Spatio-temporal analysis of flowering using LiDAR topography

• Published in: Journal of geographical sciences 2017, Vol.27 (1), p.62-78
• Main Authors: Hart, Samantha, Mikhailova, Elena, Post, Christopher, McMillan, Patrick, Sharp, Julia, Bridges, William
• Format: Article
• Language: English
• Subjects: Accumulation; Cluster analysis; Earth and Environmental Science; Flowering plants; Forest ecosystems; Geographical Information Systems/Cartography; Geography; Global positioning systems; GPS; Lakes; Lidar; Low flow; Nature Conservation; Physical Geography; Remote Sensing/Photogrammetry; Sandy loam; Soil types; Topography
• ISSN: 1009-637X; 1861-9568
• DOI: 10.1007/s11442-017-1364-x

Spatio-temporal patterns of flowering in forest ecosystems are hard to quantify and monitor. The objectives of this study were to investigate spatio-temporal patterns (e.g. soils, simple slope classes, slope aspect, and flow accumulation) of flowering around Lake Issaqueena, South Carolina (SC, USA) using plant-flowering database collected with GPS-enabled camera (stored in Picasa 3 web albums and project website) on a monthly basis in 2012 and LiDAR-based topography. Pacolet fine sandy loam had the most flowering plants, followed by Madison sandy loam, both dominant soil types around the lake. Most flowering plants were on moderately steep (17%–30%) and gently sloping (4%–8%) slopes. Most flowering plants were on west (247.5°–292.5°), southwest (202.5°–247.5°), and northwest (292.5°–337.5°) aspects. Most flowering plants were associated with minimum and maximum flows within the landscape. Chi-square tests indicated differences in the distributions of the proportions of flowering plants were significant by soil type, slope, aspect, and flow accumulation for each month (February-November), for all months (overall), and across months. The Chi-square test on area-normalized data indicated significant differences for all months and individual differences by each month with some months not statistically significant. Cluster analysis on flowering counts for nine plant families with the most flowering counts indicated no unique separation by cluster, but implied that the majority of these families were flowering on strongly sloping (9%–16%) slopes, on southwest (202.5°–247.5°) aspects, and low flow accumulation (0–200). Presented methodology can serve as a template for future efforts to quantify spatio-temporal patterns of flowering and other phenological events.