Intraspecific variation in seed dormancy and germination among populations of Stellaria media and Galium aparine: a modelling approach to investigate the effects of environmental and management factors

Abstract

Weeds are a substantial threat to arable crop production, reducing crop yields and/or reducing crop quality and market value. Worldwide expenditure on herbicides outstrips that on either insecticides or fungicides. However falling gram prices are causing farmers to reconsider the economics of herbicide use and public concern about countryside conservation and demand for organically produced food is generating increased pressure to minimise pesticide use. With increasing interest in more integrated approaches to weed control there is a need to better understand the ecology and biology of weed species. This thesis therefore aims to describe aspects of tire ecology of two common winter annual weed species, Stellaria media and Galium aparine, in order to support the development of more sustainable weed control measures in arable crops. As effective broad-leaved weed control tends to be associated with early removal of weed seedlings, this thesis adopts a modelling approach to investigate tire potential benefits derived from predications of weed seedling emergence

Seedling emergence can be divided into three stages: seed domrancy, seed germination and seedling pre-enrergence growth. As seed dormancy and germination are particularly variable this thesis focuses on these key processes. Models for the timing and extent of germination for S. media and G.aparine are developed and variability is addressed by investigating the extent of intraspecific variation between populations and according to maternal effects. Intraspecific variation between populations was investigated using contrasting populations selected from a large-scale screening.

The seed dormancy and germination models were developed following a review of existing models and identification of data requirements. This initiated field experiments to quantify seasonal changes in S. media seed dormancy and laboratory experiments to quantify the combined effects of temperature and water potential on seed gemination for S. media and G. aparine. Also according to the data requirements identified by review, further laboratory experiments were conducted to analyse the effects of light and nitrate on S. media germination. For G. aparine. additional data were derived, as required, from published literature.

For S. media significant variation between populations was recorded in initial and seasonal patterns of seed dormancy, and in the timing and extent of seed germination according to temperature, water potential and light. For G. aparine significant variation was recorded between populations for initial seed dormancy and seed gemination according to temperature and water potential.

In general, for both S. media and G. aparine, seed dormancy was released in autumn and induced in early summer. Dormancy induction in the summer for S. media was complete but for G. aparine,restricted germination to a narrower range of temperatures. Optimum temperatures for gemination of S. media tended to 20 °C and for G. aparine to 10 °C (range assessed 5 to 30 °C). Germination extent decreased and the time to gemination increased as water potential decreased, but G. aparine was relatively more resilient to reduced water potential over the range 0 to -0.4 MPa, at optimal and sub- optimal temperatures (range assessed 0 to -0.8 MPa).

The timing and extent of germination in the soil seedbank was modelled in terms of the factors influencing dormancy (soil temperature) and germination (temperature and water potential). The dormancy models were developed from existing models for summer annual species. The germination models were developed independently following examination and assessment of existing hydrothermal time models. Both models were developed in a stochastic framework and parameters were estimated separately for the different populations.

Maternal effects were not included in the model, but experiments showed this additional variation would also significantly impact on model application. This was illustrated using seeds produced following application of the herbicide, fluroxypyr, at a range of reduced rates. For S. media the effect of fluroxypyr application on subsequent seed germination varied between populations, and higher rates reduced the extent of seed germination for two of the three populations. For G. aparine effects similarly varied between populations, though the higher rates of fluroxypyr application consistently reduced the extent of seed germination and lengthened the timing of germination. For S. media additional experiments investigating the effect of nitrogen environment during seed production showed no significant effects on subsequent seed germination.

Variability between populations and variability between seeds produced in different maternal environments complicates tire task of predicting tire extent and timing of S. media and G. aparine germination. However these models do serve a useful role as a research tool in summarising current understanding of how different environmental and management factors may effect weed seed dormancy and germination.