Progress in Physical Geography

Progress in Physical Geography http://ppg.sagepub.com/

Progress in the use of remote sensing for coral reef biodiversity studies Anders Knudby, Ellsworth LeDrew and Candace Newman Progress in Physical Geography 2007 31: 421 DOI: 10.1177/0309133307081292

The online version of this article can be found at: http://ppg.sagepub.com/content/31/4/421

Published by:

http://www.sagepublications.com

Additional services and information for Progress in Physical Geography can be found at:

Email Alerts: http://ppg.sagepub.com/cgi/alerts

Subscriptions: http://ppg.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

Citations: http://ppg.sagepub.com/content/31/4/421.refs.html

>> Version of Record - Aug 29, 2007

What is This?

Downloaded from ppg.sagepub.com at SIMON FRASER LIBRARY on August 15, 2013 Progress in Physical Geography 31(4) (2007) pp. 421–434 Ã

Progress in the use of remote sensing for coral reef biodiversity studies

Anders Knudby,* Ellsworth LeDrew and Candace Newman

Department of Geography, University of Waterloo, 200 University Avenue West, Waterloo, Ontario, N2L3G1, Canada

Abstract: Coral reefs are hotspots of marine biodiversity, and their global decline is a threat to our natural heritage. Conservation management of these precious ecosystems relies on accurate and up-to-date information about ecosystem health and the distribution of species and habitats, but such information can be costly to gather and interpret in the fi eld. Remote sensing has proven capable of collecting information on geomorphologic zones and substrate types for coral reef environments, and is cost-effective when information is needed for large areas. Remote sensing-based mapping of coral habitat variables known to infl uence biodiversity has only recently been undertaken and new sensors and improved data processing show great potential in this area. This paper reviews coral reef biodiversity, the infl uence of habitat variables on its local spatial distribution, and the potential for remote sensing to produce maps of these habitat variables, thus indirectly mapping coral reef biodiversity and fulfi lling information needs of coral reef managers.

Key words: coral reefs, ﬁ sh biodiversity, habitat characteristics, lidar, remote sensing, rugosity.

I Introduction and coral diseases at the local scale to world- Coral reefs are the most biodiverse marine wide ocean warming, acidiﬁ cation, and sea- ecosystems on the planet, estimated to level rise. Most of these threats are expected harbour nearly one million species globally to worsen their impact in the coming decades (Reaka-Kudla, 1997: 93). They can be places (Hoegh-Guldberg, 1999; Kleypas et al., 1999; of extraordinary beauty, and are essential Pittock, 1999), leaving an uncertain future for to the livelihoods of people who depend on coral reefs. The decline of coral reefs, both them for food, coastal protection, tourism- past and projected, is of much more than aca- based income and more (Birkeland, 1997: 2). demic interest; it is a serious threat to global However, the health of coral reefs is declining biodiversity, an important part of our natural at a global scale (Wilkinson, 2004: 7), and heritage. the threats that have precipitated this decline Because of the complexity of coral reef range from overﬁ shing, nutrient enrichment ecosystems and the multitude of threats,

*Author for correspondence. Email: [email protected]

