Data Collection and Organization
Flowering Data Collection and Organization:
A lot of data already exists on flowering of dipterocarps in Southeast Asia. I gathered flowering data from the following studies:
1. Appanah, S. (1993). Journal of Biosciences 18(4): 457-474.
2. Ashton, P. S. (1988). Annual Review of Ecology and Systematics 19: 347-370.
3. Brearley, F. Q., J. Proctor, et al. (2007). Journal of Ecology 95(4): 828-839.
4. Cannon, C. H., L. M. Curran, et al. (2007). Ecology Letters 10(10): 956-969.
5. Curran, L. M., I. Caniago, et al. (1999). Science 286(5447): 2184-2188.
6. Hamann, A. (2004). Journal of Ecology 92(1): 24-31.
7. Naito, Y., M. Kanzaki, et al. (2008). Journal of Plant Research 121(1): 33-42.
8. Numata, S., M. Yasuda, et al. (2003). American Journal of Botany 90(7): 1025-1031.
9. Sakai, S. (2002). Biological Journal of the Linnean Society 75(2): 233-247.
10. Sakai, S., R. D. Harrison, et al. (2006). American Journal of Botany 93(8): 1134-9.
11. Sun, I. F., Y. Y. Chen, et al. (2007). Journal of Ecology 95(4): 818-827.
The sampling units represent flowering intensity observed at each site. The predictor variables are climatic factors. Al the variables are continuous.
Flowering data was recorded for each site and the flowering intensity was converted to an ordinal scale of 0 to 5 to standardize the flowering intensity at each site. Values of 0-5 correspond to a percent of trees flowering in a community at the time of a GF event in a study site. They were broken down as follows:
0: 0-10%
1: 11-20%
2: 21-40%
3: 41-60%
4: 61-80%
5: 81-100%
The reason that the break down was not equal in value is because absence of a GF (or the value of 0) does not necessary mean 0% flowering as sporadic flowering can occur in a dipterocarp forest. Flowering of below 20% is considered a "sporadic flowering event". The classification of a proper GF even starts when flowering intensity reaches above 20 % after which they are broken down in 4 equal subgroups.
Figure 1 below shows the sites where flowering data was collected. Diagram 1 below shows the way flowering data was orginized and aggregated.
A lot of data already exists on flowering of dipterocarps in Southeast Asia. I gathered flowering data from the following studies:
1. Appanah, S. (1993). Journal of Biosciences 18(4): 457-474.
2. Ashton, P. S. (1988). Annual Review of Ecology and Systematics 19: 347-370.
3. Brearley, F. Q., J. Proctor, et al. (2007). Journal of Ecology 95(4): 828-839.
4. Cannon, C. H., L. M. Curran, et al. (2007). Ecology Letters 10(10): 956-969.
5. Curran, L. M., I. Caniago, et al. (1999). Science 286(5447): 2184-2188.
6. Hamann, A. (2004). Journal of Ecology 92(1): 24-31.
7. Naito, Y., M. Kanzaki, et al. (2008). Journal of Plant Research 121(1): 33-42.
8. Numata, S., M. Yasuda, et al. (2003). American Journal of Botany 90(7): 1025-1031.
9. Sakai, S. (2002). Biological Journal of the Linnean Society 75(2): 233-247.
10. Sakai, S., R. D. Harrison, et al. (2006). American Journal of Botany 93(8): 1134-9.
11. Sun, I. F., Y. Y. Chen, et al. (2007). Journal of Ecology 95(4): 818-827.
The sampling units represent flowering intensity observed at each site. The predictor variables are climatic factors. Al the variables are continuous.
Flowering data was recorded for each site and the flowering intensity was converted to an ordinal scale of 0 to 5 to standardize the flowering intensity at each site. Values of 0-5 correspond to a percent of trees flowering in a community at the time of a GF event in a study site. They were broken down as follows:
0: 0-10%
1: 11-20%
2: 21-40%
3: 41-60%
4: 61-80%
5: 81-100%
The reason that the break down was not equal in value is because absence of a GF (or the value of 0) does not necessary mean 0% flowering as sporadic flowering can occur in a dipterocarp forest. Flowering of below 20% is considered a "sporadic flowering event". The classification of a proper GF even starts when flowering intensity reaches above 20 % after which they are broken down in 4 equal subgroups.
Figure 1 below shows the sites where flowering data was collected. Diagram 1 below shows the way flowering data was orginized and aggregated.
Figure 1: Study sites
Figure 1 shows all the sites used in the analysis. The sites were grouped into three sub-regions of Peninsular Malaysia (Malaysia), Borneo, and the Philippines. Red dots represent specific sites of flowering data and red ractangles represent regional flowering data.
Diagram1: Flowering Intensity aggregation procedure
The climate variables came from the Tyndall Centre World Climate Database. Climate variables were extracted to the closest estimated Latitude and Longitude location corresponding to specific study sites using SAS statistical software. Since flowering data was recorded yearly and not monthly, climate data was averaged yearly as well. However, dipterocarp flowering usually occurs in the months of March, April, and May and the cycle of flowering and fruiting completes by late September for the three regions. For this reason, climate data needed to include a period before flowering induction in dipterocarps occurs in order to trace the cue. Therefore, regular climate years were transformed into “dipterocarp years” which start in October of the previous years and runs until September of the current year, when flowering completes its cycle. Please refer to Diagram 2 below to see the steps that were taken in climate data organization.
Diagram 2: Climate variables aggregation procedure
Finally, Table 1 below shows that data table that was used in the analysis. The data table below represents only one region, but similar tables were constructed for all three regions for the analysis.
Table1: Sample data table
