). Banks steeper than 45° were
too steep to sample exuviae safely. This denition of suit-
able sites is not applicable to the articial channel “Nav-
iglio Langosco”, which has vertical concrete walls, and is
always suitable for sampling on its entire length. Particu-
lar attention was paid to the hydrological situation of the
rivers to be sampled and we selected days during which
the water level was expected to be stable or falling and we
avoided days directly after abundant rainfall.
Exuviae were collected in the years 2011 to 2015 (Ta-
bles 1 & 2) by searching and collecting all exuviae found
in the rst meters of riverbanks (about 2-4 meters from the
shore), while walking slowly for 30 minutes in the same
direction. If non-suitable areas (e.g. sandbanks, steep and
slippery banks, overhanging or fallen trees, banks cov-
ered by leaves) (Fig. 2) were encountered, the timer was
stopped and once a new stretch of riverbank suitable for
sampling had been reached, the timer was re-started. Eve-
ry site was visited at least 5 times, with intervals as evenly
distributed as possible, between 23.V. to 03.VIII. (G. a-
vipes) and 12.VI. to 24.VIII. (O. cecilia).
All exuviae were collected in plastic jars, later air
dried and determined using Carchini (1983) and Gerken &
Sternberg (1999). All exuviae are deposited in the collec-
tion of the Centro Nazionale per lo Studio e la Conservazi-
one della Biodiversità Forestale Carabinieri “Bosco Fon-
tana”, Marmirolo, Mantova – Italy.
Results
During the study we collected a total of 1176 exuviae of G.
avipes (11 samping sites) and 79 exuviae of O. cecilia (6
sampling sites) (Tables 3 & 4) with average numbers per
site being 106 exuviae for G. avipes and 13 for O. cecil-
ia. The rivers Adda, Oglio (Gazzuolo), Po, and Terdoppio
ooded during the study and therefore it was impossible
to respect the initial sampling scheme in one year and the
samplings for these sites were completed during the fol-

57
sampling method for the monitoring of Odonata
lowing year. The highest number of exuviae found in 30
minutes was 97 for G. avipes (08.VI.2015, Po, Motteggi-
ana) and 22 for O. cecilia (19.VI.2012, Terdoppio, Pieve
Albignola).
The eld work revealed that a method based on a pre-
dened stretch of river bank would have led to practical
problems, as in a number of sites the stretch initially sam-
pled were not suitable later during the year. For example
in the site Motteggiana (Po) on 28.VII.2015 approximately
half of the river bank was occupied by a large sand bank,
which had emerged due to low water level. On the same
day the entire riverbank of the site Ostiglia (Po) was cov-
ered by a dense layer of fallen leaves (Figure 2), which
were absent only some hundred meters downstream. In
cases when the sampling could be effected without en-
countering unsuitable stretches we estimated that we sam-
pled river bank sections between 130 and 190 m in length
in 30 minutes.
The results of the eldwork were used to calculate the
abundances of the two species at the sites investigated.
First the lowest count was removed, as proposed for other
insect species (Trizzino et al. 2013; Hardersen et al. 2017).
This should allow to eliminate eventual outlier and to re-
duce the variability of the nal value. The remaining four
Table 1 – Sites in the Lombardy region (Italy) where Gomphus avipes was monitored by means of collecting exuviae (years 2011 to
2015).
Table 2 – Sites in the Lombardy region (Italy) where Ophiogomphus cecilia was monitored by means of collecting exuviae (years 2011
to 2013).