Downloaded from ppg.sagepub.com at SIMON FRASER LIBRARY on August 15, 2013 422 Progress in Physical Geography 31(4) identification of the important threats and II Coral reef biodiversity the necessary management for a given reef Although reefs have been a recurrent phe- is diffi cult. Nevertheless, the global area of nomenon throughout the history of life on coral reef under some form of management Earth, modern coral reefs only developed in is growing, and so is the need for information the Triassic (Newell, 1972). The evolution on which to base management measures of the Symbiodinium group of dinofl agellates (Wilkinson, 2004: 1). A manager of a protected and their incorporation into scleractinian area needs assessments of various aspects corals produced the symbiotic organisms that of reef health, which provide both a basis are still the primary reef-building organisms from which zonation plans and management (Webb, 1998). Important groups of fauna regulations can be developed, and a baseline evolved thereafter, notably sea urchins, reef from which changes can be assessed. fi shes and currently widespread coral genera Reef health is an intangible concept, and is such as Acropora, Porites and Pocillopora typically mapped and monitored using a num- (Wood, 1998). Today, despite covering only ber of proxies. Live coral cover is often used for between 0.1% and 0.5% of the total area of practical reasons (Mumby et al., 2004b), and the oceans (Moberg and Folke, 1999), coral so is diversity or abundance of ecologically reefs harbour more species than any other important or vulnerable species (Hodgson marine ecosystem (Dubinsky, 1990: 251; et al., 2004). However, owing to the diffi culties Reaka-Kudla, 1997: 93). The largest groups and expense associated with conducting ex- of known coral reef species are fishes and tensive fi eld surveys under water, spatially sponges, the reef-building organisms them- distributed data of the appropriate type and selves being relatively few in number. Com- detail are rarely available. Remote sensing prehensive species lists do not exist for any technologies have therefore been used to map reef area of the world (Paulay, 1997); it is coral reefs since the early days of Landsat apparent that only a fraction of the existing (Smith et al., 1975), and research into the use fauna has been described, and it is likely that of remote sensing technology continues with small cryptic species outnumber the num- the advent of new sensors and data processing erous described fi shes and sponges (Moran methods (Kutser et al., 2006). Classifi cation of and Reaka, 1988). broad substrate types is now routinely possible Coral reefs are generally limited to shal- in clear and shallow water, and water depth low and clear water with a mean water can be derived from a variety of data sources temperature of 18°C or higher, and are thus with varying accuracy. The interference of the largely confi ned to the tropics (Yonge, 1940). water column, however, continues to pose Although reefs exist wherever conditions problems for classifi cation accuracy, and so do permit, there is profound variation in the the similarities in spectral signatures between number of different species that make up the important substrate types. New technologies reef fauna in different parts of the world. At show promise for mapping aspects of coral a regional scale, three factors have had a large reef health beyond substrate types, including infl uence on today’s biogeographic patterns. water quality and reef structural complexity, First, the current distribution of land masses thus providing complementary information and oceans has resulted in upwelling at the for mapping of coral reef biodiversity. In this eastern margins of the two major oceans, paper we aim to review coral reef biodiversity the Atlantic and the Indo-Pacific, and the and its spatial distribution, the infl uence of dominant equatorial currents running east habitat characteristics on biodiversity, and to west (Veron, 1995: 99). Cold nutrient- remote sensing approaches to mapping coral rich water from upwelling is detrimental to reef habitats, with a focus on mapping habitat reef growth, and western margins of oceans variables known to infl uence biodiversity. therefore house larger reef areas than eastern

Downloaded from ppg.sagepub.com at SIMON FRASER LIBRARY on August 15, 2013 Anders Knudby et al.: Remote sensing for coral reef biodiversity studies 423 margins (Hubbard, 1997: 52). Second, the across taxonomic groups. This tends to be closing of the Isthmus of Panama separated the case for fish rather than plants (Ward the Atlantic and Pacifi c oceans between 3.1 et al., 1999) or coral (Beger et al., 2003), which and 3.5 million years ago (Coates and Obando, is unfortunate because fi sh are not directly 1996), which resulted in different species now mappable using remote sensing. Fortunately, being present on either side of the Isthmus. distinguishable habitat categories such as reef, Third, the large area of shallow tropical seas in seagrass and sand can predict biodiversity the Indo-West Pacifi c has resulted in greater equally well or better than fi sh, and are more biodiversity of coral reef fauna here than in easily mappable (Ward et al., 1999). any other of the recognized biogeographic regions: the Western Atlantic, Eastern Pacifi c III Habitats as indicators of coral reef and Eastern Atlantic, in order of decreasing biodiversity biodiversity (Briggs, 1999; Bellwood and As noted, several environmental variables Hughes, 2001). Within the Western Pacifi c, have been shown to infl uence the biodiversity the species richness of scleractinian corals of a given habitat. Mapping such habitat themselves is greatest in the coral triangle, variables could indicate the likely spatial formed by Indonesia, the Philippines and distribution of biodiversity at a local scale and New Guinea, with more than 450 described. suggest priority areas for conservation, at least Species richness gradually declines towards for the species for which habitat-biodiversity the Indian Ocean or the Pacifi c (Veron, 1995: relationships have been identifi ed. A survey of 140). A similar biogeographic pattern is found the literature relating biodiversity and habitat for reef fi shes (Bellwood and Hughes, 2001; variables yields a complex picture. Studies Bellwood and Wainwright, 2002), and for all have focused on a variety of taxonomic or other taxa for which data are available (Paulay, functional groups, have employed different 1997: 303). measures of biodiversity, and have measured At smaller spatial scales, communities different habitat variables in different ways. form as a subset of the available species of the A brief overview is presented in Table 1. De- region, determined by the local physical and tailed relationships are obscured by the biological environment. Within a given coral number of different variables used, and the reef ecosystem several spatial biodiversity different temporal and spatial scales studied patterns have been shown to exist. Biodiversity (Jones and Syms, 1998). It can thus be argued of corals is greatest at intermediate depth that a species-specific approach is more (Huston, 1994: 385; Karlson, 1999: 31), with appropriate, as habitat associations of many intermediate disturbance history (Connell, individual species are well known and well 1978) and fi sh biodiversity increases close to defi ned (Allen et al., 2003). For a few highly the reef edge (Friedlander and Parrish, 1998), signifi cant species such an approach may be in areas with high coral cover (Chabanet et useful but, with approximately 10,000 species al., 1997), and high structural complexity of described fi sh, and a total of one million (McCormick, 1994). These relationships, species in all taxonomic groups estimated however, are not uniform across taxonomic and to exist on coral reefs, a species-specific functional groups, and are sometimes heavily approach is unfeasible for studies of general infl uenced by stochastic recruitment events biodiversity patterns. (Sale, 1991: 203). Studies have demonstrated Despite these problems, several con- spatial covariance of biodiversity for different clusions can be drawn from the literature: taxonomic groups, and have confirmed (i) several habitat characteristics influence that species with high spatial variation in local biodiversity; (ii) the number of studied biodiversity are good indicators of biodiversity biodiversity and habitat variables is large

