The Western Ghats of India are a globally important biodiversity hotspot, but around 90 % of the land has been converted to agriculture. Little is known about how bats respond to the conversion of native rainforest to different plantation types in the Western Ghats. This thesis examines the response of bats to coffee and tea plantations, and to riparian habitats, in the southern Western Ghats. Most bat assemblages in the tropics have been studied by catching bats, but many studies have shown that catching alone can give biased and incomplete results. In order to use acoustic data as well as catching data in this landscape I made a library of the echolocation calls of fifteen echolocating species in the landscape. Comparisons of the data from each method showed that combining catching and acoustic data gave the most complete picture of the assemblage, but that acoustic data alone detected more species than catching data alone. Acoustic and catching data were used to build habitat suitability models for ten species. Scales of 100 m – 500 m were the most important for predicting bat presence. Several species showed a positive response to habitats containing native trees and habitat richness, and a negative response to tea plantations and distance to water. Coffee plantations did not differ significantly from forest fragments in terms of bat species or abundance, but did differ in species composition. Bat assemblages in coffee plantations were functionally very similar to those in forest fragments. Tea plantations had the lowest bat species richness of all habitats and differed in species composition from all other habitats. Bat assemblages in tea plantations had lower functional richness and specialisation than other habitats, and the bats remaining were open-adapted species. Rivers with riparian corridors did not have significantly greater bat species richness than rivers without corridors, but differed functionally in several ways. Rivers without riparian corridors had reduced functional specialisation and functional divergence. Rivers with riparian corridors supported more forest adapted species than rivers without riparian corridors.

Plantations are widespread throughout temperate regions, and the area of plantation land cover is predicted to get larger in the future. Interest in ensuring sustainable plantation management is also growing, as it is increasingly recognised that productive areas should play a role in biodiversity conservation. Plantation landscapes can comprise the majority of forested cover in some countries, but taxon-specific guidance can be lacking, due to plantations often being under surveyed. Therefore, despite substantial incentives existing to ensure that plantations meet various ecological criteria, plantation managers lack the information necessary to implement effective management plans. Many bat species have undergone widespread declines in recent decades, attributed to habitat loss and fragmentation, particularly of forested habitat. In many temperate countries, historical deforestation has resulted in very low native tree cover, and subsequently, considerable replanting with non-native commercial coniferous plantations has taken place. Species specific habitat surveys have often demonstrated avoidance of conifer plantations by bats, which has been attributed to a lack of roosts and low invertebrate prey abundance. Furthermore, widespread lepidopteran declines have been partly attributed to afforestation with non-native conifer, but moth associations with commercial coniferous plantations are usually only studied for pest species. Bats present a particular challenge in plantation landscapes; tree cover is important to many species to a greater or lesser extent, and in the United Kingdom, destruction of a roost site is illegal, regardless of whether it was deliberate or accidental. However, the extent to which bats associate with non-native commercial plantations is relatively unexplored. This is the first study to explicitly test bat associations with Picea sitchensis plantations (using acoustic detectors, trapping and radio tracking), and shows that, contrary to expectations, they may be an important habitat for breeding populations of Pipistrellus spp., particularly P. pygmaeus. High levels of activity were recorded for both P. pygmaeus and P. pipistrellus, despite little difference in dipteran abundance between different stand types, both species preferentially foraged in felled or less dense stands. This suggests that bats preferentially forage in areas with less acoustic and physical clutter, which will increase foraging efficiency. The impacts of felling in non-native commercial coniferous plantations on foraging activity was tested, for the first time, using a Before – After – Control – Impact experimental design. Bat activity (specifically P. pipistrellus and Nyctalus) increased after felling, particularly in smaller stands. In contrast felling had significant, negative impacts on moth abundance, species richness and diversity, and these effects remain after constraining for functional trait similarity. Reductions in richness and diversity in response to felling were similarly large for both rare and abundant species. Therefore, while bats may benefit from clear fell practices, albeit as long as the size of patches is small, moth populations could benefit from a shift towards other forestry methods, iv such as continuous cover forestry. These results also have implications for the recent, but increasing practice of siting wind turbines in commercial coniferous plantations, as pre-installation preparation involves clearing small patches of forest which may attract foraging bats; post felling monitoring should be carried out to examine potential impacts on bat populations. The presence of broadleaf trees in and around plantations significantly increased moth richness, mostly through increased occurrence of rare species. Broadleaf woodlands (defined as land spanning more than 0.5 ha, with trees higher than 5m and a combined cover of shrubs, bushes and trees above 10%), also had higher functional redundancy than plantation sites. For a diverse moth population to persist in plantation landscapes, preserving remnant patches of broadleaf trees is essential. There was little difference in bat activity between broadleaf woodlands and plantation sites. However, bat abundance, particularly that of reproductively active females, was greater in broadleaf sites compared to plantations. This was particularly true for Myotis and Nyctalus spp., very few of which were trapped in commercial plantations. Therefore, although reproductively active female Myotis bats are present in the surrounding landscape, they do not appear to associate with plantations themselves. This may reflect a lack of roost availability; both P. pygmaeus and P. pipistrellus preferentially form large maternity colonies in buildings, but for Myotis and Nyctalus spp. which roost switch regularly and often use trees, it is unlikely many suitable roosts exist within the plantations themselves. Many substantial P. pygmaeus maternity colonies were identified in and around Galloway forest, with some holding more than 500 individuals. All maternity colonies were in buildings, and most inhabited (and one uninhabited) buildings within the plantation contained a roost. Although females occasionally used old or dead deciduous trees as temporary roosts, there was no evidence of roosting in crop trees such as P. sitchensis. During this study, the Forestry Commission installed 36 bat boxes; within 6 months over 90% had been used, with a number of harems found inside. This fast uptake compared with bat use of boxes in other locations reflects the paucity of appropriate structures for either roost or harem use in commercial plantations. Twelve bats were captured while foraging, tagged with small radio transmitters, and followed for between 2 and 6 nights during 2014 and 2015. All but one tagged female preferentially foraged within the plantation, with individuals selecting equally riparian habitats and felled stands. Tagged females which roosted furthest from the plantation had the largest home ranges; one individual flew nearly 40km each night to reach foraging areas distant from her roost, suggesting that the food availability within the plantation was sufficient to render such a long journey energetically viable. v These results have important implications for bat populations in and around commercial coniferous plantations. Far from being avoided by bats, plantation landscapes may constitute an important habitat type for both P. pygmaeus and P. pipistrellus, likely due to the high abundance of nematoceran diptera in plantation woodlands. Furthermore, plantation forests support a similar richness of moth species to urban and agricultural woodlands, including a number of declining species of special conservation concern. A list of management recommendations to benefit both bat and moth populations in commercial plantations is presented at the end of this thesis.