Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Pome fruit (apple and pear) is a popular component of healthy nutrition. Pome fruit production constitutes an important contribution to local economies such as in the fruit growing region in the East of the Flanders region of Belgium. However, the commercial production systems are vulnerable to meteorological hazards such as hail, drought and spring frost. The latter led to moderate to full harvest losses in 2017, 2019, 2021, and is the focus of this thesis.The resistance of pome fruit trees to cold temperatures decreases in spring, as buds develop into flowers to subsequently form fruits. Thereby the timing of the development stages (phenology) is conditioned by ambient temperatures and thus potentially affected by a changing climate. This relationship between phenology and temperature can be used to predict changes in phenology for decades to come and to better understand the hazard of spring frosts. In a warming climate, pome fruit tree flowering occurs earlier in the year, but so do the last frost days.The first objective of this research was to find out whether frosts would occur in Belgium more or less often during the flowering stage of pear trees in the coming decades. In Chapter 2, two plant development models were fit to 70 years of flowering observations for several cultivars and air temperature records at the Flemish fruit research station PCFruit. In Chapter 3, the best performing model was applied on future temperature projections provided by regional climate models (CORDEX) for the pear cultivar Conference. Comparing the decades 2020-2070 to 1970-2020, the projected flowering started around 7.5 to 10.8 days earlier on average, 2020 (for low or high CO2 emission scenarios, respectively), while the last frost date advanced by 12.8 to 17.9 days. For most models and most locations in Belgium, we detected fewer frosts during the flowering until 2070 than in the past. Regions in the south of Belgium remain frost-prone, while the northern regions become less frost prone. However, occasional destructive frost events could not be ruled. Hence damage prevention and mitigation remains importantA second research objective was to review the literature to identify the most effective damage prevention and mitigation technique(s). A peer reviewed protocol (Chapter 4) was the basis for a multilingual and systematic review of the relevant literature (Chapter 5). According to predefined criteria, all relevant literature from major academic and specialized databases for the most important temperate fruits since 1900 was identified. The conditions for higher prevention efficiency were investigated, as were the effects of the techniques on air temperatures, bud survival and yields. The identified research was restricted to only a few fruit types and countries, that do not reflect the global production. The highest damage protection was recorded for sprinkling systems and certain foliar sprays, and the largest increases in field temperatures were found to be achieved by wind machines. However, the variability of findings was large. Two-thirds of the studies were likely subject to biases and nearly all studies lacked details on experimental set-up and quantitative observations. Beneficial conditions related to the local environmental setting could therefore not be identified clearly, although statistical differences between the studies were apparent. Improved reporting standards in frost protection research are recommended to allow transfer research findings for evidence-based decision making.In conclusion, the reduction in late spring frosts is more pronounced than the advancement of the onset of flowering. The risk for frost damage is reduced in low lying northern half of Belgium, however, spring frosts remain considerable. More transferable research on resource efficient prevention and mitigation remains necessary.

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

More than 20% of the Flemish territory consists of grasslands. Due to climate change, factors influencing grassland yields will change. CO2 concentrations and temperatures will rise and the number of extreme weather events such as droughts will increase. In addition, imposed fertilisation restrictions will further reduce the amount of nutrients. These fertilisation standards had a negative influence on total yields at the end of the 20th Century. To get a better picture of the weather impact on grassland yields, the effect on yields was investigated for the years 2016, 2017 and 2018. Characteristics of grassland locations were first examined. A spatial comparison was made between grasslands and other agricultural crops in Flanders. Permanent grasslands generally occur on soils associated with lower yields such as heavy clay soil textures in the Polder region and sandy soil textures in the Campine region. In addition, significantly more poorly drained soils were found under permanent pastures with P-values less than 0.001. Average slopes were also generally steeper for grassland plots. Similar results were obtained for the comparison between permanent and temporary grassland with known yields. An explanation can be found in the Common Agricultural Policy of the European Union. Farmers nowadays have to fulfil several preconditions to receive subsidies. One of these is to maintain the ratio of permanent grassland to arable land, which benefits general biodiversity and ecosystem functions. Farmers therefore often use their least productive parcels as permanent grassland. In addition, grass is often planted to combat erosion, which occurs mainly on parcels with steeper slopes. A geodata set was then created that linked yield data from plots to their location. These locations were linked to weather data from the corresponding year and location. In addition, the characteristics of the location of the plot were added. A distinction was made between grasslands cultivated before or after the main arable crop and permanent grasslands. The permanent grasslands were further subdivided according to the type of harvest: silage and hay were distinguished. These data were further split into three agricultural regions with different characteristics, i.e. the Polder region, the Campine region and the Sandy loam region. Models were created to predict harvest based on average maximum temperature, total precipitation and cumulative precipitation deficit, calculated over the growing season from early March to late October. Rising temperatures generally have a positive impact on grassland yields. Five out of nine models showed a significantly positive relationship with P-values less than 0.05. Yields decreased as the integral of the cumulative rainfall deficit increased, indicating that yields decreased under drier conditions. Five of the nine models showed a significantly negative relationship with P-values less than 0.05. No uniform relationship could be shown between total precipitation and yields, indicating that yields were influenced by precipitation patterns rather than total precipitation. Positive effects of increasing maximum temperature were smaller than negative effects of increasing cumulative rainfall deficit. R-squared values ranged between 0.036 and 0.455. In addition, 9 models were created that also included location characteristics. R-squared values varied between 0.064 and 0.982. In particular, interaction effects were important. Significan