Browsing by Subject "Modelling"

Now showing 1 - 9 of 9
  • Thumbnail Image
    Publication
    Adaptation of herd simulation models to predict the efficiency of the use of resources in tropical ruminant production systems
    (2020) Bateki Adjogo, Christian; Dickhöfer, Uta
    Agricultural systems in the (Sub-)Tropics are under increasing pressure to produce more food and satisfy the growing demand of a rapidly growing and more affluent human population for agricultural products. With growing rates of urbanization in these regions and the associated dietary changes, the demand for calories from animal-based foods like milk, meat, and eggs could increase by 74 to 114 % between 2010 and 2050. Ruminant livestock have the potential to contribute to satisfying the demand for animal-based foods in the (Sub-)Tropics, but also raise considerable environmental concerns, amongst others due to their emissions. The use of simulation models is a holistic approach to identify how to sustainably harness the potential of ruminants for animal-based food production in the (Sub-)Tropics. Although several ruminant herd models are relevant for studying tropical ruminant production systems, most of them were developed using data that quantify and characterize biological processes of ruminants in temperate regions. Therefore, the present thesis identified and adapted an existing ruminant livestock herd model to adequately predict resource use and the potential outputs from production systems in the (Sub-)Tropics. The present thesis showed that state-of-the-art ruminant livestock herd models used to simulate tropical production systems need further development to enable them to address the modelling needs identified. Instead of developing new models to address these modelling needs, existing simulation models could be adapted using the increasingly available data that quantify and characterize biological processes in ruminants in these regions. This approach will ensure that not only the direction of change for different management strategies will be identified for tropical ruminant production systems, but also the correct magnitude of resources use and productive and reproductive performance.
  • Thumbnail Image
    Publication
    Agro-economic policy analysis with the regional production model ACRE : a case study for Baden-Wuerttemberg
    (2011) Henseler, Martin; Dabbert, Stephan
    Since its introduction the Common Agricultural Policy (CAP) of the European Union (EU) has undergone several reforms in order to adapt policy instruments and enable the agricultural sector to fulfil multiple functions with respect to economic, supply and environmental objectives. In the German federal state Baden-Wuerttemberg agricultural production is characterized by regional heterogeneity. Therefore it is important to estimate the impacts resulting from changes in the CAP at a detailed regional level. In this study the agricultural policy model ACRE (Agro-eConomic pRoduction model at rEgional level) has been used to simulate different policy scenarios and to analyze regional economic, production and environmental impacts. In particular the study aims to address the following research questions: What are the regional impacts of different policy measures in the German federal state Baden-Wuerttemberg with respect to economic, production and environmental objectives? How suitable are the simulated policy measures for achieving the policy objectives of the CAP 2003 reform, as well as the objectives of subsidy reduction, promotion of energy crop production, reduction of environmental pollution and promotion of agro-environmental measures? How suitable is the regional supply model ACRE as a tool for policy analysis and policy decision support? In order to address the research questions, ACRE has been updated, adapted and extended to simulate agricultural production in the federal state Baden-Wuerttemberg at NUTS3 level. The policy scenarios simulated in this study are defined to cover recent discussions on the future development of the CAP and their results are analysed according to a regional framework for NUTS3 counties, farm types and the complete model region. The simulation of the reference year (REF) implies the policy reform Agenda 2000 in the simulation year 2000. Thus, REF represents the observed situation of regional agricultural production on whose statistical data ACRE is calibrated. The scenario CAP2003 simulates the policy measures of the CAP 2003 reform in the simulation year 2015. Assumptions of increased yields and prices as well as harmonized direct payments for arable land and grassland result in an increase in income as well as in an increase of subsidy volume. In the entire model region Baden-Wuerttemberg cereal production increases while the production of fattening bulls and pigs decreases. Increases in crop production intensity result in an increase in environmental pollution. The scenario CAP2003 is used as the baseline scenario to compare the results of simulated policy scenarios which are delineated in the following paragraphs only with the most important results for the complete model region Baden-Wuerttemberg. In two subsidy reduction scenarios the simulated policy instruments aim to reduce subsidy volume by reducing Pillar 1 payments by 60% and by shifting 70% of the money from Pillar 1 to Pillar 2 respectively. Both scenarios result in the positive impact of a decrease in subsidy volume, but show a negative impact, especially an increase of abandoned land. In two energy crop scenarios the production of energy maize is simulated under the assumption that different situations in energy policy and energy markets result in different competitiveness between production of energy maize and food. In both scenarios energy crop production partially replaces cereal production, although the extent varies according to the high or small level of competitiveness between production of energy maize and food. Impacts on agricultural income and subsidies are small while increased environmental pressure is expected in the event of a significant expansion in energy