Site name
Site name
Province
Province
UTM
UTM
Sessions
Sessions
ID
ID
River: Adda
1
River: Oglio
2
River: Terdoppio
3
Articial channel: Naviglio Langosco
4
River: Po
5
6
7
8
9
10
11
River: Adda
1
River: Oglio
12
13
2
River: Terdoppio
3
Articial channel: Naviglio Langosco
4
Pizzighettone
Gazzuolo
Pieve Albignola
Tromello
Cava Manara
Senna Lodigiana
Stagno Lombardo
Viadana
Motteggiana
Ostiglia
Felonica
Pizzighettone
San Michele
Canneto sull’Oglio
Gazzuolo
Pieve Albignola
Tromello
Cremona
Mantova
Pavia
Pavia
Pavia
Lodi
Cremona
Mantova
Mantova
Mantova
Mantova
Cremona
Mantova
Mantova
Mantova
Pavia
Pavia
32T 560703 5005880
32T 625782 4989516
32T 497107 4996812
32T 488051 5005460
32T 510714 4995736
32T 543549 4997721
32T 582256 4991506
32T 615616 4976989
32T 638616 4989328
32T 670427 4990419
32T 686314 4983398
32T 560703 5005880
32T 623654 4996132
32T 608993 4999915
32T 625782 4989516
32T 497107 4996812
32T 488051 5005460
30.V.2012/19.VI.2012/29.VI.2012/20.VII.2012/03.VIII.2012
24.V.2012/19.VI.2012/29.VI.2011/12.VII.2011/26.VII.2011
30.V.2012/19.VI.2012/29.VI.2012/20.VII.2012/03.VIII.2012
23.V.2013/07.VI.2013/21.VI.2013/05.VII.2013/22.VII.2013
04.VI.2014/05.VI.2015/24.VI.2014/14.VII.2015/29.VII.2015
04.VI.2014/05.VI.2015/24.VI.2014/14.VII.2015/29.VII.2015
04.VI.2014/05.VI.2015/24.VI.2014/14.VII.2015/29.VII.2015
04.VI.2014/08.VI.2015/23.VI.2014/13.VII.2015/28.VII.2015
05.VI.2014/08.VI.2015/23.VI.2014/13.VII.2015/28.VII.2015
05.VI.2014/08.VI.2015/23.VI.2014/13.VII.2015/28.VII.2015
05.VI.2014/08.VI.2015/23.VI.2014/13.VII.2015/28.VII.2015
19.VI.2012/29.VI.2011/12.VII.2011/20.VII.2011/12.VIII.2011
12.VI.2013/27.VI.2013/17.VII.2013/06.VIII.2013/24.VIII.2013
12.VI.2013/27.VI.2013/17.VII.2013/06.VIII.2013/24.VIII.2013
19.VI.2012/29.VI.2011/12.VII.2011/20.VII.2011/12.VIII.2011
19.VI.2012/29.VI.2011/12.VII.2011/20.VII.2011/12.VIII.2011
21.VI.2013/05.VII.2013/22.VII.2013/03.VIII.2013/17.VIII.2013

58
Hardersen & Toni
values were used to calculate the average number of exuvi-
ae collected in each session. Based on these values classes
of abundance are proposed (Table 5).
Discussion
The time-based sampling of exuviae, repeated ve times,
allowed for the collection of a large number of exuviae of
G. avipes and O. cecilia in the selected sites. It is com-
mon for rivers that high water levels preclude sampling
of exuviae during some parts of the summer (Hunt et al.
2010; Farkas et al. 2012; Janák et al. 2015). Also during
this study some rivers ooded, which made it impossible
to respect the initial sampling scheme. Thus, one impor-
tant outcome of this study is that it was not possible to col-
lect exuviae during ve evenly spaced sampling sessions
in 10-11 weeks in a single year in all sites. We therefore
suggest to allow for two seasons when planning the stand-
ardized sampling of G. avipes and O. cecilia. A further
important result is the fact that it was not always possible
to sample the same stretches of the single sites as the dy-
namic nature of the rivers and uctuations in water lev-
el lead to some river banks becoming unsuitable for sam-
pling during some visits. In these cases the time-based ap-
proach was advantageous, as the methods did not need to
be modied in response to the original bank section be-
coming unsuitable. The above two points were revealed
because we attempted to apply the standard methods over
a number of years in numerous sites. Only this “testing” of
the method allowed to evaluate its practicability, to iden-
tify shortcomings and to propose improvements. Similar-
ly, Campanaro et al. (2017) found that not all monitoring
methods tested for the cerambycid beetle Rosalia alpina
Fig. 2 – On 28.VII.2015 the riverbank of the site Ostiglia (Po)
was covered by a dense layer of fallen leaves and rendered this
site unsuitable for sampling exuviae.
