UNIVERSITE DE PERPIGNAN VIA DOMITIA

MASTER "BIOLOGIE INTEGREE : MOLECULES, POPULATIONS ET DEVELOPPEMENT DURABLE"

MENTION PROFESSIONNELLE "BIODIVERSITE ET DEVELOPPEMENT DURABLE"

EFFECTS OF FOREST

COMPOSITION ON

ACTIVITY AND

STRUCTURE OF BAT

COMMUNITIES. LANDES DE GASCOGNE FOREST, SOUTHWESTERN FRANCE

GAÜZERE PIERRE

ANNEE UNIVERSITAIRE 2011 -2012

SOUS LA DIRECTION DE LUC BARBARO ET YOHAN CHARBONNIER

BIOGECO – UMR 1202 INRA BORDEAUX 1

REMERCIEMENTS

« Une âme délicate est gênée de savoir qu'on lui doit des remerciements, une âme grossière, de

savoir qu'elle en doit. »

Friedrich Nietzsche

Au risque de passer pour un grossier personnage, je tiens à remercier tous ceux qui m’ont fait

confiance et m’ont accompagné pour me permettre de réaliser cette étude dans les meilleures

conditions qui soient.

Ainsi je tiens à remercier particulièrement Luc Barbaro et Yohan Charbonnier. Premièrement pour

m’avoir offert l’opportunité de réaliser mon stage de fin d’étude sur un sujet aussi passionnant que

celui ci. D’autre part pour avoir bénéficié d’un encadrement et d’un appui pédagogique très

enrichissant qui m’ont permis de progresser dans de nombreux domaine, et d’appréhender mon

travail de la meilleure des façons. Enfin, pour m’avoir laissé une marge de manœuvre et d’initiative

conséquente, nécessaire à l’épanouissement de tout stagiaire. Merci surtout pour votre bonne

humeur, votre disponibilité, votre gentillesse, et vos conseils !

D’une manière plus générale, merci à l’ensemble de l’équipe Ecologie des Communautés pour son

accueil et son soutien. En particulier un grand merci à Jean Yves Barnagaud pour son aide

précieuse, sa disponibilité et sa pédagogie à toute épreuve, ainsi qu’à Fabrice Vétillard pour sa

réactivité et son efficacité, qui ont grandement participé au confort dans lequel ce travail a pu être

réalisé.

Pour finir, merci aux nombreuses chauves-souris qui se sont prêtées au jeu et n’ont eu de cesse de

crier sous nos détecteurs pour révéler quelques uns de leurs secrets intimes. Ce modèle d’étude tout

à fait nouveau pour moi s’est révélé particulièrement grisant, frustrant parfois, mais toujours

passionnant.

CONTENTS

1 Introduction ........................................................................................................................ 1

2 Materials & Methods ......................................................................................................... 4

2.1 Study area and design ............................................................................................................ 4

2.1.1 The Landes de Gascogne forest ..................................................................................... 4

2.1.2 Design ............................................................................................................................ 5

2.2 Data collection ....................................................................................................................... 5

2.2.1 Bat sampling .................................................................................................................. 5

2.2.2 Echolocation call analysis .............................................................................................. 6

2.2.3 Habitat characterization ................................................................................................. 6

2.3 Data analysis.......................................................................................................................... 8

2.3.1 Species activity patterns during night ............................................................................ 8

2.3.2 Effects of abiotic factors on bat activity ........................................................................ 8

2.3.3 Effects of habitat variables effects on bat activity ......................................................... 8

2.3.4 Community assemblages among plots ........................................................................... 9

3 Results .............................................................................................................................. 10

3.1 Bat community sampling..................................................................................................... 10

3.2 Species activity patterns during the night............................................................................ 11

3.3 Effects of abiotic factors on bat activity .............................................................................. 12

3.4 Habitat variables effects on bat activity .............................................................................. 12

3.5 Community assemblage among plots .................................................................................. 14

4 Discussion ........................................................................................................................ 16

4.1 Bat community sampling..................................................................................................... 16

4.2 Species activity patterns during the night............................................................................ 17

4.3 Effects of abiotic factors on bat activity .............................................................................. 17

4.4 Influence of forest composition ........................................................................................... 18

4.5 Influence of habitat structure ............................................................................................... 20

4.6 Management ........................................................................................................................ 21

4.7 Conclusion ........................................................................................................................... 22

5 Bibliography..................................................................................................................... 23

1

1 INTRODUCTION

The contribution of bats to ecosystem services has been widely overlooked until recently. Their role

on the control of pests is significant, as they provide essential top-down control of herbivorous

insects (Williams-Guillén et al., 2008; Böhm et al., 2011; Kunz et al., 2011). For instance, herbivore

populations are more intensely affected by bats than by birds in forests (Kalka et al., 2008).

The 92/43/CEE Directive of 21 May 1992 on the conservation of natural habitats and wild fauna

and flora requires the strict protection of all European bat species (listed in Annex IV), and the

designation of special conservation areas for the 12 species listed in Annex II. More than half of the

34 species of western Europe are forest bats (Arthur & Lemaire, 2009). Chiropterans are

particularly vulnerable to human disturbance and environmental changes. Loss or modifications to

nursery sites, roosts and foraging grounds can lead to dramatic decrease in bat populations (Boyd &

Stebbings, 1989). Thus, human - induced modifications in land use is known as one of the several

causal factors responsible for drastic population declines for some European bat species (Stebbings

& Griffith, 1986; Arthur & Lemaire, 2009).

In particular, intensive logging simplifies vegetation structure and composition, and removes dead

trees. Such treatments reduce roost and food availability, thus affecting distribution, diversity and

density of bats (Mccomb et al., 1986; Graves, 1999; Russo et al., 2010). Hence, deciphering finely

the factors that influence bat occurrence and predation pressure is therefore essential to evaluate

their conservation status as well as their contribution as providers of services such as biological

control to silviculture.

As for many taxa, the broad-scale spatial distribution of bat species are primarily controlled by

climatic factors, while landscape and habitat features influence fine-scale abundance patterns

(Ethier & Fahrig, 2011; Razgour et al., 2011). At landscape and local scales, habitat features are

known to influence fine-grained patterns in animal populations in general (Fahrig, 2003) and

specifically for bat distribution (Ethier & Fahrig, 2011). Landscape composition is an important

factor influencing bat abundance within a region (Walsh & Harris, 1996a), but connectivity

between suitable habitats are also preponderant (Walsh & Harris, 1996b). At very local scales,

structure and composition of forest stands strongly influence habitat quality (Kalcounis et al., 1999;

Kanuch et al., 2008; Adams et al., 2009; Plank et al., 2011; Jung et al., 2012) by affecting the

availability and accessibility of food, roosts, protection from predators (Baxter et al., 2006) and

2

micro-climatic conditions (Chen et al., 1999). First, food availability (Walsh & Harris, 1996b;

Müller et al., 2012; Dodd et al., 2012) and roost density (Walsh & Harris, 1996a; Brigham et al.,

1997; Menzel et al., 2002; Elmore et al., 2004; Russo et al., 2005, 2010) play a major role in

structuring forest-dwelling bat communities. It is thus essential to address the effects of forest

management on bats through its effects on food availability. Forest architecture also determine

wildlife movements, which is particularly decisive for the distribution of flying vertebrates such as

birds or bats (Jung et al., 2012). Morphology highly influences bat habitat selection and foraging

grounds. Wide wingspan species have fast and straight flight, while smaller species have a slow and

sinuous flight (Norberg & Rayner, 1987). Therefore, larger species are restricted to open habitats

while more maneuverable bats use natural resources from more closed habitats (Aldridge &

Rautenbach, 1987; Crome, 1988; Krusic et al., 1996).

Managers therefore need to be informed about bat responses to forest structure and composition. In

the French forest of Landes de Gascogne (see figure 1.), stands are dominated by monospecific

maritime pine (Pinus pinaster) plantations (Maizeret, 2005), resulting in a low stand complexity

(Lindenmayer et al., 2006). Presence of deciduous trees, generally as shrub in the understorey layer,

represents the most common structural complexification and is generally related to low logging

intensity and old stands. Multi-taxa studies conducted in the Landes de Gascogne forest (Barbaro et

al., 2005, 2007; van Halder et al., 2008) highlighted the benefits of deciduous and mixed stands on

abundance and diversity of birds, carabids, spiders and butterflies.

Potential benefits of mixed-species production forests over monocultures are largely documented by

an increasing body of scientific literature. These benefits involve (i) higher habitats diversity for

biodiversity, (ii) improved growth rates, (iii) better recreational value, (iv) improved soil conditions,

(v) reduced damage to focal tree species by grazers pest insects and fungal pathogens and (vi)

reduced risk of wind and fire damage (Hartley, 2002; Lindenmayer et al., 2006; Nichols et al.,

2006; Felton et al., 2010; Richards et al., 2010)

While the effects of mixed species forest are well documented in European birds (Adams &

Edington, 1973; Felton et al., 2010), we are not aware of any published study highlighting the

effects of polycultural logging on bats in temperate forest. However, Kalcounis et al., (1999) report

a higher activity in mixed wood compared with monocultures in nearctic boreal forest.