crop production. Two nitrogen reduction scenarios simulate policy measures according to the water framework directive (WFD) and the OSPAR convention. The scenario according to the WFD (limitation of organic nitrogen input to a maximum of 170kg nitrogen per hectare) does not result in any impacts. In contrast, the scenario according to the OSPAR convention (reduction of nitrogen input quantities by 10%) results in a decrease in environmental pollution and is accompanied by a reduction of income and reduction of agricultural production under land abandonment. In the scenario of mandatory agri-environmental measures (AEM) it is assumed that the area with applied AEM is extended. The increase of AEM area results in a decrease in cereal production and a reduction of environmental pollution, while income decreases only slightly. Two combined scenarios simulate a mix of different policy and market situations which provoke an intensive and an extensive agricultural production. The results of these scenarios illustrate the interaction of the single policy measures. The measures of subsidy reduction have similar reducing impacts on income and subsidy volume in both scenarios. In the intensive production scenario high competitive energy crop production and a less restrictive nitrogen restriction result in a compensation effect of land abandonment by extension of energy crop area. In the extensive production scenarios, less competitive energy crop production and a high restrictive nitrogen constraint result in reduced agricultural production, increased land abandonment and reduced environmental pressure. In order to evaluate the impact of the simulated policy measures on the achievement of policy objectives the results of all scenarios are compared and ranked according to their impact on the policy objectives. The analyses of the model results show impacts of policy measures which are likely to be expected. However, the analyses at NUTS3 as well as farm types' level reveal that the impacts of the policy measures can be regionally quite different. Thus the detailed regional model results clearly show that (and where) the implementation of agricultural policy measures requires a regional specific evaluation and monitoring. In order to discuss the study with regard to the methods applied and the outcome, a final strengths and weaknesses analysis was conducted. The analysis highlights the strengths of the study (e.g. the model validation, the regional analysis of different policy scenarios, the possibility of cooperation with regional stakeholders). The validation and the results of the study also show that ACRE is a suitable tool for regional agricultural policy analysis and policy decision support. Supplementary work could help to overcome single shortcomings and caveats and to further develop the model. However, ACRE can already be used now as a useful tool for the regional agricultural policy analysis of the CAP in Baden-Wuerttemberg.
  • Thumbnail Image
    Publication
    Benefits and trade-offs of legume-led crop rotations on crop performance and soil erosion at various scales in SW Kenya
    (2021) Koomson, Eric; Cadisch, Georg
    Soil erosion and land fragmentation threaten agricultural production in large parts of the Western Kenyan Highlands. In Rongo watershed, maize–common bean intercropping systems, which dominate the agricultural landscape, are vulnerable to soil degradation, especially on long slope lengths where ground and canopy cover provision fail to protect the soil from the disruptive impact of raindrops. The inclusion of soil conservation measures like hedgerows, cover crops or mulch can reduce soil erosion, but compete with crops for space and labour. Knowledge of critical slope length can minimise interventions and trade–offs. Hence, we evaluated maize–common bean intercrop (MzBn) regarding runoff, erosion and crop yield in a slope length trial on 20, 60 and 84 m plot lengths, replicated twice on three farms during one rainy season in Rongo, Migori County. Additionally, we investigated systems of MzBn (farmers’ practice), MzBn with 5 Mg ha-1 Calliandra calothyrsus mulch (Mul), Arachis hypogaea (Gnt), Lablab purpureus (Lab) and Mucuna pruriens (Muc), regarding their impact on infiltration, runoff, soil loss, soil C and N loss during three rainy seasons (long and short rains, LR and SR, 2016, and LR 2017). Measured field data on soil, crop, spatial maps and meteorology were used as input datasets to parameterize and calibrate the LUCIA model. The calibrated and validated model was then used to simulate agronomic management scenarios related to planting date (planting with first rain vs baseline) and vegetation cultivar (short duration crop) to mitigate water stress. Based on the measurements, groundcover was most influential over rainfall intensity (EI30) and plant canopy cover in predicting soil loss. Dense groundcover of Mul at the beginning of the rainy seasons was decisive to significantly (p<0.05) lowering overall seasonal average runoff by 88, 87 and 84% over MzBn, Lab and Gnt, respectively, whereas, soil loss under Mul was reduced by 66 and 65% over Gnt and Lab, respectively. The high proportion of large soil aggregates (> 5mm) in the topsoil under Mul at the end of SR 2016 significantly (p<0.05) increased infiltration rates (420 mm hr-1) in LR 2017 compared to Lab (200 mm hr-1) and Gnt (240 mm hr-1). Average C and N concentrations in eroded sediments were significantly reduced under Mul (0.74 kg C ha–1, 0.07 kg N ha–1) during the LR 2016 as compared to MzBn (3.20 kg C ha–1, 0.28 kg N ha–1) and Gnt (2.54 kg C ha–1, 0.23 kg N ha–1). Likewise, in SR 2016 Mul showed significantly lowered C and N losses of 3.26 kg C ha–1 and 0.27 kg N ha–1, respectively, over Lab (9.82 kg C ha–1, 0.89 kg N ha–1). Soil loss over 84 m slope length was overall significantly higher by magnitudes of 250 and 710% than on 60 and 20 m long plots, respectively, which did not differ significantly among each other (p<0.05). For runoff, 84 m plot length differed significantly from 60 and 20 m, but in the opposite trend as for soil loss. Across