Study site and transplantation experiments
In low-flowering years, few or no tillers flower in the field, and even in high-flowering years only a minority (<20%) of all tillers on each plant flower (Rees et al., 2002). To improve our chances of having both flowering and vegetative samples, we took leaf samples from unmanipulated plants in the field, and from manipulated plants that were transplanted to lower-elevation (warmer) or higher-elevation (colder) sites to make flowering, respectively, more or less likely (Table 1). From 2016 to 2018, unmanipulated C. pallens plants at the main Mt Hutt site at 1070 m elevation (43⁰32ʹ S, 171⁰ 33ʹ E) were selected with leaf samples taken from three marked tillers on each of 10 marked plants. Only large (full-sized) tillers were selected and tagged for sampling, to avoid smaller tillers that might have been young and reproductively immature, but it was not possible to tell visually which tillers might be going to flower. None of the tagged tillers at the control plot flowered in 2017 and 2018, but some did in 2019 (Table 1).
Several transplantation experiments to higher or lower altitudes were carried out to manipulate flowering in C. pallens . Ten separate plants from the vicinity of the control plot were moved to near sea-level at the University of Canterbury (UC, Christchurch; 43° 31ʹ S, 172° 35ʹ E, 15 m above sea level) for the inductive summer period in December 2016 to March 2017 (17Hot transplants) and then transplanted back to the 1070 m site in March. Leaf samples from tagged tillers (both transplants and control plants) were collected at different times throughout the year (January 2017, March 2017, October 2017 and January 2018; between 11:00 am and 2:00 pm) and were used for the gene expression studies. Leaf samples collected during the inductive summer period (January 2017) were also used for the transcriptomic analysis as described below. Two other sets of 10 plants each were permanently transplanted in 2015 from the control site at 1070 m to different altitudes: to 1520 m on Mt Hutt (16Cool) and to UC (16Hot). Leaf samples were collected at different seasonal time-points (between 11:00 am and 2:00 pm), including summer (January 2016), autumn (March 2016), spring (October 2016) and autumn (post flowering; March 2017).
In all the experiments, all leaf samples were collected and put directly into vials on dry ice within 20 seconds and stored at -80 ºC until further analysis. The tagged tillers and the leaf samples from which they were taken were correspondingly labelled and subsequent tiller behaviour (flowering or not) was recorded in the following year. The stored leaf samples could then be identified as being from tillers that subsequently flowered or remained vegetative, and suitable samples were selected for downstream analysis.