Designing profitable and climate-smart farms using virtual reality

We have developed a linear logic model that combines carbon datasets with pasture pro- duction data so that the user can assess the interaction between GHG emissions and profitabil- ity under various scenarios. With future developments, a variety of similar science-based data (e.g., spatial layers, lookup tables, linear equations) can be incorporated into the VR applica- tion. For instance, tree planting strategies are expected to affect a range of other farm outcomes that are in addition to the two considered currently (i.e., carbon sequestration and farm profit- ability). Visual assessment of aerial imagery of the farm suggests there is significant overlap between areas that have the lowest estimated production (Fig 6A) and areas where there is evi- dence of soil erosion (Fig 6B).While planting trees in these locations will increase carbon sequestration with the least impact on profitability, mitigating hillslope erosion is also likely [34]. Furthermore, the ability to mitigate the impact of climate change on local surface temper- ature might also be considered in future steps [9, 35–40]. The planting design that maximises both carbon sequestration and profitability (i.e. Fig 6A), suggest a combination of small exotic forest patches distributed across the farm. However, under the current New Zealand ETS system, planted areas that are less than 1 ha are not eligi- ble for carbon credits [41]. In this context, we believe that, in order to incentivise the simulta- neous delivery of multiple ecosystem services by rural landscape and avoid the large-scale

conversion of productive pasture to exotic forest, areas that are less than 1 ha should be made eligible for carbon credits. The prototype application described here simulates the immediate consequences of the planting activities on farm cash flow, including loss in profit compared with the original land use and establishment costs. The sale of carbon credits is not considered here because it is assumed that the carbon sequestration arising from the planted trees is instead used to offset on-farm carbon emissions that originate from livestock. Furthermore, in New Zealand, profit from forestry is only realised after approximately 25 years once the tree stand has matured and Fig 6. Planting scenarios. (A) Map of the farm border (red line). Map of the farm border (red line). Grey shaded areas represent the footprint (area) that would need to be planted for the farm to be carbon neutral with the assumption that the area producing the least pasture are converted first. When considering pine trees, only the dark grey area would need to be planted for the farm to be carbon neutral. When considering native trees, because their sequestration rate is smaller than pine trees, all the shaded areas (dark and light grey) would need to be planted for the farm to be carbon neutral. As expected from the results presented in Fig 5 the highlighted locations are mostly clustered around the steepest slopes of the farm. The black box at the northwest (top left) of the map highlights the spatial extent of the aerial

imagery presented on (B). (B) High-resolution (300mm) aerial imagery of the location highlighted in (A). From a visual analysis, the bright yellow areas are bare soil on steep slopes suggesting very active land erosion that has also been confirmed during a visual assessment in the field. These areas overlap the planting zones suggested by the map on (A) and also represented by the hatched symbology on (B). Across the farm, converting the least productive areas from pasture to forest might help mitigating soil erosion in addition to generating carbon sequestration. https://doi.org/10.1371/journal.pone.0286723.g006 PLOS ONE Designing profitable and climate-smart farms using virtual reality PLOS ONE | https://doi.org/10.1371/journal.pone.0286723 June 2, 2023 9 / 16