driven by environmental influences (Charles-Edwards, 1982). A genotype's ultimate performance is determined Seed yield in pea (Pisum sativum L.) is a physiologically complex by how it integrates genotype and environmental influtrait that is strongly influenced by both genotype and environment. Seed yield can be described in terms of its components, which include ences. The end result is seed yield, which has often been plant number per unit area, seed weight, and seed number. These described as the product of its components: number of yield components show interdependence and plasticity in response to plants per unit area, number of seeds per unit area environment; therefore, selection based on a single component is un-(number of pods per plant, number of seeds per pod), likely to succeed in increasing yield in breeding programs. To improve and mean seed weight (Moot and McNeil, 1995). These our understanding of the genetic basis of seed yield determination in yield components show interdependence or plasticity pea and to identify genetic loci involved, quantitative trait loci (QTL) (Wilson, 1987). For example, compensation is observed for yield per se, yield components (seed weight, seed number, and between the number of pods per plant and number of harvest index [HI]) and developmental traits (node of first flower [NFF], seeds per pod (Moot and McNeil, 1995), or between seed number of flowering nodes [NFN], total node number [TNN]) were number and seed weight (Sarawat et al., 1994). Since mapped. The QTL mapping was conducted using F 2 -derived families of a cross between Primo, a marrowfat cultivar, and OSU442-15, a selecting for a particular yield component may be inblue pea breeding line. Linkage maps containing 108 loci on 11 linkage effective as a means of increasing yield per se, alternate groups (LGs) were constructed for 227 families derived from this approaches have been proposed. Hedley and Ambrose cross. Traits were measured in three replicated field trials conducted (1985) suggested selecting for plants that produce a high, in New Zealand during the summers

Abstract. The effects of straw disposal by burning and incorporation on soil and crop nitrogen (N) supply, were investigated on two light textured soils in central (ADAS Gleadthorpe) and eastern England (Morley Research Centre) over the period 1984 to 1995. Nitrogen balance calculations showed that after 11 years of contrasting straw incorporation versus burn treatments, the cumulative N returns in straw were c. 570kg/ha at Gleadthorpe and c. 330 kg/ha at Morley However, these N returns via straw incorporation were not reflected in increased total soil N levels in autumn 1994. There were no differences (P > 0.05) between straw disposal treatments in autumn soil mineral N supply, readily mineralizable N or organic carbon. Similarly, there were no consistent differences between the treatments in terms of crop yield, crop N uptake or optimum fertilizer N rates. Fertilizer N applications of 200 kg N/ha/y increased topsoil organic carbon from 1.18 to 1.28% and total N content from 0.091 to 0.102% on the loamy sand textured soil at ADAS Gleadthorpe, but not at Morley. Previous fertilizer N applications increased the quantity of nitrate‐N leached in drainage water by c. 20 kg/ha at Gleadthorpe and c. 60 kg/ha at Morley overwinter 1994/95, and by 10–20 kg/ha at both sites overwinter 1995/96. There was some indication overwinter 1994/95 that straw incorporation reduced nitrate‐N leaching by 10–25 kg/ha, but there were no differences between treatments overwinter 1995/96.

Annual dry matter (DM) production and botanical composition of six dryland pasture combinations, grown under sheep grazing at Lincoln University for 5 years, are presented. In 4 years, lucerne produced the highest DM yields (13.1-18.5 t/ha/yr) through higher daily growth rates, compared with grass based pastures, particularly during periods of water stress in summer and autumn.

The effects of temperature, water and nitrogen on pasture production of an 8 year old 'Wana' cocksfoot pasture were quantified at Lincoln University, Canterbury. The maximum dry matter (DM) yield was 22.0 t/ha/yr when neither water nor N were limiting. Crude protein yield of +N pastures was 3.2-4.2 t/ha/yr compared with 1.0 t/ha in -N pastures. Metabolisable energy averaged 178*103 MJ ME/ha/yr for the +N pastures compared with 69*103 MJ ME/ha/yr for -N pastures. Seasonal differences in pasture production were explained in relation to thermal time with 7.0 kg DM/oCd for N fertilised pastures and 3.3 kg DM/oCd when no N was applied. During periods of water stress, relative yield decreased at a rate of 1.4%/mm when the soils critical limiting deficit of 78 mm was exceeded. Keywords: Dactylis glomerata, irrigation, nitrogen, thermal time.

