LiDAR, habitat structure and the ecology of ungulates in a landscape of fear

Karen Lone successfully defended her doctoral thesis, “LIDAR, habitat structure and the ecology of ungulates in a landscape of fear”, on February 13, 2015.

The topic for the trial lecture was “Do females and males perceive the environment differently? Sex-specific predation risk and antipredator strategies in birds and mammals”. We congratulate!

Thesis abstract

The expression ‘landscape of fear’ describes a situation in which prey behavior and space use across the landscape are shaped by spatial and temporal patterns in predation risk. Yet such behavioral and indirect effects of predation have been much less studied than its numerical and direct effects. One of the challenges in doing so is characterizing habitat at the resolution and extent necessary to effectively study this at the landscape level. In this thesis, I investigated the spatial risk patterns created by natural predators and human hunting and how these two predators influence the habitat selection and ecology of forest ungulates (roe deer, red deer, and moose) across several study systems in Norway.

I used airborne LiDAR (light detection and ranging), an emerging technology, to obtain detailed data on habitat structure over large scales. The three-dimensional LiDAR data were used to 1) elucidate how understory vegetation density, canopy cover, and other habitat characteristics modify predator-prey interactions and 2) predict wildlife forage availability. Whereas structural information from LiDAR was sufficient to model predation risk from lynx and hunters well, it needed to be integrated with auxiliary data, particularly plant species information, to predict forage satisfactorily.

Most studies address a situation with only a single prey-predator pair. Yet, with recolonization of large carnivores, ungulates are frequently facing multiple predators with contrasting hunting styles. An example of such an understudied multiple-predator situation is the roe deer in southern Norway facing two predators (lynx and humans). I documented how these predators create contrasting risk patterns due to their different hunting methods, as lynx risk increased and hunting risk decreased with increasing understory cover. I discussed how this could lead to lower non-consumptive effects of predation in such systems compared to systems with a single predator. Against this background, I investigated dynamic responses in space use to temporally structured predation risk. Diurnal and seasonal variation was studied in roe deer, and adaptive responses to the onset of hunting were studied in a heavily hunted, but otherwise predator-free red deer population. Roe deer shifted their habitat selection towards safer areas during periods of high predator activity, but have a weaker response, possibly no response, to lynx risk during winter. The latter is likely due to harsh winter conditions leading to a more severe trade-off between safety and energetic considerations that constrained roe deer habitat selection during this season. In the study on red deer, deer that were shot during the following hunting season differed from survivors in their habitat use. Whereas all males used similar habitat in the days before the hunting season, the onset of hunting induced an immediate switch to habitat with more concealing cover in surviving males, but not in males that were later shot. Responding to the onset of hunting appeared to be adaptive, given that it was linked to increased survival, an important fitness component. A final case study expanded habitat characterization to include forage availability and related moose habitat selection to forage and two LiDAR variables capturing variation in concealment and thermal conditions: understory cover and canopy cover. All three functional gradients were important for moose habitat selection on landscape and home range scales, both during winter and summer. Including habitat structural characteristics directly derived from LiDAR datasets in habitat selection studies can be meaningful and successful, while also requiring less effort than alternative predictive approaches.

This thesis elucidates the spatial patterns and temporal nature of spatial antipredator responses in ungulates, and the importance of trade-offs in that regard. My work demonstrates possible applications of LiDAR-derived data on habitat structure to ecological studies at large scales. It establishes that hunters and carnivores impact Norwegian forest ungulates not only numerically, but also behaviorally by creating a ‘landscape of fear’. The magnitude of indirect effects of human and natural predators on prey populations and the question of whether predation and hunting cause behaviorally mediated trophic cascades deserve further attention.

Supervisors

Professor Leif Egil Loe (main supervisor) (INA, NMBU)
Professor Terje Gobakken (INA, NMBU)
Professor Atle Mysterud (CEES, Dept. of Biosciences, University of Oslo)

Evaluation committee

Associate Professor Kerri Vierling (Fish and Wildlife Resources, University of Idaho)
Assistant Professor Simone Ciuti (Dept. of Biometry and Environmental System Analysis, University of Freiburg)
Professor Geir Sonerud (INA, NMBU)

Reference

  • Lone, K.. ((2015). Lidar, habitat structure and the ecology of ungulates in a landscape of fear.). PhD Thesis.
    [Bibtex]
    @PhdThesis{Lone2015,
    Title = {LiDAR, habitat structure and the ecology of ungulates in a landscape of fear},
    Author = {Lone,K.},
    School = {Norwegian University of Life Sciences},
    Year = {2015},
    Owner = {hanso},
    Timestamp = {2016.03.02},
    Url = {https://static02.nmbu.no/mina/forskning/drgrader/2015-Lone.pdf}
    }