Ground cover in temperate native perennial grass pastures. I. A comparison of four estimation methods

Abstract

Studies were conducted to compare visual estimates of ground cover and canopy cover by both inexperienced and experienced observers and to compare those estimates with those from more objective methods in native pastures in the high rainfall, temperate rangelands of northern NSW. Ground cover and canopy cover of 60 quadrats was estimated using visual, mapped area, digital image analysis and photo point quadrat methods. Inexperienced observers were trained by estimating ground cover of reference quadrats. Differences between mean visual estimates of ground cover and canopy cover for experienced and inexperienced observers were not significant (P>0.05). Mean ground cover estimates by the mapped area, digital image analysis and point quadrat methods were also not different from each other. The overall relationship between mean visual estimate and mean objective estimate of ground cover was non-linear (second order polynomial, R² = 0.93), observers tending to underestimate in the mid-range (20 to 80%) of cover compared with objective methods. Mean visual estimate of ground cover was 73.7% compared with the mean objective estimate of 83.7%. Visual estimates of canopy cover (mean 34.6%) were highly correlated (R² = 0.90) with those of the mapped area method (mean 34.3%) and the relationship was linear. Measurement of ground cover is a standard technique used in many pasture ecology and management studies and is increasingly being used by land managers to monitor pasture production and sustainability. Inexperienced observers were trained quickly and easily to estimate ground cover and canopy cover with sufficient accuracy to identify ranges of cover using visual estimation, indicating that the visual estimation technique should be suitable for estimating ground cover in land management research.