| Popis: |
Global land use is changing at an unprecedented rate and has been identified as a key driver of habitat loss, fragmentation and species decline in the natural environment. Understanding how land use influences spatial patterns in species abundance, and habitat connectivity at a landscape scale is critical for the survival of wildlife populations. The focal species of my research is the rare greater horseshoe bat (Rhinolophus ferrumequinum), which was once widespread across southern England and Wales. However, owing to changes in agricultural land management and the expansion of urban areas, its range has contracted considerably over the last century. Using a series of ecological techniques, including a novel predictive modelling approach, field experiments and social network analysis, this thesis aims to identify which ecological factors affect their activity and movement at a landscape scale. The work also provides conservation practitioners the ability to identify the locations of these impacts, pinch-points, in the wider environment; where strategic planning and mitigation measures can be applied to increase their overall occurrence and abundance in the wider environment. \ud Using a field experiment, I examined how traffic noise can influence the relative activity levels of free-living bats. Overall, I showed that traffic noise can significantly reduce the activity levels of R. ferrumequinum, as well as other bat species, along linear feature. Using a separate field experiment, I determined that the sonic spectrum had a greater negative effect on bat activity than the ultrasonic spectrum. These results therefore suggest that the mode of action is likely to be through general deterrence and avoidance rather than through the masking of echolocation calls. R. ferrumequinum are widely considered to be dependent on linear landscape features such as woodland edges and hedgerows. My research supported this view, and highlighted the particular importance of treelines, which were associated with greater activity than even sympathetically managed hedgerows. However, an important novel finding from my research was that about a third of all activity recorded at paired detectors was derived from the middle of fields. It is therefore important to consider these more open habitats, as well as hedgerows, treelines and woodland edges, when designing and conducting ecological impact assessments for future developments. \ud Bats use the landscape at a large spatial scale, and responses to any particular challenge (such as a new lighting scheme or urban development) are likely to depend on their context within a landscape. For example, the disruption of a commuting route is likely to have a greater impact where only one suitable route exists, compared with scenarios where there are numerous alternatives. To understand better how R. ferrumequinum interacts with the British landscape, I therefore created models of predicted functional connectivity around four maternity roosts using Circuitscape software. Using non-invasive static bat detectors as a method of ground validation, I created robust models predicting R. ferrumequinum movement; which allowed the identification of pinch-points in the landscape, either those areas limiting species movement or highly important for the species conservation. \ud While most research, and the work of ecological practitioners, focuses on the maternity season, hibernation ecology has received much less attention. Yet given that horseshoe bats are known to move roost location frequently during the hibernation period, it is important to understand more about this behaviour. Using social network analysis, I demonstrate that adult males are significantly more central (connected to a higher number of individual bats) in the network during the hibernation period. I found that movements between hibernacula were associated with both age and degree centrality of individual bats, with those more geographically isolated hibernacula playing an important role for the movement of certain individuals at a landscape scale. This highlights that despite low activity in some of these smaller roost locations, they are a conservation priority to decrease the risk of fragmentation and loss of connectivity within the wider landscape.\ud The results of my meta-analysis, which was based on 22 studies, demonstrated the significant negative effects endectocides on Aphodiinae dung beetles. My results suggest that ivermectin has the highest negative effect on the abundance of both adult and large Aphodiinae dung beetles. However, contrasting results were observed for dung beetle occurrence, with adult beetles showing an attraction to dung with endectocides and larvae showing the complete opposite, with poor survival rates and impaired development. Over time this could have significant negative effects on dung beetle populations. \ud The results of this thesis indicate that the landscape-scale conservation of R. ferrumequinum is complex. Considerations need to be given to a suite of factors ranging from the prey items they consume to the physical habitat structures which they utilise. From this research, specific locations and features which have impacts on their movement and activity can be identified, allowing the outputs to be used by decision-makers as a tool to inform local management strategies. The prioritisation of conservation activity for the species can be aided by spatially-explicit models, such as the one I developed using Circuitscape, which bring together multiple input layers to create outputs readily interpretable to practitioners. However, to achieve a successful outcome for this priority species, collaborative efforts from many stakeholders, across boundaries, are required. |
