For site- or embayment-level mapping, use high-resolution imagery sources such as drone and side scan sonar as the most accurate and eelgrass-inclusive mapping tools (which can be significantly enhanced with diver ground-truthing of the edge).

For greater spatial coverage, use imagery acquired from airplane or high-resolution satellites. However, since these methods miss more eelgrass, resources should be allocated to enhance ground-truthing of the edge.

Regardless of remote sensing survey method used, apply a management buffer if precise diver-measured edge surveys are not conducted. At a minimum, the mean edge error for each survey method at each depth could be used, however this is not a fully protective buffer. The maximum edge error observed—approximately 120 m across all survey methods—is recommended for the most protective buffer.

Mapping programs should collect eelgrass canopy height and distribution data during ground-truthing efforts to help track meadow change and detectability over time.

For eelgrass mapping programs using aerial imagery, it is strongly recommended that sub-meter resolution satellite imagery be further explored, given that MassDEP’s 0.25 m resolution aerial imagery performed similarly to satellite imagery or coarser (3 m) resolution.

Site-specific conditions greatly influenced results. Sites with a high degree of macroalgae, and those with expansive, diffuse patchy edges are best surveyed by SCUBA divers.

Drop-frame photo ground-truthing is not reliable in detecting very low-density eelgrass and should be augmented by SCUBA diver surveys.

This study did not investigate the ability of remote sensing methods to detect and map new or previously unmapped meadows—it specifically targeted the edge of known meadows. Given that all remote sensing methods in this study performed poorly in patchy, low density, and shorter areas of the meadow, mapping programs are likely to consistently miss these eelgrass areas. Thus, diver, video transects, and/or dropframe ground-truthing surveys should be implemented in places where eelgrass was previously mapped via remote sensing but has since disappeared, or where impacts to highly suitable habitat are proposed.