Downloaded from ppg.sagepub.com at SIMON FRASER LIBRARY on August 15, 2013 424 Progress in Physical Geography 31(4)

Table 1 Studies demonstrating spatial correlationsA between habitat and biodiversity variables Habitat variable Biodiversity variable Source Depth Fish species richness Luckhurst and Luckhurst, 1978 Huston, 1994 Fish species diversity Friedlander and Parrish, 1998 Herbivore fi sh species richness Ormond et al., 1996 Structural complexity Fish species richness Friedlander and Parrish, 1998 McCormick, 1994 Gratwicke and Speight, 2005 Fish species diversity Luckhurst and Luckhurst, 1978 Gastropod abundance Kohn, 1968 Gastropod species richness Kostylev et al., 1997; 2005 Live coral cover Fish abundance, species richness Jones et al., 2004 and diversity Chabanet et al., 1997 Garpe and Ohman, 2003 Corallivore fi sh species richness Friedlander and Parrish, 1998 Sea urchin abundance McClanahan, 1988 Branching coral cover Fish species richness Chabanet et al., 1997 Fish abundance Sale and Dybdahl, 1975 Number of holes, total hole Fish species richness, fi sh Friedlander and Parrish, 1998 volume, mean hole volume abundance, fi sh species diversity Hixon and Beets, 1993 Distance to reef edge Fish species richness and diversity Friedlander and Parrish, 1998 Distance to river mouth Species richness and diversity Friedlander and Parrish, 1998 Total area Fish species diversity Molles, 1978 and relationships are not restricted to a few heterogeneity and refuge for prey species – variables; (iii) some habitat variables, namely spaces big enough for them to enter but too depth, live coral cover and reef structural small for their predators (Friedlander and complexity, influence more biodiversity Parrish, 1998). This illustrates the important variables than others, and show stronger relation between the body sizes of organisms correlations than others; and (iv) common and the spatial scale of the structure providing biodiversity measures, including species refuge. Habitat influences on biodiversity abundance, richness and diversity, show also extend to the temporal domain; loss of correlations with these habitat variables. fish biodiversity can often be attributed to Several causative mechanisms have been loss of coral cover (Jones et al., 2004) or loss proposed for the relations between habitat of physical structure following hurricanes or and biodiversity. The infl uence of depth has severe bleaching (Connell et al., 1997; Garpe been cited as an example of intermediate et al., 2006; Graham et al., 2006). Mapping disturbance positively influencing species coral reef habitats can therefore provide coral richness at intermediate depth (Huston, reef managers with important information on 1994: 383), whereas the infl uence of live coral the likely spatial distribution of biodiversity cover has been related to larval settlement in their area and, in the absence of frequent success and to survival of corallivorous and field surveys, can warn about changes in coral-dwelling species (Jones et al., 2004). biodiversity to be expected from changing Reef structural complexity provides physical habitats.

