Biology subjects

Tylianakis, J.

Publications and source records attributed to Tylianakis, J..

3 recordsLinked to original sources

bioRxiv2020

Predicting direct and indirect non-target impacts of biocontrol agents using machine-learning approaches

Biological pest control (i.e. biocontrol) agents can have direct and indirect non-target impacts, and predicting these effects (especially indirect impacts) remains a central challenge in biocontrol risk assessment. The analysis of ecological networks offers a promising approach to understanding the community-wide impacts of biocontrol agents (via direct and indirect interactions). Independently, species traits and phylogenies have been shown to successfully predict species interactions and network structure (alleviating the need to collect quantitative interaction data), but whether these approaches can be combined to predict indirect impacts of natural enemies remains untested. Whether predictions of interactions (i.e. direct effects) can be made equally well for generalists vs. specialists and across different habitat types is also untested for consumer-prey interactions, though previous work on mutualist networks suggests that interactions among generalists may be more difficult to predict. Here, we used two machine learning techniques (random forest and k-nearest neighbour; KNN) to test whether we could accurately predict empirically-observed quantitative host-parasitoid networks using trait, abundance, and phylogenetic information. Then, we tested whether the accuracy of machine-learning-predicted interactions depended on the generality of the interacting partners or on the source (habitat type) of the training data. Finally, we used these predicted networks to generate predictions of indirect effects via shared natural enemies (i.e. apparent competition), and tested these predictions against empirically observed indirect effects between hosts. We found that random-forest models predicted host-parasitoid pairwise interactions (which could be used to predict attack of non-target host species) more successfully than KNN, and this predictive ability depended on the generality of the interacting partners, but not the source (habitat type) of data used to train the models. Further, although our machine-learning informed methods could significantly predict indirect effects, the explanatory power for both direct and indirect interactions was reasonably low. Combining machine-learning and network approaches provides a starting point for reducing risk in biocontrol.

ecology↗

bioRxiv2019

Land-use and related pressures have reduced biotic integrity more on islands than on mainlands

Tracking progress towards biodiversity targets requires indicators that are sensitive to changes at policy-relevant scales, can easily be aggregated to any spatial scale and are simple to understand. The Biodiversity Intactness Index (BII), which estimates the average abundance of a diverse set of organisms in a given area relative to their reference populations, was proposed in 2005 in response to this need. A new implementation of BII was developed as part of the PREDICTS project in 2016 and has been adopted by GEO BON, IPBES and CBD. The previous global models for BII estimation could not account for pressures having different effects in different settings. Islands are a setting of particular interest: many are home to a disproportionate number of endemic species; oceanic islands may have relatively low overall species diversity because of their isolation; and the pattern and timing of human pressures can be very different from that seen on mainlands. Here, we test whether biotic integrity - as estimated by BII - has decreased more severely on islands than mainlands. We update methods previously used to estimate BII globally (Newbold et al., 2016) to allow pressure effects to differ between islands and mainlands, while also implementing some other recent improvements in modelling. We estimate BII for islands and mainlands by combining global models of how two aspects of biodiversity - overall abundance, and compositional similarity to minimally-impacted sites - have been affected by human pressures. We use these models to project high-resolution ([~]1km2) global maps of BII for the year 2005. We calculate average BII for island and mainland biomes, countries, IPBES regions and biodiversity hotspots; and repeat our analyses using a richness-based version of BII. BII on both islands and mainlands has fallen below the values proposed as safe limits across most regions, biomes and biodiversity hotspots. Our BII estimates are lower than those published in 2016, globally, within all biodiversity hotspots and within most biomes, and show greater spatial heterogeneity; detailed analysis of these differences shows that they arise mostly from a combination of improvements to the modelling framework. Average BII does not strongly differ between islands and mainlands, but richness-based BII has fallen by more on islands. It seems native species are more negatively affected by rising human population density and road development on islands than mainlands, and islands have seen more land conversion. Our results highlight the parlous state of biodiversity native to islands.

ecology↗

bioRxiv2018

Worldwide impacts of past and projected future land-use change on local species richness and the Biodiversity Intactness Index

Although people have modified the world around us throughout human history, the Great Acceleration has seen drivers such as land conversion, exploitation of natural populations, species introductions, pollution and human-induced climate change placing biodiversity under increasing pressure. In this paper we examine 1) how terrestrial species communities have been impacted over the last thousand years of human development and 2) how plausible futures defined by alternative socio-economic scenarios are expected to impact species communities in the future. We use the PREDICTS (Projecting Responses of Ecological Diversity In Changing Terrestrial Systems) database to model impacts of land-use change and human population on local species richness, community abundance, and biodiversity intactness using a mixed-effects modelling structure. Historical impacts are inferred through projection of model results onto maps of historical land use, provided by the land-use harmonization project, and gridded human population density (HYDE 3.1). Future impacts are explored using the Shared Socio-economic Pathway (SSP) scenarios. These scenarios detail five plausible global futures based upon socio-economic factors such as wealth, population, education, technology, and reliance on fossil fuels, and can be combined with Representative Concentration Pathway (RCP) scenarios to consider climate mitigation strategies. We project model results onto the gridded outputs of six SSP/RCP scenario combinations: SSP1/RCP2.6, SSP2/RCP4.5, SSP3/RCP7.0, SSP4/RCP3.4, SSP4/RCP6.0, and SSP5/RCP8.5. Historical trend lines show that most losses in local biodiversity are relatively recent, with 75% of all loss in both abundance-based Biodiversity Intactness Index and species richness occurring post-1800. Stark regional differences emerge in all future scenarios, with biodiversity in African regions undergoing greater losses than Oceania, North America and the European regions. Although climate change is expected to have severe detrimental impacts to biodiversity - which are not quantified in these results - it is important to consider how the climate change mitigation itself may also impact biodiversity. Our results suggest that strong climate change mitigation through biofuel production will detrimentally impact biodiversity: SSP4/RCP3.4 (with high biofuel mitigation) is predicted to see two times the decrease in abundance-based biodiversity intactness and three times the decrease in local species richness between 2015-2100 as is predicted for SSP4/RCP6.0 (with lower levels of mitigation). SSP4/RCP3.4 forecasts the greatest impact to average local species richness of all the SSP/RCP combinations with an average loss of 13% of local species richness projected to have occurred by 2100. SSP3/RCP7.0 - a scenario describing a globally segregated, and economically protectionist future with low climate change mitigation - has the worst impacts on abundance-based biodiversity intactness with an average loss of 26% of intactness by 2100. However, a brighter future is possible; SSP1/RCP2.6 describes a more sustainable future, where human populations are provided for without further jeopardising environmental integrity - in this scenario we project that biodiversity will recover somewhat, with gains in biodiversity intactness and species richness in many regions of the world by 2100.

ecology↗