The water use efficiencies (WUE) of a range of temperate pasture species were calculated from measurements on several different dryland and irrigated pastures in Canterbury. The annual WUE ranged from 6.7 kg DM/ ha/mm for a dryland cocksfoot pasture to 40 kg DM/ha/ mm for a dryland lucerne crop grown on a Wakanui silt loam soil. Keywords: Cichorium intybus, Dactylis glomerata, Lolium perenne, Medicago sativa, Trifolium ambiguum, T. michelianum, T. pratense, T. repens, T. subterraneum

Six dryland pastures were established at Lincoln University, Canterbury, New Zealand, in February 2002. Production and persistence of cocksfoot pastures established with subterranean, balansa, white or Caucasian clovers, and a perennial ryegrass-white clover control and a lucerne monoculture were monitored for nine years. Total annual dry-matter (10.0-18Á5 t DM ha À1 ) and sown legume yields from the lucerne monoculture exceeded those from the grass-based pastures in all but one year. The lowest lucerne yield (10 t ha À1 yr À1 ) occurred in Year 4, when spring snow caused ungrazed lucerne to lodge and senesce. Cocksfoot with subterranean clover was the most productive grassbased pasture. Yields were 8Á7-13Á0 t DM ha À1 annually. Subterranean clover yields were 2Á4-3Á7 t ha À1 in six of the nine years which represented 26-32% of total annual production. In all cocksfoot-based pastures, the contribution of sown pasture components decreased at a rate equivalent to 3Á3 AE 0Á05% per year (R 2 = 0Á83) and sown components accounted for 65% of total yield in Year 9. In contrast, sown components represented only 13% of total yield in the ryegrasswhite clover pastures in Year 9, and their contribution declined at 10Á1 AE 0Á9% per year (R 2 = 0Á94). By Year 9, 79% of the 6.6 t ha À1 produced from the ryegrasswhite clover pasture was from unsown species and 7% was dead material. For maximum production and persistence, dryland farmers on 450-780 mm yr À1 rainfall should grow lucerne or cocksfoot-subterranean clover pastures in preference to ryegrass and white clover. Inclusion of white clover as a secondary legume component to sub clover would offer opportunities to respond to unpredictable summer rainfall after sub clover has set seed.

No abstract

The pre- and post-grazing dry matter yields from six individual paddocks in a grazed lucerne rotation were compared at Ashley Dene, Lincoln University, over 5 years. The mean annual yield was ~12 000 kg DM/ha and unaffected by the rotation order in which paddocks were first grazed in spring. On an annual basis the post-graze residual was about 1000 kg DM/ha lower from Paddock 1 than Paddock 6. First grazing usually occurred with ewes and lambs in the third week of September when the pre-grazing height was 15-25 cm. Rapid spring growth of lucerne meant Paddock 6 was sometimes over 40 cm tall with a pregrazing mass of 4000 kg DM/ha upon first entry 4-5 weeks later. The grazing interval between the first and second rotation in spring was between 29 and 36 days. There was an average 1800 kg DM/ha more on offer in Paddock 6 than Paddock 1 in the first rotation. On average, the ewes and lambs consumed ~2100 kg DM/ ha at ~75% utilisation from each paddock in the first rotation. Weaned lambs offered similar pre-grazing yields consumed less herbage (~1700 kg DM/ha) at ~60% utilisation. A lower pre-graze target for lambs (2000-2500 kg DM/ha) is therefore recommended, particularly if there is no other class of stock available to consume the residual. This may require a quicker rotation with weaned lambs than the 34 days averaged in this study. For ewes and lambs the target of ~3000 kg DM/ha (~30-35 cm pre-grazing height) used for this study seemed appropriate rather than a fixed number of days or particular phenological stage. Keywords: alfalfa, lucerne, Medicago sativa L., rotational grazing, stock class, utilisation