all three farms, slope gradient and slope length were the variables with highest explanatory power to predict soil loss. At the individual farm level, under homogeneous slope and texture, slope length and profile curvature were most influential. Considering results of slope length experiments, plot lengths less than 50 m appear to be preferential considering soil loss, sediment load, and soil loss to yield ratio under the given rainfall, soil and slope conditions. Our results call for integrating slope length options and cropping systems for effective soil conservation. We recommend planting Mucuna and Calliandra–hedgerows as buffer strips below the critical slope length, and legume cash crops and maize uphill. Such approaches are critical in the backdrop of land fragmentation and labour limitation in the region to sustainably maximise land area. In the modelling exercise, crops planted one and three weeks after the baseline planting date increased Maize and Muc grain yield over the baseline during the three cropping seasons, the three weeks treatment in particular. This could be due to more favourable weather conditions during the shifted vegetation period. Increased grain yield corresponded to high water use efficiency (WUE). The short duration crop planted three weeks after the baseline planting date (PD3WL+SDC10) showed the highest grain yield after PD3WL (three weeks late plaing with BL variety). The use of cultivars with short growth cycle offers the flexibility of planting again where crops failed due to crop water stress or where the rains delay, ensuring completion of the growth cycle before the season ends. Given that short growth duration crops produce less grain yield compared to their counterpart full season crops, due to the length of their cycles, breeding programs must prioritize traits that can enhance the size of the grain-filling sink. At the plot level, management systems that reduce evaporation and retain soil moisture, e.g. mulching, application of farmyard manure etc., must be promoted to reduce evapotranspiration.
  • Thumbnail Image
    Publication
    Exploring and modelling the influence of spectral light composition on soybean (Glycine max (L.) Merr.)
    (2019) Hitz, Tina; Graeff-Hönninger, Simone
    The development of soybean cultivars for the climatic conditions in Europe is an urgent need in order to increase the European production and to decrease the dependence of imported soybean. A speed breeding system can accelerate the process of developing new cultivars by growing more generations per season in climate chambers. The project MoLED-Plant aimed towards the development of a speed breeding system for soybean under LED lighting. The major objectives of this thesis were to: (i) construct a three dimensional model of an LED chamber to simulate micro-light climate, (ii) develop a functional-structural plant (FSP) model of soybean and derive a blue photon flux density (BPFD) response curve from simulations, (iii) apply the FSP model with the integrated response curve for spectral optimization, (iv) explore the influence of BPFD under constant photosynthetic photon flux density (PPFD), and (v) disentangle the influence of red to far-red ratio (R:FR) and PPFD on the shade avoidance response (SAR). The objectives were fulfilled with a combination of FSP modelling in the Growth Grammar-related Interactive Modelling Platform (GroIMP) and plant experiments under multiple spectra in LED chambers. The presented LED chamber model was the first three dimensional environment, which was developed for spectral optimizations in indoor farming using FSP modeling. Measurements performed with a spectrometer in multiple heights and orientations were compared to simulations recorded with a virtual sensor at the same locations. The model was evaluated as a tool for assessment of spectral light heterogeneity under an alternative placement of the LED modules. Applying the model can assist in choosing the best chamber design and placements of LEDs to assure homogeneous light conditions. Subsequently, static soybean plants were incorporated into the chamber model. Comparison of light simulations and measurements from below the soybean canopy in four reconstructed scenarios assured a good simulation of micro-light climate. The model was applied to simulate the effect of an increased plant density in an experiment in the chamber. The simulations of light homogeneity in the experimental setup can assist in choosing the optimal design. The developed dynamic FSP model of soybean within the chamber model represents the first FSP model with an integrated response to BPFD. The soybean model was calibrated with data from BPFD experiments. From simulations, a common response curve of internode elongation to the perceived BPFD was derived for the second and third internode. The response curve was integrated in the model and was applied for spectral optimization in a chamber scenario with an alternative high reflective bottom material. The soybean response to BPFD under constant PPFD and the influence of R:FR and PPFD on SAR was explored by designing specific spectra from LEDs. Soybean experiments were performed under six levels of BPFD (60-310 µmol m-2 s-1) and constant PPFD (400 µmol m-2 s-1). Plant height and biomass decreased, leaf mass ratio increased and the ratio of stem biomass (internode plus petiole) translocated to the internode decreased under high BPFD. Three soybean cultivars were grown under nine light treatments to disentangle the effect of R:FR and PPFD. Internode elongation responded mainly to low PPFD with an additive effect from low R:FR, whereas petiole elongation was influenced to a great extent by low R:FR. In the context of SAR, petiole elongation can be seen as the main response to the threat of shade (high PPFD and low R:FR) and both petiole and internode elongation as the response to true shade (low PPFD and low R:FR). This thesis showed how PPFD, BPFD and R:FR work both independently, antagonistically and synergistically on the physiology and morphology of soybean. The increased insight in these responses