Small scale habitat preferences of Myotis daubentoniiPipistrellus … · The aim of this study was to investigate habitat preferences of both M. daubentoniiand P. pipistrel-lus, and

INTRODUCTION

Accurate descriptions of bat habitat requirementsare a key part of their conservation management(Walsh and Harris, 1996b; Zeale et al., 2012).Whilst variation in geographical patterns in batspecies abundance on large scales, such as altitudeand predominant habitat type are well known(Walsh et al., 1995; Walsh and Harris, 1996a), fewerstudies have investigated relationships betweenfiner scale landscape patterns and species abun-dance, habitat usage, and foraging patterns (Warrenet al., 2000; Downs and Racey, 2006; Ávila-Caba -dilla et al., 2012). Moreover, given that many insec-tivorous bats are opportunistic feeders (Vaug-han, 1997; Levin et al., 2009), and are quick to exploit swarms of insects (de Jong and Ahlén,

1991), factors affecting distribution and abundanceof prey must be considered when discussing bat habitat requirements, especially in a management sense.

Rivers and riparian habitat are hugely importantforaging sites for Daubenton’s bat, Myotis dauben-tonii (Furmankiewicz and Kucharska, 2009) andcommon pipistrelle, Pipistrellus pipistrellus (Downsand Racey, 2006). These habitats are interdiscipli-nary in terms of their management, and are under in-creasing agricultural, hydrological, and recreationalpressure. Assessment of the relative importance ofsmall-scale habitat characteristics, such as river sur-face state and extent of bank-side trees, should allowclarification of management guidelines for the ben-efit of bat species in upland river systems (Polasky,2008; Linke et al., 2011).

Acta Chiropterologica, 19(2): 255–272, 2017PL ISSN 1508-1109 © Museum and Institute of Zoology PAS

doi: 10.3161/15081109ACC2017.19.2.004

Small scale habitat preferences of Myotis daubentonii, Pipistrellus pipistrellus,

and potential aerial prey in an upland river valley

VICTORIA LOUISE GEORGIA TODD1, 2, 4 and DEAN ANDREW WATERS3

1Ocean Science Consulting Ltd. Spott Road, Dunbar, East Lothian, EH42 1RR, Scotland, United Kingdom2School of Media Arts and Technology, Southampton Solent University, East Park Terrace, Southampton SO14 0YN,

United Kingdom3Environment Department, University of York, Wentworth Way, Heslington, York YO10 5NG, United Kingdom

4Corresponding author: E-mail: [email protected]

Distribution and abundance of two temperate-zone insectivorous bats, Daubenton’s (Myotis daubentonii) and common pipistrelle(Pipistrellus pipistrellus), and their potential prey were studied along an altitudinal river gradient in relation to environmentalvariables including air temperature, wind speed, water surface state, and presence or absence of bank-side trees. Using a Latin squaredesign at ten different habitat combination types, ultrasound recordings and insect sampling were carried out to quantify bat habitatpreferences and potential prey abundance and classification. Myotis daubentonii and P. pipistrellus activity was significantly higherover smooth water river sections with trees on either or both banks while cluttered and rapid water sections were avoided.Conversely, insect abundance was not related to water surface condition or the presence or absence of bank-side trees. Nematocerandipterans made up 98% of insect numbers, with small numbers of brachycerans and cyclorrhaphans. The most common insectfamilies were Chironomidae and Ceratopogonidae. There was no correlation between bat activity and aerial insect activity,suggesting that aerial prey availability is not the sole driver of bat habitat choice. Bat and insect abundance were each correlatedpositively with night-time air temperature. No bat passes or flying insects were recorded at temperatures < 4°C. At 5°C, only M. daubentonii were observed foraging, and at 6ºC there were more M. daubentonii present than any other bat species. Nocorrelation was found between number of bat passes hr-1 and wind speed, moon visibility, moon phase, and percentage cloud cover.Rain did not affect M. daubentonii, but P. pipistrellus preferred to forage on dry nights. Bats were predicted to forage preferentiallywhere aerial insect abundance was highest but this was found to not be case, and other aspects such as detection of prey againstclutter may have an important role to play in habitat choice.

Key words: Daubenton’s bat, common pipistrelle, insects, habitat preferences, temperature, riparian, echolocation constraints

The aim of this study was to investigate habitatpreferences of both M. daubentonii and P. pipistrel-lus, and their potential aerial prey along the upperreaches of the River Wharfe (UK). This study exam-ines ways in which habitat type, insect abundance,and distribution interact with environmental vari-ables that affect insect ecology to influence ulti-mately distributions of these bats along a river gra-dient. Based on current knowledge M. daubentoniiand P. pipistrellus foraging ecology (e.g., Warren etal., 2000; Todd and Waters, 2007), and of specieswith similar ecology such as soprano pipistrelle P. pygmaeus (Abbott et al., 2009; Scott et al., 2010)and little brown bat Myotis lucifugus (Vindigni etal., 2009; Clare et al., 2011), we test the hypothesisthat higher densities of bats will be found foragingalong smooth, tree-lined water sections. Further -more, we test that trees should act as a windbreak forinsects (Lewis, 1969), resulting in increased insectabundance in such localities and that bat activityshould correlate with insect abundance.

MATERIALS AND METHODS

Study Area

The study area was a 27 km stretch of a U-shaped glaciatedvalley in Wharfedale (North Yorkshire, UK). Encompassedwithin the valley is the river Wharfe, which passes through thehamlets of Yockenthwaite and Hubberholme, and villages ofStarbotton, Kettlewell, Conistone, Grassington, and Burnsall,decreasing in altitude respectively (Fig. 1 and Table 1). FromYockenthwaite to Hubberholme the river is a high-energy envi-ronment characterised by confined bedrock with very few poolsand little sediment. From Hubberholme onwards there are ex-tensive areas of ponded, tree-lined areas until Grassington andBurnsall where the river widens. Valley sides are generallysteep-sided along the entire length of the catchment, incised fre-quently by small high-energy streams (colloquially termed gills)with numerous waterfalls. Small deciduous wooded areas aresporadic (often restricted to borders of gills); ca. 20–25% of theWharfe is bounded by riparian woodland, mainly in the lowerriver reaches.

At the time of this study in 1999, Wharfedale and theWharfe were part of an Environmentally Sensitive Area (ESA),but this has been superseded by the Environmental Stewardshipscheme. Much of the Wharfe’s upper reaches are designated asSites of Special Scientific Interest (SSSIs) (Newson and Warner,2007), alongside surrounding terrestrial areas which are alsolisted as Special Protection Areas (SPA), and Special Areas ofConservation (SAC) (Fig. 1). Such designations are, i.a. attrib-uted to abundance and diversity of bankside flora (Bradley,1985), which in turn support riparian specialist fauna.

Habitat Mapping

Observations commenced on 24th April 1999 and termi-nated on 27th June 1999, with 40 nights sampling in total. Study

area was divided into four roughly equidistant sections of differ-ing altitude (Table 1). At each altitude, nine habitat types, as de-fined by Warren et al. (2000), plus a control were sampled.Habitat types were defined by presence of three major riverinefeatures: smooth, rapid (white water and heavy riffles) and clut-tered (projecting rocks and riffles) water surfaces with associ-ated sub-categories based on the presence or absence of bank-side trees (Table 2). The control site was situated over opengrassland, no more than 10 m away from the riverbank, andnever close to any linear landscape elements along which batsare known to forage (Verboom and Huitema, 1997; Downs andRacey, 2006; Morris et al., 2010; Boughey et al., 2011b). Batsgenerally avoid foraging over grassland or semi-improvedgrassland (Walsh et al., 1995; Walsh and Harris, 1996b), whichjustifies choice of the grass control habitat.

At all four altitude sections, predominant vegetation typeswere mapped. Bank height and extent of bank undercuttingalong the Wharfe were also measured at the different habitattypes according to the River Habitat Quality (RHQ) surveymethod (Raven et al., 1998). Each sampled habitat was delim-ited with reflective tape (3M Scotchlight) for nightly re-loca-tion. To avoid the possibility of sampling the same population ofboth bats and insects, contiguous habitats were never selected.All sampled habitats were a minimum of 200 m away from eachother to reduce possibility of any auto-correlative effects orpseudoreplication (Hurlbert, 1984).

Altitude sampling order was determined by random numbertables (Fowler and Cohen, 1990), and a Latin square design wasused to define sampling order for both bats and insects at eachhabitat (Table 3). Starting at the first selected altitude, at 30 minafter sunset on the first night, habitats were sampled from one toten such that all habitat combinations were sampled within eachnight. On the second night, sampling started with habitats ninethen ten, then habitats one to eight etc. (Table 3). This ensuredthat throughout the five nights of observations, each habitat wassampled at the full range of times throughout the night, such asat the beginning of the evening and later throughout the night tocontrol for time of night effects. The full set of five nights wasthen repeated using the same sampling schedule. Number ofnights for each altitude totalled ten (five experimental and fivereplicated). The same procedure was repeated at the next se-lected altitude and so forth. Total number of nights for all fouraltitudes thus totalled 40 including all 20 replicate runs giving400 separate samples of bat and insect activity. Nights weresampled contiguously when possible, but due to weather vari-ability, some nights were separated by one to ten days. Samplingdid not take place if wind speed exceeded 6 ms-1 or in heavyrain, conditions known to affect the abundance of bats (Nyholm,1965) and insects (Peng et al., 1992a). Sampling was alsoavoided under spate conditions.

Insect sampling and bat activity recordings began 30 minafter sunset when M. daubentonii are known to arrive at forag-ing sites (Todd and Waters, 2007). Northern England BritishSummer Time (BST) sun and moon set/rise (phase) times wereobtained from the 1999 Whitaker’s Almanack (Marsden, 1999).

Environmental Variables

At each habitat type, commencing 30 min after sunset andthroughout the night, the following environmental conditionswere recorded: air temperature (ºC), wind direction (º) andspeed (ms-1), percentage cloud cover, moon visibility and phase,and general weather conditions (rain, fog etc.). Temperature was

256 V. L. G. Todd and D. A. Waters

taken by suspending a glass thermometer from a pole 2 m abovethe ground located as close to the river as possible and nevermore than three m away from the water surface. Wind directionand wind speed readings were taken with a compass and Wil -helm Lambrecht 34 anemometer vane. Wind conditions were al-ways recorded at time of insect netting, as gusts of wind affectinsect distribution on very short time-scales (Verboom and Spo -elstra, 1999). Percentage cloud cover was estimated, as cloudcover has been shown to affect the behaviour of bats (Eisen -traut, 1952). Moon phase was taken from the 1999 Whitaker’sAlmanac (Marsden, 1999).