Downloaded from ppg.sagepub.com at SIMON FRASER LIBRARY on August 15, 2013 Anders Knudby et al.: Remote sensing for coral reef biodiversity studies 425

IV Mapping coral reefs with remote ically expressed as the number of classes that sensing can be discriminated and the accuracy of the Most coral reef remote sensing research has classifi cation, depends on the platform and not been carried out directly in the context sensor type and on environmental factors of biodiversity, but has focused on mapp- such as water depth and turbidity, the state ing geomorphologic zones (Smith et al., of the sea surface, and the atmosphere 1975; Andréfouët and Guzman, 2005) and (Mumby et al., 2004b). Early studies that characteristic substrate types (Mumby et al., used Landsat TM and SPOT HRV sensors, 1997; Andréfouët et al., 2003). Nevertheless, for example, only allowed discrimination of remote sensing has become an important broad categories such as coral, sand, sea- tool for coral reef studies and the existing grass and algae. Even with such broad research has built a foundation from which classes, distinguishing between the coral biodiversity studies can benefi t. In addition, and algae classes was difficult because the new remote sensing technologies have im- refl ectance properties of both are dominated proved the accuracy and spatial scales at by chlorophyll a; the differences are caused which coral reefs can be mapped, so products by accessory pigments whose contribution more relevant for biodiversity research can to the spectral refl ectance is relatively small be developed. (Fang et al., 1995; Hedley and Mumby, 2002; Kutser et al., 2003). In addition, light 1 Mapping geomorphology and substrate types attenuation with depth reduced differences Mapping of geomorphologic zones was the in spectral refl ectance between highly refl ect- fi rst use of remote sensing on coral reefs (Smith ing substrates at depth and less reflecting et al., 1975) and, despite developments in substrates in shallow water. Landsat TM and automated systems (Suzuki et al., 2001), geo- SPOT HRV sensors with broad bandwidths morphologic zones are typically still mapped (>65 nm full-width half maximum (FWHM) manually – outlined on a plot of original or for all water-penetrating wavelengths) were classifi ed data by an expert user (Andréfouët not designed to register the small differences et al., 2001; Andréfouët and Guzman, 2005). between these substrate types, and classifi - Classes can be limited to the major geomor- cation accuracies are therefore rarely better phologic zones – forereef, reef crest, lagoon, than 60–70% for the four main classes of backreef and patch reefs – or they can be coral, sand, seagrass and algae (Green et al., arranged in a hierarchy incorporating the 2000), and as low as 30% for a detailed major geomorphologic zones and slope, classification using more than 10 classes depth, or other variables deemed important (Mumby and Edwards, 2002). The infl uence (Mumby and Harborne, 1999). Because of variable depth can be partly mitigated by geomorphologic zones themselves infl uence correcting radiance values for its influence biodiversity, their mapping can provide a fi rst (Lyzenga, 1978; Stumpf et al., 2003), or by insight into likely areas of high biodiversity stratifying substrate types into depth intervals (Andréfouët and Guzman, 2005). in the classifi cation (Turner and Klaus, 2005). Specular refl ection off the water surface is 2 Mapping substrate types another source of noise that can be partly More research effort has gone into mapping eliminated by using radiance values in the substrate types (Mumby et al., 2004b; Kutser near-infrared wavelength range to correct et al., 2006). The spectral refl ectance proper- values in the visible range (Hochberg et al., ties of various substrate types differ, which 2003a; Hedley et al., 2005). Another problem enables multispectral and hyperspectral with the use of data from Landsat TM and instruments to discriminate between them. SPOT HRV sensors has been their pixel The level of detail that can be obtained, typ- sizes, 30 m and 20 m, respectively, which