Moreover, our understanding of European forest-dwelling bats ecology and behavior is still poor.

Some studies (Krusic et al., 1996; Menzel et al., 2002; Jung et al., 2012) highlight the use of habitat

and roosts needs, or foraging patterns and response to prey abundance (Müller et al. 2012) in

3

European forests. Given their ecological importance, bats have to be included in future forests

conservation and management plans and should also be considered in agricultural management

strategies based on natural pest control. Even so, they are still an overlooked faunal assemblage, and

habitat management is limited by the poor understanding of their ecology.

We investigated the effects of forest composition and structure on bat species richness and activity.

We specifically tested the following predictions:

i. Activity and taxonomic diversity increase with deciduous tree proportion at stand scale

ii. This increase is driven by food and/or roost site availability and/or vegetation structure

iii. Vegetation structure, roost site and food availability increase with deciduous tree proportion.

We used automatic bat call sampling in 21 stands of a French forest, along a pine-oak mixture

gradient. We also monitored vegetation structure, availability and diversity of prey and roost site

availability. Besides the value of exploring the stand-scale responses of a sensitive and little known

taxa to environmental variables in a responsible forest management perspective; the issue is

particularly sensitive in the Landes de Gascogne forest where bats could be crucial regulators of

herbivorous insect pests.

4

2 MATERIALS & METHODS

2.1 STUDY AREA AND DESIGN

2.1.1 THE LANDES DE GASCOGNE FOREST

Study took place in the north of the Landes de Gascogne forest. This area consists of an intensively

managed forest spanning c.10 000 km² (44°40’N to 44°44’N, 0°57’W to 0°46’W, see figure1.).

Coniferous Mixt Deciduous

Sampling plots

Figure 1.Sampling plots in the Landes de Gascogne Forest. A=Situation of the Landes de Gascogne forest in

Southwestern France. B=Maritime pine plantation of the Landes de Gascogne forest bringed out in purple.

C = position of the 21 sampling plots in study area..

A B

C

5

Landscape is largely dominated by stands of native maritime pine (Pinus pinaster) with a rotation

cycle of 40-50 years, creating mosaic landscapes composed of different even-aged stands from

clearcuts to mature pine stands. Remnants of deciduous woodlands dominated by Quercus robur,

Quercus pyrenaica and Betula pendula represent the only smalls (<5ha) and isolated patches of

semi-natural habitats within this plantation forest (Barbaro et al., 2007). The area is under influence

of a thermo-atlantic climate (mean annual temperature 12°C, mean annual rainfall 700 mm) and

the elevation is low (c. 50 m a.s.l.). Podzols established on a sandy substrate account for most soils

(Maizeret, 2005).

2.1.2 DESIGN

Experimental units consist of 50 m ray circular forest plot with homogenous forest composition and

structure. In order to avoid any bias in the monitoring of bat activity, each plot respected the

absence of crossing paths, water bodies, buildings and a minimal distance of 1.6km from a major

road (>30000 vehicles per day, Berthinussen & Altringham, 2012) . A number of 21 plots were

distributed in forest stands with variable proportions of pedunculate oaks (Quercus robur) and

maritime pines (Pinus pinaster) to achieve the most possible continuous gradient of oak-pine

mixture (see section 2.2.3.1). The geographical distribution of the plots was carried out randomly

and according to availability in aimed stands.

2.2 DATA COLLECTION

2.2.1 BAT SAMPLING

Automatic ultrasound bat detector systems (Sound Meter SM2Bat + platform, Wildlife acoustics)

fitted with multidirectional microphones (SMX-US weatherproof ultrasonic microphone, Wildlife

acoustics) were used for bat activity sampling. Devices were calibrated so as to detect all bat calls

while avoiding noise recording (sampling rate: 38.4 kHz, channel = MonoL, compression=WAC0,

Dig HPF left = fs/48, effective parameters remained constant throughout the study). Timers were

set to record the full night (1 h before sunset until 1 h after sunrise). Bat detector systems were

deployed in the center of each circular plot. Devices range detection is extremely variable,

depending on vegetation, obstacles, humidity, temperature, air pressure, wind, and frequency ,

amplitude, loudness and directionality of the bat call itself. According to Wildlife acoustics (see

http://www.wildlifeacoustics.com/support-resources/frequently-asked-questions), most bat species

can be detected well over 30m, with a likely maximum of about 100m. As all sampling plots are

6

within forest, a guestimate is that maximal range detection of the SM2 Bat+ detector should be

around 50 m. This distance corresponds to the ray of experimental unit (see 2.1.2.) Each plot was

investigated for two consecutive nights during May and June 2012. We did not sample on nights

with rain, when wind was >30 km/h or when the ambient temperature fell below 10.0°C (Hayes,

1997; Bat Conservation Trust, 2007).

2.2.2 ECHOLOCATION CALL ANALYSIS

Recorded sequences were stored, listened and measured with the software BatSound 4.1 (Pettersson

Elektronik AB, Uppsala, Sweden). This application dedicated to acoustic analysis allows to

transform acoustics data into sonograms. Following the Barataud method (Barataud, 2012), all

sequences and sonograms were analyzed by a trained operator to spot bat calls and identify the

species or the lowest possible Operational Taxonomic Unit (OTU) behind the recorded pass.

Number of bat passes and number of minutes sampled were recorded. A pass was defined as a

series of two or more consecutive calls with a maximum duration of 5 seconds, separated from

other calls by one second or more (e.g. Thomas, 1988; Hayes, 1997; Wilson & Barclay, 2006).

Walsh et al., (2004) suggested that bat activity (number of bat passes per species at a site) may be

used to quantify bat relative abundance provided that:

(i) there is no change in equipment sensitivity over time;

(ii) there is no trend in species detectability across time or sites;

(iii) passes can be reliably and consistently identified to species

(iv) survey points are consistent from site-to-site.

Since our sampling protocol meets these assumptions, we used bat activity as a measure of relative

abundance. Moreover, characteristics of echolocation calls permit to assess foraging activity.

During predation, bats emit short echolocation pulses at high repetition rate just before contact with

the prey. These pulses are known as feeding buzzes (Griffin et al., 1960). For each recorded pass,

we noted presence or absence of a feeding buzz.

2.2.3 HABITAT CHARACTERIZATION

2.2.3.1 Mixture gradient assessment

Mixture gradient was represented by the percent of basal area occupied by deciduous species trees

in each plot. Measures of basal area of the stems in a stand were performed following the angle

count method. (Bitterlich, 1984). Scores from deciduous or coniferous species were separately

noted, thus permitting calculation of deciduous species basal area, coniferous species basal area and

total basal area. Three tallying were performed from randomly distributed points within each plot.

7

From this gradient, 3 modalities were also defined to segregate plots: “Pine” (<25% deciduous

basal area), “Mixt” (25% - 75% deciduous basal area) and “Deciduous” (>75% deciduous basal

area).

2.2.3.2 Vegetation structure

To define vegetation structure within forest plots, we used a method inspired from Prodon &

Lebreton (1981), as the difference that only 4 strata were distinguished (see Table 1). Within each

layer, the relative degree of cover (Cv) was estimated by comparison with a reference figure (see

Appendix I). Cover of a layer is defined as the percentage of ground covered by the vertical

projection of the aerial portion of plants from the layer in a 0.2ha buffer. Small openings in the

canopy and intraspecific overlap are excluded (Prodon & Lebreton, 1981; Anderson, 1986). Cover

variables of each strata were estimated visually by the same two observers within each experimental

units during may 2012.

Mean height of canopy was also measured using a rangefinder. Three measures were recorded

within the plot and averaged to obtain mean canopy height value.

Cover of Herbaceous plants (<0.5m) Cv0.5%

Shrub Cover (50cm-2m) Cv2%

Tree cover (2m-16m) Cv16%

High tree cover (16m-32m) Cv32%

Vegetation height Vh (m)

Number of layer with Cv% > 25% Nlayer

Index of stratification diversity (Shannon and Weaver)

Mean canopy height Height

Table 1. Vegetation structure variables

2.2.3.3 Roost site availability

To roughly assess the susceptibility of each plot to contain suitable roosts, we noted each tree

corresponding to the 1 to 7 decay stages, based on the British Columbia wildlife-tree classification

system (see Appendix II). This system is based on characteristics such as the percentage of bark

remaining, number of limbs present, condition of the top, heartwood and sapwood (Vonhof &

Barclay, 1996; Brigham et al., 1997). We also noted all trees presenting cavities and/or peeled bark.