Insect Sampling

Insects were collected using a fine mesh (1 mm2) whitesweep net (Philip Harris, Leicestershire) with a diameter of 355 mm, attached to a 1.2 m aluminium pole. Sweep-netting isre-commended as a method for sampling Diptera (Grootaert etal., 2010), which are the primary prey item for M. daubentonii, P. pipistrellus and P. pygmaeus (Vaughan, 1997). Netting beganas soon as the channel target position had been reached andhead lamps were switched off to reduce swarming effects in-duced by positive phototaxis (Jayanthi and Verghese, 2013).

Temperate bat and insect riparian habitat preferences 257

FIG. 1. River Wharfe hydrological catchment and contiguous watercourses alongside four main sampling areas within the YorkshireDales National Park. AMSL = Above Mean Sea Level. (A) Special Protected Areas (SPA’s), (B) Special Areas of Conservation (SAC’s)

and (C) Sites of Special Scientific Interest (SSSI’s) within the Yorkshire Dales National Park (YDNP)

Control sites were sampled in the same way, ensuring that thenet was above the level of the grass to only sample aerial insects.

Commencing with the first habitat, 40 180º sweeps of thenet were made approximately 1 m above the water surface asclose to the centre of the river channel as possible, avoidingtrees and river banks, to sample accurately foraging space usedby M. daubentonii. Identical sampling location was used eachnight for each habitat.

Once in the net, insects were sprayed immediately with 70%alcohol, and extracted gently with a soft brush into pre-labelledglass vials. Care was taken not to extract insects that had beenattracted to the net post sampling. Insects were counted andidentified with a binocular microscope (Karl Zeiss, Germany)under ×10 magnification. Identification was completed to a mi -ni mum of order or sub-order using Unwin (1981), Chinery(1993) and Armitage et al. (1995).

Bat Activity

An ultrasound detector (Tranquility II, Courtpan DesignLtd, Cheltenham, UK) was set up on a tripod ca. 1 m away fromthe water surface directly facing the desired area. Detection ofbats using typical Myotis and Pipistrellus frequencies are gener-ally limited to a range of 20 m or less (Adams et al., 2012) ensuring that only the habitat selected, and not adjacent habi-tats were sampled. For control sites, the detector was placed fac-ing away from the river. The highly directional response of similar designs of detector (Waters and Walsh, 1994) ensuredthat bats foraging over the river would not be detected at thecontrol site, and only bats using the meadow would be record-ed. The detector, configured to record 40 ms in time-expansionmode, was connected via the right channel to a stereo cassetterecorder (Genexxa SCR-59). Detector high frequency outputwas time-expanded (10×) for analysis, as recommended by Jones et al. (2000). Detector microphone had a frequency response of 12–200 kHz (± 20 dB), and the cassette record-er ca. 400 Hz–12 kHz (± 3 dB). Recordings were stored on 90 min normal position tapes (TDK, Luxembourg D-IEC/type I).

At each site, time, location, date and habitat type wererecorded on the left channel of the cassette recorder via the internal microphone. Recording lengths varied from site to site,as they were conducted during insect sampling, and were termi-nated once all insects had been removed from the net, but a min-imum of five minutes was set for each site as in the studies ofRacey et al. (1998) and Rydell et al. (1999).

On this same river system, Senior et al. (2005) found thatwhile M. daubentonii may fly some kilometres to a foragingsite, once there, they tended to forage repeatedly in one patch ofaround 100 m or less. While this ensures that we measured batactivity specific to the chosen habitat type, it is not possible to


Major category Sub-category Habitat No.

Smooth water Trees present on both banks 1surface Trees present on one bank 2

No trees on either bank 3Rapid water Trees present on both banks 4

surface Trees present on one bank 5No trees on either bank 6

Cluttered water Trees present on both banks 7surface Trees present on one bank 8

No trees on either bank 9

Control Grass 10

TABLE 2. Major physical features of the river Wharfe accordingto Warren et al. (2000). Trees were more than five metres highwith no gaps between adjacent trees; each tree was more thanfive metres in width including branches

disentangle bat activity, as measured by bat passes, from num-bers of individual bats.

Spectrographic analysis of time-expanded calls was per-formed initially using a shareware dual-channel audio spectrumanalyser (Spectrogram version 4.2.12 for Windows 95) on a PC.Spectrograms of calls were constructed using a 512 point FastFourier Transform (FFT) on a scope display in real time. Batspecies identification was confirmed by analysing associatedtime-expanded audio sequence using Batsound (Peters-son Electronic) on a PC and observing the call spectrogram (512 point FFT, Hamming Window). Calls of M. daubentoniiwere separated from those of Pipistrellus species by the lowerterminal frequency and the lack of a constant frequency tail at the start of approach phase. While separation of calls of M. daubentonii from those of other Myotis species is problem-atic (Walters et al., 2012), previous work at this site has shownthat M. dau bentonii was by far the commonest Myotis spe-cies (Warren at al., 2000), and so all Myotis calls were attributedto M. daubentonii unless different significantly in expected parameters.

Data Analysis

Parametric statistical procedures were carried out on all nor-mally distributed data or transformed non-normal data. Non-parametric statistics were used when transformation failed. Datawere tested for differences due to the factor of altitude, andwhere this had no significant effect, altitudes were pooled.Mann-Whitney rank sum tests were used to test for activity ofbats and insects in relation to wind, moon phase and cloudcover. One-way ANOVAs were used to test for temperature dif-ferences between habitats 1–10 at any one location. Windroseswere generated in Matlab (version 5.111) to determine predom-inant wind direction and speed. Mean wind directions for thefour locations were calculated using circular statistics (Bats -chelet, 1981) programmed into Microsoft Excel® 97 (5.0). Kru -s kal-Wallis tests determined significant differences in maximumwind speed recorded between different habitat types and habitatpreferences of both bats and insects. Post hoc Dunn’s tests formultiple comparisons determined where differentiation in bathabitat selection occurred. Spearman’s rank order correlationswere used to examine the relationship between bat activity andair temperature, and bat and insect activity and wind speed. A Pearson’s product moment correlation was carried out on the

TABLE 1. Altitude Above Mean Sea level (AMSL), length (L)and area (A) of all habitats combined at all four study locations(assuming rectangular shape)

Location Altitude (m) L (km) A (km2)

Yockenthwaite 270 1.316 0.145Starbotton 210 1.783 0.242Grassington 165 7.495 0.492Burnsall 150 2.991 0.189

log-transformed number of insects in relation to temperature.For graphical presentation and statistical analysis, all bat passeswere normalised for time (bat hr-1).

RESULTS

Habitat Mapping

Individual habitat sizes varied between the fourlocations; Grassington contained the largest habitatpatches in comparison to all other locations. Pre -dominant trees present at all habitats were alder(Alnus glutinosa), smooth-leafed elm (Ulmus car-pi nifolia), sycamore (Acer pseudoplatanus) andhawthorn (Crataegus monogyna) with willow (Salixspp.) present mainly in lowland areas. Trees weresparser in upland areas, including Starbotton andYockenthwaite, than at lower altitudes in Gras sing -ton and Burnsall, where the majority of river reacheswere tree-lined. Left and right bank heights at all theriverine habitat types at all altitudes were generallyabove 1 m (Yockenthwaite 0 ± SD: 2.0 ± 0.74 m,Starbotton: 1.3 ± 0.52 m, Grassington: 5.8 ± 9.24 mand Burnsall: 2.9 ± 5.42 m). Inflated means atGrassington and Burnsall were attributed to the riverrunning through a scar at Burnsall and a steep valleyin Grassington where the bank height was essen-tially the scar/valley height. Highest proportion ofundercutting was observed at Yockenthwaite (44.4%of banks were undercut), followed by Star bot-ton (38.9%) with least undercutting observed at Gras sing ton (16.7%) and then Burnsall (22.2%). At Burn sall, extensive areas of one side of the bankwere reinforced concrete and devoid of riparian veg-etation, even where trees were present.

Environmental Variables

Most air temperatures were between 7°C and12°C, with few very warm or very cold nights. Itwas not possible to compare the same night temper-ature differences at the different habitat types, be-cause of the effects of time and temperature drop offwith increasing night length; however, given thatover the 40 nights of sampling, each habitat typewas sampled within the entire range of timesthroughout the night, data were pooled for location.Habitat types within one location did not influencetemperature with no significant differences in tem-perature between habitats 1–10 at any one location(one-way ANOVA: F = 0.165, d.f. = 9, P = 0.98).

Wind in the upper Wharfe valley was omni-directional, varying considerably from night to nightat the two highest altitudes, Yockenthwaite and


TABLE 3. Latin square design for habitat sampling, e.g. for onelocation. In bold — habitat rotations by two

Star botton (Fig. 2). Grassington and Burnsallshowed the most polarised wind directions with anoverall westerly direction at Grassington andsoutherly to southwesterly at Burnsall. This corre-sponds to the general orientation of the valley inthese two areas. On nights where wind was present,there was no significant difference in maximumwind speed record ed between different habitat types(Kruskal-Wallis test: H = 8.21, d.f. = 9, P = 0.51).

Insect activityTotal numbers of insects captured per night is

presented. No corrections were made for duration ofthe night or for hours of trapping, since insect activ-ity occurs predominantly at dusk and dawn.

Insect habitat preferencesGiven that there was no overall difference in the

pattern of insect abundance for all four altitudes,data were pooled (combined) for all altitudes(Kruskall-Wallis H = 1.84, d.f. = 3, P = 0.61). In -sects expressed no particular preference to any onehabitat type (Fig. 3). Kruskal-Wallis tests showedthat there were no significant differences in numberof insects flying above the water surface at the 10 different habitat types at any of the altitudes either pooled for season (H = 16.24, d.f. = 9, P = 0.062) or within each altitude (Burnsall, H = 5.46,P = 0.79; Grassington, H = 7.90, P = 0.55; Star -botton, H = 7.18, P = 0.62, and Yockenthwaite, H = 7.40, P = 0.60; in all cases d.f. = 9).