Downloaded from ppg.sagepub.com at SIMON FRASER LIBRARY on August 15, 2013 426 Progress in Physical Geography 31(4) result in most pixels containing a mix of sub- both universal and mappable with current strate types. The better spatial resolution of technology, these studies require intensive the IKONOS and Quickbird satellites has fi eldwork to determine the relation between reduced, but not eliminated, this problem, and the mapped geomorphologic zones and classifi cation accuracies have increased to the substrate types and their species composition. 75–85% range using broad substrate classes The remotely sensed information thus func- and around 50% for more detailed classes tions more as geolocation of fi eld observations (Mumby and Edwards, 2002; Andréfouët than as the primary information source, and et al., 2003). The relationship between the classification algorithms cannot be applied sensor type, the number of classes, and the beyond the studied fi eld site. classifi cation accuracy is well developed and Habitat variables known to directly infl u- can be used to design an optimum mapping ence biodiversity have been mapped sur- strategy for a given application (Mumby et al., prisingly rarely. Adjeroud et al. (2000) used 2004b). The development of hyperspectral SPOT HRV data to map pinnacle density, instruments has improved the degree to surface area, and hydrodynamic aperture of which accessory pigments can be used to nine atolls in French Polynesia, and found separate detailed classes, and they have that these explained part of the between-atoll therefore enabled mapping of detailed variation in species richness of investigated classes while retaining satisfactory mapping taxa (corals, molluscs, echinoderms and algae). accuracy (Mumby et al., 2004b; Kutser et al., Andréfouët and Guzman (2005) have found 2006). However, the dominant features of a non-significant correspondence between the spectral signatures used to discriminate geomorphologic zones and biodiversity in between typical coral reef substrates are in Panama, which they attribute to variation in the part of the visible spectrum where water reef structural complexity rather than geo- penetration is lowest, namely 550–700 nm morphology per se. However, no study has (Mobley, 1994; Kutser et al., 2003). This attempted to use remote sensing directly to rapidly reduces the number of distinct sub- predict spatial variation in biodiversity from strate types that can be distinguished as habitat variables. depth increases (Holden and LeDrew, A review of the literature showed that 1999; Capolsini et al., 2003; Hochberg and depth, live coral cover and reef structural Atkinson, 2003). In a best-case scenario, complexity all show correlations with several mapping shallow reefs with the airborne biodiversity variables. Depth is routinely hyperspectral CASI sensor, Mumby et al. mappable using a variety of remote sensing (1997) mapped nine substrate classes with an methods, and algorithms to derive live coral overall accuracy of 81%. cover from airborne hyperspectral data have been developed. Using remote sensing to V Mapping habitat variables and map coral reef structure has only recently biodiversity been explored, and different spatial scales A few recent studies have related the mapped of remote and in-situ measurements remain geomorphologic zones and substrate types an unresolved issue. Other habitat variables to the presence of specific species. These with known infl uences on species richness (eg, studies do not map individual species directly, distance to the reef edge or a nearby river) are but rather reef zones in which particular also routinely mappable. There is thus ample species are known from fi eld observations to scope for further exploring the possibility to be dominant (Turner and Klaus, 2005; Purkis predict the spatial distribution of biodiversity et al., 2006). As these species assemblages on coral reefs using habitat maps derived from do not have unique properties that are proven remotely sensed data.

Downloaded from ppg.sagepub.com at SIMON FRASER LIBRARY on August 15, 2013 Anders Knudby et al.: Remote sensing for coral reef biodiversity studies 427

1 Remote sensing of depth operate at depths up to 70 m, depending on Water depth can be mapped using acoustic, water optical quality (LaRocque et al., 2004; passive optical, or active optical instruments. Tenix, 2005). Both acoustic and active optical Acoustic instruments located at or below methods are both so well developed and the sea surface emit a sound pulse toward tested that in-situ verifi cation is typically un- the sea floor and measure the temporally necessary (IHO, 1998). resolved return pulse (the echo). With know- Water depth can also be mapped with pas- ledge of the speed and direction of the pulse, sive optical remote sensing when no better the vertical distance between the instrument data sources are available. Methods rely on the and the seafl oor can be calculated (Preston, wavelength dependency of light attenuation 2004). Nadir-pointing acoustic instruments in water; within the visible spectrum, longer emit pulses at beam angles of 15–25%, and wavelengths attenuate more rapidly (Mobley, register depth based on the earliest received 1994), hence substrates located in deeper water echo assumed to be directly underneath the will show a greater proportion of reflected instrument (Foster-Smith and Sotheran, light in shorter wavelengths (Lyzenga, 1978). 2003). Multibeam instruments emit fan- Variation in substrate spectral reflectance shaped pulses and employ multiple narrow introduces error, but can, at least in theory, be fi eld-of-view receivers to register the return adjusted for using hyperspectral data (Hedley pulse from several reflecting points on the and Mumby, 2003). Water optical properties sea floor (Mitchell and Clarke, 1994). The and substrates with very low refl ectance intro- movement of the instrument as it is towed duce additional complications that can also behind a vessel provides areal coverage and be partly mitigated (Philpot, 1989; Stumpf onboard GPS provides each measurement et al., 2003). Depth mapping with passive with a coordinate in space. Point spacing is optical remote sensing always requires in-situ determined by the speed of the towing vessel, calibration, and is rarely possible at depths water depth, and the proximity of surveyed greater than 25 m (Stumpf et al., 2003). tracks. Acoustic instruments can function in all but the shallowest depths (<50 cm) 2 Remote sensing of live coral cover (Mumby et al., 2004b). Mapping of live coral cover has been partly Active optical (lidar) instruments employ addressed by the substrate classification a similar principle, instead emitting and re- techniques mentioned above, but the presence ceiving a laser pulse of water-penetrating of more than one substrate type within a pixel wavelength, typically 532 nm, using an is problematic. Spectral unmixing, routinely instrument mounted underneath an aircraft used in terrestrial environments to deal with (Guenther, 2001). The temporally resolved subpixel heterogeneity, is complicated by return pulse (the waveform) displays refl ection the effects of the water column in addition peaks from both the water surface and the sub- to the strong similarity between the spectral strate. Areal coverage is provided by pulses signatures of different substrate types. If the emitted at frequencies up to 3 kHz, a spinning diffuse attenuation coeffi cient of the water mirror distributing light pulses across the column is known, linear spectral unmixing fl ight path, and the movement of the aircraft. can be modifi ed to include the infl uence of The spatial location of each reflecting sur- the water column, as long as all endmember face is determined by an onboard kinematic spectra are known (Hedley and Mumby, GPS and an Inertial Measurement Unit, 2003). However, spatially distributed values combined with knowledge of the direction of of the diffuse attenuation coefficient are the light pulse and its speed in air and water. rarely, if ever, known, and in practice end- Most commercial instruments can achieve member spectra show strong similarity be- a minimum point spacing of 2 m, and can cause of the dominance of chlorophyll a. This