can e.g. support crop breeding and spectral optimization in indoor farming. Furthermore, interesting and important objectives for future research were identified. These experiments should include physiological measurements for a deeper understanding of interactions and underlying mechanisms. Spectral optimization of indoor farming depends on the purpose of the production. For instance, a high BPFD of 260 µmol m-2 s-1 was optimal for speed breeding, whereas an intermediate BPFD would be preferable to increase biomass. Comprehensive investigation of spectral influence on plant physiology and morphology is necessary to fully utilize the spectral flexibility of LED lighting. The developed FSP model of soybean in a virtual LED chamber represents an important step towards utilizing the advantages of FSP modelling by simulation of a wide variety of scenarios. The model can be adjusted or extended depending on the purpose of the model. It can be calibrated for other crop species and the setting of the chamber dimensions can be changed.
  • Thumbnail Image
    Publication
    Investigation of fluidised bed coating : measurement, optimisation and statistical modelling of coating layers
    (2017) van Kampen, Andreas; Kohlus, Reinhard
    Fluidised bed coating describes a process to encapsulate particles. The coating layer is applied in order to protect the core material from chemical reactions with the environment, to control the release of drugs or to mask bad taste. Depending on the application, the coating layer must fulfil various quality requirements, such as completeness, homogeneity and minimum layer thickness. The measurement of the coating layer thickness is therefore necessary in order to determine appropriate parameters for an optimal coating process. This, however, is difficult in the investigated core particle size range of 100 to 500 μm with a coating layer thickness of around 10 μm. Fluorescent imaging of sliced particles or imaging of optical slices using confocal laser scanning microscopy are possible ways to make the coating layer visible and to measure the coating layer thickness using image analysis techniques. This leads to detailed images of the coating layer and an accurate description of the coating layer thickness distribution, but is rather time consuming due to tedious sample preparation and long image acquisition times. Consequently only relatively few particles are measured and used to draw conclusions on the population. Other methods like measurement of the change of particle size using laser diffraction or assessment of the volume ratio of coating to core material usually only deliver the mean thickness and no information on completeness and homogeneity of the coating. In the first part of this thesis a quick method for coating thickness measurement was developed based on a dissolution test. Sodium chloride was used as a core material and maltodextrin DE21 was used as a coating material. When dissolved in deionised water, sodium chloride raises the conductivity in contrast to maltodextrin. Therefore, the measurement of conductivity can be used to assess the dissolution curve of the core material. The coating layer delays the dissolution of the core and by comparison with the dissolution curve of pure sodium chloride the coating thickness distribution can be assessed by deconvolution. It was shown that this method is well reproducible and delivers reliable results comparable to other methods. The method is fast, which enables the measurement of many samples with replicates and using appropriate sample division should provide a good representation of the population. The shape of the thickness distribution allows the quantification of the three aforementioned quality parameters. The method was therefore used in the second part of this thesis in order to investigate the coating process using design of experiments. The four factors spray rate, air temperature, air velocity and concentration of the coating solution were investigated using a central composite design of experiments. The dissolution method was used to assess the coating quality. The particle size distribution was measured in order to quantify the agglomeration rate and the mass of deposited coating material was assessed by quantifying a tracer colour in order to assess the efficiency of the process. Significant quadratic models were fitted to all response variables. These were successfully used to find a local optimum within the investigated parameter space which allowed the formation of an optimal coating layer within a short time frame. The results of the previous investigations showed that the thickness distribution can be well described by a Weibull distribution. Furthermore, it was possible to confirm effects that were previously described in the literature, i.e. that a low concentration of the coating solution leads to more homogeneous coating layers. In order to give a general description of the coating layer, a statistical model of the coating thickness distribution was developed in the third part of this thesis and verified by a Monte-Carlo simulation. The model reproduces the experimentally determined effect of the concentration of the coating solution qualitatively and is able to calculate the mean thickness distribution with given concentration, contact angle, sprayed mass and core particle and droplet size. Appropriate adjustments of these parameters lead to a good agreement between the model and measured thickness distributions of real experiments. It was concluded that predominant spray drying of small droplets and an increase of concentration of the remaining droplets due to pre drying negatively affects the homogeneity of the coating layer. It was further confirmed that the Weibull distribution can be used to describe the coating layer thickness in the investigated thickness range. The thickness distribution transitions from the Weibull distribution to a normal distribution as the coating becomes thicker. Thin coatings with defects can be described by a clinched Weibull distribution containing the uncoated area fraction as an offset.
