The grain legume grass pea (Lathyrus sativus L.) is adapted to drought-prone environments, but the extent and mechanisms of its tolerance are not well understood. In a pot experiment, water deficit was imposed on plants by withholding water from first flowering until predawn leaf water potential (LWP pd ) was )3.12 MPa. Water deficit reduced dry matter, seed yield, harvest index and water use efficiency by 60 %, 87 %, 67 % and 75 %, respectively, when compared with the controls. Flower production stopped when LWP pd fell to )1.8 MPa. At LWP pd )1.5 MPa, only 25 % of flowers resulted in filled pods (compared with 95 % filled pods in the control) with the rest aborted as flowers (48 %) or pods (27 %). Filled pods had more aborted ovules than controls, resulting in 29 % less seeds per pod. Water deficit reduced pollen viability, germination and the number of pollen tubes reaching the ovary by 13 %, 25 % and 31 %, respectively. Emergence from seeds produced from water-deficient plants was 21 % less than controls, but seedling shoot dry mass was 18 % higher, in accordance with the 19 % higher seed mass. The sensitivity of flowering to drought limited pod numbers but enabled plants to retain existing pods and develop near-normal seeds with low b-N-oxalyl-l-a-b-diaminopropionic acid toxin concentrations. This trait is useful for farming systems reliant on harvested seed for the next crop and in cases where seed size influences the value of the product.

No abstract

No abstract

Many rural households in the developing world are subject to chronic seasonal food insecurity. Often cited among coping strategies is foraging for wild foods. This study was to understand the role of wild food in household food security in Timor-Leste. Information on wild food use comes from three sources: a) a longitudinal study of food consumption among 14 subsistence farmer households across four districts in 2006-2007; b) a survey in 2011 of 1,800 farmer households across all 13 districts in Timor-Leste; and c) a survey of 64 households from eight community seed groups in three districts in 2013. The consumption of wild food fluctuated widely across the year with consumption much reduced in the wet season (December -April) compared to the dry season (May -November). Wild food use in a normal year -2006-2007 and a fooddeficit year -2011, characterized by a severe hungry season, were compared. In the normal year the maize grain store was exhausted in 50 % of households (expressed as cumulative percentage of 'at risk' households) during the month of August and the percent of households consuming wild food was 9.2 %. By contrast in the food deficit year, maize grain stores were typically exhausted 2 months earlier than in a normal year, and 50 % of all interviewed households were foraging for wild food by May in four districts. Clearly in the food deficit season wild food foraging was dramatically increased compared to a normal year, particularly by poor at-risk households. The early depletion of grains store stocks under food insecurity resulted in greater coincidence of the period of wild food foraging with the hungry season than in a normal year. In two surveys there were clear differences between districts in the extent of wild food foraging. The most widely eaten wild plant foods in Timor-Leste are lesser yam (Dioscorea esculenta considered wild in Timor-Leste) followed by elephant's foot yam (Amorphophallus paeoniifolius) and bitter bean (Phaseolus lunatus). Looking to the future in Timor-Leste, the wild food resource will continue to provide an important food buffer in deficit years especially to poor at-risk households, increasing their resilience and reducing their vulnerability.