Insect activity in relation to temperatureBelow 4°C no insects were observed flying, and

most insects were caught at temperatures between9–14ºC. Above 4°C, numbers of aerial insects in-creased with increasing air temperature (Fig. 4).

1 2 3 4 5

1 9 7 5 32 10 8 6 43 1 9 7 54 2 10 8 65 3 1 9 76 4 2 10 87 5 3 1 98 6 4 2 109 7 5 3 1

10 8 6 4 2

Nights

Hab

itat

A Pearson’s product moment correlation on the log-transformed number of insects revealed that therewas a positive correlation between temperature andnumbers of insects caught when data from all alti-tudes were pooled (r = 0.303, n = 400, P < 0.001).The same analysis at all individual altitudes revealedthe same pattern. There was also a strong correlationbetween maximum number of insects caught on eachnight with preceding maximum day temperature(Pearson’s product moment correlation: r = 0.578, n = 40, P < 0.001).

Insect activity in relation to wind, cloud cover, moonphase and rain

There was a weak negative correlation betweenminimum wind speed (ms-1) and number of insectsat all altitudes (Fig. 5). As wind speed increased,number of insects captured decreased, with most insects being caught at 0.0 ms-1 and least between1.8–4.0 ms-1 (Spearman’s rank order correla-tion: Rs = -0.16, n = 400, P < 0.05). There were no

significant differences in number of insects caughtwith different levels of moon visibility, phase andpercentage cloud cover (Mann-Whitney rank sumtests and Kruskal-Wallis tests, P = 0.19 and P = 0.42, respectively). There were significantlyfewer insects flying in the rain (0 ± SD = 1.8 ± 3.37)than during dry conditions (12.9 ± 45.80; Mann-Whitney rank sum test, T = 6509.0, P = 0.015).

Insect diversitySix orders and 24 families of insects were re -

corded during netting sessions. Of these, dipteranswere the most prevalent group (98% of all orders).Of the three dipteran sub-orders, 99.7% were Nema -tocera, with small numbers of Brachycera andCyclorrhapha. All other orders were present in lownumbers. Of nematocerans, Chironomidae and Ce -ratopogonidae were by far the most common fam -ilies (Fig. 6). Chironomids swarmed mainly in smallpatches, typically over objects projecting out of thewater, or under canopies of overhanging trees.


FIG. 2. Wind direction (degrees) expressed as windroses at for 10 nights from highest (a) to lowest (d) altitudes along river Wharfe

Bat activityFive species of bat were recorded at all altitudes:

M. daubentonii (n = 3,611 counts), P. pipistrel-lus (n = 4,259) and, to a considerably lesser extent,the soprano pipistrelle P. pygmaeus (n = 517 ) andtwo noctule (Nyctalus) spp. (n = 393), which were not differentiated between the common noc-tule (N. noctula) or the lesser noctule (N. leisleri).Both Nyc talus spp. and P. pygmaeus were exclud-ed from statistical analysis, as sample sizes were too small. On most occasions following hiatuses in bat activity, P. pipistrellus and P. pygmaeusappeared at the foraging sites simultaneously and

the two species could be observed foraging in close proximity to each other. All species arrivedat foraging sites from both upstream and down-stream.

Bat habitat preferencesBoth M. daubentonii and P. pipstrellus showed

significant differences in number of bat passes ineach habitat at all four altitudes (Kruskall-Wallis H = 153.77, d.f. = 9, P < 0.05) with the exception ofP. pipi strellus at the lowest altitude site whichshowed no strong habitat associations. There werealso differences in numbers of bats with altitude


FIG. 4. Relationship between aerial insect number (pooled for altitude and all habitat types) with air temperature (r² = 0.092)

FIG. 3. Mean ± SD of number of aerial insects caught with sweep net above water surface (and over grass = control), pooled for all four altitudes (n = 40 nights)

pooled across habitats (H = 30.16, d.f. =3, P < 0.05for M. daubentonii and H = 30.38, d.f. = 3, P < 0.05for P. pipistrellus) with a trend for higher numbersof bat passes at the lowest altitude. Dunn’s pair-wisemultiple comparisons test showed these differencesto occur only between Altitude 1 (lowest) and allother altitudes for M. daubentonii (P < 0.05) show-ing that the lowest altitude had a statistically signif-icantly higher number of bat passes than all higheraltitudes, but that there were no differences betweenaltitudes 2, 3 and 4. For P. pipi strellus there were

significantly fewer bats at the highest site, altitude 4,than all other altitudes (Dunn’s test P < 0.05) but nodifferences between altitudes 1, 2 and 3.

To isolate differences in habitat selection,Kruskall-Wallace tests were undertaken betweenhabitat types within each altitude, followed byDunn’s pair-wise comparisons in the event of a significant difference being found in bat passes between habitat. For the lowest altitude, only M. dau bentonii showed a preference for any habitat,with habitats 1 (smooth, trees both banks) and 2


FIG. 5. Relationship between numbers of aerial insects caught with a sweep net (pooled for all altitudes and habitat types) above the water surface with minimum wind speed

FIG. 6. Percentage of most prevalent aerial insects found during the 40 nights of sweep net sampling along river Wharfe. Samples been pooled for altitude and habitat type. Insect families represented by < 0.1 % not shown

(smooth, trees on one bank) being preferred over allothers. At altitude 2, the same pattern was repeatedwith only habitats 1 and 2 showing any higher activ-ity than the others for M. daubentonii. There wasonly a slight preference shown for P. pipistrellus forhabitats 1 and 3 (smooth water, no trees). At altitude3, habitat 1 showed significant higher activity thanall other habitats for M. daubentonii with a muchless marked selection for any habitats by P. pipi-strellus. At altitude 4, the highest site, habitats 1 and2 were again preferred by M. daubentonii over allothers with no other habitats being selected for.Again, P. pipistrellus showed a slightly higher number of passes comparing habitat 1 with someother habitats but the pattern was not as clear as forM. daubentonii.

As the pattern for habitat selection was similarbetween altitudes, it is shown pooled between alti-tude for clarity (Fig. 7). It is clear from the analysisthat both species preferred smooth water sectionswith trees on either both or one banks (habitats 1 and2, respectively) but that the preference is mostmarked in M. daubentonii. Rapid (habitats 4, 5 and6) and cluttered (habitats 7, 8, and 9) were not se-lected for by either species. Throughout 40 nights ofsampling, M. daubentonii activity was recorded atevery type 1 habitat (smooth water, trees present onboth sides) location.

Bat activity in relation to environmental variablesThere was a positive correlation between num-

ber of both M. daubentonii and P. pipistrellus and airtemperature (Spearman’s rank order correlation: M. daubentonii, Rs = 0.207, n = 393, P < 0.001 andP. pipistrellus, Rs = 0.283, n = 393, P < 0.001). Nopasses were recorded at temperatures of 4°C orbelow. At 5°C, only M. daubentonii were recorded(0 ± SD = 62 ± 151.8 bat passes hr-1), and whileboth species were recorded at 6ºC, there were more

M. daubentonii (66 ± 113.6 bat passes hr-1), than P. pipistrellus (47 ± 140.9 bat passes hr-1).

There was no correlation between numbers of batpasses hr-1 and wind speed for either M. daubentoniior P. pipistrellus (Spearman’s rank order correlation:P = 0.22 and P = 0.72, respectively). Furthermore,moon visibility, phase and percentage cloud coverhad no significant effect on the numbers of batpasses recorded for either species (Mann-Whitneyrank sum tests and Kruskal-Wallis test, P = 0.75 andP = 0.42, respectively). There was no significant difference (P = 0.88) in the numbers of bat passes hr-1 on dry (0 ± SD = 72.4 ± 125.88) and rainy(107.0 ± 186.68) nights for M. daubentonii, butthere were significantly more bat passes hr-1

recorded on dry (97.8 ± 186.42) than on rainy nights(43.3 ± 86.82) for P. pipistrellus (Mann-Whitneyrank sum tests, T = 6320.5, P < 0.01).

Bat activity in relation to insect activityThere was no significant correlation between

numbers of bat passes hr-1 and aerial insects for M. daubentonii (P = 0.18) but there was a weak, butsignificant positive correlation between number ofbat passes hr-1 and aerial insects for P. pipistrellus(Spearman’s rank order correlation, Rs = 0.120, n = 400, P = 0.003).

DISCUSSION

Habitat Mapping and Environmental Variables

This study found that habitat type did not in-fluence air temperature within three m of the river; how ever, given that observations were not de-signed specifically to measure temperature varia-tions between habitats, microscale vegetation andtopographic temperature-induced effects outwith the edge of each habitat type might not have been


FIG. 7. Bat activity expressed as 0 ± SD bat passes hr-1 in relation to habitat type (water state) of both M. daubentonii andP. pipistrellus, pooled for all four altitudes

A B

discerned (Moore et al., 2005). This study focuseson insects and bats over the river, so temperaturesmeasured are more likely to reflect conditions overthe river itself than measurements taken in the mid-dle of habitat types to avoid edge effects (see Mele -ason and Quinn, 2004; Moore et al., 2005 for studiesof how riparian forests influence environmentalconditions in stream ecosystems).

Wind was stronger and more omni-directional inthe valley’s upper reaches in comparison to lower al-titudes, characteristic of more exposed upland areas.When comparing habitat types with and withouttrees, no difference in wind speed was observed.Local topography and tree cover have been shown toinfluence wind characteristics at both microscale(Lewis and Stephenson, 1966; Lewis and Dibley,1970) and mesoscale (Danard, 1977) levels. Pres -ence of microscale fluctuations may have beenmasked by edge effects i.e. being only 3 m from theriver. Additionally, given variation of wind speedsover short time scales, a vigorous comparison be-tween habitat types would require simultaneousmeasurements (Davies-Colley et al., 2000), whichwas clearly unfeasible in this study.

Insect Diversity

Of the six orders and 24 families of insectscaught, nearly all were dipterans and 98% of thosewere nematocerans of the families Chironomidaeand Ceratopogonidae. This corresponds to findingsof many other riparian insect studies (e.g., Nelson,1965; Belwood and Fenton, 1976; Anthony andKunz, 1977; Jónssen, 1987; de Jong and Ahlén,1991; Gray, 1993; Racey et al., 1998; Manel et al.,2000; Warren et al., 2000) and is typical of Euro -pean upper course streams. A number of low-flyingdipterans, such as dolichopodids (Peng et al.,1992b), trichopterans (Solem, 1978; Peng et al.,1992b), tipulids (Service, 1973) and psychodids(Peng et al., 1992b) were all caught regularly,though at low levels, in the sweep samples. Whilethere are inherent biases in taxonomic groups caughtdue to sampling methods, sweep-netting is an estab-lished method for Diptera (Grootaert et al., 2010)that make up the bulk of the diets of M. daubentoniiand P. pipistrellus and do represents the insect preyavailable to these bats at these locations.

