Abstract
Field survey and developed gardening trials under this study were implemented in Phulpur upazila (sub-district) of Mymensingh district during March 2003 to September 2004. The study endeavoured at examining the influence of technologies and utilization of resources on gardening output performances, and contrasting the contributions of traditional and developed gardening to household food security. A system dynamics model was built and simulated to distinguish the behavioural pattern of household food security system, and to test different policy options. An optimal crop plan was designed by using linear programing model to maximize the gardening output performance. The survey results revealed that the studied households had a homestead area accounting for around two-fifths of their owned area. About 68 percent of homestead area could be treated as potential for gardening. Only 30 percent of the potential area was essentially being utilized under traditional gardening, and the rest was remaining underutilized. The household potential garden area fairly increased linearly with the increase of household homestead area (R2 = 0.8628). The actual garden area increased non-linearly at a low rate with power curve fit (R2 0.3812). The regression models demonstrated that more than 86 percent of the gardening output performances obtained by the gardening systems (R2 = 0.8630), crop varieties (R2 = 0.8620) and garden irrigation conditions (R2 = 0.8649) could be explained in terms of the area, labour, irrigation and capital in conjunction with the aforesaid technologies. The effects due to developed gardening were as high as 277.5 percent on using irrigation water and as low as 4.1 percent on utilization of homestead area. The developed gardening increased input-use efficiencies by 127.1, 15.3, 34.8 and 13.8 percent for land, labour, irrigation and capital, respectively. It enhanced the household food security status to 0.94 from 0.92 with a weighted average increase of 2.2 percent. It reduced the number of food insecure household by 4.5 percent, and increased contribution to household consumption requirement by 5.3 percent along with decreasing household food purchase requirement by 3.9 percent. There were three leverage points identified by simulating the household food security system behaviour. The leverage points referred to the food insecure periods: lasting from October to December, 2nd week of February to 1st week of April and 2nd week of August to September. The combination of doubling the garden food production and cash transfer through cash for work activities was the best policy intervention, which augmented the food secure period from 5.25 to 11.5 months. The linear programing model predicted an optimum gardening output as high as 283,039 kcal, which accounted for increments of 181.3 and 23.4 percent over outputs of traditional and trial gardening, respectively.