  • Thumbnail Image
    Publication
    Measuring and modelling of soil water dynamics in two German landscapes
    (2018) Poltoradnev, Maksim; Streck, Thilo
    The soil water regime is focus of various disciplines including agricultural sciences, hydrology, weather forecast and climate modelling. As an inherent part of land surface exchange processes, the dynamics of soil water content (SWC) is simulated in distributed hydrological models and land surface models (LSM). The accuracy of the simulated SWC directly influences the simulation outcome and its performance. Biases in modelled temporal SWC dynamics and its spatial distribution lead to errors in evapotranspiration, runoff, cloud and precipitation simulations. The main objective of my thesis was to study the factors that control the SWC dynamics and its spatial variability. Long-term measurements from the soil moisture networks Kraichgau (KR) and Swabian Alb (SA) provided the data basis of this study. SWC was sensed based on the Time Domain Transmission (TDT) technique. In each region, 21 measuring locations were distributed across three spatial domains: an inner domain 3 km × 3 km (5 stations), a middle 9 km × 9 km (8 stations), and an outer domain 27 km × 27 km (8 stations). The sizes of the three domains correspond with typical grid sizes of coupled atmosphere-LSM models. All stations were mounted on cropped agricultural sites. Each station was equipped with a TDT sensor, installed 15 cm deep into the soil, a rain gauge and a remote transfer unit. After adjusting the sensor networks, an in-situ field calibration was performed to derive pedotransfer and site-specific calibrations for TDT soil moisture sensors. The chemical and physical analysis of soil samples collected at each station revealed that soil bulk density influences in both regions the TDT readings. Moreover, the pedotransfer calibrations included electrical conductivity in KR and silt fraction and organic nitrogen content on SA. These variables are relatively easy to measure. Accordingly, the pedotransfer calibrations derived in this study are a quick possibility to calibrate TDT sensors in areas with similar soil properties as in KR and SA. Nevertheless, the site-specific calibrations performed the best and were therefore used for further data analysis. In the second study, a three-year record of SWC and rainfall was evaluated. The response of the regional mean (theta) of SWC to a rain event was influenced by the seasonal water balance (SWB). In KR, the relation was more pronounced for positive SWB and less for neutral and negative SWB. On SA, where SWB was highly positive in all three years, the response of theta to rainfall was always strong. At the seasonal scale, the relationship between the spatial standard deviation of SWC (sigma) and theta was investigated through sigma-theta phase-space diagrams. The results show that with decreasing SWC sigma-theta data pairs are approaching sigma at the permanent wilting point (sigma-thetawp). With increasing SWC, in contrast, sigma-theta data pairs are moving towards sigma at saturation (sigma-thetas). These two points were termed anchor points. The sigma-theta relationships formed combinations of concave and convex hyperbolas reflecting the variability of soil texture and depending on sigma in relation to the anchor points. At the event scale, hysteresis in the sigma-theta was observed. Most sigma-theta clockwise hysteresis cases occurred at an intermediate and intermediate/wet state of SWC. Among the factors that trigger the initiation of a sigma-theta hysteretic loop, the present study revealed the following: rainstorms with spatially highly variable intensities (threshold rainfall intensity of 1.1 ± 0.6 mm and 2.9 ± 2.8 mm for KR and SA, respectively), preferential flow and, possibly, hysteresis in soil water retention curves. Based on these results, the following hypothesis was formulated: sigma-theta phase space diagrams are useful to test whether hydrological models or land surface models (LSMs) capture the realistic range of spatial soil water variability. The concept was tested with the Noah-MP LSM. Observations obtained from KR and SA soil moisture networks over a three-year period from 2010 to 2012 were used to build up the sigma-theta phase-space. The study included two different setups used to compute the hydraulic conductivity and the diffusivity: 1) the default setting: the Clapp and Hornberger approach, and 2) the van Genuchten-Mualem functions. The default model parameterization was stepwise substituted with site-specific rainfall, soil texture, leaf area index (LAI) and green vegetation fraction (GVF) data. The atmospheric forcing was obtained from eddy covariance stations located in the regions. Although the model matched observed temporal theta dynamics fairly well for the loess soils of KR, it performed poorly in the case of the shallow, clayey and stony soils of SA. The best match was achieved with the van Genuchten-Mualem functions and site-specific rainfall, soil texture, GVF and LAI. Nevertheless, the Noah-MP LSM failed to represent the spatial variability of SWC. In most cases, the simulated sigma-theta data points were located below the bottom edge of the envelope, which indicates that the model smooths spatial variability of soil moisture. This smoothing can be mainly attributed to missing topography and terrain information, inadequate representation of the spatial variability of soil texture and hydraulic parameters, and the model assumption of a uniform root distribution.