Insect Habitat Preferences

Insects exhibited no habitat preferences at any altitude, were not affected by water surface state or

clutter, and did not select for trees; however, thevery high levels of variability (Fig. 4) suggest thatany underlying differences would be masked by thehigh levels of variation, probably due to the ‘snap-shot’ approach to our sampling methodology.

Several studies, using various insect capturetechniques, have reported contrasting insect distri-butions and habitat preferences. For example, Ry -dell et al. (1996) captured more insects over openwater than water overhung with a tree canopy.Ekman and de Jong (1996) also reported higher in-sect abundance in open areas compared to forestededges; however, Grindal and Brigham (1999) andFuentes-Montemayor et al. (2013) found that insectscongregate in high abundance at the margins of for -ested areas, and increased concentrations of volantdipterans were found in close proximity to wood-land hedgerows by Peng et al. (1992b). This studyfocused on insect abundance in the central riverchannel, so abundance of insects at riparian treeedges were not meas ured, but it was found that in-sect abundance was greater over the river than overterrestrial controls, which supports previous find-ings by Barclay (1991) and Rydell et al. (1996).

In a study on the River Wharfe, Warren et al.(2000) found aerial insects in higher densities oversections of smooth water with contiguous ripariantrees, whereas Rydell et al. (1996) reported that aer-ial insects were most abundant over ripples on a river in north-eastern Scotland. Low-flying insectshave been suggested to benefit from the boundarylayer effect (Taylor, 1958, 1960, 1974), in which a layer of relatively still air is found directly abovethe water surface (Taylor, 1960). This may explainwhy insects in many studies are found in increasedconcentrations near trees, the ground, or water sur-faces. Aerial insects found in increased abundanceover ripples by Rydell et al. (1999) may have beenbenefitting from reduced air turbulence immediatelyabove ripples.

Height of the river channel’s banks (> 1 m) mayhave also contributed to a boundary layer effect byreducing wind speed, thus protecting insects directlyabove the water surface in a similar manner towooded hedgerows, as reported by Lewis (1969).Proportion of bank undercutting was generallygreater at all habitat types in the Wharfe’s upperreaches, reflecting the higher energy environment ofthe watercourse at elevated altitudes. Increased bankundercutting correlates generally with increased sur-face ripples, reduced air turbulence, and increasedboundary layer effects (Nanson and Hickin, 1986).Moist ledges associated with undercut banks may


be important sources of refuge for adult insects inopen meadows, where there is little overhead cover(Myers and Resh (2000).

To determine insect habitat preferences fully,species-level identification of insects would be re-quired, rather than to family-level as in this, andmany other studies. Additionally, the netting proce-dure was completed in the centre of the channel, sovariations in insect abundance across the channel orvertically through the air column above it may havenot been detected, and certain insect families mayhave been under-sampled as a result, despite theirpresence and potential availability as prey to forag-ing insectivorous bats such as M. daubentonii and P. pipistrellus.

Insect Activity in Relation to Temperature andEnvironmental Variables

In general, this study found that aerial insectabundance was correlated positively with nightly air temperature, as observed by Bridcut (2000),Arbuthnott and Brigham (2007), Ciechanowski etal. (2007) and Kusch and Schmitz (2013). Relation -ship between insects in flight and temperature iscomplex, as temperature affects both species’ flightability and population levels through breeding(Taylor, 1963), which may explain why correlationsbetween night temperature and insect abundance,though positive, were weak. Peak number of insectswas correlated strongly and positively with preced-ing maximum day temperature, similar to a findingby Peng et al. (1992a). Temperature thresholds arespecies- (Waringer, 1991) and sex-specific (Ander -sen, 1979), although, below 4°C no aerial insectswere caught, and thus no aerial insects were avail-able to aerial hawking bat species in Wharfedale.

This study found a negative correlation betweenwind speed and number of insects. Peng et al.(1992a) observed comparable results in nocturnaldipteran families with the minimum wind speed dic-tating abundance. Large insects are more adept inmaintaining control of their flight track than smallerindividuals (Lewis, 1967; Johnson, 1969; Taylor,1974). Trichopterans, for example, are relatively largeinsects and therefore not affected greatly by windspeed (Waringer, 1991). Considering that most in-sects caught in this study were small midges, it ishardly surprising that strong wind dissipated swarms.

Significantly fewer insects were recorded flyingin rain conditions than on dry nights, as precipitationmay deter insects from flying (Burles et al., 2009).This is in agreement with Peng et al. (1992a), who

reported that many volant dipteran families includ-ing Cecidomyiidae and Anisopodidae express pref-erence for dry conditions; however, this is in con-trast to other studies that did not find precipitation tohave a significant effect on aerial dipteran activity(Waringer, 1991), particularly when controlled forair temperature (Cucco and Malacarne, 1996).

Cloud cover had no significant effect on insectactivity, supported by Kovats et al. (1996). Lang etal. (2006) and Yela and Holyoak (1997) both sug-gest that lunar phobia induced crypsis occurs inmany insect species; however, there was no evi-dence that insect activity differed with moon phasewithin this study.

Bat Activity and Habitat Preferences

Bat species encountered most frequently forag-ing along the Wharfe were P. pipistrellus (49.8% of bat passes) and M. daubentonii (42.2%) and to a much lesser extent, P. pygmaeus (5.8%), and Ny -ctalus spp. (3.2%). Both Warren et al. (2000) andBellamy et al. (2012) used transects to quantifypresence of M. daubentonii and Pipistrellus spp. inthe Yorkshire Dales National Park (YDNP), andwhilst methodologies differed, relative proportion ofbats recorded in these studies was similar to that ofthe present investigation. Pipistrellus pygmaeus isassociated commonly with riparian habitat (Abbottet al., 2009; Scott et al., 2010), so the low proportionpresent in the YDNP is surprising. In the UK, P. pi -pi strellus has larger foraging ranges and is less se-lective with foraging sites than P. pygmaeus. The twosympatric pipistrelle species have shown evidenceof habitat partitioning (Davidson-Watts et al., 2006;Nicholls and Racey, 2006a, 2006b), including mutu-ally exclusive feeding grounds (Nicholls and Racey,2006a, 2006b). This niche partitioning, coupled withthe lack of major water bodies in the YDNP may explain the low proportion of P. pygmaeus (Bellamyet al., 2012).

Myotis daubentonii preferred smooth water withtrees on both or one side of the river, and to a lesserextent smooth water without trees. Individuals ac-tively avoided both rapid and cluttered water sec-tions. Habitat preferences of M. daubentonii pre-sented in this study are typical of this species, whichfeeds mainly over smooth water on lakes, ponds, orrivers (Nyholm, 1965; Jones and Rayner, 1988; Kal -ko and Schnitzler, 1989; Rydell et al., 1999; Toddand Waters, 2007). Warren et al. (2000) found simi-lar habitat preferences expressed by M. daubento-nii in upper Wharfedale although individuals in that


study avoided areas without trees. This may havebeen because that study found significantly more insects in sections with trees, and differences could perhaps be explained in terms of bats select-ing for prey abundance, rather than presence oftrees. Numerous studies have reported higher bat activity near trees (Holloway and Barclay, 2000;Lloyd et al., 2006; Boughey et al., 2011a; Fuentes-Monte mayor et al., 2013) and various other bat species express preference for water and riparian habitat (Fellers and Pierson, 2002; Biscardi et al.,2007; Johnson et al., 2010; Kniowski and Gehrt,2014).

Myotis daubentonii takes refuge occasionallywithin trees whilst foraging (e.g., Nyholm, 1965;Childs and Aldhous, 1995; P. Richardson, personalcommunication; VLGT personal observation fromradio tracking in Wharfedale), possibly for rest, con-sumption of prey between excursions, protectionfrom crepuscular predators (Speakman, 1991; Fukuiet al., 2006) or night roosting and going into torporwhen temperatures are unfavourable. Proximity today roosts may also contribute to bats’ preferencefor habitats with riparian trees; higher activity of M. daubentonii in areas closest to roost sites hasbeen reported (Childs and Aldhous, 1995; Kapfer etal., 2008). Bare banks are likely to be less attractivedue to higher risk of predation, an increased energyexpenditure during take-off (Nor berg and Rayner,1987), and potentially longer commute times fromtree roosts which may result in lost foraging time(Drury and Geiser, 2014). Myotis dau bentonii mayalso select tree-lined river sections due to increasedprey abundance. Insects have been shown to occur inhigh abundance around woodland (Fuentes-Montemayor et al., 2013) and congregate along treeedges (Grindal and Brigham, 1999), correlating withincreased activity in other bat species (Morris et al.,2010). Similar habitat preferences were observed inthe trawling bat Myotis capaccinii by Biscardi et al.(2007), in which selection of riparian vegetation wasprioritised over smooth water state.

