Scale dependency in seagrass dynamics: how does the neighboring effect vary with grain of observation?

Abstract

Although the importance of spatial scale in ecology has been increasingly recognized, the effects on ecological processes of changing the grain size of the observation have rarely been tested for empirical populations. A seagrass bed is an ideal system to study scaledependency because it occurs in two-dimensional shallow soft-bottoms and can be monitored on a broader scale by using remote-sensing techniques. To investigate the grain dependency of seagrass spatial dynamics, we analyzed the effect of neighboring vegetation on the annual transition between vegetated and unvegetated states in a seagrass meadow in Futtsu, Tokyo Bay. The presence or absence of seagrass vegetation was observed at different grains from aerial photographs taken annually over 17 years. We detected the presence of a neighboring effect both in the increasing process (transition from the unvegetated to the vegetated state) and the decreasing process (vice versa) of vegetation. In the increasing process, the intensity of the neighboring effect was positive with the small grain, but the effect decreased to 0 with grain of ca. 20 m. In the decreasing process, the neighboring effect was negative with the small grain and increased to 0 with grain of ca. 30 m. The observed grain dependency in the neighboring effects also varied among different positions of the bed and among different years. The grain dependency in the increasing process cannot solely be explained by shoot elongation of the seagrass, which can cause positive neighboring effects only at small grain (B6 m). The neighboring effect at the greater grain can be regarded as an emergent property.

Asynchronous local dynamics contributes to stability of a seagrass bed in Tokyo Bay

“…Analyses incorporating information on the landscape properties of seagrass beds have great potential for understanding how seagrass dynamics are affected by multiple factors operating at various spatio-temporal scales (Bell et al 1999, Robbins and Bell 2000, Kendrick et al 2005. Recent developments in computer and image analysis techniques have enabled researchers to analyze changes in seagrass beds at a landscape level over broad spatial scales and long time scales (Kendrick et al 2002, Yamakita and Nakaoka 2009).…”

Application of deep learning techniques for determining the spatial extent and classification of seagrass beds, Trang, Thailand

“…Visual interpretation by experts can be an accurate way to construct benthic cover maps, particularly when the images are of high spatial resolution. Supervised and unsupervised classification based on the value of each pixel in an image can be performed using remote sensing software packages (Yamakita and Nakaoka 2009). Although these are popular methods for determining benthic cover classifications, they generate small fragments of noise along with correctly classified pixels depending on the condition or resolution of the images.…”

Aquatic Biodiversity Conservation and Ecosystem Services

“…Methods for mapping such patchy seagrass by supervised classification using aerial photography and satellite imagery have been proposed (e.g., Wabinaz et al 2008;Yamakita and Nakaoka 2009;Waycott et al 2009;Sakuno and Kunii 2013). Seagrass inhabits shallow areas of a few meters' depth.…”