Assessing the ecological application of lasergrammetric techniques to measure fine-scale vegetation structure

In the study of habitat use by animals, it is important to sample vegetation structure or the degree of cover at different heights or in vegetation strata. Commonly-used techniques to estimate vegetation cover (e.g. point height intercept) are time-consuming and can be especially subjective in tall forest environments. We investigated the application of a Terrestrial Laser Scanner (TLS) to quantify forest structure, in terms of evenness and functional diversity. We also explored the suitability of the developed lasergrammetric technique (LGT) in the monitoring of nest site preferences of translocated populations of New Zealand endemic birds, the South Island saddleback (Philesturnus carunculatus) and the New Zealand robin (Petroica australis). At an acquisition rate of up to 6000 points per second and with an estimated error of less than 2 cm, the LGT technique utilises terrestrial scanners that allow collection and processing of field data rapidly and precisely. Further, LGT is able to account for small structural elements such as minor branches, leaves and vine stems. The derived measures (mean height of canopy (MCH), indices of structural evenness (Evar) and functional diversity (FDvar)) can quantified vertical variations in vegetation density but appeared highly dependent on the vertical scale of measurement. When applied to an ecological case study, results suggested that the LGT method provides complementary quantitative estimates to summary methods when assessing vegetation structure for micro-scale studies. However, further validations of the LGT accuracy are needed more specifically in relation to density of obstructing vegetation and when used in the measuring of small vegetative objects and total biomass estimates. The development of new software packages that easily convert large amounts of three-dimensional data collected using laser scanners, into a form ready for biostatistical analysis, also is strongly encouraged.