Projected Wine Grape Cultivar Shifts Due to Climate Change in New Zealand.

Climate change has already been affecting the regional suitability of grapevines with significant advances in phenology being observed globally in the last few decades. This has significant implications for New Zealand, where the wine industry represents a major share of the horticultural industry revenue. We modeled key crop phenological stages to better understand temporal and spatial shifts in three important regions of New Zealand (Marlborough, Hawke's Bay, Central Otago) for three dominant cultivars (Merlot, Pinot noir, and Sauvignon blanc) and one potential new and later ripening cultivar (Grenache). Simulations show an overall advance in flowering, véraison, and sugar ripeness by mid-century with more pronounced advance by the end of the century. Results show the magnitude of changes depends on the combination of greenhouse gas emission pathway, grape cultivar, and region. By mid-century, in the Marlborough region for instance, the four cultivars would flower 3 to 7 days earlier and reach sugar ripeness 7 to 15 days earlier depending on the greenhouse gas emission pathway. For growers to maintain the same timing of key phenological stages would require shifting planting of cultivars to more Southern parts of the country or implement adaptation strategies. Results also show the compression of time between flowering and véraison for all three dominant cultivars is due to a proportionally greater advance in véraison, particularly for Merlot in the Hawke's Bay and Pinot noir in Central Otago. Cross-regional analysis also raises the likelihood of the different regional cultivars ripening within a smaller window of time, complicating harvesting schedules across the country. However, considering New Zealand primarily accommodates cool climate viticulture cultivars, our results suggest that late ripening cultivars or extended ripening window in cooler regions may be advantageous in the face of climate change. These insights can inform New Zealand winegrowers with climate change adaptation options for their cultivar choices.

Mapping floral resources for honey bees in New Zealand at the catchment scale.

Honey bees require nectar and pollen from flowers: nectar for energy and pollen for growth. The demand for nectar and pollen varies during the year, with more pollen needed in spring for colony population growth and more nectar needed in summer to sustain the maximum colony size and collect surplus nectar stores for winter. Sufficient bee forage is therefore necessary to ensure a healthy bee colony. Land-use changes can reduce the availability of floral resources suitable for bees, thereby increasing the susceptibility of bees to other stressors such as disease and pesticides. In contrast, land-based management decisions to protect or plant bee forage can enhance pollen and nectar supply to bees while meeting other goals such as riparian planting for water-quality improvement. Commercial demand for honey can also put pressure on floral resources through over-crowding of hives. To help understand and manage floral resources for bees, we developed a spatial model for mapping monthly nectar and pollen production from maps of land cover. Based on monthly estimated production data we mapped potential monthly supply of nectar and pollen to a given apiary location in the landscape. This is done by summing the total production within the foraging range of the apiary while subtracting the estimated nectar converted to energy for collection. Ratios of estimated supply over theoretical hive demand may then be used to infer a potential landscape carrying capacity to sustain hives. This model framework is quantitative and spatial, utilizing estimated flight energy costs for nectar foraging. It can contribute to management decisions such as where apiaries could be placed in the landscape depending on floral resources and where nectar limited areas may be located. It can contribute to planning areas for bee protection or planting such as in riparian vegetation. This would aid managed bee health, wild pollinator protection, and honey production. We demonstrate the methods in a case study in New Zealand where there is a growing demand for manuka (Leptospermum scoparium) honey production.

Mapping of Escherichia coli Sources Connected to Waterways in the Ruamahanga Catchment, New Zealand.

Rivers and streams in New Zealand are natural with free access and used by many people for swimming and fishing. However, pastoral farming with free grazing animals is a common land use in New Zealand and faecal microorganisms from them often end up in waterways. These microorganisms can seriously affect human and animal health if ingested. This paper describes spatial modeling using GIS of Escherichia coli sources in a large catchment (350 000 ha), the Ruamahanga. By examining the pathway of water over and through soils, it is possible to determine whether E. coli sources are connected to waterways or not. The map of E. coli sources connected to waterways provides useful context to those setting water quality limits. This approach avoids the complexity of modeling the fate and transport of E. coli in waterways, yet still permits the assessment of catchment-wide mitigation and best management practice. Fencing of waterways would minimize E. coli sources directly defecated to water and would reduce total E. coli sources by approximately 35%. Introduction of dung beetles would minimize sources connected to waterways by overland flow and would reduce total E. coli sources by approximately 35%. Construction of dairy effluent ponds would minimize sources connected to waterways through high bypass flow in soils and would reduce total E. coli sources by approximately 25%.

Upscaling NZ-DNDC using a regression based meta-model to estimate direct N2O emissions from New Zealand grazed pastures.