8

2.3 DATA ANALYSIS

2.3.1 SPECIES ACTIVITY PATTERNS DURING NIGHT

For each hour, percent of total activity recorded on all nights was calculated to bring up activity

patterns of bats species throughout the night. Patterns were calculated only with species with more

than 30 contacts.

2.3.2 EFFECTS OF ABIOTIC FACTORS ON BAT ACTIVITY

We analyzed the effects of abiotic variables on bat activity with Generalized Linear Mixed Models

(GLMM) (Bolker et al., 2009) in R 2.14.1 (R Development Core Team, 2010) and package lme4

(Bates, 2005). The response variable was the number of call sequences in each sampling night

(Actn). The explanatory variables were Proportion of visible moon (moon), Minimal night

temperature (temp.min in °C) and Session of recording (session, May or June). We accounted for

the hierarchical structure of the data (repeated sampling period) by adding simple random “plot”

effect to the model intercept.

EQUATION (1)

Αctn = αp + β*moonn + γ*temp.minn + δ*sessionn + (1|plot)

Model validation was assessed by “residuals repartition” and “deviation from normality”(normal Q-

Q) interpretation plots (see Appendix III). The model was heteroscedastic and unbiased. We fitted

all possible fixed effect structures nested within maximum model (1) and ranked them on the basis

of their Akaike’s Information Criterion adjusted for small samples (AICc; Burnham & Anderson,

2002,). We considered that the average of all models differing by less than 2 AICc units was the

best possible fixed effects model structure given our data. We considered only relative measures of

model performance as provided by the AICc, as no accurate measure of fit exists for mixed models

(Orelien & Edwards, 2008).

2.3.3 EFFECTS OF HABITAT VARIABLES EFFECTS ON BAT ACTIVITY

2.3.3.1 Community level analysis

We analyzed the effects of local habitat in the same way as 2.3.2. The response variable was the

number of call sequences in each hour/sampling night/plot (Acti,j,h). The explanatory variables were

Cover of Herbaceous plants (cv0.5), Low shrubs Cover (cv2), Tree cover (cv16), High tree cover

(cv32), Vegetation height (vh, in m), Indice of structure diversity (isw) and Number of dead tree

(dt). We accounted for the hierarchical structure of the data by adding nested random “session” (s),

“plot” (p) and “hour” (h) effects to the model intercept.

9

EQUATION (2)

Αcti,j,h=αs,p,h + β*cv0.5p + γ*cv2p + δ*cv16p + ε*cv32p +ζ*Vhp + η*Iswp +

θDtp + (1| session/plot/hour)

We fitted all possible fixed effect structures nested within maximum model (2) and ranked them on

the basis of their Akaike’s Information Criterion adjusted for small samples (AICc; Burnham &

Anderson, 2002,) as for precedent model (1).

2.3.3.2 Species level analysis

The same model (2) was also fitted on species or Operational Taxonomic Unity (OTU, see 3.1) with

enough contact to ensure receivable model structure (total number of contact > 190).

2.3.4 COMMUNITY ASSEMBLAGES AMONG PLOTS

A sampling plots x species matrix was created, where each cell encode the presence (1) or absence

(0) of the bat species. Two plots with less than two occurrences of species were excluded from the

analysis. Considering this matrix, Non-Metric Multi Dimensional Scaling (NMDS) using Bray-

Curtis dissimilarities was performed with R 2.14.1 (R Development Core Team, 2010) by

metaMDS function from vegan R package, (Holland, 2008; Oksanen, 2011). In order to fit

ecological variables onto species ordination among plots, environmental fitting from vegan R

package was calculated with all environmental variables (see Table 1). Non-metric

multidimensional scaling is an ordination technique that differs in several ways from nearly all

other ordination methods.

(i) Contrary to other ordination methods, a small number of axes are explicitly chosen prior to

the analysis and the data are fitted to those dimensions; there are no hidden axes of variation.

(ii) In contrast to other ordination methods are analytical and therefore result in a single unique

solution to a set of data, NMDS is a numerical technique that iteratively seeks a solution and stops

computation when an acceptable solution has been found.

(iii) MDS is not an eigenvalue-eigenvector technique like principal components analysis or

correspondence analysis that ordinates the data such that axis 1 explains the greatest amount of

variance. As a result, an MDS ordination can be rotated, inverted, or centered to any desired

configuration.

10

3 RESULTS

3.1 BAT COMMUNITY SAMPLING

We recorded 14738 bats calls in 926 hours of recording divided into 84 nights. More than 79% of

all bat calls were identified to species level. Otherwise, calls were assigned to an OTU combining

two or more species:

- Nyctaloid potentially including N. leisleri and Eptesicus serotinus

- Myotis sp. potentially including all Myotis species

- Plecotus sp. potentially including all Plecotus austriacus and Plecotus auritus

- P.kuhnat potentially including P.kuhlii and P.nathusius

P. nathusius is considered as a rare species in Aquitaine, and P. kuhlii as a species widely

distributed. There is a good chance that contacts attributed to the OTU P.natkuh reflect P. kuhlii

activity. (see Appendix VI.).

Altogether, we detected at least twelve bat species. Activity was almost six times higher in May

than in June. Four species or OTU represented 97.5 % of total activity. Pipistrellus pipistrellus was

by far the most contacted species with 58.6 % of activity. Generally, the most contacted species

were present in a large part of plots. P.pispistrellus was totally ubiquitous. P.kuhnat, E.serotinus

and P.kuhlii was also widely distributed through sampling sites.

OTU May June Overall % Activity % Presence

Pipistrellus pipistrellus 7397 1236 8633 58.6 100.0

P.kuhnat 2561 329 2890 19.6 85.7

Eptesicus serotinus 1257 231 1488 10.1 90.5

Pipistrellus kuhlii 1206 149 1355 9.2 85.7

Nyctaloid 61 73 134 0.9 76.2

Myotis sp. 48 44 92 0.6 76.2

Chiroptera sp indet. 7 25 32 0.2 47.6

Barbastellus barbastellus 7 19 26 0.2 23.8

Plecotus sp. 17 9 26 0.2 28.6

Nyctalus leislerii 6 11 17 0.1 19.0

Pipistrellus pygmaeus 9 8 17 0.1 14.3

Nyctalus noctula 1 9 10 0.1 14.3

Pipistrellus nathusius 6 4 10 0.1 19.0

Rhinolophus hipposideros 5 1 6 < 0.1 14.3

Rhinolophus ferruquimenum 1 1 2 < 0.1 9.5

Sum 12589 2149 14738 100 100

Table 2.Number of bat calls recorded in May, June and overall study for each species or OTU. Proportion of

representation (% activity) and proportion of sampling plots were species or OTU is present (% presence).

11

3.2 SPECIES ACTIVITY PATTERNS DURING THE NIGHT

0

5

10

15

20

25

30

20 21 22 23 00 01 02 03 04 05 06

% t

ota

l bat

act

ivit

y

Myostis sp

20 21 22 23 00 01 02 03 04 05 06

Pipistrellus kuhlii

0

5

10

15

20

25

20 21 22 23 00 01 02 03 04 05 06

% t

ota

l bat

act

ivit

y

Pipistrellus pipistrellus

20 21 22 23 00 01 02 03 04 05 06

Pipistrellus pygmaeus

20 21 22 23 00 01 02 03 04 05 06

Plecotus sp

P.Kuhlii & P.pipistrellus

Pipistrellus kuhlii

20 21 22 23 00 01 02 03 04 05 06

Pipistrellus pipistrellus

20 21 22 23 00 01 02 03 04 05

Night time (hour)

Barbastellus barbastellus

0

10

20

30

40

50

60

70

20 21 22 23 00 01 02 03 04 05 06

% t

ota

l bat

act

ivit

y

Night time (hour)

Eptesicus serotinus

0

5

10

15

20

25

20 21 22 23 00 01 02 03 04 05 06

Night time (hour)

ALL SPECIES

A

D

CB

E F

G H I

Specific activity patterns along nightFigure 2. Species night activity pattern, representing by proportion of total activity for each hour of night. Only species

with more than 30 contacts are presented.

Bimodal pattern was observed on total activity. The first peak was during the first hour after sunset

and the second around two hours before sunrise. The activity dropped in the middle of night.

Substantial variability was observed among species, with “early night bats” like Eptesicus serotinus

(G) or Barbastella barbastellus (H), “middle night bats” like Myotis sp.(D), Pipistrellus pipistrellus

(A), Plecotus sp. (E) and “late night bats” like Pipistrellus kuhlii (B) and Pipistrellus pygmaeus (F).

Activity patterns of P.pipistrellus and P.kuhlii seems to be particularly opposed, as P.kuhlii peak tie

in a drop of P.pipistrellus activity. Special attention should be focused on OTU with low rates of

contact (B.barbastellus, Myotis sp., Plecotus sp, P. pygmaeus.), as activity pattern could be biased

by the lack of data.