Downloaded from ppg.sagepub.com at SIMON FRASER LIBRARY on August 15, 2013 428 Progress in Physical Geography 31(4) approach has not yet been developed into an the distance between the same two points as operational tool (Hedley et al., 2004), though measured in a straight line (Risk, 1972). Other it has been proven useful in conditions of in-situ measures of structural complexity shallow clear water (Mumby et al., 2004a). include vertical relief, fractal dimension, Disregarding proportions of non-coral number and volume of holes, percent cover substrate types, Isoun et al. (2003) designed a of branching corals, and ordinal scales based classifi cation scheme based only on percentage on visual estimation. Rugosity is dependent live coral cover, and achieved 77% overall on two scales of its measurement: the total classification accuracy using three narrow horizontal distance between the two end bands (10 nm FWHM) from an airborne sensor points and the detail with which the contour- optimized for coral reef benthos discrim- following length is measured. These scales ination. Going one step further, Joyce (2004) are referred to as spatial extent and grain, used an index-based approach to investigate respectively, and are key parameters in all correlations between live coral cover and spec- quantification of spatial structure (Wiens, tral refl ectance ratios and derivatives, using 1989; Marceau, 1999; Hay et al., 2001). Finer hyperspectral data from in-situ and airborne grain will register more roughness on a sur- measurements. She found that the optimum face and greater extent will incorporate band ratio and derivative varied between larger-scale topography, both leading to ‘blue’ and ‘brown’ coral types (Hochberg higher rugosity values. It is therefore unfortu- et al., 2003b), and depended on resampling nate that with few exceptions (Luckhurst of the hyperspectral data set, depth and and Luckhurst, 1978; Kostylev et al., 1997; water quality. From a survey on Heron Reef, 2005) this dependence seems to have gone Australia, she obtained a correlation of unnoticed in the literature, and that the two r = –0.76 between live coral cover and the scales of measurement vary between studies. ratio of refl ectances in the 529 nm and 439 Nevertheless, a review of the literature reveals nm bands of the airborne sensor. that the measure of rugosity is used more often than any other measure, and that the 3 Remote sensing of reef structure extent and grain are typically in the regions of The structural complexity of coral reefs has 3–10 m and 1–10 cm respectively (Table 2). received far less research attention from It therefore seems reasonable that remote remote sensing scientists than water depth sensing of reef structure for biodiversity and live coral cover. Structural complexity studies should operate close to these scales. is quantified in several ways and exists at Structural complexity can be quantified scales of varying importance for biodiver- from remote sensing data by developing sity, ranging from regional distribution of digital elevation models from depth data, reef complexes to the intricate coral skeleton and the spatial scale of these models is structure. To explore the potential for remote determined by the instruments used for data sensing to map structural complexity as it infl u- collection. Acoustic instruments can obtain ences biodiversity, it is necessary to review data sets with point spacing in the range of the quantification methods and spatial centimetres (Weber, 1996), and they are scales used in fi eld studies that have estab- therefore well suited for detailed analyses of lished the relationship. reef structure. Although such studies have not Structural complexity on coral reefs has yet been conducted, it is likely that remotely predominantly been measured using the sensed rugosity values would correspond chain-and-tape method. This method yields closely to those measured in situ, and thus the measure of rugosity, calculated as the be useful for biodiversity studies. However, ratio between the contour-following dis- this level of detail is achieved at the expense tance between two points on the reef and of areal coverage, particularly in the shallow