Trawling bats may avoid areas of rough waterand rapids due to physical risk of colliding with therippling surface or manoeuvrability constraints(Rydell et al., 1999); however, it is more likely thatsuch areas of turbulent water are avoided by batsdue to ultrasonic noise generated by fast-flowingwater interfering with their use of echolocation(Mackey and Barclay, 1989; Rydell et al., 1999;Gillam and McCracken, 2007). Even small ripplesmay generate brief ultrasonic pulses, adding‘acoustic clutter’ to echolocation signals (Kalko and

Schnitzler, 1993), and being mistaken potentially for echoes from insect prey (Rydell et al., 1999),leading to reduced foraging efficiency. Physicalclutter and obstacles around the water surface havea similar acoustic impairment effect, creating back-ground echoes which must be distinguished fromprey echoes (Mackey and Barclay, 1989; Zsebok etal., 2013). For example, presence of duckweed on a pond was found to interfere with capture ofprey by M. daubentonii, by scattering echolocationpulses and creating acoustic clutter (Boonman et al.,1998). Furthermore, it has been suggested that a smooth water surface can be used as an acousticmirror to increase foraging efficiency (Zsebok et al.,2013). Myotis daubentonii has been found to con-centrate its foraging activities over calm surfaces,avoiding the adjacent area with small ripples (Rydellet al., 1999), and Nyholm (1965) reported that M. dau bentonii was never observed foraging overrapid water, consistent with the findings of thisstudy. Similar behaviour has also been reported inthe trawling/gaffing and aerial hawking large-footedbat Myotis adversus (Jones and Rayner, 1991).Foraging activity of both M. lucifugus and the long-legged myotis, M. volans is affected negatively bythe presence of artificial physical clutter, probablydue to the combination of both increased acousticclutter effect as discussed above and the increasedcosts associated with flying through a more complexenvironment (Brigham et al. (1997b). From an ener-getics perspective, foraging in close proximity tosmooth surface water increases a bat’s flight effi-ciency due to ground effect (reduced drag and en-hanced lift), consequently reducing energy expendi-ture (Aldridge, 1988). Additionally, when aerialhawking species concentrate their activity in openhabitats free of physical clutter or obstacles, thestringent demands of slow, manoeuvrable flight arerelaxed (Aldridge and Rautenbach, 1987). Foragingin open riparian habitats is likely to lead to more ef-ficient echolocation for M. daubentonii compared tohabitats requiring navigation around physical obsta-cles and the associated acoustical clutter.

Pipistrellus pipistrellus showed a preference forsmooth water habitat with trees on either side, butunlike M. daubentonii, utilised all other habitattypes, including rapid and cluttered water. Warren etal. (2000) also found that P. pipistrellus preferredsmooth water with trees on both banks; however,bats were found to select against habitats 3 (smoothwater, no trees), 5 (rapid water, trees one bank) and 6 (rapid water, no trees), although bats were detected in all habitat types. The findings of this


study conform with the known ecology of P. pipi -strellus, which displays more generalist behaviourthan M. daubentonii and its sister species P. pygmae -us (Bellamy and Altringham, 2012; Kusch andSchmitz, 2013). Pipistrellus pipistrellus makes useof a range of habitats, but is often associated withedge habitats (e.g. woodland edges) and linear land-scape features such as hedgerows and tree lines(Dav idson-Watts et al., 2006; Nicholls and Racey,2006b; Boughey et al., 2011b; Dunn and Waters,2012).

Although P. pipistrellus is an aerial hawker andforages generally in the air space above the < 1 mused by M. daubentonii (Nyholm, 1965; Jones andRayner, 1988), both species have been observed for-aging low above the water surface (Racey and Swift,1985; Jones and Rayner, 1988), so energetic andacoustic constraints discussed above may also affectP. pipistrellus. Kalko and Schnitzler (1993) defined‘cluttered space’ as being within 5 m of any objectproducing acoustic clutter. High frequency sound at-tenuates rapidly in air; however, acoustic clutter as-sociated with sections of rapid or very structurallycluttered water may extend to the higher air spaceused by P. pipistrellus, leading to avoidance of rapidwater sections for acoustic reasons (Kalko andSchnitzler, 1993) rather than physical collision risks.

Bat Activity in Relation to Temperature and OtherEnvironmental Variables

Activity of M. daubentonii and P. pipistrellus in-creased generally with temperature, a result foundcommonly in other studies (e.g., Negraeff andBrigham, 1995; Walsh and Harris, 1996a; Hayes,1997; Hamilton and Barclay, 1998; Rachwald et al.,2001; Arbuthnott and Brigham, 2007; Ciechanowskiet al., 2007; Kusch and Schmitz, 2013). In thisstudy, increased bat activity was due likely to in-creased abundance of aerial prey with temperatureas suggested by both Arbuthnott and Brigham(2007) and Ciechanowski et al. (2007). Decreasingtemperature correlated with decreased activity inboth species and their prey. Myotis daubentoniiand P. pipistrellus were not observed foraging be-low 4°C and 6°C respectively, agreeing with Rie ger(1996) who reported that M. daubentonii forage intemperatures between 3.5°C and 25°C. As with P. pi pi strellus, Rydell (1989) reported little activityin the Northern bat (Eptesicus nilssoni) and theirprey below 6°C, which is accepted generally asbeing the temperature at which foraging ceases inmost insectivorous bat species that do not gaff/trawl.

This study was not able to determine proportionof feeding buzzes that were attributed to aerialhawking or water surface trawling/gaffing prey cap-ture events; however, Todd and Waters (2007) deter-mined that while aerial hawking in M. daubentoniiaccounts for 86% of all prey capture attempts, de-spite aerial insect availability falling close to zerofor much of the night, prey density on the water sur-face can be an order of magnitude higher than aerialprey density, and increases through the night due toaquatic invertebrate drift. That study demonstratedthat M. daubentonii is able to switch prey capturestrategy to reflect change in prey availability on thewater’s surface. It is likely that similar processesmay have been operating in Wharfedale, whichmeans that M. daubentonii is less likely to rely on airtemperature as a means of determining aerial preyavailability, since it can exploit non-aerial prey onthe water surface, when other species such as P. pi -pi strellus stop foraging due to falling temperaturesand prey availability.

Bat activity was not affected by wind speed, similar to findings by Rieger and Alder (1993) andBoonman et al. (1998) for M. daubentonii, and Ver boom and Spoelstra (1999) for P. pipistrellus,while Rydell (1989) found that foraging activity inE. nilssoni was correlated negatively to wind speed,attributed to increased energetic expenditure.

Lunar light has been suggested to depress activ-ity of insectivorous bats through increased predationrisk and/or decreased insect prey abundance (Langet al., 2006); however, such observations are gener-ally anecdotal, with insufficient quantitative data tosupport or refute them (Negraeff and Brigham,1995). This study found that moon phase did not sig-nificantly affect foraging activity of bats. Thesefindings comply with numerous studies such asKarlsson et al. (2002) and Thomas and Jacobs(2013) for a range of bat species, with Negraeff andBrigham (1995) for M. lucifugus; Brigham et al.(1997a) for the lesser long-eared bat (Nyctophilusgeoffroyi) and Gould’s long-eared bat (N. gouldi)¸and Gaisler et al. (1998) for M. daubentonii and P. pipistrellus. In Wharfedale, lunar light was notvisible frequently at habitats with overhanging trees,thus any potential effect on foraging bats was likelyreduced.

Rain did not affect activity of M. daubentonii,comparable to findings reported by Rieger (1996),Boonman et al. (1998) and Thomas and Jacobs(2013). Rydell (1989) also reported little effect oflight rain on foraging activity of E. nilssoni and likewise for M. lucifugus (Fenton et al., 1983).


There were, however, fewer P. pipistrellus passesrecorded on rainy nights than on dry nights comple-mentary to the findings of Erickson and West (2002)and Kusch and Schmitz (2013). This was probablydue to reduced availability of aerial insects. Myotisdau bentonii may still be able to continue foragingby exploiting insects on surface water through trawl-ing/gaffing, whereas exclusively aerial hawking batslike P. pipistrellus cannot exploit such resources inrainy conditions (Todd and Waters, 2007). Addition -ally, insulation provided by bats’ fur is reducedwhen wet, leading to higher energetic costs whenflying in the rain (Burles et al., 2009), which maynot be offset if foraging when insect abundance islowered. Reduced heat insulation may be more of a problem for the smaller P. pipistrellus.

Bat Activity in Relation to Insect Activity

Only a very weak correlation was found betweenthe number of bat passes and aerial insect abundancefor P. pipistrellus, with no correlation found for M. daubentonii. While we did not measure insectdrift on the water surface in this study, Todd andWaters (2007) found significant insect availabilityon the water surface on a river in Wales and demon-strated the ability of M. daubentonii to strategy-switch from hawking to gaffing depending on timeof night, prey availability, temperature, and alti-tude. It was clear, however, that both bat speciesceased to forage below 4°C, most likely due to thelack of prey availability, as this was also the temper-ature threshold below which no aerial insects werecaught, and was presumably too cold for trawling/gaffing.

In addition to trawling/gaffing from the surface,M. daubentonii aerial hawks mainly in the airspace1 m above the water surface (Nyholm, 1965; Jonesand Rayner, 1988) whereas P. pipistrellus tends toforage in the airspace above that of M. daubentonii.This vertical differentiation in airspace use likely af-fects the aerial prey species available to the two batspecies, particularly when considering that Dipteracommonly express strong vertical spatial distribu-tion (Peng et al., 1992b). A number of low-flyingdipterans, such as dolichopodids (Peng et al.,1992b), trichopterans (Solem, 1978; Peng et al.,1992b), tipulids (Service, 1973), and psychodids(Peng et al., 1992b) were all caught frequently,though at low levels, in the sweep samples. Conver -sely, anisopodids, which are known to exploit higherairspace (Peng et al., 1992b), were not present.Insects captured in this study corresponded to those

expected to be found in the foraging airspace usedby M. daubentonii.

Prey size and encounter rate are important fac-tors influencing the degree to which a predator’s dietdiffers from available food. Prey selection strategiesvary amongst insectivorous bats, with some speciesselecting actively the most profitable prey (Koselj etal., 2011; Gonsalves et al., 2013) whilst others areopportunistic (Zeale et al., 2011; Cryan et al., 2012)with changes in diet reflecting changes in availabil-ity. Given the abundance and wide diversity of sim-ilar-sized aquatic insects over freshwater, it is un-likely that M. daubentonii is a particularly selectivefeeder, supported by diet analyses by Vesterinen etal. (2013). Despite the suggestion by Taake (1992)that M. daubentonii selects only insects larger than 3 mm, overall, M. daubentonii appears to expressopportunism, feeding primarily on the most abun-dant insects, aquatic Diptera, mainly Chironomidae(Armitage et al., 1995) and Ceratopogonidae (Vaug -han, 1997). Pipistrellus pipistrellus expresses verysimilar prey preferences to M. daubentonii, feedingmainly on aquatic Diptera (Zeale et al., 2011), againsuggesting opportunistic behaviour rather than prof-itable prey selection (Swift et al., 1985).

Conclusions

Considering that bats avoided rapid and clutteredwater sections, where prey activity and availabilitywere similar to smooth water sections, it can be con-cluded that habitat preferences of M. daubentoniiand P. pipistrellus are not due entirely to abundanceand distribution of aerial prey, as predicted origi-nally. Instead, avoidance of such areas may be dueto echolocation and energetic constraints restrictingefficient foraging capacity, and from M. dauben-tonii’s ability to trawl/gaff prey from the water sur-face. In turn, aerial insect prey distribution may notbe dependent on trees for shelter from wind. Pro -portion of undercut banks may well provide suitablehabitat refuge for insects and act as a wind shelter,as may boundary layer effects present above thewater surface.