  • Thumbnail Image
    Publication
    Measuring and modelling resource use competition at the crop-soil-hedge interface on a hillside in Western Thailand
    (2015) Hussain, Khalid; Cadisch, Georg
    Thailand’s western uplands are facing severe soil loss and runoff problems due to intensive cultivation of cash crops for high food, feed, fiber, and fuel demand by an increasing population. Thus the Land Development Department and the International Board for Soil Research and Management in Thailand are promoting the use of soil conservation measures such as contour hedgerows, grass barriers and agroforestry systems based on fruit trees and annual crops. Although such measures have been shown to be effective in controlling soil erosion, farmers often are reluctant to adopt such practices as inclusion of trees reduces the cropped area and yields competition for available resources with crops. Hence, a better understanding of the underlying processes at the crop-soil-hedge interface is needed to design soil conservation systems that are more attractive for farmers. It was hypothesized that soil conservation systems with hedgerows and intercropping will induce spatial patterns of resource use which can be linked to competition while planting patterns affect canopy characteristics and light distributions. This study focused on the following objectives; (i) to improve understanding of competition at the crop soil hedge interface by combining stable isotope discrimination, electrical resistivity tomography and time domain reflectometry, (ii) to identify the effects of intercropping and hedgerows on maize biomass accumulation, nitrogen concentration and light use efficiency, (iii) to evaluate the competition between maize hedges at crop-soil-hedge interface, (iv) to explore possible mitigating options to cope with competition between hedge and maize by using a modelling approach. A field trial was laid out in randomized complete block design with three replicates at Queen Sirikit research farm, Ban Bo Wi village (13°28´ N and 99°15´ E), Suan Phueng District, Ratchaburi province in western Thailand with 20-25% slope magnitude. The experiment was established in 2009 while the research presented here was carried out during the 2010 and 2011 maize growing seasons. Six cropping treatments with following management practices were investigated: (T1, control) maize [Zea mays L.) monocrop, tillage, with fertilizer application (farmer’s practice; (T2) maize intercropped with chili (Capsicum annuum L. cv. Super Hot), tillage and fertilizer application; (T3) maize intercropped with chili, minimum tillage, fertilizer application, and Jack bean (Canavalia ensiformis (DC) L.) relay cropping; (T4) maize intercropped with chili, minimum tillage, fertilizer application, Jack bean relay cropped, and leucaena (Leucaena leucocephala (Lam.) de Wit) hedgerows; (T5) and (T6) as (T3) and (T4), respectively, but both without fertilizer application. Tillage was carried out manually by hoe to around 0-20 cm depth. Plots were 13 x 4 m. Fertilizer was applied to maize at a rate of 62, 22, 36 kg ha-1 of N, P, and K, respectively. Urea (N) application to maize was done in two splits as 31 kg ha-1 of N one month after sowing maize and another split of 31 kg ha-1 of N two months after sowing maize Chili received 92 kg N ha-1 at the time of transplanting and 92 kg ha-1 N as top dressing one month after transplanting. The impact of competition at the crop-soil-hedge interface was studied in 2011, two years after establishment of soil conservation measures, to exclude the establishment period of leucaena with a potentially weak impact on maize. At this time, highest above ground biomass (AGB) production of maize of 1364 g m-2 was witnessed in T2 being statistically different from all other treatments, except T4 and T3; while lowest above ground biomass production of 1034 g m-2 was observed in T5. In hedgerow treatments, maize rows planted distant to hedges produced 46% and 73% higher AGB than maize growing in rows close to the hedgerow (p≤0.0001) in T4 and T6, respectively. Similar effects were observed for plant height, grain nitrogen concentration and grain yield. Mean grain δ13C was significantly higher in T6 (-9.32‰) than in T4 (-10.49‰) and T1 (-10.55‰). Generally, higher δ13C signals mean higher water availability; however the higher δ13C signals in unfertilized T6 treatment imply that lack of nutrients may have led to increased δ13C values. Similarly in T4, δ13C signals were significantly higher in maize grains originating form rows planted close to hedges (10.33‰, p≤0.0001) than samples obtained from maize rows distant to hedges (-10.64‰). These results point out that competition at the crop-soil-hedge interface was driven by nutrient rather than water limitation. The electrical resistivity tomography (ERT) imaging further supported this finding showing that maize monocrop showed higher soil moisture depletion patterns than hedge intercrop with fertilizer (T4) treatment, while hedge intercrop without fertilization (T6) depleted soil moisture least. Light use efficiency (LUE) for maize above ground biomass production was statistically higher LUEAGB (1.56 g DM MJ-1) in maize and chilli intercrop (T2) than in maize sole cropping LUEAGB (1.23 g DM MJ-1). In T4 and T6 maize rows planted close to hedgerows had lower LUEAGB than rows distant to hedgerows. The land equivalent ratio showed that inclusion of hedgerows with fertilizer application in the intercropping treatment enhanced land utilization by 21%. The Water, Nutrient and Light Capture in Agroforestry (WaNuLCAS) model simulated AGB with R2= 0.83, RMSE=6.3, EF=0.82 and CD=1.4 during calibration while model validation also showed promising results with R2= 0. 