Downloaded from ppg.sagepub.com at SIMON FRASER LIBRARY on August 15, 2013 Anders Knudby et al.: Remote sensing for coral reef biodiversity studies 429

Table 2 Measures and scales of measurement in the quantiﬁ cation of structural complexity

Measure of structure Extent (m) Grain (cm) Source Rugosity 3 1.3 Friedlander and Parrish, 1998 Cover of branching corals n/a Chabanet et al., 1997 Rugosity 3 10 and n/a McCormick, 1994 Number and size of holes n/a Hixon and Beets, 1993 Microhabitats n/a Ormond et al., 1996 Rugosity 3 1.5 Luckhurst and Luckhurst, 1978 Rugosity ? ? Garpe and Ohman, 2003 Rugosity 10 Nylon line McClanahan, 1999 Rugosity 14 ? Chapman and Kramer, 1999 Fractal dimension 0.14/0.15 0.1 Kostylev et al., 1997; 2005 Rugosity 1–3.64 n/a Gratwicke and Speight, 2005 Vertical relief n/a Syms and Jones, 2000 Rugosity 10 ? McClanahan, 1988 Visual estimation n/a Kohn, 1968; Kohn and Leviten, 1976 Visual estimation 6-point scale Lara and Gonzalez, 1998 Rugosity 10 (transect) 1.3 Rogers et al., 1991 Vertical relief n/a Bainbridge and Reichelt, 1988 waters in which most coral reefs are found it is questionable whether this may be due to (Moyer et al., 2005; Riegl and Purkis, 2005). the 80 cm point spacing being inappropriate Most lidar instruments in operation have for this purpose (Knudby and LeDrew, 2007). been developed for mapping of bathymetry, Nevertheless, Wedding et al. (unpublished for which spatial detail is of less concern than data), working in Hawaii, have demonstrated depth penetration and areal coverage. As that lidar-based rugosity, captured with 4 m mentioned above, these instruments therefore point spacing, can predict fish biodiversity provide data at a minimal point spacing of 2 m. variables very accurately, rugosity values NASA’s Experimental Advanced Airborne explaining more than 60% of the variability Research Lidar (EAARL) is a research lidar in ﬁ sh biomass between different areas of a able to obtain a point spacing of 80 cm, the reef. It is likely that the obtainable correlations best spatial resolution currently offered by and the optimal spatial scale of rugosity an airborne lidar (USGS, 2006). Brock et al. measurements depend on the environment (2004) used transects of depth measurements under investigation, and the biodiversity from this instrument to quantify the structural variable investigated (Kuffner et al., 2007). complexity of patch and bank reefs in Florida, In terrestrial environments, lidar waveforms and later showed that abrupt depth variations have been analysed individually to investigate indicate presence of massive stony corals vertical structure of the surface within the (Brock et al., 2006), the main structural ‘footprint’ of each pulse. Forest structural elements on the investigated reefs. These variables such as canopy height, aboveground studies show that airborne lidar can provide biomass and stem basal area have all been information about reef structure at the scale estimated from measures of the shape of of large coral colonies, the main structural lidar waveforms. In marine environments, elements on coral reefs. However, correlation waveform analysis has been used to derive between the measures produced by Brock the Inherent Optical Properties (IOPs) of the et al. (2004) and biodiversity variables have water column (Kopilevich et al., 2005) and proven very weak (Kuffner et al., 2007), and improve estimates of substrate reﬂ ectance

Downloaded from ppg.sagepub.com at SIMON FRASER LIBRARY on August 15, 2013 430 Progress in Physical Geography 31(4)

