ABSTRACT
Summary Growth and persistence of many tropical pasture legumes can be limited by flooding or waterlogging. Selection of tolerant species is important for pasture development in the humid lowland tropics. The aim of this project was to compare flooding tolerance of seven tropical pasture legumes and to determine whether the pot technique we developed was reproducible and relevant to the field situation. Plants were grown in pots and at the start of flowering were flooded to 15 cm above soil level. Water temperature and oxygen content were monitored regularly. Sets of four uniform pots were removed from flooding after 7, 14, or 21 days. After 7 days of recovery the dry weight of tops, roots, and nodules of the flooded and matched sets of nonflooded controls were compared. The ratio of dry weight flooded: dry weight control was an index of tolerance. The species were ranked by index values as follows: Macroptilium lathyroides (1.04 and 0.87), Desmodium intortum (0.85), Pueraria phaseoloides (0.76), Stylosanthes guianensis (0.62), M. atropurpureum (0.44), Centrosema pubescens (0.37), and D. uncinatum (0.34). Root growth was more affected by flooding than shoot growth. Nodulation was drastically reduced in all species except M. lathyroides. The adaptation of M. lathyroides was related to a rapid production of adventitious roots from the submerged stem and rapid renodulation of these and other roots. D. intortum also had rapid adventitious root production but no renodulation. The screening technique appeared to yield reproducible results, and rankings were related to field observations.