76, P<0.001, RMSE=4.6 and EF=0.69. Simulations also pointed to major nutrient limitation between maize rows planted close to hedgerows. Simulations showed that negative impacts of hedgerows on crops can be reduced by managing fertilizer application in hedge adjacent maize rows leading to a successful application of agroforestry systems on a long-term basis not only for soil conservation but also for sustainable crop production in tropical uplands. The study figured out the scope of stable isotopic discrimination, ERT, light use efficiency and modelling approaches for evaluating resource use competition at crop-soil-hedge interface on hillside agriculture. The combination of isotopic discrimination and ERT measurements provided useful information for identification of cause-impact relationships. Spatial LUE patterns provided insights for canopy light harvest under various soil conservation options. Furthermore, light use data was also helpful in validation of WaNuLCAS model which did not only provide multiple options to figure out resource use competition at crop-soil-hedge interface but also allowed to test mitigation options for sustainable crop production in tropical uplands. Model scenarios showed that negative impacts of hedgerows on crops growing close to hedges can be reduced by applying minute additional doses of fertilizer only to the crop rows planted close to hedgerows, leading to a sustainable crop production along with soil conservation. Productivity evaluation of investigated cropping systems showed that inclusion of hedgerows and intercropping in tropical hillside agriculture is promising in enhancing crop production and thus can be adopted by farmers with yield advantage.
  • Thumbnail Image
    Publication
    Modelling of particulate matter and ammonia emissions from German agriculture
    (2016) Beletskaya, Olga; Zeddies, Jürgen
    Agricultural production comes along with numerous environmental effects, such as contribution to climate change, harmful to health emission impacts as well as eutrophication and acidification of soils and waters. Political regulations and environmental protection measures at the national and international level shall support development of sustainable agriculture. The intention of this work is to analyze the alterations of particulate matter, ammonia, and greenhouse gas losses from German agriculture arising due to adaptations in agricultural and environmental policy, and to find out efficient PM and NH3emission abatement options. To show, how certain economic and political conditions and their adjustment over time do affect amount of NH3, PM, and GHG released from agriculture and to evaluate emission mitigation options, economic-ecological static integer linear model, EFEM (Economic Farm Emission Model), has been developed. Following exogenous parameters have been integrated into the model: emission factors and the activities data stemming from FADN (Farm Accountancy Data Network) and census databases. In EFEM farm structure, production activities and extrapolation tool are represented in the system of interrelated modules for five farm types, i.e., arable farms, forage-growing, mixed and intensive livestock farms (one with the emphasises on pig husbandry and another one specializing in poultry production). The modelling is done for three German counties and each of them has focus regions, which are exemplary for important sources of PM, NH3, and GHG emissions in agriculture. Thus, following study regions have been chosen: Baden-Württemberg characterized by forage growing prevailing there, Lower Saxony marked by intensive livestock productions and Brandenburg due to its sandy arable sites at risk of erosion. The individual scenarios analyse abatement and financial efficiency of the adjustments of emission sources. Emission sources are adjusted in the framework of the emission relevant agricultural production practice, as exclusion of urea from mineral fertilization practices, switching from slurry to solid manure based livestock housing systems, introduction of crude protein reduced feeding by pigs and poultry, environmentally friendly slurry storage and land application, reduced tillage, and installation of exhaust air treatment systems in pig barns. Relatively efficient abatement of NH3 results from the exclusion of urea from fertilization practice at farms and in regions with higher land endowments, the switch from liquid to solid manure based housing system for cattle and the injection of liquid manure into the soil and covering manure storage with granulate at farms and in regions with comparatively high livestock density. Efficient PM and GHG emission reductions, mainly due to carbon sequestration, follows the introduction of the reduced tillage. Installation of Exhaust Air Treatment Systems, such as one-stage and multiple-stage chemical scrubbers, results in relatively expensive but more efficient measure for the reduction of both NH3 and PM losses. Net benefit, as the difference between reduced costs of damage for human health and environment and mitigation costs, gives the insight into the effect of abatement measure for the overall economy. Among all scenarios analysed in this study, the emission abatement options assuring maximal net benefits and emissions reduction are combined together and suggested as the abatement strategy at the farm and policy level. The scenario results are compared with national emission abatement ratios proposed by the European Commission for the years between 2020 and 2029. However, this efficiency of each individual abatement measure varies for study regions due to their individual conditions. Increasing attention to environmental problems at the regional and global level requires higher contribution of scientists from all over the world to the definition of pollution and emission abatement status. This study demonstrates the relevance of further investigation of PM and NH3 emissions in and from agriculture and of the ways to abate them.