From a conservation standpoint, maintenance ofnaturally meandering watercourses with smoothwater sections alongside riparian vegetation is likelyto promote prime habitat for both M. daubentoniiand P. pipistrellus. Conservation strategies that aimto maintain natural processes of erosion and deposi-tion in naturally dynamic riverine habitats are alsobeneficial to a number of other mammal and birdspecies (Florsheim et al., 2008).


ACKNOWLEDGEMENTS

We would like to extend our gratitude to Dr. Jorge LeonelLeón-Cortés formerly of the University of Leeds for substantialassistance in the field and aid with insect identification. Thanksto Dr. Ruth Warren, also formerly of the University of Leeds forher support. Thank you to Peter Katic of the National Trust forland-access permission and for free accommodation each sum-mer in Wharfedale. Thank you to Dr. Ingo Rieger, Prof. DavidJacobs, and Phil Richardson for their advice and provision of lit-erature, to Dr. Melanie Orr (OSC New Zealand) for initial im-provements to the first draft, and to Edward Lavallin (OSC) forassistance with the GIS map. The manuscript was improved byhelpful comments from two anonymous referees. Work wasfunded by a Natural Environment Research Council (NERC)Studentship with extra support from the National Trust andCountryside Council for Wales (CCW).

LITERATURE CITED

ABBOTT, I. M., D. P. SLEEMAN, and S. HARRISON. 2009. Bat ac-tivity affected by sewage effluent in Irish rivers. BiologicalConservation, 142: 2904–2914.

ADAMS, A. M., M. K. JANTZEN, R. M. HAMILTON, and M. B.FENTON. 2012. Do you hear what I hear? Implications of de-tector selection for acoustic monitoring of bats. Methods inEcology and Evolution, 3: 992–998.

ALDRIDGE, H. D. J. N. 1988. Flight kinematics and energetics in the little brown bat, Myotis lucifugus, with reference tothe influence of ground effect. Journal of Zoology (Lon -don), 216: 507–517.

ALDRIDGE, H. D. J. N., and I. L. RAUTENBACH. 1987. Morphol -ogy, echolocation and resource partitioning in insectivorousbats. Journal of Animal Ecology, 56: 763–778.

ANDERSEN, T. 1979. Some caddis flies (Trichoptera) in WesternNorway, and their arrival pattern in light traps. Fauna Nor -vegica, 26B: 12–17.

ANTHONY, E. L., and T. H. KUNZ. 1977. Feeding strategies of thelittle brown bat, Myotis lucifugus, in southern New Hamp -shire. Ecology, 58: 775–786.

ARBUTHNOTT, D., and R. M. BRIGHAM. 2007. The influence of a local temperature inversion on the foraging behaviour ofbig brown bats, Eptesicus fuscus. Acta Chiropterologica, 9:193–201.

ARMITAGE, P., P. S. CRANSTON, and L. C. V. PINDER. 1995. TheChironomidae: the biology and ecology of non-bitingmidges. Chapman & Hall, London, 572 pp.

ÁVILA-CABADILLA, L. D., G. A. SANCHEZ-AZOFEIFA, K. E. STO -NER, M. Y. ALVAREZ-AÑORVE, M., QUESADA, and C. A.PO R TILLO-QUINTERO. 2012. Local and landscape factors de-termining occurrence of phyllostomid bats in tropical se con -dary forests. PLoS ONE, 7: e35228.

BARCLAY, R. M. R. 1991. Population structure of temperate zoneinsectivorous bats in relation to foraging behavior and en-ergy demand. Journal of Animal Ecology, 60: 165–178.

BATSCHELET, E. 1981. Circular statistics in biology. AcademicPress, London, 371 pp.

BELLAMY, C., and J. D. ALTRING HAM. 2012. Mapping bat forag-ing habitat suitability in the Yorkshire Dales National Park:Habitat Suitability Modelling report. University of Leeds,Leeds, UK, 20 pp.

BELLAMY, C., A. TORSNEY, E. BROWN, A. GLOVER, and

J. AL TRINGHAM. 2012. Bat foraging habitat suitability mapsfor the Yorkshire Dales National Park. Habitat SuitabilityModelling report, Part II December 2012. University ofLeeds, Leeds, UK, 31 pp.

BELWOOD, J. J., and M. B. FENTON. 1976. Variation in the diet ofMyotis lucifugus (Chiroptera: Vespertilionidae). CanadianJournal of Zoology, 15: 1674–1678.

BISCARDI, S., D. RUSSO, V. CASCIANI, D. CESARINI, M. MEI, andL. BOITANI. 2007. Foraging requirements of the endanger-ed long-fingered bat: the influence of micro-habitat struc-ture, water quality and prey type. Journal of Zoology (Lon -don), 273: 372–381.

BOONMAN, A. M., M. BOONMAN, F. BRETSCHNEIDER, and W. A.VAN DE GRIND. 1998. Prey detection in trawling insectivo-rous bats: duckweed affects hunting behaviour in Dau -benton’s bat, Myotis daubentonii. Behavioral Ecology andSociobiology, 44: 99–107.

BOUGHEY, K. L., I. R. LAKE, K. A. HAYSOM, and P. M. DOL-MAN. 2011a. Effects of landscape-scale broadleaved wood-land configuration and extent on roost location for six bat species across the UK. Biological Conservation, 144:2300–2310.

BOUGHEY, K. L., I. R. LAKE, K. A. HAYSOM, and P. M. DOLMAN.2011b. Improving the biodiversity benefits of hedgerows:How physical characteristics and the proximity of foraginghabitat affect the use of linear features by bats. BiologicalConservation, 144: 1790–1798.

BRADLEY, C. 1985. Sites of special scientific interest: riverWharfe, North Yorkshire in N. England. Section 23 of theWild life and Countryside Act, 1.

BRIDCUT, E. E. 2000. A study of terrestrial and aerial macroin-vertebrates on river banks and their contribution to driftingfauna and salmonid diets in a Scottish catchment. Hydro -biologia, 427: 83–100.

BRIGHAM, R. M., R. L. FRANCIS, and S. HAMDORF. 1997a. Mi -cro habitat use by two species of Nyctophilus bats: a test ofecomorphology theory. Australian Journal of Zoology, 45:553–560.

BRIGHAM, R. M., S. D. GRINDAL, M. C. FIRMAN, and J. L. MO -RIS SETTE. 1997b. The influence of structural clutter on activ-ity patterns of insectivorous bats. Canadian Journal of Zool -ogy, 75: 131–136.

BURLES, D. W., R. M. BRIGHAM, R. A. RING, and T. E. REIM -CHEN. 2009. Influence of weather on two insectivorous batsin a temperate Pacific Northwest rainforest. Canadian Jour -nal of Zoology, 87: 132–138.

CHILDS, J., and T. ALDHOUS. 1995. Chemi-luminescent markingof Daubenton’s bats at Stockgrove Country Park, Bed ford -shire. Bat News, 36: 3–4.

CHINERY, M. 1993. Insects of Britain and Northern Europe. Har -p er Collins Publishers, London, 320 pp.

CIECHANOWSKI, M., T. ZAJĄC, A. BIŁAS, and R. DUNAJSKI. 2007.Spatiotemporal variation in activity of bat species differingin hunting tactics: effects of weather, moonlight, food abun-dance, and structural clutter. Canadian Journal of Zoology,85: 1249–1263.

CLARE, E., B. BARBER, B. SWEENEY, P. HEBERT, and M. B. FEN TON. 2011. Eating local: influences of habitat on the dietof little brown bats (Myotis lucifugus). Molecular Ecology,20: 1772–1780.

CRYAN, P. M., C. A. STRICKER, and M. B. WUNDER. 2012. Evi -dence of cryptic individual specialization in an opportunis-tic insectivorous bat. Journal of Mammalogy, 93: 381–389.


CUCCO, M., and G. MALACARNE. 1996. Reproduction of the pal-lid swift (Apus pallidus) in relation to weather and aerial insect abundance. Italian Journal of Zoology, 63: 247–253.

DANARD, M. 1977. A simple model for mesoscale effects of to-pography on surface winds. Monthly Weather Review, 105:572–581.

DAVIDSON-WATTS, I., S. WALLS, and G. JONES. 2006. Differ entialhabitat selection by Pipistrellus pipistrellus and Pipistrelluspygmaeus identifies distinct conservation needs for crypticspecies of echolocating bats. Biological Conser v a tion, 133:118–127.

DAVIES-COLLEY, R. J., G. W. PAYNE, and M. VAN ELSWIJK. 2000.Microclimate gradients across a forest edge. New ZealandJournal of Ecology, 24: 111–121.

DE JONG, J., and AHLÉN, I. 1991. Factors affecting the distribu-tion pattern of bats in Uppland central Sweden. HolarcticEcology, 14: 92–96.

DOWNS, N. C., and P. A. RACEY. 2006. The use by bats of habitatfeatures in mixed farmland in Scotland. Acta Chiropterolo -gica, 8: 169–185.

DRURY, R. L., and F. GEISER. 2014. Activity patterns and roost-ing of the eastern blossom-bat (Syconycteris australis). Aus -tralian Mammalogy, 36: 29–34.

DUNN, J. C., and D. A. WATERS. 2012. Altitudinal effects on hab -itat selection in two sympatric pipistrelle species. Mam -malia, 76: 427–434.

EISENTRAUT, M. 1952. Beobachtungen über Jagtroute und Flug -beginn. Bonner Zoologische Beitrage, 3: 200–211.

EKMAN, M., and J. DE JONG. 1996. Local patterns of distributionand resource utlization of four bat species (Myotis brandti,Eptesicus nilssoni, Plecotus auritius and Pipistrellus pipi -strellus) in patchy and continuous environments. Journal ofZoology (London), 238: 571–580.

ERICKSON, J. L., and S. D. WEST. 2002. The influence of region -al climate and nightly weather conditions on activity pat -terns of insectivorous bats. Acta Chiropterologica, 4: 17–24.