12

3.3 EFFECTS OF ABIOTIC FACTORS ON BAT ACTIVITY

ABIOTIC VARIABLES ESTIMATES

(Intercept) moon Session (May) Temp.min

Chiroptera 1.451 0.091 0.299 0.306

Table 3. Fixed effects structure of best fitted GLMM model explaining effects of abiotic variables on total bat activity.

Abiotic effects seem to explain a part of variability observed among sampling nights. Activity in

May was significantly higher than activity in June. Activity increased substantially with Minimal

night temperature. Although selected in model structure, percentage of visible moon seems to have

a negligible influence on activity.

3.4 HABITAT VARIABLES EFFECTS ON BAT ACTIVITY

3.4.1 COMMUNITY LEVEL ANALYSIS

HABITATS VARIABLES ESTIMATES

(Intercept) % deciduous cv2 cv16 cv32 isw height dt

Chiroptera 1.146 0.333 0.099 -0.0850 0.148 0.073 0.051 -0.235

Table 4. Fixed effects structure of best fitted GLMM model explaining effects of environmental variables on total bat

activity.

The magnitude of the model coefficient related to forest composition (%deciduous) was the highest

one, indicating a strong effect of forest composition on total bat activity (Table 4.). Activity

decreased as Number of dead tree per plot (dt) increased. Vegetation structure variables seem to

have lesser influence, but were retained in model structure. Activity decreased as a function of tree

cover (cv16), while structural diversity indice (isw), shrub cover (cv2) and canopy tree cover (cv32)

influenced activity positively.

13

3.4.2 SPECIES LEVEL ANALYSIS

HABITATS VARIABLES ESTIMATES

OTU (Intercept) % deciduous cv2 cv16 cv32 isw height dt

E.serotinus -3.968 0.212 -0.190 -0.101 -0.097 -0.248 0.413 -0.469

P.kuhnat -1.469 -0.605 -0.113 -0.262 -0.942 -0.135 -- 0.161

P.pipistrellus 0.276 0.637 0.385 -0.409 0.347 0.247 -0.255 -0.254

Table 5.Fixed effects structure of best fitted GLMM model explaining effects of environmental variables on activity of

species or OTU modelised

Species level responses to habitat factors were variable depending on species. Responses patterns of

P.kuhnat and E.serotinus were similar. For these species, activity decreased with increase of

deciduous proportion and vegetation structure complexification all strata confounded. P.pipistrellus

activity increased with proportion of deciduous tree, shrub cover (cv2), high tree cover (cv32) and

structural diversity indice (isw). Tree cover (cv16) had negative effect on P.pipistrellus activity.

E.serotinus activity increased as a function of tree height, P.pipistrellus activity decreased when

tree height increased. No effect of tree height was observed for P.kuhnat.

Figure 3. Barplots presenting distribution of activity among stand modalities. All species with more than 30 contacts

were plotted; underlined names indicate tested in GLMM, p<0.05

14

As seen in 3.4.1, responses of bat activity to forest composition were pronounced and variable from

one species to another. For the entire community, activity clearly increased with proportion of

deciduous along the mixture gradient. Only Plecotus sp. seems to show higher activity in “Pine”

plots.

P.nat/kuh shows clear preference for “Mixt” plots. P.pipistrellus was more active in “Deciduous”

than “Mixt” and “Pine”. Activity of Myotis sp and Nyctaloid OTUs did not show strong

differences between modalities. These results could be explained by (i) the low number of contact

and (ii) that these OTUs represent several bat species with different ecological requirements.

3.5 COMMUNITY ASSEMBLAGE AMONG PLOTS

Figure 4. Biplot based on non-metric multidimensional scaling (NMDS) of bat assemblages among sites. Results of

environmental fitting for variables with significant association (see Table 6.) are presented in blue vectors. The angle and length

of vector loadings indicate the direction and strength of associations, respectively. A = Distances between sampling sites (colored

points) on the ordination reflect dissimilarity. Modalities of plots are presented by ellipses and colors of points ( black =

“Deciduous”, red=”Mixt”, green=”Pine”). B = Distances between bat species reflect (orange text) dissimilarity in composition.

A B

15

VECTORS

FACTORS

NMDS1 NMDS2 r2 Pr(>r)

r2 Pr(>r)

deciduous 0.279 0.960 0.435 0.007 **

modalities 0.245 0.046 *

cv0.5 0.384 -0.923 0.139 0.301 cv2 -0.064 0.998 0.091 0.472 cv16 0.557 0.830 0.178 0.219

NMDS1 NMDS2 cv32 -0.887 -0.463 0.146 0.289

Deciduous 0.142 0.230

tree_height -0.999 -0.046 0.066 0.595

Mixt -0.017 -0.043 dt -0.994 0.109 0.202 0.149

Pine -0.062 -0.082

isw -0.132 0.991 0.023 0.852 basal_area 0.057 -0.998 0.320 0.050 *

Table 6. Results of environmental fitting calculated on NMDS

Plotting vectors of habitat metrics illustrated that deciduous gradient (p <0.01, Table 6.) structured

the ordination of species with Rhinolophus hipposideros (rhihip), Nyctalus noctula (Nycnoc),

Nyctalus leisleri(Nyclei) associated to high proportion of deciduous tree. And the ordination of

plots, with a substantial segregation of “Deciduous” plots (black ellipse) from “Mixt” and “Pine”

(respectively green and red ellipse) (p < 0.05, Table 6.).

P.kuhnat (pipnatkuh), P.pipistrellus (pippip), E.serotinus (eptser), Myotis sp. (myospp) and

Nyctaloid (nyctaloid) were the most ubiquitous species as they present in more than 75% of plots

(see 3.1)

Long-eared bats Plecotus sp. (plespp) seems associated to “Pine” plots, as other species only

appears in “Mixt” or “Deciduous” plots.

Impact of Basal area in the ordination was also retained as significant (p<0.05, Table 6.), but its

effects on bat distribution seem less obvious. Plotting of value surface on sites space ordination

(see Appendix IV .) reveals that there was no linear relationship between basal area and ordination.

16

4 DISCUSSION

4.1 BAT COMMUNITY SAMPLING

Our results must be viewed in the light of biases inherent to the use of ultrasonic detectors to assess

bat activity and habitat selection. It is actually impossible to distinguish between recordings of the

same individual flying in front of the bat detector several times and several individuals flying in

front of the bat detector only once (O’Farrell & Gannon, 1999). Nevertheless, this method has been

used convincingly by other researchers studying relative habitat use by bats (Barclay, 1999).

As described by Hayes (1997) activity fluctuated significantly among sampling nights (mean

number of bat pass per night = 181, sd=556) and among species (mean number of bat pass per

species =1215, sd=2627). Total bat activity was strongly influenced by overwhelming presence of

P.pipistrellus in recorded calls. Three nights with extremely high number of bat passes were

excluded from data analysis. Very high levels of bat flight activity were recorded in plot 060A (May

16, 2012 – 3493 passes) and plot 095A (May 11, 2012 – 3344 passes and May 12, 2012 – 1937

passes). As described in other studies (Ceľuch & Kropil, 2008), these high activities are probably

due to exceptional insect swarming.

At least 12 species were contacted out of 25 bat species known in Aquitaine. According to the

number of forest-dwelling species potentially represented by OTU Myotis spp (6 species),

taxonomic diversity was surprisingly high for the Landes de Gascogne plantation forests. The high

number of collected data shed new light on the occurrence and distribution of bats in the regions

that still remains poorly informed (see Appendix VI.)

P. pipistrellus was by far the most often contacted bat species, which is consistent with its

widespread distribution, being the most common bat in the study area (see Appendix VI.). Its high

flexibility in habitat requirements; with a preference for forests explain its spatial pervasiveness.

High activity recorded for P.kuhlii and Eptesicus serotinus were surprising since this species are

considered as scarce in closed forest habitats (Arthur & Lemaire, 2009). However, mature pine

plantation forests have relatively low tree density, allowing some open habitat species to penetrate

into forest stands as demonstrated for other taxa in the same study area (Barbaro et al., 2005; van

Halder et al., 2008).

17

4.2 SPECIES ACTIVITY PATTERNS DURING THE NIGHT

The bimodal pattern of activity observed is typical for the majority of insectivorous bats, well

known and largely described in literature (e.g Krusic et al., 1996; Rydell et al., 1996; Hayes, 1997).

However, activity patterns show variability depending on the species (see figure 3.). Overnight

activity pattern of P.kuhlii and P.pipistrellus were clearly different and opposed (see figure 1.C.)

This segregation was surprising as this species are quite similar on habitat and diet requirements

(Arthur & Lemaire, 2009). This could therefore represent a temporal segregation in ecological niche

and/or interspecific competition for foraging.

