Development of a generic, model-based approach to optimize light distribution and productivity in strip-intercropping systems

摘要

Due to a growing world population, an extension of bioenergy production and the larger proportion of meat and dairy products in the human diet, with the latter particularly in India and China, the demand for agricultural products will further increase. Under decreasing resources and negative environmental impacts related to past intensification, more sustainable agricultural production systems need to be developed in order to meet the future demand for agricultural products.China, as the most populous nation with an enormous economic growth since the end of the 1970’s, plays a major role in global agricultural production. On a national level, agricultural production has to be increased by 35% during the next 20 years. However, land and water resources in China are very limited.With this in mind, the Sino-German International Research Training Group (IRTG) entitled ‘Modeling Material Flows and Production Systems for Sustainable Resource Use in Intensified Crop Production in the North China Plain’ was initiated by the Deutsche Forschungs-Gemeinschaft (DFG) and the Chinese Ministry of Education (MOE). The present doctoral thesis was embedded in the IRTG and focused, in particular, on exploring combinations of different crops produced on the same land at the same time, known as intercropping. In general, the higher productivity in intercropping, compared with monocropping, arises from the complementary use of resources (radiation, water, and nutrients) over space and time by crops that differ in physiology, morphology and phenology. The decisive question is how to optimize intercropping systems over space and time. To address this question, the present doctoral thesis combined field experiments with modeling approaches with the following aims: (i) to investigate the light availability on high temporal and spatial resolutions; (ii) to develop and validate a model that simulates the light availability for the smaller crop and accounts for the major aspects of cropping design; (iii) to determine the effect of the modified light availability on growth of maize and the smaller, shaded crop; (iv) to evaluate the plant growth model CROPGRO for its ability to simulate growth of the smaller, shaded crop; (v) to investigate the interactions between maize cultivar, cropping design and local growth conditions; and, (vi) to identify promising cropping designs and detect future research needs to increase the productivity of strip-intercropping systems.For this purpose, field experiments comprising of strip-intercropping with maize (Zea mays L.) and smaller vegetables, including bush bean (Phaseolus vulgaris L. var. nana), were carried out over three growing seasons from 2010-2012 in southwestern Germany and in the North China Plain. Growing the crops in strips facilitates mechanized management, addressing the ongoing decrease of intercropping in China due to labor scarcity in rural areas. The crop combination of maize, a tall C4-crop with erectophile leaves, and bush bean, a small, N-fixating C3-crop with a more horizontal leaf orientation, was chosen due to the large potential for a complementary resource use. Special emphasis was given on the competition for light as it plays a major role in this cropping system due to the large height differences between the crops. In this context, measurements of the photosynthetically active radiation (PAR) were conducted on high spatial (individual rows across the strip) and temporal resolutions (five-minute intervals) at the top of the bush bean canopy over a two-month co-growing period with maize. The collected data formed the basis of the simulation study towards investigating competition for light and its influence on plant growth with modeling approaches.Experimental results showed that maize yields increased in the border rows of the strip due to a higher lateral incoming radiation in years with a sufficient water supply. On average, maize yields calculated for strips consisting of 18 to four rows increased by 3 to 12% and 5 to 24% at the German and Chinese sites, respectively. Analysis of yield components revealed that yield increases in the border rows of the maize strip were mainly determined by a larger number of kernels per plant. On the other hand, shading by the taller adjacent maize induced considerable shade adaptations of bush bean, such as larger canopy dimensions and a substantially increased leaf area index due to thinner, larger leaves. These shade adaptations increased light interception, and indicated that bush bean could tolerate shading up to 30%, resulting in a total and pod dry matter similar to that of monocropped bush bean. These results suggested that there is a good potential for utilizing bush bean in strip-intercropping systems in combination with taller crops. However, higher shade levels (>40%) resulted in considerable decreases of total and pod dry matter.The high temporal and spatial resolution of the PAR measurements clearly revealed a highly heterogeneous diurnal distribution of PAR across the bush bean strip. The developed light model simulated this heterogeneity with a high accuracy under both clear and cloudy conditions. Comparison of simulated and observed hourly values of PAR across several rows within the strip of bush bean showed a root mean square error (RMSE) ranging between 47 and 87 μmol m-2 s-1 and a percent bias (PBIAS) ranging between -3.4 and 10.0%. Furthermore, the model reasonably captured the influence of different widths of the bush bean strip, strip orientations and maize canopy architecture (height, leaf area index, and leaf angle distributions). Simulations run for different latitudes and sky conditions, including different strips widths, maize canopy heights and leaf area indices (LAI), indicate that: (i) increasing the strip width might only reduce shading in the border rows of the smaller crop at lower latitudes under a high fraction of direct radiation; (ii) at higher latitudes, the selection of a maize cultivar with reduced height and LAI are suitable options to increase the light availability for the smaller crop.The present doctoral thesis presents the first approach to use the monocrop plant growth model CROPGRO to simulate growth of a legume crop grown in an intercropping system. The CROPGRO model was chosen because it provides an hourly simulation of leaf-level photosynthesis, and algorithms that account for the effects of radiation intensity on canopy dimensions and specific leaf area. CROPGRO, calibrated on data of monocropped bush bean, captured, quite well, the effects of the strongly reduced radiation on leaf area, and total and pod dry matter in the most shaded bush bean row. This indicated the models’ applicability on other intercropping systems exhibiting high levels of shading. Under a lower level of shading, cultivar and ecotype parameters had to be calibrated individually for a respective row within the bush bean strip to achieve a high accuracy of the simulations. Model simulations aided in explaining the effects arising from different shares of direct and diffuse radiation on canopy photosynthesis. This is a very important point to be further explored as diffuse radiation remains a part of light distribution and photosynthesis hardly studied in general; and, in particular, becomes more important with the increasing impact of shading.The simulation of the light availability, plant growth and yield formation within the strip of maize can be handled in a similar way as described for the smaller crop, bush bean. Modifications of the light model and a suitable plant growth model are presented and discussed.In conclusion, the main outcomes of this thesis indicate that the selection of cultivars adapted to the modified light environment have the largest potential to increase the productivity of strip-intercropped maize and bush bean. The most important characteristics of suitable maize cultivars include: (i) a high potential of kernel set; (ii) a higher water stress tolerance; and, (iii) reduced canopy height and LAI. The importance given to each of the components would subsequently be determined by the local weather and management conditions and the shade tolerance of the neighboring crop. On the other hand, to optimize yields of the smaller shaded crop, we present two options: (i) to modify the co-growing period of the intercrops temporarily to alleviate light competition during shade-sensitive growth stages; and, (ii) to modify the cropping design spatially and/or select different maize cultivars to reduce shading to the tolerated degree during the respective growth stage of the smaller crop. When the shade tolerance during the respective growth stages is determined, the light model developed can be used to optimize the cropping system temporarily and spatially.In this thesis, a promising approach, which combines a specific light partitioning model with process-oriented monocropping plant growth models, was developed. All models included in the approach can be applied at any location, and their generic nature also facilitates the integration of other crops. These attributes present a highly valuable contribution to intercropping research as their future optimization will depend strongly on the efficiency of the research efforts given: (i) the complexity of the underlying processes that determine the productivity; and, (ii) the minor share of time and money invested in intercropping research. Intercropping research has to prevent reinventing the wheel by identifying aspects in common with and already studied in monocropping systems and focus on aspects particularly inherent to intercropping systems.