FELLERS, G. M., and E. D. PIERSON. 2002. Habitat use and for-aging behavior of Townsend’s big-eared bat (Corynorhinustownsendii) in coastal California. Journal of Mammalogy,83: 167–177.

FENTON, M. B., H. G. MERRIAM, and G. L. HOLROYD. 1983. Batsof Kootenay, glacier, and Mount Revelstoke National Parks in Canada — identification by echolocation calls, dis-tribution, and biology. Canadian Journal of Zoology, 61:2503–2508.

FLORSHEIM, J. L., J. F. MOUNT, and A. CHIN. 2008. Bank erosionas a desirable attribute of rivers. Bioscience, 58: 519–529.

FOWLER, J., and L. COHEN. 1990. Practical statistics for field biology. John Wiley & Sons, Chichester, 227 pp.

FUENTES-MONTEMAYOR, E., D. GOULSON, L. CAVIN, J. M. WAL -LACE, and K. J. PARK. 2013. Fragmented woodlands in agri-cultural landscapes: The influence of woodland characterand landscape context on bats and their insect prey. Agricul -ture, Ecosystems & Environment, 172: 6–15.

FUKUI, D., M. MURAKAMI, S. NAKANO, and T. AOI. 2006. Effectof emergent aquatic insects on bat foraging in a riparian for-est. Journal of Animal Ecology, 75: 1252–1258.

FURMANKIEWICZ, J., and M. KUCHARSKA. 2009. Migration ofbats along a large river valley in southwestern Poland. Jour -nal of Mammalogy, 90: 1310–1317.

GAISLER, J., J. ZUKAL, Z. ŘEHÁK, and M. HOMOLKA. 1998. Hab -itat preference and flight activity of bats in a city. Journal ofZoology (London), 244: 439–445.

GILLAM, E. H., and G. F. MCCRACKEN. 2007. Variability in theecholocation of Tadarida brasiliensis: effects of geographyand local acoustic environment. Animal Behaviour, 74:277–286.

GONSALVES, L., B. BICKNELL, B. LAW, C. WEBB, and V. MONA -MY. 2013. Mosquito consumption by insectivorous bats:does size matter? PLoS ONE, 8: e77183.

GRAY, L. J. 1993. Response of insectivorous birds to emergingaquatic insects in riparian habitats of a tallgrass prairiestream. American Midland Naturalist, 129: 288–300.

GRINDAL, S. D., and R. M. BRIGHAM. 1999. Impacts of forestharvesting on habitat use by foraging insectivorous bats atdifferent spatial scales. Ecoscience, 6: 25–34.

GROOTAERT, P., M. POLLET, W. DEKONINCK, and C. VAN ACHTER -BERG. 2010. Sampling insects: general techniques, strategiesand remarks. Pp. 377–399, in Manual on field recordingtechniques and protocols for All Taxa Biodiversity In -ventories (J. EYMANN, J. DEGREEF, H. HÄUSER, J. C. MONJE,Y. SAMYN, and D. V. SPIEGEL, eds.). Belgian National FocalPoint to the Global Tax onomy Initiative, Brussels, Volume8, Part 2, 653 pp.

HAMILTON, I. M., and R. M. R. BARCLAY. 1998. Ontogenetic in-fluences on foraging and mass accumulation by big brownbats (Eptesicus fuscus). Journal of Animal Ecology, 67:930–940.

HAYES, J. P. 1997. Temporal variation in activity of bats and thedesign of echolocation-monitoring studies. Journal of Mam -malogy, 78: 514–524.

HOLLOWAY, G. L., and R. M. R. BARCLAY. 2000. Importance ofprarie riparian zones to bats in southeastern Alberta. Eco -science, 7: 115–122.

HURLBERT, S. H. 1984. Pseudoreplication and the design of ecol ogical field experiments. Ecological Monographs, 54:187–211.

JAYANTHI, K., and A. VERGHESE. 2013. Phototactic response ofsapota seed borer, Trymalitis margarias Meyrick, a clue forintegrated pest management. Pest Management in Horti -cultural Ecosystems, 15: 68–69.

JOHNSON, C. G. 1969. Migration and dispersal of insects byflight. Methuen, London, xxii + 766 pp.

JOHNSON, J. B., W. M. FORD, J. W. EDWARDS, and M. A. MENZEL.2010. Bat community structure within riparian areas ofnorthwestern Georgia, USA. Folia Zoologica, 59: 192–202.

JONES, G., and J. M. V. RAYNER. 1988. Flight performance, for-aging tactics and echolocation in free-living Daubenton’sbats Myotis daubentoni (Chiroptera: vespertilionidae). Jour -nal of Zoology (London), 215: 113–132.

JONES, G., and J. M. V. RAYNER. 1991. Flight performance, for-aging tactics and echolocation in the trawling insectivorousbat Myotis adversus (Chiroptera: Vespertilionidae). Journalof Zoology (London), 225: 393–412.

JONES, G., N. VAUGHAN, and S. PARSONS, 2000. Acoustic identi-fication of bats from directly sampled and time expand-ed recordings of vocalizations. Acta Chiropterologica, 2:155–170.

JÓNSSEN, E. 1987. Flight periods of aquatic insects at LakeEsrom, Denmark. Archiv für Hydrobiologie, 110: 259–274.

KALKO, E. K. V., and H.-U. SCHNITZLER. 1989. The echolo-ca tion and hunting behaviour of Daubenton’s bat, Myotisdaubentoni. Behavioral Ecology and Sociobiology, 24:225–238.

KALKO, E. K. V., and H.-U. SCHNITZLER. 1993. Plasticity inecho location signals of European pipistrelle bats in search


flight — implications for habitat use and prey detection.Behavioral Ecology and Sociobiology, 33: 415–428.

KAPFER, G., T. RIGOT, L. HOLSBEEK, and S. ARON. 2008. Roostand hunting site fidelity of female and juvenile Daubenton’sbat Myotis daubentonii (Kuhl, 1817) (Chiroptera: Vesper ti -lio nidae). Mammalian Biology, 73: 267–275.

KARLSSON, B.-L., J. EKLÖF, and J. RYDELL. 2002. No lunar pho-bia in swarming insectivorous bats (family Vespertilioni -dae). Journal of Zoology (London), 256: 473–477.

KNIOWSKI, A. B., and S. D. GEHRT. 2014. Home range and habi-tat selection of the Indiana bat in an agricultural landscape.Journal of Wildlife Management, 78: 503–512.

KOSELJ, K., H.-U. SCHNITZLER, and B. M. SIEMERS. 2011.Horse shoe bats make adaptive prey-selection decisions, in-formed by echo cues. Proceedings of the Royal Society,278B: 3034–3041.

KOVATS, Z. K., J. J. H. CIBOROWSKI, and L. D. CORKUM. 1996.Inland dispersal of adult aquatic insects. Freshwater Bi ol -ogy, 36: 265–276.

KUSCH, J., and A. SCHMITZ. 2013. Environmental factors affect-ing the differential use of foraging habitat by three sym-patric species of Pipistrellus. Acta Chiropterologica, 15:57–67.

LANG, A. B., E. K. V. KALKO, H. RÖMER, C. BOCKHOLDT, and D.K. DECHMANN. 2006. Activity levels of bats and katydids inrelation to the lunar cycle. Oecologia, 146: 659–666.

LEVIN, E., Y. YOM-TOV, and A. BARNEA. 2009. Frequent summernuptial flights of ants provide a primary food source forbats. Naturwissenschaften, 96: 477–483.

LEWIS, T. 1967. The horizontal and vertical distribution of flyinginsects near artificial windbreaks. Annals of Applied Biol -ogy, 60: 23–31.

LEWIS, T. 1969. The distribution of flying insects near a hedge -row. Journal of Applied Ecology, 6: 443–452.

LEWIS, T., and G. C. DIBLEY. 1970. Air movement near wind-breaks and a hypothesis of the mechanism of the accumula-tion of air-borne insects. Annals of Applied Biology, 66:477–484.

LEWIS, T., and J. W. STEPHENSON. 1966. The permeability of artificial windbreaks and the distribution of flying insects inthe leeward sheltered zone. Annals of Applied Biology, 58:355–363.

LINKE, S., E. TURAK, and J. NEL. 2011. Freshwater conservationplanning: the case for systematic approaches. FreshwaterBiology, 56: 6–20.

LLOYD, A., B. LAW, and R. GOLDINGAY. 2006. Bat activity on ri-parian zones and upper slopes in Australian timber produc-tion forests and the effectiveness of riparian buffers. Biol -ogical Conservation, 129: 207–220.

MACKEY, R. L., and BARCLAY, M. R. 1989. The influence ofphysical clutter and noise on the activity of bats over water.Canadian Journal of Zoology, 67: 1167–1170.

MANEL, S., BUCKTON, S. T., and ORMEROD, S. J. 2000. Testinglarge-scale hypothesis using surveys: the effects of land useon the habitats, invertebrates and birds of the Himalayanrivers. Journal of Applied Ecology, 37: 756–770.

MARSDEN, H. 1999. Whitaker’s almanack. J. Whitaker & SonsLtd., London, 1280 pp.

MELEASON, M. A., and J. M. QUINN. 2004. Influence of riparianbuffer width on air temperature at Whangapoua Forest,Coro mandel Peninsula, New Zealand. Forest Ecology andManagement, 191: 365–371.

MOORE, R. D., D. SPITTLEHOUSE, and A. STORY. 2005. Riparian

microclimate and stream temperature response to forest harvesting: a review. Journal of the American Water Re -sources Association, 41: 813.

MORRIS, A. D., D. A. MILLER, and M. C. KALCOUNIS-RUEPPELL.2010. Use of forest edges by bats in a managed pine forestlandscape. Journal of Wildlife Management, 74: 26–34.

MYERS, M. J., and V. H. RESH. 2000. Undercut banks: a habitatfor more than just trout. Transactions of the AmericanFisher ies Society, 129: 594–597.

NANSON, G. C., and E. J. HICKIN. 1986. A statistical analysis ofbank erosion and channel migration in western Canada.Geo logical Society of America Bulletin, 97: 497–504.

NEGRAEFF, O. E., and R. M. BRIGHAM. 1995. The influence ofmoonlight on the activity of little brown bats (Myotis lucifu-gus). Zeitschrift für Säugetierkunde, 60: 330–336.

NELSON, J. M. 1965. A seasonal study of aerial insects close toa moorland stream. Journal of Animal Ecology, 34: 573–579.

