Changes of coral communities over 35 years: Integrating in situ and remote-sensing data on Saint-Leu Reef (la Réunion, Indian Ocean)

Abstract

Monitoring coral reef communities at a decadal scale is necessary to understand and project their dynamics to provide a basis for reef management in light of disturbances and climate change. Complementing infrequent and localised in situ observations, time-series of aerial photographs and remotely sensed satellite images provide a means of monitoring the position and extent of reef-top coral communities, vegetation and abiotic substrata at the whole reef scale. The aim of the present study was to map such changes on Saint-Leu Reef (La Réunion) between 1973 and 2007. The period included two cyclones (1989, 2002) and one severe bleaching event (2002). Vertical images of the reef were recorded in five aerial photographs (1973, 1978, 1989, 1997, and 2003) and two Quickbird satellite images (2002, 2006) during that time. Quantitative in situ observations of parts of the reef-top were also available both to document ecological and substratum characteristics that produce the color and texture observable in the photos and satellite images. Coral communities were mapped on all images using manual delineation of polygons identified according to color and texture contrast. The 2006 Quickbird image was used as the mapping base and 15 types of coral communities were identified from a reef survey conducted in 2007. This hierarchical typology used coral growth forms, live and dead coral cover, macro-algae, substratum (sand, rubble and platform) and, to a lesser extent, coral taxonomy. The polygons at date t were over-laid onto image t − 1 across the whole series and their boundaries were manually edited to match the pattern on the earlier image. Labelling of polygons was guided by field-survey data and maps. For coral-dominated patches, six successive pairs of maps from 1973 to 2006 were compared to produce ‘coral community change maps’ for that period. Despite the multiple disturbance events, the coral community distribution and composition in 2006 on Saint-Leu Reef did not display major differences compared to 1973. This suggests a high degree of coral resilience at the site, led by rapid recovery of compact branching corals. The mapping techniques overcame challenges due to different image quality and the sparsity of in situ observations in time and space. Our results demonstrate the potential for further application of reef monitoring protocols based on complementary in situ and remote-sensing data to help understand the dynamics of reef-top coral reef communities and geomorphology over years to decades.

Introduction

Around the world, many coral reefs have become degraded (Pandolfi et al., 2003, Souter and Lindén, 2005, Bruno and Selig, 2007, Obura et al., 2008) as anthropogenic stresses and natural disturbances (cyclones and bleaching) have increased in strength and frequency (Hughes et al., 2003). Coral communities respond with different levels of sensitivity to these stressors (Lirman and Fong, 1997). Current and future management programmes also need to be cognisant of normal successional changes on reefs (Pichon, 1974, Kleypas et al., 2001, Perry et al., 2008) driven by natural catastrophes such as cyclones and flood plumes, whose influence on the state of corals at a particular place reflects both the normal environment and the site-specific history of disturbance (Done, 1995, Connell et al., 1997). Monitoring programmes attempt to capture such changes in the spatial distribution and composition of coral communities, and to understand the dynamics associated with degradation and recovery at different spatial and temporal scales, from the global (e.g. Wilkinson, 2008) to the local (e.g. Done et al., 2007). However, most studies have focused on spatially limited quadrats or transects within a set of particular reefs, and were often established in response to a single event. Consequently, they do not include description of a historical baseline of undisturbed communities and rarely provide synoptic (i.e. reef-scale, regional-scale km² – 1000 km²), spatially continuous observations, helpful for the development of spatially explicit conservation and management plans (Purkis and Riegl, 2005).

Remotely sensed images and maps derived from them can provide synoptic and long-term observations of reef-top habitats. Medium and high spatial resolution (30 m to 1 m), multi-spectral digital satellite images are now routinely available at moderate cost, on demand, or from a regular collection schedule (Andréfouët and Riegl, 2004, Elvidge et al., 2004). For many places, black and white or color analog aerial photographs provide synoptic views of reefs before the satellite era, often extending back to the 1930s, well before coral reef monitoring programmes expanded in the 1990s. Aerial photographs may be the only source of observations of the state of a coral reef before many major disturbance events (Lewis, 2002). The combination of satellite and airborne images provides long-term and synoptic observations on map reef-top features and detects changes therein (Yamano et al., 2000, Cuevas-Jiménez et al., 2002, Lewis, 2002, Palandro et al., 2003, Purkis and Riegl, 2005).

Manual delineation in photo-interpretation has proven to be a successful mapping technique for characterising patch dynamics in the seagrass environment (McKenzie et al., 2001, Hernandez-Cruz et al., 2006). In the coral reef environment, this technique applied at a fine geomorphological scale allowed detection of structural reef loss in response to a cyclone in the West Indies (Lewis, 2002).

Some changes in reef-top features detected from satellite-image-based maps can be used to infer dynamic characteristics, both ecological and geomorphic. For example, Purkis and Riegl (2005) detected episodes of coral community development in the Arabian Gulf after recurrent coral mass mortality due to bleaching. Palandro et al. (2008) also documented habitat changes in the Florida Keys using remotely sensed imagery and in situ surveys at two different scales. These examples suggest that the right combination of historical imagery and in situ observations at suitable spatial scales and temporal sampling intervals enables mapping of the dynamic reef-top communities and substrata over periods of years to decades. In particular, observations at suitable regular intervals, yearly or less, over a longer term (>10 years), may capture cycles of reef degradation and recovery, or long-term changes of state (McClanahan et al., 2002). This provides essential support information for reef science, monitoring and management.

The goal of this study was to combine available imagery (satellite and airborne) and in situ data to map coral community and other reef-top changes over 35 years on Saint-Leu Reef, La Réunion. More specifically, we aimed to use the imagery to scale up localised in situ observations to characterise the dynamics of an entire shallow reef. The challenges were both technical (processing of disparate imagery across time with limited coincident field-surveys) and thematic (identifying a relevant coral community typology trackable in imagery and compatible with in situ monitoring data).