4.3 EFFECTS OF ABIOTIC FACTORS ON BAT ACTIVITY

4.3.1 MOON

Our results confirm findings of many studies, that is moonlight has no overall effect on activity by

forest-dwelling insectivorous bats (Negraeff & Brigham, 1995; Hecker & Brigham, 1999; Kalka &

Kalko, 2006). Furthermore, other studies anecdotally report that lunar illumination suppresses

activity (Fenton et al., 1977) or could be responsible of a shift in habitats. These behaviors could be

linked either by increasing risk of predation or changing distribution of insects (Williams, 1936).

Both mechanisms has been actually proved relevant for other nocturnal feeding vertebrates such as

storm petrels (Mougeot & Bretagnolle, 2000).In the present study, as light penetration is limited in

forest stands, effects of moonlight appear to be more limited here than in open fields.

4.3.2 SESSION

Foraging activity changes over the course of the reproductive season. Jong & Ahl, (1991) observed

changes in the Common pipistrelle (P.pipistrellus) habitat use between early and late summer. As

noticed in our study, forest areas were preferentially used in spring, and then bats gradually shifted

to open environments during the summer. Early summer is presumably one of the critical periods

for bats, just after hibernation. In particular, lactating females have high energy demands and must

increase their prey intake accordingly. This can be accomplished by an increase in foraging time

(Barclay, 1989), but may also be facilitated by increased foraging efficiency following parturition

(Kunz, 1974), and/or increased insect abundance in the environment (Anthony & Kunz, 1977; Swift

et al., 1985). An abundance of food at this time implies early reproduction which may be important

for subsequent survival of young (Swift et al., 1985). This fact shows the importance of forest

habitat in early summer as critical feeding sites for the local bat fauna. Later in the summer, an

overall increase in bat activity is expected when juvenile bats begin flying and foraging on their

own (Wilson, 2004).

18

4.3.3 TEMPERATURE

Temperature effect on bat activity was clear, but must be seen through its indirect influence on

insect prey abundance. As endothermic animals, bats are not limited by night temperature for

foraging activities. Furthermore, cold temperatures inhibit the flight activity of insects (Taylor,

1963), and are generally associated with low levels of bat activity (Hickey & Fenton, 1996;

Wilkinson & Barclay, 1997).

4.4 INFLUENCE OF FOREST COMPOSITION

Results from global linear mixed models support the prediction that total bat activity increase with

deciduous tree proportion at stand scale. A number of small-scale studies have already identified

semi-natural broadleaved woodland sites as preferred foraging areas for vespertilionids bats (Swift

et al., 1985; Furlonger et al., 1987) and more particularly for the Common pipistrelle (Pipistrellus

pipistrellus) and the Northern bat Eptesicus nilssonii (Jong & Ahl, 1991). Otherwise, fewer bats are

recorded in coniferous plantations than deciduous in Walsh & Harris (1996b), indicating that it

represents a less optimal woodland type. Some studies also report that old growth and natural

conifer forest can be prime habitat for some forest dwelling bats (Jung et al., 1999), whereas other

studies indicate that conifer and mixed stands has less dense understories allowing bats to forage

more easily (Kalcounis et al., 1999). Our results tend to be in favor of the first theory, indicating

that broad leaved forest stands represent more optimal habitat for bats in general. In a landscape of

intensely managed pine forest, deciduous patches are the only remnants of semi-natural habitats.

P. pipistrellus, P.kuhnat and Eptesicus serotinus covered near than 90% of total bat calls. As these

species almost exclusively use buildings as day roosts, their use of forest is restricted to foraging

activity. Hence higher activity recorded in deciduous habitats should be widely driven by higher

food disponibility and foraging suitability. Different forest types also provide different food and

habitat for insects (Molles Jr., 1982). Although we have no data on insect abundance, we suspect

that abundance and diversity of insects were lower in pure maritime pine stands, less structurally

complex than mixed pine-oak or pure deciduous stands. Previous research has suggested that

coniferous forests provide relatively few foraging opportunities for bats (Thomas, 1988; Jong &

Ahl, 1991) and that bat activity recorded in coniferous forests primarily reflects roosting and

commuting bats (Thomas, 1988; Patriquin & Barclay, 2003). Forest types may also differ in insect

diversity and abundance because of the production of resins synthesized by conifers as a defensive

secretion against insect attack (Funk & Croteau, 1994). Finally, previous studies conducted in the

Landes de Gascogne Forest has shown that conifer plantations contain lower numbers of insect

19

species (Barbaro et al., 2005) than deciduous woodlands patches acting as refuges and source

habitats for arthropods (Barbaro & van Halder, 2009).

As the activity of dominant P.pipistrellus widely influenced results extrapolated from the all-

species-confounded activity, we also need to examine other species responses. Results from GLMM

report that activity of P.kuhnat - that could be attributed to the Kuhl's Pipistrelle (Pipistrellus

kuhlii) (see 4.1) - was generally reduced along composition gradient. Results from modalities

plotting (see figure 2.) indicate that mixed stands hosted four times more activity than pure

deciduous or pine stands. Beyond the effect of deciduous proportion tree along the mixture gradient,

Kalcounis et al. (1999) also showed that deciduous-conifer mixedwood forest stands harbored more

activity than pure pine or pure deciduous stands. Furthermore, mixed woodlands could have also

higher insect diversity and abundance than monospecific pine stands (Kalcounis et al., 1999;

Barbaro et al., 2005). We could therefore make the assumption that mixed woodlands provide more

foraging opportunity than pure stands for this species. This support the hypothesis of improved

habitat quality in mixed than monospecific stands. Moreover, it also shows that responses can be

strongly various among closely related species, as pointed out by Jung et al., (1999).

Results from the NMDS highlight the significant role played by the high proportion of deciduous

tree on bat community composition. Pure deciduous woodlands were characterized by the presence

of species not or rarely found in pine or mixed stands (Nyctalus noctula, Nyctalus leisleri,

Rhinolophus hipposideros). Common Noctule (N.noctula) and Leisler's Bat N.leisleri generally

have clear preference for deciduous stands, but also use conifer forests for foraging. However, they

only use deciduous trees for summer roosts (Arthur & Lemaire, 2009). Trees have several potential

roost sites, including foliage and bark, crevices beneath loose bark, abandoned hollows of

woodpeckers, and naturally formed cavities (Brigham et al., 1997). In pure maritime pine stands,

trees presenting these characteristics are scarcer than in mixedwood and deciduous stands. As roost

density play a major role in the organization of forest-dwelling bat communities (Walsh & Harris,

1996a; Brigham et al., 1997; Menzel et al., 2002; Elmore et al., 2004; Russo et al., 2005, 2010), the

presence of Nyctalus species in our plots appears logically driven by deciduous tree proportion

through suitable roost availability.

This suggests that both roost sites and foraging opportunities increase with deciduous tree

proportion at the stand scale, but probably also at the landscape scale (Barbaro et al., 2007)

20

4.5 INFLUENCE OF HABITAT STRUCTURE

As deciduous tree proportion, several variables describing vegetation structure were retained in the

models, but had weaker effects. Vertical stratification is known to play a substantial role for bat

activity (Kalcounis et al., 1999; Plank et al., 2011). Especially, canopy tree cover seems to

significantly influence total activity in our study area.

Vegetation structure can have effect on foraging through (i) its impact on foraging accessibility and

(ii) its impact on insect prey availability.

(i) Morphology determines the selection of foraging habitats in bats, and differences

in morphology will result in the partitioning of spatial resources (Aldridge &

Rautenbach, 1987). Larger-winged bats are not adapted to closed habitat, and the

most cluttered forests can only host forest specific bats (e.g. Myotis sp.).

Reduction of the amount of structural volume in the understory to mid-canopy

provide more suitable habitat for foraging bats (Titchenell et al., 2011).

(ii) As indicated in Luque et al. (2007), forest architecture (i.e. the relative coverage

by the different vegetation levels) is important for the composition of moth

communities. Forest structure and especially differences at the shrub level can

play a role on the densities of some moth species. Generally, higher structural

heterogeneity within the stand is related to higher diversity and abundance in

moths.

Our results shows that total bat activity increased with shrub (0.5m-2m) and high tree (+16m) cover,

but decreased with tree (2m-16m) cover. In the light of theories outlined above, we can presume that

more dense stands seem to improve moth prey abundance, but opened tree layer (2m-16m) are

important for bat movement during hunting. This is remarkably consistent with previous studies on

compared prey availability and accessibility for another nocturnal aerial feeding vertebrate, the

European nightjar Caprimulgus europeus (Sierro et al., 2001).

P.kuh/nat and the the serotine bat (E.serotinus) activities were clearly driven by low vegetation

cover values in all strata. These results are not surprising in the sight of their ecologic requirements.

