Potted satsuma mandarin trees (Citrus unshiu Marc. cv. Okitsu Wase) budded onto trifoliate orange rootstocks were exposed to different air and soil temperatures in the greenhouse or growth chambers and then transferred to constant temperature growth chamber.
In Experiment I, the greenhouse air temperature was not controlled but ambient, while the pot soil temperatures were kept at below 15°C, or above 30°C or ambient. Four-year-old trees, exposed to these conditions for three months until December 1, were transferred to a growth chamber kept at 30°C. Trees exposed to soil temperature below 15°C bloomed profusely, whereas, those in pots kept at above 30°C produced much less flowers.
In Experiment II, the treatments. which lasted 2.5 months until December 16, combined air and soil temperatures of 15° and 30°C, followed by a constant 30°C in the growth chambers. These four-year-old trees kept at the air/soil temperatures of 30/15°C bore some flowers after they transferred to constant 30°C, but those exposed to 30/30°C were barren of flowers. Trees kept at 15/15°C bore more flowers than did those kept at 15/30°C.
In Experiment III, temperature treatments were the same as those in Experiment II except that they lasted three months until December 1. The trees were then defoliated before being transferred to a 25°C of growth chamber. Analysis for gibberellin-like activity in summer leaves and for carbohydrate, nitrogen, and free proline contents in spring leaves were conducted. Observation on the flower development revealed that as in Experiment II, air/soil temperatures of 15/15°C promote flowering, whereas, 30/30°C did not. Gibberellin-like activity was the lowest in the trees that bore the greatest number of flowers, and the highest in leaves of trees that barren of flowers. Starch content of leaves was higher at soil temperature of 15°C than it was at 30°C; sugar content was unaffected by air/soil temperatures. Nitrogen content in leaves was greater at a soil temperature of 30°C than it was at 15°C. The C/N ratio was higher at soil temperature of 15°C than it was at 30°C; it was unaffected by air temperatures. A positive correlation was found between proline content of leaves and the number of flowers produced.
Experiment IV conducted in 1989 was a replicate of Experiment III; the results of the two seasons were similar.