Browsing by Subject "REGRESSION TREES"

Sort by: Order: Results:

Now showing items 1-2 of 2
  • Reside, April E.; VanDerWal, Jeremy; Moilanen, Atte; Graham, Erin M. (2017)
    With the high rate of ecosystem change already occurring and predicted to occur in the coming decades, long-term conservation has to account not only for current biodiversity but also for the biodiversity patterns anticipated for the future. The trade-offs between prioritising future biodiversity at the expense of current priorities must be understood to guide current conservation planning, but have been largely unexplored. To fill this gap, we compared the performance of four conservation planning solutions involving 662 vertebrate species in the Wet Tropics Natural Resource Management Cluster Region in north-eastern Australia. Input species data for the four planning solutions were: 1) current distributions; 2) projected distributions for 2055; 3) projected distributions for 2085; and 4) current, 2055 and 2085 projected distributions, and the connectivity between each of the three time periods for each species. The four planning solutions were remarkably similar (up to 85% overlap), suggesting that modelling for either current or future scenarios is sufficient for conversation planning for this region, with little obvious trade-off. Our analyses also revealed that overall, species with small ranges occurring across steep elevation gradients and at higher elevations were more likely to be better represented in all solutions. Given that species with these characteristics are of high conservation significance, our results provide confidence that conservation planning focused on either current, near-or distant-future biodiversity will account for these species.
  • Salonen, J. Sakari; Korpela, Mikko; Williams, John W.; Luoto, Miska (2019)
    We test several quantitative algorithms as palaeoclimate reconstruction tools for North American and European fossil pollen data, using both classical methods and newer machine-learning approaches based on regression tree ensembles and artificial neural networks. We focus on the reconstruction of secondary climate variables (here, January temperature and annual water balance), as their comparatively small ecological influence compared to the primary variable (July temperature) presents special challenges to palaeo-reconstructions. We test the pollen-climate models using a novel and comprehensive cross-validation approach, running a series of h-block cross-validations using h values of 100-1500 km. Our study illustrates major benefits of this variable h-block cross-validation scheme, as the effect of spatial autocorrelation is minimized, while the cross-validations with increasing h values can reveal instabilities in the calibration model and approximate challenges faced in palaeo-reconstructions with poor modern analogues. We achieve well-performing calibration models for both primary and secondary climate variables, with boosted regression trees providing the overall most robust performance, while the palaeoclimate reconstructions from fossil datasets show major independent features for the primary and secondary variables. Our results suggest that with careful variable selection and consideration of ecological processes, robust reconstruction of both primary and secondary climate variables is possible.