These species are known to be scarce in closed habitats (see 4.), but occurred frequently in a wide

part of forest stands. However, their activity seems to be more important in low-structured stands.

E. serotinus is a wide wingspan bat (c.350mm), so its morphology and echolocation call design fit it

in the “open land” or “woodland edge foragers” guild. This bat is also generalist and opportunist,

so its abundance and ubiquity conceal its large ecological requirements. We presume that general

high activity recorded for this species in forest stands is due to increased food availability, but low

21

cluttered stands are more suitable in regard of their morphology. This results are consistent with

Titchenell et al., (2011), demonstrating that the reduction of the amount of structural volume in the

understory to mid-canopy provide suitable habitat for foraging bats.

Number of dead trees was used to roughly assess the susceptibility of each plot to contain suitable

roosts. Unlike Brigham et al. (1997), our presumption that activity should increase with the number

of dead tree was not verified. Our results show weak and negative effect of number of dead tree on

activity. Several factors may be responsible for this :

(i) The dominant the species contacted do not use roosts in tree but in buildings.

(ii) Most species contacted use pine stands only for foraging and do not roost in pines, so

that dead tree availability do not affect bat occurrence within pine stands, mainly

driven by prey availability

(iii) All dead trees listed were maritime pine (Pinus pinaster). As this tree presents a thin

and straight trunk, it should not be a potential roost source once dead.

Our study demonstrated that other habitat variables than deciduous tree proportion were also

important for bat communities, since they influenced bat distribution patterns in forests. This

influence is based on the effects of vegetation structure through its effects on prey availability and

flying suitability. Several other habitat variables not tested in this study could explain bat

community composition in plantation forest, notably suitable roost availability. Above all,

management of forest stands stay to be the most important features in forest-dwelling bat

distribution and conservation.

4.6 MANAGEMENT

Due to current and potential future declines of bat populations, it has become increasingly important

for natural resource managers to understand how forest management practices influence bat habitat

relationships. Our results show the sensitivity of bats to forest composition and structure. In

industrial plantations these features are greatly influenced by scale stand management

(Lindenmayer et al., 2006). As many other studies (Jung et al., 1999; Kalcounis et al., 1999; Jaberg

et al., 2006), we conclude that deciduous stands are of fundamental ecological importance for many

chiropteran species, especially in monospecific plantation forests. The differences observed

between deciduous woodlands and pine stands are due to their distinct tree composition but also to

differential management. Plantation stands are typically characterized by a uniform and intensive

management, whereas management of deciduous woodlands is more variable in time and space

allowing a greater structural diversity. Maintenance of mature deciduous wood lands would

22

increase the availability of roosts (Brigham et al., 1997) used by bats depending on old deciduous

trees.

Consequently, these measures could improve diversity and abundance of bats communities. A

comparative approach on the quantification of leaf damage highlighted the importance of local

attributes such as tree age, forest composition and species richness of vertebrate predators for

control of arthropod herbivory in temperate forests (Böhm et al., 2011). As bat predation both

directly reduces arthropod abundance on plants and indirectly reduce herbivory, they are likely key

regulators of insect pests affecting productivity in the Landes de Gascogne forest. Given their

ecological importance, bats should be included in future forest management strategies based on

biological pest control.

4.7 CONCLUSION

The present study assesses the importance of mixedwood and deciduous forest stands for bats in a

monospecific and intensely managed pine plantation landscape. Both activity and diversity of bat

communities were positively influenced by deciduous tree cover along a pine-oak composition

gradient. Indeed, the integration of deciduous trees in pine plantations appear to create high value

habitats for foraging, and could represent critical feeding sites when lactating females have to

increase their prey intake in early summer. Others explanatory variables measured at the local scale

(i.e vegetation structure) explained a significant part of data variability, and are more or less linked

to the mixture gradient. Logging should take into account stand management impact on bats, as they

could play a key role in insect pest regulation.

As other animal forest vertebrate communities such as birds, multi-scale factors are required to

explain distribution and organization of bat communities (Miles et al., 2006; Kanuch et al., 2008).

Our study, as well as those of (Krusic et al., 1996) and (Grindal & Brigham, 1998), suggested that

forest-dwelling bats may require a matrix of forest types on a landscape, which includes older

forests. Some old-growth deciduous stands should be maintained across a landscape in forest

management planning, and others planned for the future via lengthening of rotation cycles to enable

old-growth stand characteristics to develop. Harvesting that creates a mosaic of patches with

different tree densities is likely to satisfy the requirements of more species than a system with less

diverse harvesting styles. Further studies should focus on effects of landscape structure and

composition, and assess importance of semi-natural habitat remnants like deciduous tree islands

within plantation (Lindenmayer et al., 2006). As foreground providers of essential top-down control

of herbivorous insects (Kalka et al., 2008), the conservation of vulnerable and endangered taxa like

chiropteran has to be at the forefront of biodiversity consideration in forest management.

23

5 BIBLIOGRAPHY ADAMS, M.D., LAW, B.S. & FRENCH, K.O. (2009) Vegetation

structure influences the vertical stratification of open- and

edge-space aerial-foraging bats in harvested forests. Forest Ecology and Management, 258, 2090–2100.

ADAMS, M.W. & EDINGTON, J.M. (1973) A comparison of song-bird populations in mature coniferous and broadleaved woods.

Forestry, 46, 191–202.

ALDRIDGE, H. & RAUTENBACH, I.L. (1987) Morphology,

echolocation and resource partitioning in insectivorous bats.

The Journal of Animal Ecology, 56, 763–778.

ANDERSON, E. (1986) A guide for estimating cover. Rangelands, 8,

236–238.

ANTHONY, E.L.P. & KUNZ, T.H. (1977) Feeding strategies of the

little brown bat, Myotis lucifugus, in Southern new

Hampshire. Ecology, 58, 775–786.

ARTHUR, L. & LEMAIRE, M. (2009) Les chauves-souris de France, Belgique, Luxembourg et Suisse. In Biotope, Mèze

(Collection Parthénope) pp. 136–233. Biotope;

BARATAUD, M. (2012) Écologie acoustique des chiroptères

d’Europe: identification des espèces, étude de leurs habitats et

comportements de chasse (avec DVD). Biotope.

BARBARO, L. & VAN HALDER, I. (2009) Linking bird, carabid beetle

and butterfly life-history traits to habitat fragmentation in mosaic landscapes. Ecography, 32, 321–333.

BARBARO, L., PONTCHARRAUD, L. & VETILLARD, F. (2005) Comparative responses of bird , carabid , and spider

assemblages to stand and landscape diversity in maritime pine

plantation forests. Ecoscience, 12, 110–121.

BARBARO, L., ROSSI, J.-P., VETILLARD, F., NEZAN, J. & JACTEL, H.

(2007) The spatial distribution of birds and carabid beetles in pine plantation forests: the role of landscape composition and

structure. Journal of Biogeography, 34, 652–664.

BARCLAY, M.R. (1989) The effect of reproductive condition on the

foraging behavior of female hoary bats, Lasiurus cinereus.

Behavioral Ecology and Sociobiology, 24, 31–37.

BARCLAY, M.R. (1999) Bats are not birds: a cautionary note on using

echolocation calls to identify bats: a comment. Journal of Mammalogy, 80, 290–296.

BAT CONSERVATION TRUST (2007) Bat surveys: good practice guidelines. In Europe. Bat Conservation Trust.

BATES, D. (2005) Fitting linear mixed models in R. R news, 5, 27–

30.

BAXTER, D.J.M., PSYLLAKIS, J.M., GILLINGHAM, M.P. & O’BRIEN, E.L. (2006) Behavioural Response of Bats to Perceived

Predation Risk While Foraging. Ethology, 112, 977–983.

BERTHINUSSEN, A. & ALTRINGHAM, J. (2012) The effect of a major

road on bat activity and diversity. Journal of Applied Ecology, 49, 82–89.

BITTERLICH, W. (1984) The relascope idea. Relative measurements

in forestry. Commonwealth Agricultural Bureaux.

BOLKER, B.M., BROOKS, M.E., CLARK, C.J., GEANGE, S.W.,

POULSEN, J.R., STEVENS, M.H.H. & WHITE, J.-S.S. (2009)

Generalized linear mixed models: a practical guide for ecology and evolution. Trends in ecology & evolution, 24,

127–35.

BOYD, I. & STEBBINGS, R.E. (1989) Population changes of brown

long-eared bats (Plecotus auritus) in bat boxes at Thetford

Forest. Journal of Applied Ecology, 26, 101–112.

BRIGHAM, R.M., VONHOF, M.J., BARCLAY, M.R. & GWILLIAM, J.C.

(1997) Roosting behavior and roost-site preferences of forest-dwelling California bats (Myotis californicus). Journal of

Mammalogy, 78, 1231–1239.