Fig. 1 – Sites in Lombardy region (Italy) where the monitoring of Gomphus avipes (squares) and Ophiogomphus cecilia (triangles) was
carried out between 2011 and 2015 (1 = Pizzighettone; 2 = Gazzuolo; 3 = Pieve Albignola; 4 = Tromello; 5 = Cava Manara; 6 = Senna
Lodigiana; 7 = Stagno Lombardo; 8 = Viadana; 9 = Motteggiana; 10 = Ostiglia; 11 = Felonica).

59
sampling method for the monitoring of Odonata
Table 3 – Number of G. avipes exuviae collected during 5 monitoring sessions carried out in the Lombardy region (Italy) years 2011
to 2015.
Table 4 – Number of O. cecilia exuviae collected during 5 monitoring sessions carried out in the Lombardy region (Italy) years 2011
to 2013.
Session 1
Session 1
Session 2
Session 2
Session 3
Session 3
Session 4
Session 4
Session 4
Session 4
TOT
TOT
Site name
Site name
Date
Date
Date
Date
Date
Date
Date
Date
Date
Date
N.
Exuviae
N.
Exuviae
N.
Exuviae
N.
Exuviae
N.
Exuviae
N.
Exuviae
N.
Exuviae
N.
Exuviae
N.
Exuviae
N.
Exuviae
Cava Manara
Senna Lodigiana
Stagno Lombardo
Viadana
Motteggiana
Ostiglia
Felonica
Gazzuolo
Pizzighettone
Pieve Albignola
Naviglio Langosco
TOT
San Michele
Canneto Sull’Oglio
Naviglio Langosco
Pizzighettone
Pieve Albignola
Gazzuolo
TOT
04.VI.2014
04.VI.2014
04.VI.2014
04.VI.2014
05.VI.2014
05.VI.2014
05.VI.2014
24.V.2012
30.V.2012
30.V.2012
23.V.2013
12.VI.2013
12.VI.2013
21.VI.2013
19.VI.2012
19.VI.2012
19.VI.2012
05.VI.2015
05.VI.2015
05.VI.2015
08.VI.2015
08.VI.2015
08.VI.2015
08.VI.2015
19.VI.2012
19.VI.2012
19.VI.2012
07.VI.2013
27.VI.2013
27.VI.2013
05.VII.2013
29.VI.2012
29.VI.2012
29.VI.2011
24.VI.2014
24.VI.2014
24.VI.2014
23.VI.2014
23.VI.2014
23.VI.2014
23.VI.2014
29.VI.2011
29.VI.2012
29.VI.2012
21.VI.2013
17.VII.2013
17.VII.2013
22.VII.2013
20.VII.2012
20.VII.2012
12.VII.2011
14.VII.2015
14.VII.2015
14.VII.2015
13.VII.2015
13.VII.2015
13.VII.2015
13.VII.2015
12.VII.2011
20.VII.2012
20.VII.2012
05.VII.2013
06.VIII.2013
06.VIII.2013
03.VIII.2013
03.VIII.2012
03.VIII.2012
20.VII.2011
29.VII.2015
29.VII.2015
29.VII.2015
28.VII.2015
28.VII.2015
28.VII.2015
28.VII.2015
26.VII.2011
03.VIII.2012
03.VIII.2012
22.VII.2013
24.VIII.2013
24.VIII.2013
17.VIII.2013
16.VIII.2012
16.VIII.2012
12.VIII.2011
13
13
8
7
20
39
9
33
0
0
0
142
0
0
0
0
1
1
2
36
8
15
6
97
24
41
45
1
0
0
273
1
1
0
0
22
6
30
14
30
18
12
34
25
29
38
13
4
0
217
0
1
2
0
13
3
19
52
49
72
9
90
66
28
62
14
0
0
442
2
2
2
0
9
2
17
7
14
29
1
17
9
11
1
7
4
2
102
0
3
1
2
4
1
11
122
114
142
35
258
163
118
179
35
8
2
1176
3
7
5
2
49
13
79
Table 5 – Proposed classes of abundance based on the average number of exuviae collected in 30 minutes during the four collections
with the highest counts.
