It was found that areas with larger-sized landholdings have
more trees/ha (13.1 trees/ha) than areas with smaller-size
ones (10.9 trees/ha) (Table 4), which is a significant difference (|p| B 0.05) according to Mann–Whitney U test.
Although the association between landholding size and tree
density is relatively weak, the mean number of trees/ha is
modestly higher in cells having large-sized landholdings
and somewhat lower in cells with small-sized holdings
(Table 4). It was also found that only one-third of all cells
with small-sized holdings have high mean tree densities,
whereas one-half of cells with large-sized holdings have
high densities (Table 5).This finding is in keeping with an
earlier study in Northeast Thailand by Vityakon et al.
(1996) which found that tree densities were higher on
larger-sized farms. This appears to be a common pattern in
agroforestry systems on a global basis. Thus, a meta-analysis of 32 empirical studies of factors influencing adoption
of agroforestry in the tropics, found that almost twice as
many studies reported a positive association between plot
size and tree planting than reported a negative association
(Pattanayak et al. 2003). The positive association between
farm size and tree planting may reflect the fact that farmers
with larger plots simply have more space available for trees
and may also have a greater capability to absorb the risks
associated with growing trees (Sood and Mitchell 2009).
However, in central India it was found that smaller farms
had much higher tree densities in paddy fields than larger
farms, reflecting the fact that a great deal of labor was
required to prune the canopies and roots of the Acacia
nilotica trees, making it difficult for farmers with larger
farms to maintain high densities (Viswanath et al. 2000). In
Northeast Thailand, however, farmers expend little time or
effort in managing trees in their paddies.
352 Environmental Management (2014) 53:343–356
123