Abstract:

Among the three Globba species in this study, Globba barthei and G. schomburgkii propagated only by bulbils and with no seed set while G. racemosa propagated by both seeds and bulbils under natural conditions. The aim of this study was to understand the effect of nutrition supply on the relative levels of sexual and asexual reproduction of these three species in both natural and artificial conditions and to determine whether there is compensatory mechanism that enhances sexual reproduction when asexual reproduction is constrained and vise versa.
For G. barthei and G. schomburgkii, young seedlings propagated by bulbils were planted in flowerpots (30 cm in diameter) and placed under a rain and insect proof greenhouse in XTBG. The soil medium in the flowerpots was composed of forest soil (2), coarse sand (2), peat (2) and scattered brick (1). Treatments in the experiment included: 1) high-level fertilizer supply consisted of 150 ml per pot of a 5% mixed fertilizer (N:P:K=15:15:15) once every two weeks; 2) high-level fertilizer supply (same as treatment (1)) plus removal of flowers buds everyday; 3) high-level fertilizer supply (same as treatment (1)) plus removal of bulbils once a week; 4) low-level fertilizer supply (only water supplied) 5) low-level fertilizer plus removal of flower buds everyday; 6) low-level fertilizer plus removal of bulbils once a week. The results showed that 1) high-level fertilizer supply significantly increased bulbil dry weight ((1.11±0.10) gvs. (3.08±0.69) g,p=0.010 7), bulbil number ((30.58±2.92) vs. (74.74±9.73),p=0.000 4), bulbil size ((10.5±1.48) mg vs. (53.50±11.42) mg,p=0.001 5) and vegetative organ dry weight ((3.99±0.23) g vs. (9.67±1.17) g,p=0.000 2) of G. barthei and the bulbil number ((51.74±3.64) vs. (108.71±21.52),p=0.018 1) of G. schomburgkii; 2) different levels of fertilizer supply did not significantly affect the flower number of the two species (p>0.05); 3) removing bulbils or flowers had no significant effect on the number of flowers or the bulbils in the current year (p>0.05).
ForG. racemosa, we collected 30 individuals from six native populations throughout Yunnan Province and measured the biomass and size of sexual and asexual organs of each individual. Meanwhile, we collected three soil samples from each of the six populations and analyzed the physical and chemical characteristics of the soil samples. The bulbil size, bulbil number, bulbil dry weight, fruit dry weight, seed number and vegetative organ dry weight among the six populations varied significantly (p<0.001). Bulbil number was positively correlated to vegetative organ dry weight (r=0.901, p<0.05), bulbil dry weight was positively correlated to bulbil size (r=0.849, p<0.05). Fruit dry weight was positively correlated to seed number(r=0.998; p<0.05). A principal component analysis on soil factors of the six populations indicated that the three principal factors could explain 50.58%, 18.20% and 11.88% of the total variation. Total nitrogen, available nitrogen and soil organic matter wese the main factors of the first principal component, and the loading values were 0.959, 0.885 and 0.821, respectively. Available phosphorous was the main factor for the second factor with a loading value of 0.824. Vegetative organ dry weight had a significant positive relationship with available phosphorous (r=0.906, p<0.01), bulbil size was positively related to total nitrogen (r=0.798, p<0.05), available nitrogen (r=0.780, p<0.05) and soil organic matter(r=0.821, p<0.05). Fruit dry weight and seed number were not significantly correlated to any soil factors.
As a whole, the asexual organs of the threeGlobba plants were more sensitive to different levels of nutrient supply whereas the sexual organs did not show any significant response. Compensatory mechanisms between sexual and asexual reproduction were not detected in these three species.

Key words: Globba, Sexual reproduction, Asexual reproduction, Vegetative growth, Soil factors