Population size Abundant Frequent Rare
Gomphus avipes
Ophiogomphus cecilia
> 30
> 6
> 3-30
> 1-6
0,25-3
0,25-1
were applicable in all sites. These two examples underline
the importance of testing monitoring methods thoroughly
before publishing them.
The periods during which the exuviae were sampled
seem appropriate, as the lowest numbers of exuviae sam-
pled were the sessions 1 and 5 for both species and peak
emergence is likely to occur in the center of the periods
chosen. Farkas et al. (2012) and Horváth (2012) reported
peak emergence for G. avipes in June, whereas Müller
(1995) reported a peak in June and July, with considerable
variation between years. This is in keeping with our data,
which showed the highest number of exuviae collected in

60
Hardersen & Toni
July. Müller (1995) and Farkas et al. (2012) reported peak
emergence for O. cecilia in late May-early June, which is
in contrast with our nding as we found high numbers on-
ly from late July onwards. Also Boudot & Kalkman (2015)
indicated that the ight period for O. cecilia is later than
that of G. avipes and Torralba-Burrial et al. (2012) sug-
gested for Spain that the best time period for sampling exu-
viae of O. cecilia is 23.05 to 30.07. When considering the
best sampling period for G. avipes and O. cecilia it is im-
portant to realize that the emergence period of these spe-
cies can vary substantially between years (Müller 1995)
and any standard method necessarily needs to allow for
such variability. On the basis of our results we suggest for
Northern Italy to carry out the monitoring for the two spe-
cies during the following periods: G. avipes from 30.05 to
30.07 and O. cecilia from 15.06 to 15.08. Our suggestion
conrm the period advanced by Trizzino et al. (2013) for
O. cecilia, and slightly changes the period for G. avipes.
As the overall objective of the Habitats Directive is
to achieve and maintain favorable conservation status for
habitats and species of community interest, monitoring
must lead to a clear picture of the actual conservation sta-
tus. It is therefore necessary to repeat the monitoring at
regular intervals and to evaluate population size of the spe-
cies at numerous sites. The information on biodiversity,
collected locally, is needed to report trends at local, na-
tional or international scales and to support policies at all
these scales (Mace et al. 2005). To facilitate classication
of population size and conservation status of local popu-
lations it is necessary to provide quantitative classes. The
classes proposed here are based on the average number of
exuviae sampled during the four collections with the high-
est numbers of exuviae (Table 5). The collection with the
lowest number is not considered, in an attempt to reduce
variability of the data, following Trizzino et al. (2013). We
emphasize that the proposed abundance classes are a rst
proposal. It is clear that they are based on monitoring a
limited number of sites during a few years and that they
will probably modied in the future when more monitor-
ing data become available. Data collected over a 6-year
period at the river Oder (Germany) showed that abundanc-
es of exuviae of G. avipes and O. cecilia varied up to
83% and 57% between consecutive years (Müller 1995).
Similarly, Farkas et al.. (2012) reported a 2.6 fold differ-
ence in the number of G. avipes exuviae collected over
a period of ve years in Hungary. Thus it is important to
allow for natural uctuations in population sizes when
evaluating the conservation status of a species. Moreover,
it seems likely that abundance classes vary also in other
geographical areas. Even given the above limitations, the
abundance classes proposed here should be useful for fu-
ture comparisons.
Even if the different methods employed do not allow
for a direct comparison, it is clear that the highest aver-
age number of exuviae found in 30 minutes (G. avipes:
60,25 (Po, Motteggiana) and O. cecilia: 12 (Terdoppio,
Pieve Albignola)) are indicative of abundant populations
as often densities of exuviae of these two species are much
lower (Schiel & Hunger 2006; Kolozsvári et al. 2015).