BURNHAM, K.P. & ANDERSON, D.R. (2002) Model selection and

multimodel inference: a practical information-theoretic

approach.

BÖHM, S.M., WELLS, K. & KALKO, E.K.V. (2011) Top-down control

of herbivory by birds and bats in the canopy of temperate broad-leaved oaks (Quercus robur). PloS one, 6, e17857.

CELUCH, M. & KROPIL, R. (2008) Bats in a Carpathian beech-oak forest ( Central Europe ): habitat use , foraging assemblages

and activity patterns. Methods, 57, 358–372.

CHEN, J., SAUNDERS, S., CROW, T. & NAIMAN, R. (1999)

Microclimate in forest ecosystem and landscape ecology.

BioScience, 49, 288–297.

CROME, F. (1988) Bats and gaps: microchiropteran community

structure in a Queensland rain forest. Ecology, 69, 1960–

1969.

DODD, L.E., LACKI, M.J., BRITZKE, E.R., BUEHLER, D. A., KEYSER, P.D., LARKIN, J.L., ET AL. (2012) Forest structure affects

trophic linkages: How silvicultural disturbance impacts bats

and their insect prey. Forest Ecology and Management, 267, 262–270.

ELMORE, L.W., MILLER, D. A. & VILELLA, F.J. (2004) Selection of diurnal roosts by red bats (Lasiurus borealis) in an intensively

managed pine forest in Mississippi. Forest Ecology and

Management, 199, 11–20.

ETHIER, K. & FAHRIG, L. (2011) Positive effects of forest

fragmentation, independent of forest amount, on bat abundance in eastern Ontario, Canada. Landscape Ecology,

26, 865–876.

FAHRIG, L. (2003) FRAGMENTATION ON OF HABITAT

EFFECTS BIODIVERSITY. Annual Review of Ecology,

Evolution, and Systematics., 34, 487–515.

FELTON, A., LINDBLADH, M., BRUNET, J. & FRITZ, Ö. (2010)

Replacing coniferous monocultures with mixed-species production stands: An assessment of the potential benefits for

forest biodiversity in northern Europe. Forest Ecology and Management, 260, 939–947.

24

FENTON, M., BOYLE, N., HARRISON, T. & OXLEY, D. (1977) Activity

patterns, habitat use, and prey selection by some African

insectivorous bats. Biotropica, 9, 73–85.

FUNK, C. & CROTEAU, R. (1994) biosynthesis in conifers:

characterization of two cytochrome P450-dependent monooxygenases and an aldehyde dehydrogenase involved in

abietic acid biosynthesis. Archives of biochemistry and

biophysics, 308, 258–266.

FURLONGER, C., DEWAR, H. & FENTON, M. (1987) Habitat use by

foraging insectivorous bats. Canadian Journal of Zoology, 65, 284–288.

GRAVES, A.T. (1999) The Effects of Thinning Intensity on Snag and Cavity Tree Abundance in an Appalachian Hardwood Stand.

GREGORY, R.D., VORISEK, P., STRIEN, A.V.A.N., MEYLING, A.W.G., JIGUET, F., FORNASARI, L., ET AL. (2007) Population trends of

widespread woodland birds in Europe. Ibis, 149, 78–97.

GRIFFIN, D.R., WEBSTER, F.A. & MICHAEL, C.R. (1960) The

echolocation of flying insects by bats. Animal Behaviour, 8,

141–154.

GRINDAL, S. & BRIGHAM, R. (1998) Short-term effects of small-scale

habitat disturbance on activity by insectivorous bats. The Journal of wildlife management, 62, 996–1003.

VAN HALDER, I., BARBARO, L., CORCKET, E. & JACTEL, H. (2008) Importance of semi-natural habitats for the conservation of

butterfly communities in landscapes dominated by pine

plantations. Biodiversity and Conservation, 17, 1149–1169.

HARTLEY, M.J. (2002) Rationale and methods for conserving

biodiversity in plantation forests. Forest Ecology and Management, 155, 81–95.

HAYES, J.P. (1997) Temporal variation in activity of bats and the design of echolocation-monitoring studies. Journal of

Mammalogy, 78, 514–524.

HECKER, K.R. & BRIGHAM, R.M. (1999) Does Moonlight Change

Vertical Stratification of Activity by Forest-Dwelling

Insectivorous Bats? American Society of Mammalogists, 80, 1196–1201.

HICKEY, M. & FENTON, M. (1996) Behavioural and thermoregulatory responses of female hoary bats, Lasiurus cinereus(Chiroptera:

Vespertilionidae), to variations in prey availability.

Ecoscience. Sainte-Foy, 3, 414–422.

HOLLAND, S.M. (2008) Non-Metric Multidimensional Scaling.

JABERG, C., BOHNENSTENGEL, T. & AMSTUTZ, R. (2006) Utilisation

du milieu forestier par les chauves-souris (Mammalia:

Chiroptera) du canton de Neuchâtel – implications pour la gestion sylvicole. Schweiz. Z., 157, 254–259.

JONG, J.D. & AHL, I. (1991) Factors affecting the distribution pattern of bats in Uppland, central Sweden. Ecography, 14, 92–96.

JUNG, K., KAISER, S., BÖHM, S., NIESCHULZE, J. & KALKO, E.K.V. (2012) Moving in three dimensions: effects of structural

complexity on occurrence and activity of insectivorous bats in

managed forest stands. Journal of Applied Ecology.

JUNG, T., THOMPSON, I. & TITMAN, R. (1999) Habitat selection by forest bats in relation to mixed-wood stand types and

structure in central Ontario. The Journal of wildlife, 63,

1306–1319.

KALCOUNIS, M., HOBSON, K. & BRIGHAM, R. (1999) Bat activity in

the boreal forest: importance of stand type and vertical strata.

Journal of Mammalogy, 80, 673–682.

KALKA, M.B. & KALKO, E.K.V. (2006) Gleaning bats as

underestimated predators of herbivorous insects: diet of Micronycteris microtis (Phyllostomidae) in Panama. Journal

of Tropical Ecology, 22, 1–10.

KALKA, M.B., SMITH, A.R. & KALKO, E.K.V. (2008) Bats limit

arthropods and herbivory in a tropical forest. Science, 320,

71.

KANUCH, P., DANKO, Š., CELUCH, M., KRISTIN, A., PJENCAK, P.,

MATIS, Š. & ŠMIDT, J. (2008) Relating bat species presence to habitat features in natural forests of Slovakia (Central

Europe). Mammalian Biology - Zeitschrift für

Säugetierkunde, 73, 147–155.

KRUSIC, R.A., YAMASAKI, M., NEEFUS, C.D. & PEKINS, P.J. (1996)

Bat habitat use in white mountain national forest. The Journal of wildlife management, 60, 625–631.

KUNZ, T. (1974) Feeding ecology of a temperate insectivorous bat (Myotis velifer). Ecology, 55, 693–711.

KUNZ, T.H., BRAUN DE TORREZ, E., BAUER, D., LOBOVA, T. &

FLEMING, T.H. (2011) Ecosystem services provided by bats.

Annals of the New York Academy of Sciences, 1223, 1–38.

LINDENMAYER, D.B., FRANKLIN, J.F. & FISCHER, J. (2006) General

management principles and a checklist of strategies to guide

forest biodiversity conservation. Biological Conservation, 131, 433–445.

LUQUE, C., GERS, C., LAUGA, J., MARIANO, N., WINK, M. & LEGAL, L. (2007) Analysis of forestry impacts and biodiversity in two

Pyrenean forests through a comparison of moth communities

(Lepidoptera, Heterocera). Insect Science, 14, 323–338.

MAIZERET, C. (2005) Les Landes de Gascogne.

MCCOMB, W.C., BONNEY, S.A., SHEFFIELD, R.M., COST, N.D.,

WILDLIFE, S., BULLETIN, S. & SPRING, N. (1986) Snag

Resources in Florida: Are They Sufficient for Average Populations of Primary Cavity- Nesters? Wildlife Society

Bulletin, 14, 40–48.

MENZEL, M. A, OWEN, S.F., FORD, W.M., EDWARDS, J.W., WOOD,

P.B., CHAPMAN, B.R. & MILLER, K.V. (2002) Roost tree

selection by northern long-eared bat (Myotis septentrionalis) maternity colonies in an industrial forest of the central

Appalachian mountains. Forest Ecology and Management,

155, 107–114.

MILES, A., CASTLEBERRY, S. & MILLER, D. (2006) Multi-Scale

Roost-Site Selection by Evening Bats on Pine-Dominated Landscapes in Southwest Georgia. Journal of Wildlife

Management, 70, 1191–1199.

MOLLES JR., M.C. (1982) Trichopteran communities of streams

associated with aspen and conifer forests: long-term structural

change. Ecology, 63, 1–6.