  • Thumbnail Image
    Publication
    Simulating the impact of land use change on ecosystem functions in data-limited watersheds of Mountainous Mainland Southeast Asia
    (2015) Lippe, Melvin; Cadisch, Georg
    The presented PhD thesis deals with the development of new modelling approaches and application procedures to simulate the impact of land use change (LUC) on soil fertility, carbon sequestration and mitigation of soil erosion and sediment deposition under data-limited conditions, using three mountainous watersheds in Northern Thailand, Northern and North-western Vietnam as case study areas. The first study investigated if qualitative datasets derived during participatory processes can be used to parameterize the spatially-explicit, soil fertility-driven FALLOW (Forest, Agroforest, Low-value Landscape Or Wasteland?) model. Participatory evaluations with different stakeholder groups were conducted in a case study village of Northwest Vietnam to generate model input datasets. A local colour-based soil quality classification system was successfully integrated into the FALLOW soil module to test scenarios how current or improved crop management would impact the evolution of upland soil fertility levels. The scenario analysis suggested a masking effect of ongoing soil fertility decline by using fertilizers and hybrid crop varieties, indicating a resource overuse that becomes increasingly irreversible without external interventions. Simulations further suggested that the success rate of improved cropping management methods becomes less effective with increasing soil degradation levels and cannot fully restore initial soil fertility. The second case study examined the effects of LUC on the provisioning of long-term carbon sinks illustrated for a case study watershed in Northern Thailand. Based on land use history data, participatory appraisals and expert interviews, a scenario analysis was conducted with the Dyna-CLUE (Dynamic and Conversion of Land use Effects) model to simulate different LUC trajectories in 2009 to 2029. The scenario analysis demonstrated a strong influence of external factors such as cash crop demands and nature conservation strategies on the spatial evolution of land use patterns at watershed-scale. Coupling scenario-specific LUC maps with a carbon accounting procedure further revealed that depending on employed time-averaged input datasets, up to 1.7 Gg above-ground carbon (AGC) could be built-up by increasing reforestation or orchard areas until 2029. In contrast, a loss of 0.4 Gg in AGC stocks would occur, if current LUC trends would be continued until 2029. Coupled model computations further revealed that the uncertainty of estimated AGC stocks is larger than the expected LUC scenario effects as a function of employed AGC input dataset. The third case study examined the impact of land use change on soil erosion and sediment deposition patterns in a small watershed of mountainous Northern Vietnam using a newly developed dynamic and spatially-explicit erosion and sediment deposition model (ERODEP), which was further coupled with the LUCIA (Land Use Change Impact Assessment) model building on its hydrological and vegetation growth routines. Employing available field datasets for a period of four years, ERODEP-LUCIA simulated reasonably well soil erosion and sediment deposition patterns following the annual variations in land use and rainfall regimes. Output validation (i.e. Modelling Efficiency=EF) revealed satisfying to good simulation results, i.e. plot-scale soil loss under upland swiddening (EF: 0.60-0.86) and sediment delivery rates in monitored streamflow (EF: 0.44-0.93). Cumulative sediment deposition patterns in lowland paddy fields were simulated fairly well (EF: 0.66), but showed limitations in adequately predicting silt fractions along a spatial gradient in a lowland monitoring site. In conclusion, data-limited conditions are a common feature of many tropical environments such as Northern Thailand and Northern/North-western Vietnam. Environmental modellers, decision makers and stakeholders have to be aware of the trade-offs between model complexity, input demands, and output reliability. It is not necessarily the challenge of data-limitations, but rather the decision from the very beginning if the aim is to develop a new model tool or to use existing model structures to support environmental decision making. Future modelling-based investigations in data-limited areas should combine scientifically-based approaches with participatory procedures, because scientific assessment can support environmental policy making, but stakeholders’ decision will finally determine the provisioning of ecosystem functions in the long run. A generic assessment framework is proposed as synthesis of this study to employ dynamic and spatially-explicit models to examine the impact of LUC on ecosystem functions. The application of such a generic framework is especially useful in data-limited environments such as Mountainous Mainland Southeast Asia, as it not only provides guidance during the modelling process, but also supports the prioritisation of input data demands and reduces fieldwork needs to a minimum.