Certainly, not all sites where these species reproduce can
reasonably be expected to host abundant populations. One
reason is that species also reproduce in sub-optimal habi-
tats, where densities are bound to be lower. The other rea-
son is that wider rivers offer more area for larval habitats
than smaller, narrower rivers and therefore exuvial den-
sities are expected to be higher along the shore (DuBois
2015). Thus, width of the river stretch sampled and habitat
suitability need to be considered when evaluating the con-
servation status of local populations.
A further important point to consider is that densi-
ties of gomphid larvae vary locally in streams (Foidl et
al. 1993, Müller 1995, Suhling & Müller 1996, DuBois &
Smith 2016) and emergence occurs in the vicinity of lar-
val habitats (DuBois & Smith 2016). Thus, also exuviae
are unlikely to be distributed evenly along the river bank
(cfr. Müller 1995). The time-based sampling of exuviae,
proposed here, which often leaves some river-stretches un-
sampled, might result in bank-sections with exuviae to be
excluded. However, we think that this is not an important
shortcoming of the proposed method. One reason is that
some of the excluded sites, such as very steep sections,
are impossible to sample from the river bank and this also
applies to the methods proposed by Schnitter et al. (2006)
and Trizzino et al. (2013) even if the authors don’t mention
this fact. In contrast Janák et al. (2015) acknowledges the
difculties of sampling exuviae on rivers with steep, slip-
pery banks. A further point to consider is, that normally
the river sections excluded are small in comparison with
the total stretch sampled (not more than the 10% of the
total length of the transect) and in some cases no stretch-
es were excluded. Additionally, larvae of gomphids, such
as G. avipes and O. cecilia live buried in the sediment, a
behavior which reduces drift (Suhling and Müller 1996).
This has been experimentally conrmed by DuBois &
Smith (2016), who showed for Ophiogomphus rupinsulen-
sis that 97% of exuviae moved less than 60 m from their
larval habitat. It seems thus likely that a total length of 130
m to 190 m of river bank should provide a stretch long
enough to represent the larvae present locally.
We advocate to apply the proposed monitoring to doc-
ument long term trends in local population sizes, prefer-
ably investigating more sites in the selected river reaches.
Combining the long term trends of many sites, it will be
possible to evaluate changes in the conservation status of
G. avipes and O. cecilia , also at the regional and nation-
al level. Both species suffered a severe decline in the last
century, recovered since the 1990s (Boudot & Kalkman
2015) and today are protected by the Habitats Directive. It
is to be hoped that these species will not again suffer simi-
lar declines in the future. Monitoring populations, as im-
posed by the Habitats Directive, should enable us to detect
early warning signs and to act quickly. The methods pro-

61
sampling method for the monitoring of Odonata
posed are intended to provide a methodological basis for
successful long term monitoring.
Densities of exuviae of O. cecilia were generally low-
er than those of G. avipes and the former species is less
common in the Po Plains. It was thus a positive surprise
that exuviae of O. cecilia were found (in low numbers) in
six of the seven sites chosen for the monitoring of G. a-
vipes along the river Po. The highest number of exuviae of
O. cecilia was found at Cava Manara (14) while the only
site without exuviae of this species was Stagno Lombardo.
It thus seems that O. cecilia is present in more rivers in
northern Italy than currently acknowledged and it would
be important to intensify surveys to map the current distri-
bution of this protected species in Italy. This certainly, also
applies to G. avipes.
Acknowledgments – Part of the research was funded by the Ob-
servatory for Biodiversity of the region Lombardy. We thank Ga-
briele Gheza for providing the data for Naviglio Langosco and
Gianandrea La Porta for critical comments on an earlier version
of the manuscript.
References
Boudot J.P., Kalkman V.J. 2015. Atlas of the European dragon-
ies and damselies. KNNV Publishing, The Netherlands,
381 pp.
Campanaro A., Zapponi L., Hardersen S., Méndez M., Al Fu-
laij N., Audisio P., Bardiani M., Carpaneto G.M., Corezzola
S., Della Rocca F., Harvey D., Hawes C., Kadej M., Karg
J., Rink M., Smolis A., Sprecher E., Thomaes A., Toni I.,
Vrezec A., Zauli A., Zilioli M., Chiari S. 2016. European
monitoring protocol for the stag beetle, a saproxylic agship
species. Insect Conservation and Diversity, 9: 574–584.
