Accuracy of optical image analysis compared to conventional vegetation measurements for estimating morphological features of emergent vegetation. (20th March 2015)

• Record Type: Journal Article
• Authors: Lemein, Todd; Cox, Daniel; Albert, Dennis; Mori, Nobuhito
• Is Part Of: Estuarine, coastal and shelf science. Volume 155(2015)
• Page Start: 66
• Page End: 74
• Publication Date: 2015-03-20
• Subjects: estuaries -- salt marsh -- emergent vegetation -- optics -- plant morphology -- measuring methods; Estuarine oceanography -- Periodicals; Coasts -- Periodicals; Estuarine biology -- Periodicals; Seashore biology -- Periodicals; Coasts; Estuarine biology; Estuarine oceanography; Seashore biology; Periodicals; 551.461805
• DOI: 10.1016/j.ecss.2014.12.051
• Languages: English
• ISSNs: 0272-7714

Abstract

Abstract: In this paper we evaluate the effectiveness of using optical image analysis to determine plant morphology for an emergent species of salt marsh vegetation, threesquare bulrush ( Schoenoplectus pungens ), and compare the results to conventional vegetation measurement methods. We find that mean stem height may be estimated optically to within 10% of mean stem height measured with conventional methods and that total above ground biomass may be estimated by lateral percent cover values between 0 and 70% ( R 2 = 0.89, linear regression) within a 10 cm–60 cm depth of vegetation. Additionally, we show that stem height may be used to estimate stem diameter by linear regression ( R 2 = 0.94). The product of the plant stem height, stem diameter and density measured by conventional methods are shown to correspond with the horizontal two-dimensional projection of vegetation estimated by image analysis. The optical method provided robust estimates when the depth of field is between 10 cm and 60 cm. When the depth of field exceeded 60 cm, the images were saturated, resulting in a loss of information. Graphical abstract: Highlights: Optical image analysis estimates stem height, diameter, biomass. Width of vegetation to photograph is most robust between 10 cm and 60 cm. Image saturation occurs at widths of vegetation greater than 60 cm. Biomass may be estimated by percent cover within a range below saturation of image. Integrated area of lateral obstruction estimates the vegetation surface area.