and below-ground biomass of undersown crops in Finland demonstrated the capac- ity of legumes to fix N for use by sub- sequent spring cereals and the ability of grasses to take up soil mineral N during sea sons where there is a risk of increased N leaching. Thus, the first objective of the study was fulfilled.
The effects of undersowing were very spe- cies dependent. WR and winter wheat de- creased main crop yields greatly and were unsatisfactory regarding take-up of soil N in autumn. Consequently IR, which ef- fectively reduced soil NO3-N, is recom-
mended when there is a high risk of leach- ing during autumn. Reduced competition with the main crop was registered for per- ennial grasses, among which timothy was the most reliable in terms of growth and effectiveness at catching N in the subse- quent spring. The second objective was re- alized, although differences among leg- umes were greater than among grasses. It seems, however, that white clover slightly increased N leaching risk in the following spring, as did annual black medic, but the risk was lowered through use of red clover. Concerning the third objective, the net ef- fects of annually repeated undersowing of legumes on cereal grain yield were posi- tive, and that of WR negative. WR seemed not to release significant amounts of N for use by a cereal crop during annual under- sowing, which may result from low N-% and that a six-year period was possibly too short for establishing cumulative effects. However, as N-% of grasses varied among species and according to conditions, in- creased N release recorded for long-term use elsewhere, according to the references provided, has to be taken into account in annually repeated undersowing of grasses.
Crop husbandry practices affected the out- come of the undersowing system in many ways, indicating the importance of the fourth objective. Broadcasting with seed- covering equipment was a reliable meth- od to establish an undersown crop. Only moderate seeding rates were needed for grasses to avoid substantial yield decrease of the main crop, but fairly high seeding rates could be used to ensure high N yield of a clover cover crop. Biomass production of undersown legumes decreased with in- creasing N fertilizer application, thus de- creasing the residual N effect. Late autumn incorporation, which according to the lit- erature reviewed, has earlier been report- ed to be associated with N-rich residues, worked well in the context of undersow- ing. Moreover, overwintering of timothy enhanced absorption of residual N in the soil, although the effects during the subse- quent summer would require other studies. It was surmised that various factors influ- ence the implementation of undersowing. At first a strategic decision has to be tak- en ifthe target is to fix atmospheric N or catch it from the soil, or if a more diver- sified cropping and increased soil fertility are the main goals. When substantial neg- ative effects of N leaching are expected, a mixture of IR and overwintering timothy is best as an undersown crop, but fertiliz- er N costs can be reduced by undersowing clovers. The benefits from mixtures of leg- umes and non-legumes lie in between, and red clover with timothy, for example, can be recommended because it results in rela- tively reduced competition and can replace N fertilizer to some extent without increas- ing the risk of N leaching substantially. The model for an adaptive undersowing system highlighted the principles involved in using the system based on the base sit- uation and goals, indicating the possibil-
ities to realize the aims of the fifth ob- jective. However, implementation of the system in practice requires considerable thought about conditions, realized cov- er crop growth and improved soil fertility
over the long term. As for the undersow- ing system itself, the model needs contin- uous adjustment as more experience of un- dersowing systems is gained and more data are generated.
Acknowledgements
This work was carried out at the Plant Pro- duction unit of MTT Agrifood Research Finland. The experiments were undertaken at Jokioinen, Laukaa, Pälkäne and Vihti. I want to express my deep gratitude to my initial supervisor Emeritus-Professor Timo Mela for suggesting the research topic and for providing me the opportunity to car- ry out this research, which also included important, but nowadays rare, long-term field trials.
I wish to express my sincere thanks and gratitude to my supervisor Professor Pirjo Peltonen-Sainio for offering the opportu- nity to prepare this thesis and for provid- ing supportive mentoring. Pirjo patiently reminded me year after year of the possi- bility to prepare a thesis and finally it was done. I also wish to thank Professor Aarne Kurppa for his positive attitude to remote working, which crucially influenced com- pleting the thesis.
My special thanks are also due to Profes- sor Pirjo Mäkelä for her guidance during preparation of the thesis and for under- standingly recalling me of what is includ- ed in the studies.
I wish to warmly thank Ari Lahti, who over the years of experimentation was re- sponsible for the practical work. Further- more, I thank Ari and Aino & sons, for providing me with accommodation after the longest working days and for the nice breakfast moments.
I express my gratitude to Hannu Mikko- la, Mauri Räkköläinen and Martti Vuori- nen for taking care of the trials at experi- mental sites. Christian Eriksson is warmly thanked for competent tutoring in statis- tics. Dos. Jonathan Robinson is thanked for linguistic revision. Mikko Vanha-Pert- tula, who was employed in the Farming Systems project, I also thank warmly. Professor Tor Arvid Breland, of the Nor- wegian University, and Emeritus-Profes- sor Börje Lindén, of the Swedish Univer- sity are gratefully acknowledged for their valuable comments and technical revision of the thesis.
My warmest thanks are due to all col- leagues and friends at MTT and especially those in Crop Science in Jokioinen. I wish to mention Markku Kontturi in particular, who persuaded me into a scientific career and whose knowledge and thorough atti- tude to research I highly appreciate. For all of you others, I hope you will understand the list would be too long here.
The exceptionally good climate in our work community has helped me pursue the profession of scientist, with all its pleas- ures and sorrows. Workmates have also been a channel for bringing worth to my creative writing as a counterbalance and support to the scientific input, and many others at MTT have been a fine audience. Enquiring about my Ph.D. work by col- leagues over the years has decreased and finally ceased, but still in the year of com- pleting it your sympathy was strong.
This study was initially supported by the Ministry of Agriculture and Forestry and MTT Agrifood Research Finland. Writing the thesis was financially supported by the Finnish Cultural Foundation, and pub- lishing was supported by Agronomiliitto. My dearest thanks I owe to my wife Auli and our sons Eetu and Otto. As to Ph.
D. work, I owe thanks that they have re- mained just as always, wonderful in them- selves. Maybe I have got a bit more space for this work because the boys have grown up and don’t play with me anymore. This is a pity, of course.
After all, writing the thesis was just a small piece in continuous process of learning about life!