Phenological development and hardseededness of subterranean clover : A thesis submitted in partial fulfilment of the requirements for the Degree of Doctor of Philosophy at Lincoln University

Sowing an appropriate cultivar of subterranean (sub) clover is important for successful dryland farming systems. To be economically feasible, sub clover cultivars must be competitive and accumulate biomass in the initial vegetative phases for animal production. As an annual legume, the sub clover cultivars also need to produce a sufficient number of inflorescences and viable seeds for regeneration in the following season and following years. Quantification of the life cycle of sub clover is essential for its successful use in New Zealand pastoral systems. This thesis quantifies the effects of environment (e. g. temperature and photoperiod) on sub clover vegetative and reproductive development, including time of flowering, seed set and hardseededness. A reanalysis of published literature for sub clover cultivars grown in different latitudes across Oceania showed that flowering ocurred after 44 to 271 days, or 628 to 2600oCd (Tb=0). The mean thermal time requirement to flowering (TT50F) for New Zealand was 1551oCd and for the Australian datasets it was 1259oCd. Sub clover cultivars showed an extensive range of flowering time, up to 78 days (or ~1800oCd) in Australia with the strong seasonality due to the photoperiod changes. The Australian dataset allowed estimation of the rate to TT50F under a decreasing photoperiod. It was 831oCd/hour and 977oCd/hour for “early” and “late” cultivar groups, respectively. There was a limited dataset to compare cultivars under several sowing dates in New Zealand conditions. In a laboratory experiment, cardinal temperatures and thermal time requirements for germination of four new introduced cultivars were quantified. For all four the base temperature (Tb) was equal to zero. The optimum temperature (Topt) was 16oC for ‘Monti’ and ‘Narrikup’ but 22oC for ‘Antas’ and ‘Denmark’. The maximum temperature (Tmax) was 36oC for ‘Antas’ and Denmark and 39oC for Monti and Narrikup. The thermal time requirement for 50% germination was ~36oCd. The phenological development of sub clover cultivars was quantified under both controlled environment and field conditions. The sowing dates provided a contrasting environment (E) which enabled to quantify unifying relationships for both phenological and hardseededness responses. For the phenological stages, E explained 75 to 90% of total variability. The duration of vegetative development phases, from emergence (V1) to runner initiation (VR), was more influenced by environment (sowing dates) than by genotypes (cultivars). A common thermal time requirement for emergence of ~102oCd was found for all cultivars. This was followed by ~269oCd for first trifoliate-leaf appearance (V3) with a constant phyllochron ~52oCd/leaf from July to February. During autumn (March and May sowing dates) the phyllochron increased by ~30% in all cultivars. At approximately 400oCd (5 to 6 trifoliate leaves) the rate of leaf appearance changed from linear to exponential due to the appearance of the first runner and secondary leaf production.The minimum thermal time requirement (and number of days) for all cultivars to complete their life cycle (V0-R11) was 1434±125.0 oCd (123±6.0 days) for ‘Denmark’; 1379±72.0oCd (134±4.6 days) for ‘Denmark’; 1407 ± 129.0oCd for ‘Leura’; 1269±37. oCd for ‘Monti’; 1353±44.0oCd for ‘Narrikup’ and 1320±82.8oCd for ‘Woogenellup’. The differences in duration of the life cycle were explained by changes in individual phenophases in response to genotype and sowing dates. The runner extension to floral bud phase (VR-R1) was the longest within the sub clover life cycle. This represents the window appropriate for animals to graze the plants without compromising flowering and seed set.The photoperiod tested ranged from 10.0 to 16.5 hours. For increasing photoperiods, thermal time to flowering was constant at an average of 1033± 62oCd. In contrast, for decreasing photoperiods there was an increase in the time to flowering from 1041±190oCd at 13 hours to 2409±88 oCd at 16.5 hours photoperiod. The differences between cultivar groups were evident when sowing occurred in decreasing photoperiod sowing dates.The estimated rate to TT50F under the decreasing photoperiod was 458oCd/hour and 545oCd/hour for “early” and “late” cultivar groups, respectively. The base photoperiod to TT50F determined for the “early” cultivars was 11.5 hours and 12.2 hours for the “late” cultivars. Seed yield was mainly affected by sowing date. For the ungrazed monoculture subterranean clover plots, the total mean seed yield ranged from 180 kg/ha from a March sowing (‘Woogenellup’) to 2500 kg/ha for a September sowing (‘Narrikup’), when the greatest yields were observed across all cultivars. For hardseededness related variables (HSmax and HSbreak) the effect of genotype (G) was greater than for crop development and ranged from 38% to 41%. The G effect on hardseededness was characterised by the presence of soft and hard cultivars. HSmax observed in ‘Antas’ was an average of 55% from September to April sowing dates. This contrasted with the high values measured in ‘Monti’ (86%) and ‘Denmark’ (80%) which showed less difference among sowing dates. E represented 79% of the variability explaining the percentage normal seedlings, which decreased as the sowing date advanced from 87% in June (S1) to 60% in February (S6).The biochemical components of seed coat that explained the G effect were cellulose, cutin, suberin and esters. The Fourier transform infrared microspectroscopy (FTIR) and Attenuated Total Reflectance (ATR) analysis of the seed coat for these substances might assist phenotyping for hardseededness in future selection and breeding programs for sub clover and other legume pastures species. At an agronomic level, the insights from this research can improve management and usage of sub clover in New Zealand’s agricultural systems. At the research level, such understanding shows physiological and biochemical mechanisms to inform breeding and improve predictive models of pasture production.