Carchini G. 1983. Odonati (Odonata). Guide per il riconoscimen-
to delle specie animali delle acque interne italiane. Consiglio
Nazionale delle Ricerche, AQ/1/198: 80 pp.
Cardoso P. 2012. Habitats Directive species lists: urgent need
of revision. Insect Conservation and Diversity, 5: 169–174.
Cardoso P., Erwin T.L., Borges P.A.V., New T.R. 2011. The
seven impediments in invertebrate conservation and how to
overcome them. Biological Conservation, 144: 2647–55.
Clark J.A., May R.M 2002. Taxonomic bias in conservation re-
search. Science, 297: 191–192.
Corbet P.S 1993. Are Odonata useful as bioindicators? Libellula,
12: 91–102.
DuBois R.B. 2015. Detection probabilities and sampling rates
for Anisoptera exuviae along river banks: inuences of bank
vegetation type, prior precipitation, and exuviae size. Inter-
national Journal of Odonatology, 18: 205–215.
DuBois R.B., Smith W. 2016. Pre-emergent movements and sur-
vival of F-0 larvae of Ophiogomphus rupinsulensis (Odo-
nate: Gomphidae) in a northern Wisconsin river. Internation-
al Journal of Odonatology, 19: 83–93.
European Union 1992. Council directive 92/43/EEC 1992 on the
conservation of natural habitats and of wild ora and fauna.
Ofcial Journal of the European Communities, 206: 7–49.
Farkas A., Jakab T., Tóth A., Kalmár A.F., Dévai G. 2012. Emer-
gence patterns of riverine dragonies (Odonata: Gomphidae)
in Hungary: variations between habitats and years. Aquatic
Insects: International Journal of Freshwater Entomology, 34:
77–89.
Foidl J., Buchwald R., Heitz A., Heitz S. 1993. Untersuchungen
zum Larvenbiotop von Gomphus vulgatissimus Linne 1758
(Gemeine Keiljungfer; Gomphidae, Odonata). Mitteilun-
gen des badischen Landesvereins für Naturkunde und Natur-
schutz, 15: 637–660.
Gerken B., Sternberg K. 1999. Die Exuvien Europaischer Libel-
len (Insecta Odonata): The Exuviae of European Dragonies.
Arnika & Eisvogel, Höxter, 354 pp.
Hardersen S. 2008. Dragony (Odonata) communities at three
lotic sites with different hydrological characteristics. Italian
Journal of Zoology, 75: 271–283.
Hardersen S., Bardiani M., Chiari S., Maura M., Maurizi E., Ro-
versi P.F., Mason F., Bologna M.A. 2017. Guidelines for the
monitoring of Morimus asper funereus and Morimus asper
asper. Nature Conservation, 236: 205–236.
Haslett J.R. 2007. European Strategy for the Conservation of In-
vertebrates. Council of Europe Publishing, Strasbourg, 91 pp.
Horváth G. 2012. Assessment of riverine dragonies (Odona-
ta: Gomphidae) and the emergence behaviour of their larvae
based on exuviae data on the reach of the river Tisza in Sze-
ged. Tiscia, 39: 9–15.
Hunt P.D., Blust M., Morrison F. 2010. Lotic Odonata of the
Connecticut River in New Hampshire and Vermont. North-
eastern Naturalist, 17: 175–188.
Janák M., Černecký J., Saxa A. (eds.) 2015. Monitoring of ani-
mal species of Community interest in the Slovak Republic,
results and assessment in the period of 2013-2015. Banská
Bystrica: State Nature Conservancy of the Slovak Republic.
300 pp.
Ketelaar R. 2010. Recovery and further protection of rheophilic
Odonata in the Netherlands and North Rhine-Westphalia.
Brachytron, 12: 38–49.
Kolozsvári I., Szabó L.J., Dévai G.Y. 2015. Occurrence pattern
analysis of dragonies (Odonata) on the river Tisza between
Vilok and Huszt based on exuviae. Applied Ecology and En-
vironmental Research, 13: 1183–1196.
