Abstract
Anthropogenic climate change and global population growth has a detrimental effect on agricultural production, along with causing a rising food demand. While elevated CO2 levels might increase crop production, monocultural production in combination with more extreme weather events, causes biodiversity loss and unsustainable agricultural systems. Agroforestry is a diverse production system, which combines woody perennials with pasture or agricultural crops, and has been recognised as a method for creating resilient production systems for both mitigation of and adaptation to climate change. Through international research, agroforestry has been found to increase carbon (C) sequestration, biodiversity and animal welfare, while reducing nutrient loss, compared to conventional agriculture. However, due to the limited extend of agroforestry practices in Europe, research into the effects of agroforestry is limited in temperate regions. Therefore, this PhD project aimed to assess the effects of woody vegetation in a silvoarable (combining woody perennials with crops) and a silvopastoral (combining woody perennials with pasture) agroforestry system in Denmark. This was done through quantification of the ecosystem effects of production (study A), C storage (Study B), and nitrogen (N) leaching (Study C).
Data collection was carried out in two certified organic production systems in Central Jutland, which were both irrigated and fertilised. Both systems included a shelterbelt of 28-32 years located across the centre of a field in a North-South direction, consisting of three rows of deciduous trees and shrubs, covering 2.4% of the silvoarable system, and 1.6% of the silvopastoral system. Sample plots were arranged in three transects, each with a plot at the centre of the shelterbelt and at distance 225 m to the east and the west of the shelterbelt. Over three years (2021-2023) measurements were carried out of shelterbelt biomass, grain and grass yields, rain and irrigation inputs, soil moisture levels, soil conditions and soil water.
Study A found lower yields in both grain and grass near the shelterbelts. Linear mixed-effects models was applied to estimate the yield loss as a function of distance to the shelterbelt, while additionally testing for the effect of crown cover, leaf litter, weather (precipitation, daily average temperature, daily average wind speed), and soil conditions (moisture levels, bulk densities, pH, C and N concentrations). The models pointed to shade as the main limiting factor of yield, and based on a yield loss of 5%, tree-crop competition zones were found at distance up to 0.5 times the tree height (0.5H) and 1.2H in the silvoarable and silvopastoral system, respectively. Higher N and protein content found under light-limited and low-yielding conditions was not found to compensate for yield losses in terms of protein yields.
Study B found a higher soil C stock at 0-75 cm within the shelterbelt compared to further into the field. However, the soil C stock was 53% higher in the shelterbelt (215 Mg ha-1) than in the field (140 Mg ha-1) in the silvoarable system, but only 5% higher in the shelterbelt (153 Mg ha-1) than the field (132 Mg ha-1) in the silvopastoral system, where the difference was not statistically significant. Shelterbelt woody biomass added 146 and 57 Mg C ha-1 shelterbelt, resulting in a total system C stock of 146 and 134 Mg ha-1 for the silvoarable and silvopastoral system, respectively. Additionally, a higher heterotrophic respiration and CH4 oxidation found in the top soil of the shelterbelt were attributed to higher microbial activity and less disturbance compared to the field at both sites. Estimated potential C stock based on four scenarios with increased tree cover of variable design, underlines that C stock depends on tree density, age and species.
Study C found soil N stock at 0-75 cm depth to be similar between the two agroforestry systems, but with 34% lower N stock in the shelterbelt than the field in the silvoarable system, and no differences across the silvopastoral system. Annual total N leaching was 39 ± 7 kg ha-1, and reduced with up to 5% at up to 7 m distance to the shelterbelt in the silvoarable system. In the silvopastoral system, the annual N leaching was 50 ± 11 kg ha-1, with a reduction of up to 5% at up to 5 m distance to the shelterbelt.
Collectively, Study A, B and C underlines that the inclusion of woody perennials in agricultural production systems result in increased C stock and reduced N leaching, but at the cost of yield loss, likely caused by shade. No effects were found beyond the distance of 15 m to the shelterbelt in the silvoarable system, and 10 m in the silvopastoral system. Ecosystem effects in agroforestry systems depend heavily on design, and not one design can optimise the effects of all parameters at once. Thus, the study led to a list of design recommendations for optimised design of temperate agroforestry systems based on different aims.
Data collection was carried out in two certified organic production systems in Central Jutland, which were both irrigated and fertilised. Both systems included a shelterbelt of 28-32 years located across the centre of a field in a North-South direction, consisting of three rows of deciduous trees and shrubs, covering 2.4% of the silvoarable system, and 1.6% of the silvopastoral system. Sample plots were arranged in three transects, each with a plot at the centre of the shelterbelt and at distance 225 m to the east and the west of the shelterbelt. Over three years (2021-2023) measurements were carried out of shelterbelt biomass, grain and grass yields, rain and irrigation inputs, soil moisture levels, soil conditions and soil water.
Study A found lower yields in both grain and grass near the shelterbelts. Linear mixed-effects models was applied to estimate the yield loss as a function of distance to the shelterbelt, while additionally testing for the effect of crown cover, leaf litter, weather (precipitation, daily average temperature, daily average wind speed), and soil conditions (moisture levels, bulk densities, pH, C and N concentrations). The models pointed to shade as the main limiting factor of yield, and based on a yield loss of 5%, tree-crop competition zones were found at distance up to 0.5 times the tree height (0.5H) and 1.2H in the silvoarable and silvopastoral system, respectively. Higher N and protein content found under light-limited and low-yielding conditions was not found to compensate for yield losses in terms of protein yields.
Study B found a higher soil C stock at 0-75 cm within the shelterbelt compared to further into the field. However, the soil C stock was 53% higher in the shelterbelt (215 Mg ha-1) than in the field (140 Mg ha-1) in the silvoarable system, but only 5% higher in the shelterbelt (153 Mg ha-1) than the field (132 Mg ha-1) in the silvopastoral system, where the difference was not statistically significant. Shelterbelt woody biomass added 146 and 57 Mg C ha-1 shelterbelt, resulting in a total system C stock of 146 and 134 Mg ha-1 for the silvoarable and silvopastoral system, respectively. Additionally, a higher heterotrophic respiration and CH4 oxidation found in the top soil of the shelterbelt were attributed to higher microbial activity and less disturbance compared to the field at both sites. Estimated potential C stock based on four scenarios with increased tree cover of variable design, underlines that C stock depends on tree density, age and species.
Study C found soil N stock at 0-75 cm depth to be similar between the two agroforestry systems, but with 34% lower N stock in the shelterbelt than the field in the silvoarable system, and no differences across the silvopastoral system. Annual total N leaching was 39 ± 7 kg ha-1, and reduced with up to 5% at up to 7 m distance to the shelterbelt in the silvoarable system. In the silvopastoral system, the annual N leaching was 50 ± 11 kg ha-1, with a reduction of up to 5% at up to 5 m distance to the shelterbelt.
Collectively, Study A, B and C underlines that the inclusion of woody perennials in agricultural production systems result in increased C stock and reduced N leaching, but at the cost of yield loss, likely caused by shade. No effects were found beyond the distance of 15 m to the shelterbelt in the silvoarable system, and 10 m in the silvopastoral system. Ecosystem effects in agroforestry systems depend heavily on design, and not one design can optimise the effects of all parameters at once. Thus, the study led to a list of design recommendations for optimised design of temperate agroforestry systems based on different aims.
| Original language | English |
|---|
| Publisher | Department of Food and Resource Economics, Faculty of Science, University of Copenhagen |
|---|---|
| Number of pages | 244 |
| Publication status | Published - 2025 |