The availability of detailed input data frequently limits the application of process-based models at large scale. In this study, we produced simplified meta-models of the simulated nitrous oxide (N2O) emission factors (EF) using NZ-DNDC. Monte Carlo simulations were performed and the results investigated using multiple regression analysis to produce simplified meta-models of EF. These meta-models were then used to estimate direct N2O emissions from grazed pastures in New Zealand. New Zealand EF maps were generated using the meta-models with data from national scale soil maps. Direct emissions of N2O from grazed pasture were calculated by multiplying the EF map with a nitrogen (N) input map. Three meta-models were considered. Model 1 included only the soil organic carbon in the top 30cm (SOC30), Model 2 also included a clay content factor, and Model 3 added the interaction between SOC30 and clay. The median annual national direct N2O emissions from grazed pastures estimated using each model (assuming model errors were purely random) were: 9.6GgN (Model 1), 13.6GgN (Model 2), and 11.9GgN (Model 3). These values corresponded to an average EF of 0.53%, 0.75% and 0.63% respectively, while the corresponding average EF using New Zealand national inventory values was 0.67%. If the model error can be assumed to be independent for each pixel then the 95% confidence interval for the N2O emissions was of the order of ±0.4-0.7%, which is much lower than existing methods. However, spatial correlations in the model errors could invalidate this assumption. Under the extreme assumption that the model error for each pixel was identical the 95% confidence interval was approximately ±100-200%. Therefore further work is needed to assess the degree of spatial correlation in the model errors.

Investigating a method for estimating direct nitrous oxide emissions from grazed pasture soils in New Zealand using NZ-DNDC.

In this study, we developed emission factor (EF) look-up tables for calculating the direct nitrous oxide (N2O) emissions from grazed pasture soils in New Zealand. Look-up tables of long-term average direct emission factors (and their associated uncertainties) were generated using multiple simulations of the NZ-DNDC model over a representative range of major soil, climate and management conditions occurring in New Zealand using 20 years of climate data. These EFs were then combined with national activity data maps to estimate direct N2O emissions from grazed pasture in New Zealand using 2010 activity data. The total direct N2O emissions using look-up tables were 12.7±12.1 Gg N2O-N (equivalent to using a national average EF of 0.70±0.67%). This agreed with the amount calculated using the New Zealand specific EFs (95% confidence interval 7.7-23.1 Gg N2O-N), although the relative uncertainty increased. The high uncertainties in the look-up table EFs were primarily due to the high uncertainty of the soil parameters within the selected soil categories. Uncertainty analyses revealed that the uncertainty in soil parameters contributed much more to the uncertainty in N2O emissions than the inter-annual weather variability. The effect of changes to fertiliser applications was also examined and it was found that for fertiliser application rates of 0-50 kg N/ha for sheep and beef and 60-240 kg N/ha for dairy the modelled EF was within ±10% of the value simulated using annual fertiliser application rates of 15 kg N/ha and 140 kg N/ha respectively.

Tradeoffs between soil, water, and carbon -- a national scale analysis from New Zealand.

The tradeoffs between the regulation of soil erosion, provision of fresh water, and climate regulation associated with new Pinus radiata forests in New Zealand are explored using national models. These three ecosystem services for which there is strong demand are monetised as commodities (avoided soil erosion is NZ $1 per tonne; water is NZ $1 per cubic metre; and sequestered carbon is assumed to be NZ $73 per tonne). This permits their summation on a spatial basis to produce a national map of the net benefit of these ecosystem services. Net benefit is spatially variable depending primarily on the relative mix of forest growth rates and demand for irrigation water. New P. radiata forests (once mature) generally reduce mass-movement erosion by an order of magnitude. This provides significant benefits for erosion control where there are high natural rates of erosion. Benefits are especially large in catchments where high sedimentation is increasing flood risk and degrading aquatic ecosystems. The generally high growth rates of P. radiata in New Zealand (8.5 tonnesCha(-1)yr(-1) on average for existing forest) add significant environmental benefits of carbon sinks to climate regulation. However, the reduction of water yield associated with new forests (between 30% and 50%) can neutralise these benefits in catchments where there is demand for irrigation water, such as the eastern foothills of the Southern Alps and the tussock grasslands in the South Island.

Rapid mapping and prioritisation of wetland sites in the Manawatu-Wanganui region, New Zealand.

The extent of wetland in New Zealand has decreased by approximately 90% since European settlement began in 1840. Remaining wetlands continue to be threatened by drainage, weeds, and pest invasion. This article presents a rapid method for broad-scale mapping and prioritising palustrine and estuarine wetlands for conservation. Classes of wetland (lacustrine, estuarine, riverine, marine, and palustrine) were mapped using Landsat ETM+ imagery and centre-points of palustrine and estuarine sites as ancillary data. The results shown are for the Manawatu-Wanganui region, which was found to have 3060 ha of palustrine and 250 ha of estuarine wetlands. To set conservation priorities, landscape indicators were computed from a land-cover map and a digital terrain model. Four global indicators were used (representativeness, area, surrounding naturalness, and connectivity), and each was assigned a value to score wetland sites in the region. The final score is an additive function that weights the relative importance of each indicator (i.e., multicriteria decision analysis). The whole process of mapping and ranking wetlands in the Manawatu-Wanganui region took only 6 weeks. The rapid methodology means that consistent wetland inventories and ranking can now actually be produced at reasonable cost, and conservation resources may therefore be better targeted. With complete inventories and priority lists of wetlands, managers will be able to plan for conservation without having to wait for the collection of detailed biologic information, which may now also be prioritised.