MOUGEOT, F. & BRETAGNOLLE, V. (2000) Predation Risk and Moonlight Avoidance in Nocturnal Seabirds. Journal of

Avian Biology, 31, 376–386.

25

MÜLLER, J., MEHR, M., BÄSSLER, C., FENTON, M.B., HOTHORN, T.,

PRETZSCH, H., ET AL. (2012) Aggregative response in bats:

prey abundance versus habitat. Oecologia, 169, 673–84.

NEGRAEFF, O. & BRIGHAM, R. (1995) The influence of moonlight on

the activity of little brown bats (Myotis lucifugus). Zeitschrift für Säugetierkunde, 60, 330–336.

NICHOLS, J.D., BRISTOW, M. & VANCLAY, J.K. (2006) Mixed-species plantations: Prospects and challenges. Forest Ecology and

Management, 233, 383–390.

NORBERG, U.M. & RAYNER, J.M.V. (1987) Ecological morphology

and flight in bats (Mammalia; Chiroptera): wing adaptations,

flight performance, foraging strategy and echolocation. Philosophical Transactions of the Royal Society of London.

Series B, Biological Sciences, 335–427.

OKSANEN, J. (2011) Multivariate analysis of ecological communities

in R: vegan tutorial.

ORELIEN, J.G. & EDWARDS, L.J. (2008) Fixed-effect variable

selection in linear mixed models using R2 statistics.

Computational Statistics & Data Analysis, 52, 1896–1907.

O’FARRELL, M.J. & GANNON, W.L. (1999) A comparison of acoustic

versus capture techniques for the inventory of bats. Journal of Mammalogy, 80, 24–30.

PATRIQUIN, K.J. & BARCLAY, M.R. (2003) Foraging by bats in cleared, thinned and unharvested boreal forest. Journal of

Applied Ecology, 40, 646–657.

PLANK, M., FIEDLER, K. & REITER, G. (2011) Use of forest strata by

bats in temperate forests. Journal of Zoology, 286, 154–162.

PRODON, R. & LEBRETON, J. (1981) Breeding avifauna of a

Mediterranean succession: the holm oak and cork oak series

in the eastern Pyrenees, 1. Analysis and modelling of the structure gradient. Oikos, 37, 21–38.

RAZGOUR, O., HANMER, J. & JONES, G. (2011) Using multi-scale modelling to predict habitat suitability for species of

conservation concern: The grey long-eared bat as a case

study. Biological Conservation, 144, 2922–2930. Elsevier Ltd.

RICHARDS, A.E., FORRESTER, D.I., BAUHUS, J. & SCHERER-LORENZEN, M. (2010) The influence of mixed tree plantations

on the nutrition of individual species: a review. Tree

physiology, 30, 1192–208.

RUSSO, D., CISTRONE, L., GARONNA, A.P. & JONES, G. (2010)

Reconsidering the importance of harvested forests for the conservation of tree-dwelling bats. Biodiversity and

Conservation, 19, 2501–2515.

RUSSO, D., CISTRONE, L. & JONES, G. (2005) Spatial and temporal

patterns of roost use by tree dwelling barbastelle bats

Barbastella barbastellus. Ecography, 28, 769–776.

RYDELL, J., ENTWISTLE, A. & RACEY, P.A. (1996) Timing of

Foraging Flights of Three Species of Bats in Relation to Insect Activity and and predation risk. Oikos, 76, 243–252.

SIERRO, A., ARLETTAZ, R., NAEF-DAENZER, B., STREBEL, S. &

ZBINDEN, N. (2001) Habitat use and foraging ecology of the

nightjar ( Caprimulgus europaeus ) in the Swiss Alps:

towards a conservation scheme. Biological Conservation, 98,

325–331.

STEBBINGS, R.E. & GRIFFITH, F. (1986) Distribution ans status of

bats in Europe. Abbots Rip. 7–11

SWIFT, S., RACEY, P.A. & AVERY, M.I. (1985) Feeding ecology of

Pipistrellus pipistrellus (Chiroptera: Vespertilionidae) during

pregnancy and lactation. II. Diet. The Journal of Animal Ecology, 54, 217–225.

TAYLOR, L.R. (1963) Analysis of the Effect of Temperature on Insects in Flight. Journal of Animal Ecology, 32, 99–117.

THOMAS, D. (1988) The distribution of bats in different ages of Douglas-fir forests. The Journal of Wildlife Management, 52,

619–626.

TITCHENELL, M. A., WILLIAMS, R. A. & GEHRT, S.D. (2011) Bat

response to shelterwood harvests and forest structure in oak-

hickory forests. Forest Ecology and Management, 262, 980–988.

VONHOF, M.J. & BARCLAY, M.R. (1996) Roost-site selection and roosting ecology of forest-dwelling bats in southern British

Columbia. Canadian Journal of Zoology, 74, 1797–1805.

WALSH, A.L., BARCLAY, M.R. & GARY F. MCCRACKEN (2004)

Designing bat-activity surveys for inventory and monitoring

studies at local and regional scales. In Bat Echolocation Research: tools, techniques and analysis pp. 157–165.

WALSH, A.L. & HARRIS, S. (1996a) Factors determining the abundance of vespertilionid bats in Britain : geographical ,

land class and local habitat relationships. Journal of Applied

Ecology, 33, 519–529.

WALSH, A.L. & HARRIS, S. (1996b) Foraging habitat preferences of

vespertilionid bats in Britain. Journal of Applied Ecology, 33, 508–518.

WEBB, W.L., BEHREND, D.F. & SAISORN, B. (1977) Effect of Logging on Songbird Populations in a Northern Hardwood

Forest. Wildlife Monographs, 55, 3–35.

WILKINSON, L. & BARCLAY, M.R. (1997) Differences in the foraging

behaviour of male and female big brown bats(Eptesicus

fuscus) during the reproductive period. Ecoscience, 4, 279–285.

WILLIAMS, C.B. (1936) The influence of moonlight on the activity of certain nocturnal insects, particularly of the family Noctuidae,

as indicated by a light trap. Philosophical Transactions of the

Royal Society of London. Series B, Biological Sciences, 226, 357–389.

WILLIAMS-GUILLEN, K., PERFECTO, I. & VANDERMEER, J. (2008) Bats limit insects in a neotropical agroforestry system.

Science, 320, 70.

WILSON, J.M. (2004) Foraging behaviour of Insectivorous Bats

during an Outbreak of Western Spruce Budworm. Wiley

Online Library.

WILSON, J.M. & BARCLAY, M.R. (2006) Consumption of caterpillars

by bats during an outbreak of western spruce budworm. The American midland naturalist, 155, 244–249.

Appendix I. Reference pattern used for cover estimation from Prodon & Lebreton, (1981)

Appendix II. British Columbia wildlife-tree- classification system

2

0 10 20 30 40 50 60

-4-2

02

4

GLMM residuals repartition

Fitted Values

Re

sid

ua

ls

-3 -2 -1 0 1 2 3

-20

24

Normal Q-Q Plot

Theoretical Quantiles

Sa

mp

le Q

ua

ntile

s

Appendix II. GLMM validation plots

Appendix III.Surface fitting of basal area variable to ordination Appendix IV. Surface fitting of deciduous tree proportion

variable to ordination

Appendix V. Bat distribution maps in Aquitaine , from http://www.faune-aquitaine.org

4

ABSTRACT

Bats are key insectivorous vertebrates in forest ecosystems, both because of their high

conservation value and major functional role of insect predation. Here, we examined

the effects of forest stand composition and structure on bat communities in the

context of an intensively managed maritime pine forest. We sampled bat activity by

echolocation call monitoring along a pine-oak composition gradient ranging from 0 to

100% oak cover in the tree canopy (n = 21 forest stands). Data were collected in

May-June 2012 using bat-detectors (14738 bats calls recorded in 84 nights) in the

Landes de Gascogne forest (Aquitaine, France). Bat forest habitat was quantified by

10 environmental variables characterizing tree species composition, vertical

vegetation structure and roost site availability (dead trees). General Linear Mixed

Models and Non-Metric Multidimensional Scaling reveal that both activity and

diversity of bat communities were positively influenced by deciduous tree cover

along the composition gradient. This suggests that both roost site availability and

foraging opportunities increase with deciduous tree proportion at the stand scale.

Vegetation structure also explained a significant part of data variability in bat

activity. This highlighted the influence of vegetation structure on bat distribution

patterns in plantation forests through its effects on prey availability and flying

suitability. Forest-dwelling bats are known to provide essential top-down control of

herbivorous insects that could be critical in commercial forests and are also

particularly threatened due to a high sensitivity to habitat perturbations. As logging

practices strongly influence stand scale characteristics, forest management should

take into account bat conservation to benefit from ecosystem services provided by

bats.

KEY WORDS

Bat activity, Mixed-wood forest, Forest structure, Habitat-use, Foraging, Forest

management, Bat conservation.