NEWSON, M., and J. WARNER. 2007. Contrasting UK experi-ences with participatory approaches to integrated river basinmanagement. Pp. 69–94, in Multi-stakeholder platforms forintegrated water management (J. WARNER, ed.). AshgatePublishing Ltd., Aldershot, xvi + 281 pp.

NICHOLLS, B., and P. RACEY. 2006a. Contrasting home-rangesize and spatial partitioning in cryptic and sympatric pip-istrelle bats. Behavioral Ecology and Sociobiology, 61:131–142.

NICHOLLS, B., and P. A. RACEY. 2006b. Habitat selection as a mechanism of resource partitioning in two cryptic bat spe cies Pipistrellus pipistrellus and Pipistrellus pygmaeus.Eco graphy, 29: 697–708.

NORBERG, U. M., and J. M. V. RAYNER. 1987. Ecological mor-phology and flight in bats (Mammalia: Chiroptera): wingadaptations, flight performance, foraging strategy and echo -location. Philosophical Transactions of the Royal SocietyLondon, B316: 335–427.

NYHOLM, E. S. 1965. The ecology of Myotis mystacinus (Leisl.)and M. daubentoni (Leisl.) (Chiroptera). Annales ZoologiciFennici, 2: 77–123.

PENG, R. K., C. R. FLETCHER, and S. L. SUTTON. 1992a. Effectof microclimate on flying dipterans. International Journal ofBiometeorology, 36: 69–76.

PENG, R. K., S. L. SUTTON, and C. R. FLETCHER. 1992b. Spatialand temporal distribution patterns of flying Diptera. Journalof Zoology (London), 228: 329–340.

POLASKY, S. 2008. Why conservation planning needs socioeco-nomic data. Proceedings of the National Academy of Sci -ences of the USA, 105: 6505–6506.

RACEY, P. A., and S. M. SWIFT. 1985. Feeding ecology of Pi-pi strel lus pipistrellus (Chiroptera, Vespertilionidae) duringpregnancy and lactation. 1. Foraging behaviour. Journal ofAnimal Ecology, 54: 205–215.

RACEY, P. R., S. SWIFT, J. RYDELL, and L. BRODIE. 1998. Batsand insects over two Scottish rivers with contrasting nitratestatus. Animal Conservation, 1: 195–202.

RACHWALD, A., P. BORATYŃSKI , and W. K. NOWAKOWSKI. 2001.Species composition and activity of bats flying over riversin the Białowieża Primeval Forest. Acta Theriologica, 46:235–242.

RAVEN, P. J., N. T. H. HOLMES, F. H. DAWSON, P. J. A. FOX, M. EVERARD, I. R. FOZZARD, and K. J. ROUEN. 1998. Riverhabitat quality: the physical character of rivers and streamsin the UK and Isle of Man. Environment Agency SEPA/Environment and Heritage Service, Bristol, 86 pp.


RIEGER, I. 1996. Activität von Wasserfledermäusen, Myotis dau - ben toni, über dem Rhein. Mitteilungen der Naturfor schen -den Gesellschaft Schaffhausen, 41: 27–58.

RIEGER, I., and A. ALDER. 1993. Weitere Beobachtungen anWasser fledermäuse, Myotis daubentoni auf Flugstrassen.Mitteil ungen der Naturforschenden Gesellschaft Schaff hau -sen, 38: 1–34.

RYDELL, J. 1989. Feeding activity of the northern bat Eptesicusnilssonii during pregnancy and lactation. Oecologica, 80:562–565.

RYDELL, J., A. ENTWISTLE, and P. A. RACEY. 1996. Timing of foraging flights of three species of bats in relation to insectactivity and predation risk. Oikos, 76: 243–252.

RYDELL, J., L. A. MILLER, and M. E. JENSEN. 1999. Echolo cationconstraints of Daubenton’s bat foraging over water. Func -tional Ecology, 13: 247–255.

SCOTT, S., G. MCLAREN, G. JONES, and S. HARRIS. 2010. Theimpact of riparian habitat quality on the foraging and activ-ity of pipistrelle bats (Pipistrellus spp.). Journal of Zoology(London), 280: 371–378.

SENIOR, P., R. K. BUTLIN, and J. D. ALTRINGHAM. 2005. Sex andsegregation in temperate bats. Proceedings of the RoyalSociety, 272B: 2467–2473.

SERVICE, M. W. 1973. Spatial and temporal distribution of aerialpopulations of woodland tipulids (Diptera). Journal of Ani -mal Ecology, 42: 295–303.

SOLEM, J. O. 1978. Swarming and habitat segregation in thefamily Leptoceridae (Trichoptera). Norwegian Journal ofEntomology, 25: 145–148.

SPEAKMAN, J. R. 1991. Why do insectivorous bats in Britain not fly in daylight more frequently? Functional Ecology, 5:518–524.

SWIFT, S. M., P. A. RACEY, and M. I. AVERY. 1985. Feeding ecol-ogy of Pipistrellus pipistrellus (Chiroptera: Vespertilio ni -dae) during pregnancy and lactation. II. Diet. Journal ofZool ogy (London), 54: 217–225.

TAAKE, K. H. 1992. Strategien der Resourcennutzung an wald -ge wässern jagender Fledermäuse (Chiroptera: Vesper ti lio -nidae). Myotis, 30: 7–74.

TAYLOR, L. R. 1958. Aphid dispersal and diurnal periodicity.Pro ceedings of the Linnean Society of London, 169: 67–73.

TAYLOR, L. R. 1960. The distribution of insects at low level inthe air. Journal of Animal Ecology, 29: 45–63.

TAYLOR, L. R. 1963. Analysis of the effect of temperature on in-sects in flight. Journal of Animal Ecology, 32: 99–117.

TAYLOR, L. R. 1974. Insect migration, flight periodicity and theboundary layer. Journal of Animal Ecology, 43: 225–238.

THOMAS, A. J., and D. S. JACOBS. 2013. Factors influencing theemergence times of sympatric insectivorous bat species.Acta Chiropterologica, 15: 121–132.

TODD, V. L. G., and D. A. WATERS. 2007. Strategy switching inthe gaffing bat. Journal of Zoology (London), 273: 106–113.

UNWIN, D. M. 1981. A key to the families of British Diptera.AIDGAP Field Studies, 5: 513–553.

VAUGHAN, N. 1997. The diets of British bats (Chiroptera). Mam -mal Review, 27: 77–94.

VERBOOM, B., and H. HUITEMA. 1997. The importance of linear landscape elements for the pipistrelle Pipistrellus

pi pistrellus and the serotine bat Eptesicus serotinus. Land -scape Ecol ogy, 12: 117–125.

VERBOOM, B., and K. SPOELSTRA. 1999. Effects of food abun-dance and wind on the use of tree lines by an insectivorousbat, Pipistrellus pipistrellus. Canadian Journal of Zoology,77: 1393–1401.

VESTERINEN, E. J., T. LILLEY, V. N. LAINE, and N. WAHLBERG.2013. Next generation sequencing of fecal DNA reveals thedietary diversity of the widespread insectivorous predatorDaubenton’s bat (Myotis daubentonii) in Southwestern Fin -land. PLoS ONE, 8: e82168.

VINDIGNI, M. A., A. D. MORRIS, D. A. MILLER, and M. C. KAL -COUNIS-RUEPPELL. 2009. Use of modified water sources bybats in a managed pine landscape. Forest Ecology and Mana -gement, 258: 2056–2061.

WALSH, A. L., and HARRIS, S. 1996a. Factors determining theabun dance of vespertilionid bats in Britain: geographical,land class and local habitat relationships. Journal of AppliedEcology, 33: 519–529.

WALSH, A. L., and HARRIS, S. 1996b. Foraging habitat prefer-ences of vespertilionid bats in Britain. Journal of AppliedEcology, 33: 508–518.

WALSH, A. L., S. HARRIS, and A. M. HUTSON. 1995. Abundanceand habitat selection of foraging vespertilionid bats in Brit -ain: a landscape-scale approach. Symposia of the Zool ogicalSociety of London, 67: 325–344.

WALTERS, C. L., R. FREEMAN, A, COLLEN, C, DIETZ, M. B. FEN -TON, G. JONES, M. K. OBRIST, S. PUECHMAILLE, J. SATTLER,B. M. SIEMERS, et al. 2012. A continental-scale tool foracoustic identification of European bats Journal of AppliedEcology, 49: 1064–1074.

WARINGER, J. A. 1991. Phenology and the influence of meteoro-logical parameters on the catching success of light-trappingfor Trichoptera. Freshwater Biology, 25: 307–319.

WARREN, R., D. WATERS, J. D. ALTRINGHAM, and D. J. BULLOCK.2000. The distribution of Daubenton’s bats (Myotis dauben-tonii) and pipistrelle bats (Pipistrellus pipistrellus) (Vesper ti -lionidae) in relation to small-scale variation in riverine hab-itat. Biological Conservation, 92: 85–91.

WATERS, D. A., and A. L. WALSH. 1994. The influence of bat de-tector brand on the quantitative estimation of bat activity.Bio acoustics, 5: 205–221.

YELA, J. L., and M. HOLYOAK. 1997. Effects of moonlight andmeteorological factors on light and bait trap catches of noc-tuid moths (Lepidoptera: Noctuidae). Environmental Ento -mol ogy, 26: 1283–1290.

ZEALE, M. R., R. K. BUTLIN, G. L. BARKER, D. C. LEES, andG. JONES. 2011. Taxon-specific PCR for DNA barcodingarth ropod prey in bat faeces. Molecular Ecology Resources,11: 236–244.

ZEALE, M. R., I. DAVIDSON-WATTS, and G. JONES. 2012. Homerange use and habitat selection by barbastelle bats (Barba -stella barbastellus): implications for conservation. Journalof Mammalogy, 93: 1110–1118.

ZSEBOK, S., F. KROLL, M. HEINRICH, D. GENZEL, B. M. SIEMERS,and L. WIEGREBE. 2013. Trawling bats exploit an echo-acoustic ground effect. Frontiers in Physiology, 4: 65.


Received 13 November 2016, accepted 25 May 2017

Associate Editor: Wiesław Bogdanowicz

Documents

Small scale habitat preferences of Myotis daubentoniiPipistrellus … · The aim of this study was to investigate habitat preferences of both M. daubentoniiand P. pipistrel-lus, and