Counting By far the most common and least disruptive way of estimating bat numbers in summer roosts is to count the bats leaving the roost at sunset. The most successful period over which to make counts so that yearly comparisons can be made is when the colony is at its most stable. Maternity colonies continue to grow into early June as pregnant females arrive from pre-maternity roosts. Later in the summer young bats begin flying and counts at this time will include both adults and young. Just prior to birth of the young, a peak in adult female numbers is reached and many or all of the adult females emerge every night. In general, young are born between mid- to late June inmost species, although there is variation from year to year. Counts in late May to mid-June are likely to reflect the most
stable numbers. This is the time that is recommended by the National Bat Monitoring Programme (see Walsh et al., 2001; Appendix When attempting to count a colony for the first time, it is important to establish the number and location of access holes and this may require several people surrounding the roost. Once all access point have been discovered then fewer surveyors maybe required for future counts. Assign observers to a specific exit or field of view because often two exits close together maybe counted simultaneously. Be in position at the roost about 15 minutes before sunset (earlier on overcast evenings) and listen for the sound of bats moving at access holes or for squeaks as bats jostle for position. Poor weather conditions with overcast skies and rain may delay emergence and particularly bad conditions should be avoided in case bats choose to abandon foraging and remain inside the roost. Remain counting until at least 10 minutes after the last bat has emerged. As a general guide, bats may begin emerging from just before sunset onwards. The noctule is an early emerging species, whilst Natterer’s and long-eared bats are late emerging (up to 40 minutes after sunset). Do not shine white lights directly on the roost exit because this can often delay or prevent emergence. Excess noise, particularly ultrasound from keys, coins or nylon jackets, may disturb bats and inhibitor delay emergence. Emergence counts are most effective when departing bats are silhouetted against alight background (normally a clear sky or sometimes alight coloured wall. It is best to observe from the side of the emergence point(s), rather than from in front. Sometimes the structure of the roost and behaviour of the bats means this is not possible and the additional use of a bat detector is always recommended. The behaviour of bats at emergence varies between species. Some, such as noctules, tend to fly off fast and direct to foraging grounds while others, such as the horseshoe and Natterer’s bats, may exit and reenter the roost several times before departing for foraging grounds (called light sampling’ behaviour), which makes counting more difficult. In all cases, a running total of both exits and entries should be recorded so that a final net emergence figure can be calculated. This is essential when several exits are being counted simultaneously, so that bats that emerge from one and subsequently 3.4 LOC ATION OF KEY SITES AND FEEDING AREAS 33 re-enter at another maybe properly accounted for. A small tally counter is useful for clicking up bats as they emerge. Automatic counting equipment or video equipment at roost exits can be a useful substitute for observers, although even the most sophisticated systems have their problems. Most automatic counting systems are based on one or more infrared light beams that are broken as a bat emerges from the roost. A single beam is unable to distinguish between emerging and returning bats but beam breaks should be proportional to the number of bats emerging/returning and so could be calibrated by combining beam counts with observer counts. More sophisticated systems use two sets of beams so that emerging and reentering bats can be counted separately. Both types of system are unable to cope with bats that emerge without breaking the beams or when two bats break the beam simultaneously. Problems also arise when two or more species are using the same roost or if insects fly close to roost access points. When using automatic systems results must be crosschecked with simultaneous visual checks to identify errors or consistent biases. Counts of bats within summer roosts are generally more difficult, cause more disturbance and are less accurate than emergence counts. One or two visits to a breeding roost in a season would probably be acceptable provided that every care was taken to keep the disturbance to the minimum. The exception to this is for counts of nonflying young, which maybe counted on a weekly basis once all the adults have left to forage. This method is mainly applicable to horseshoe bats.