Study site
The study site is situated at Aarhus University Foulum, Tjele, Denmark (9°35´E, 56°30´N, 48 m a.s.l.). Soils at the study site are classified as Typic Hapludult and are a sandy loam texture. The average proportions of clay, silt and sand in the surface layer (0-25 cm) were 8, 11 and 78%, respectively. The average soil pH was 6.5 and the average soil bulk density was 1.2 g cm-3. The climate was characterized as wet and temperate, with mild summers and cool to cold winters. Long-term meteorological records near the study site showed a mean annual air temperature of 7.8 degree (with moderate seasonal temperature variation), an annual precipitation of 740 mm, and an annual potential evapotranspiration of 600 mm. The study site has a history of long-term cultivation with the arable crops predominantly used across Northern Europe, such as winter rye, wheat, potato and spring barley, and has been annually fertilized with slurry.
Experimental design
The long-term field experiment was established in 2012, with the primary purpose of sustainably increasing biomass yield for biorefineries. The total land area of the study site is about 2.2 ha, which was divided into four main blocks for replicates. The buffer zones between adjacent blocks were at least 8 m wide. An incomplete split-plot design was selected to minimize the disturbance and border effects inherently associated with the diversified cropping systems, such as the shading effects caused by the difference in the height and maturity date between adjacent crop species. Thus, the plot size in each block was explicitly designed according to the species-specific characteristics. For example, some perennial grasses were planted in small plots (20 x 3 m), while some annual crops were cultivated in large plots (20 x 12 m or 20 x 21 m). The buffer zones between adjacent plots were at least 1 m. Multiple cropping systems were carefully selected with the purpose of increasing photosynthetically active crop coverage across the growing season. The cropping systems can be broadly divided into: (1) two continuous monocultures of annual crops (Table S1), i.e. maize (Zea mays L.) and triticale (Triticosecale); (2) Optimized crop rotations for a longer growing season and higher resource use efficiency including sugar beet (Beta vulgaris L.), hemp (Cannabis sativa L.), winter triticale (Triticosecale), maize (Zea mays L.) and winter rye (Secale cereale L.), with winter rape (Brassica napus) or grass-clover (Lolium perenne L.-Trifolium pratense L.) as second crops. The Optimized crop rotations were designed on a four-year basis, hence it was replicated four times with a different starting crop each year and the mean of the four Optimized crop rotations was considered in this study; (3) five intensively fertilized perennial grasses, i.e. tall fescue (Festuca arundinacea Schreb.), festulolium (Festulolium pabulare), reed canary (Phalaris arundinacea L.), cocksfoot (Dactylis glomerata D.) and a low-fertilized miscanthus (Miscanthus giganteus K.); and (4) two grass-legume mixtures without nitrogen fertilization (abbreviated to DLF and SLU according to their production origin): DLF was a mixture of Trifolium repens cv. Silvester, Festuca arundinacea cv. Tower, Lolium multiflorum cv.Humbi Phleum pratense cv. Winnetou, and Festuca arundinacea cv. Laura. SLU consisted of Trifolium repens, Trifolium hybridum, Galega orientalis, Phalaris arundinacea cv. Bamse, Festuca arundinacea cv. Hykor, Dactylis glomerata cv. Donata, and Medicago sativa.
The main annual crops were mostly sown in April and May, except triticale which was sown in September. The second annual crop − winter rye − was sown either in July (after triticale) or September/October (after maize), and grass-clover was undersown in triticale the following March. The perennial grasses were seeded in May 2012, and the miscanthus was planted as rhizomes. The seeding rates followed those for the crop under commercial conditions.
Nutrient (N-P-K) fertilization rates followed the Danish agro-legislation, with necessary adjustments for specific crops according to expert evaluation. To support grass regrowth, nutrients were applied after each harvest. For miscanthus, N, P and K were applied in the range of 60-150, 8-33, 26-117 kg ha-1 yr-1 across the five years, respectively. The other perennial grasses generally received high N, P and K fertilization rates, ranging from 300−500 kg N ha-1 yr-1, 26−72 kg P ha-1 yr-1 and 118−454 kg K ha-1 yr-1. Nutrient fertilization for annual crops varied highly between crops, and specific information on each crop can be found in Tables S1 and S2. Grass-legume mixtures only received base N fertilization of 30 kg N ha-1 at the first sowing date and then received no further N to promote the biological N fixation by legumes. To meet the nutrient demands, the grass-legume mixtures were fertilized with 14−30 kg P ha-1 yr-1 and 74−352 kg K ha-1 yr-1. Irrigation (20-50 mm yr-1 on average) was carried out with overhead sprinklers from June to August, depending on the weather conditions. Herbicides were applied following Danish agro-legislation.
Biomass yield, soil respiration, and soil extracellular enzyme activity are measured in this project.
