Comparing methods to quantify cover in turfgrass research

Abstract

Turfgrass cover can be assessed qualitatively using visual ratings, but quantitative turfgrass cover measurements are desired for producing unbiased data. Digital image analysis and point intercept are two quantitative percent cover data collection methods used in turfgrass research. A potential weakness of digital image analysis is the difficulty in evaluating color variation. Considering this, a series of controlled environment and field experiments were conducted to evaluate the accuracy of the point intercept method compared to using a digital grid, digital image analysis, and visual ratings when utilized in turfgrass research. To explore this topic, four experiments were conducted in Corvallis, OR, and one in New Brunswick, NJ. Results from research conducted in Corvallis, OR, determined that the closest point intercept spacing (2.5 cm) resulted in a percent cover that was not statistically different from the known cover and the lowest variance, but the greatest amount of time required to collect the data. Digital image analysis was the most consistent method for measuring percent cover when a known percent cover was being quantified. The estimate of the percent dollar spot cover was reduced by a single pigment application regardless of the data collection method. In the absence of pigment applications, the point intercept frame, digital image analysis, and visual ratings were well‐correlated methods used to assess dollar spot cover. Digital image analysis was not correlated with the other data collection methods when used to evaluate percent anthracnose cover. The research conducted in New Brunswick, NJ, determined that the point intercept method and digital image analysis are well‐correlated methods for quantifying turfgrass establishment from seed.

Annual bluegrass and creeping bentgrass tiller response to phosphate fertilizer and soil pH

“…The lower green cover according to digital image analysis may be a result of the software quantifying darker turfgrass leaves shadowed underneath other turfgrass leaves as “non‐green,” even when turfgrass is in the pixel. Karcher and Richardson (2003) mention that shadows may influence results when using digital image analysis techniques, and this has been observed in previous research (Kowalewski et al., 2023); however, this does not impede our ability to interpret the results.…”