b = Govaerts and Häkkinen, 2006
Solar radiation and substrate
We used the nursery of the NPGRL (latitude 14.153, longitude 121,262), as a Semi-NH for germination testing. We used locations either exposed to solar radiation (‘sun’) - only lightly shaded in the fine screen house, or without direct exposure to solar radiation (‘shade’) - in an open-sided cabinet covered at the top also in the screen house. We sowed seeds in plastic trays (100x40x10 cm), using two types of substrate (clay loam soil and fine sand), and covered seeds with 5 mm of substrate. Two replicates of 200 seeds were used for each treatment combination. We recorded the temperature and relative humidity (RH), every 20 minutes for 55 days in sowing locations using data loggers (Tinytag View 2, Gemini Data Loggers, Chichester, UK). Trays were watered daily, emergent seedlings were recorded and removed weekly. The test was concluded after 55 days.
Simulated-natural environment (glass house compartments, Belgium)
Plant material
We studied germination responses in Simulated-NHs in relation to foliage-shading and seed burial-depth using two Musa species (total three taxa, Table 2). Seeds were selected from the collection of Bioversity International/KU Leuven (Leuven, Belgium) and were supplied for scientific use with a phytosanitary and origin certificate from China (Hainan), Guadeloupe, and Nigeria. Seeds were from open-pollinated accessions in living collections and were air shipped from source to Leuven as complete bunches, where they were extracted as described above, apart from bal106 which were provided as extracted and cleaned seeds transported to Leuven, Belgium. After ambient drying, seeds were placed in the refrigerator in paper bags, until 2019 when they were sealed in aluminum bags and stored in the refrigerator at approximately 6% moisture content, fresh weight basis. Viability was assessed prior to sowing (both in 2019 and 2020) using ER (method described by Kallowet al. 2020).
Foliage-shading and seed burial-depth
We selected a total of six Simulated-NHs for germination tests. Five were in three compartments of Meise Botanic Gardens glass house, Belgium (latitude 50.925, longitude 4.330), and one in the full-ground glass house of KU Leuven, Belgium (latitude 50.860, longitude 4.680). Simulated NHs were selected to represent various heating regimes in different compartments and levels of foliage-shading/exposure to solar radiation (Table 3). All compartments included living banana specimens.
Seeds were sown in square plastic pots (9x9x10 cm), at two burial-depths (1 cm and 7 cm from the surface), using potting compost (Peltracom, composition: 70 % white peat and 30 % black peat; pH: 5.5- 6.5; particle size: 0-10 mm). Two replicates of 30 seeds of each accession were sown in separate pots which were then also buried to be level with soil surface. Data loggers (Tinytag Transit 2 TG4080, Gemini Data Loggers, Chichester, UK) were buried at each location and burial-depth. Loggers were set to record temperatures every 20 minutes. Additionally in 2020, loggers were placed at the soil surface to record light intensity and surface temperature (HOBO Pendant MX2202, Onset, Cape Cod, Massachusetts, US). Germination was monitored weekly and emergent seedlings were recorded and removed. Seeds were sown in early March 2019, and again in early March 2020s, this is the end of the winter/start of the spring season in Belgium (mean outdoor temperature 6.4°C, climate-data.org). The experiment was concluded in March 2021.
Table 3 . Summary heating regimes of glass house compartments used, temperatures are the thermostat temperature below which artificial heating is instigated.