Mace G., Delbaere B., Hanski I., Harrison J., Garcia Novo F.,
Pereira H., Watt A., Weiner J. 2005. A User’s Guide to Bio-
diversity Indicators, European Academy of Sciences Advi-
sory Council, 42 pp.
Müller O. 1995. Ökologische Untersuchungen an Gomphiden
(Odonata: Gomphidae) unter besonderer Berücksichtigung
ihrer Larvenstadien. Dissertation, Mathematisch-Natur-
wissenschaftliche Fakultät I der Humboldt - Universität zu
Berlin, Institut für Biologie, Cuvillier Verlag Göttingen,
234 pp.
Rao M., Ginsberg J. 2010. From conservation theory to practice:
crossing the divide, pp 284–312, in: Sodhi N.S.; Ehrlich, P.R.
(eds) Conservation Biology for All. Oxford University Press,
New York.
Riservato E., Fabbri R., Festi A., Grieco C., Hardersen S., Lan-
di F., Utzeri C., Rondinini C., Battistoni A., Teoli C. 2014.
Lista Rossa IUCN delle libellule italiane. Comitato Italiano
IUCN e Ministero dell’Ambiente e della Tutela del Territorio
e del Mare, Roma, 39 pp.
Schiel F.J., Hunger H. 2006. Bestandssituation und Verbreitung
von Ophiogomphus cecilia in Baden-Württemberg (Odona-
ta: Gomphidae). Libellula, 25: 1–18.
Schnitter P., Eichen C., Ellwanger G., Neukirchen M., Schröder
E. 2006. Empfehlungen für die Erfassung und Bewertung
von Arten als Basis für das Monitoring nach Artikel 11 und
17 der FFH-Richtlinie in Deutschland. Berichte des Lande-
samtes für Umweltschutz Sachsen-Anhalt (Halle), Sonder-
heft, 370 pp.
Suhling F., Müller O. 1996. Die Flussjungfern Europas. Magde-
burg & Heidelberg: Westarp-Wissenschaften & Spektrum
Akademischer Verlag, 237 pp.

62
Hardersen & Toni
Svensson G.P., Larsson M.C. 2008. Enantiomeric specicity in a
pheromone-kairomone system of two threatened saproxylic
beetles, Osmoderma eremita and Elater ferrugineus. Journal
of Chemical Ecology, 34: 189–197.
Torralba-Burrial A., Ocharan F.J., Outomuro D., Azpilicueta
Amorín M., Cordero Rivera A. 2012. Ophiogomhus cecilia.
In: VV.AA., Bases ecológicas preliminares para la conser-
vación de las especies de interés comunitario en España: In-
vertebrados. Ministerio de Agricultura, Alimentación y Me-
dio Ambiente. Madrid, 50 pp.
Trizzino M., Audisio P., Bisi F., Bottacci A., Campanaro A.,
Carpaneto G.M., Chiari S., Hardersen S., Mason F., Nardi
G., Preatoni D.G., Vigna Taglianti A., Zauli A., Zilli A., Cer-
retti P. (eds) 2013. Gli Artropodi Italiani in Direttiva Habitat:
biologia, ecologia, riconoscimento e monitoraggio. Quader-
ni Conservazione Habitat, 7. CFS-CNBFVR, Centro Nazio-
nale Biodiversità Forestale. Cierre Graca, Sommacampag-
na, Verona, 256 pp.
Westermann K. 2011. Die Asiatische Keiljungfer (Gomphus
avipes) am Restrhein zwischen Weisweil (Landkreis Em-
mendingen) und Rust (Ortenaukreis) – eine neu eingewan-
derte oder bisher übersehene Art? Naturschutz am südlichen
Oberrhein, 6: 155–156.
Zamin T.J., Baillie J.E.M., Miller R.M., Rodríguez J.P., Ardid
A., Collen B. 2010. National Red Listing beyond the 2010
target. Conservation Biology, 24: 1012–1020.