Using seminatural and simulated habitats for seed germination ecology of banana wild relatives.

Ecologically meaningful seed germination experiments are constrained by access to seeds and relevant environments for testing at the same time. This is particularly the case when research is carried out far from the native area of the studied species.Here, we demonstrate an alternative-the use of glasshouses in botanic gardens as simulated-natural habitats to extend the ecological interpretation of germination studies. Our focal taxa were banana crop wild relatives (Musa acuminata subsp. burmannica, Musa acuminata subsp. siamea, and Musa balbisiana), native to tropical and subtropical South-East Asia. Tests were carried out in Belgium, where we performed germination tests in relation to foliage-shading/exposure to solar radiation and seed burial depth, as well as seed survival and dormancy release in the soil. We calibrated the interpretation of these studies by also conducting an experiment in a seminatural habitat in a species native range (M. balbisiana-Los Baños, the Philippines), where we tested germination responses to exposure to sun/shade. Using temperature data loggers, we determined temperature dynamics suitable for germination in both these settings.In these seminatural and simulated-natural habitats, seeds germinated in response to exposure to direct solar radiation. Seed burial depth had a significant but marginal effect by comparison, even when seeds were buried to 7 cm in the soil. Temperatures at sun-exposed compared with shaded environments differed by only a few degrees Celsius. Maximum temperature of the period prior to germination was the most significant contributor to germination responses and germination increased linearly above a threshold of 23℃ to the maximum temperature in the soil (in simulated-natural habitats) of 35℃.Glasshouses can provide useful environments to aid interpretation of seed germination responses to environmental niches.