(Tuell et al., 2005). Though not yet tested, it be mapped, in order to make full use of the is possible that waveform analysis can be used technology. Ultimately, the usefulness of re- to derive measures of structural complexity mote sensing technology for conservation of on coral reefs as has been done in forests, threatened coral reefs rests upon its ability to enabling such investigation to be undertaken map relevant environmental variables, pro- at a smaller spatial scale than possible so far. vide a scientifi c basis for decision-making, and For example, with an EAARL footprint dia- thereby aid effective conservation of coral meter of 20 cm, waveform analysis of individual reef biodiversity. footprints would bring the scale of structural complexity measured by remote sensing Acknowledgements much closer to that shown by fi eld studies to This research was funded through an NSERC infl uence biodiversity on coral reefs. grant to Ellsworth LeDrew, and benefited Passive optical instruments have also been from constructive discussions with John used in situ to document the effect of the Brock and Amar Nayegandhi at USGS. structure of coral colonies on their spectral reflectance, which has been attributed to References internal shading (Joyce and Phinn, 2002; Adjeroud, M., Andréfouët, S., Payri, C. and Minghelli-Roman et al., 2002). However, Orempuller, J. 2000: Physical factors of differen- tiation in macrobenthic communities between not all studies have found a signifi cant effect atoll lagoons in the Central Tuamotu Archipelago (Holden and LeDrew, 1999), and it is not clear (French Polynesia). Marine Ecology-Progress Series how upscaling to airborne or satellite meas- 196, 25–38. urements could be achieved. No studies have Allen, G.R., Steene, R., Humann, P. and DeLoach, successfully mapped reef structure at the scale N. 2003: Reef fish identification – tropical Pacific. Jacksonville: New World Publications. of coral colonies using data from airborne or Andréfouët, S. and Guzman, H.M. 2005: Coral reef spaceborne passive optical instruments. distribution, status and geomorphology-biodiversity relationship in Kuna Yala (San Blas) archipelago, VI Concluding remarks Caribbean Panama. Coral Reefs 24, 31–42. Remote sensing is now a standard tool used Andréfouët, S., Claereboudt, M., Matsakis, P., Pages, J. and Dufour, P. 2001: Typology of atoll in coral reef studies, and both sensors and rims in Tuamotu Archipelago (French Polynesia) data processing algorithms have become in- at landscape scale using SPOT HRV images. Inter- creasingly abundant and sophisticated with national Journal of Remote Sensing 22, 987–1004. time. However, no single instrument type can Andréfouët, S., Kramer, P., Torres-Pulliza, D., provide optimum data in all situations and for Joyce, K.E., Hochberg, E.J., Garza-Perez, R., Mumby, P.J., Riegl, B., Yamano, H., White, all purposes; instead, data from the different W.H., Zubia, M., Brock, J.C., Phinn, S.R., types of sensors can complement each other. Naseer, A., Hatcher, B.G. and Muller-Karger, The local distribution of coral reef biodiversity F.E. 2003: Multi-site evaluation of IKONOS data has been shown to be infl uenced by a number for classifi cation of tropical coral reef environments. of habitat variables such as depth, live coral Remote Sensing of Environment 88, 128–43. Bainbridge, S.J. and Reichelt, R.E. 1988: An assessment cover and structural complexity, each of which of ground truth methods for coral reef remote sensing can be mapped accurately with one or more data. Townsville: AIMS, 439–44. remote sensing instruments. Studies that Beger, M., Jones, G.P. and Munday, P.L. 2003: combine the strengths of each instrument Conservation of coral reef biodiversity: a comparison type and map multiple habitat variables are of reserve selection procedures for corals and fi shes. Biological Conservation 111, 53–62. needed to test our ability to predict the local Bellwood, D.R. and Hughes, T.P. 2001: Regional-scale spatial distribution of biodiversity, and to map assembly rules and biodiversity of coral reefs. Science areas of high biodiversity. In addition, studies 292, 1532–34. must investigate in greater detail the optimum Bellwood, D.R. and Wainwright, P.C. 2002: The spatial scale at which habitat variables should history and biogeography of fishes on coral reefs.

Downloaded from ppg.sagepub.com at SIMON FRASER LIBRARY on August 15, 2013 Anders Knudby et al.: Remote sensing for coral reef biodiversity studies 431

In Sale, P.F., editor, Coral reef fi shes: dynamics and Garpe, K.C. and Ohman, M.C. 2003: Coral and fi sh diversity in a complex ecosystem, San Diego: Academic distribution patterns in Mafia Island Marine Park, Press, 5–32. Tanzania: fish-habitat interactions. Hydrobiologia Birkeland, C., editor 1997: Life and death of coral reefs. 498, 191–211. New York: Chapman and Hall. Garpe, K.C., Yahya, S.A.S., Lindahl, U. and Briggs, J.C. 1999: Coincident biogeographic patterns: Ohman, M.C. 2006: Long-term effects of the 1998 Indo-West Pacifi c Ocean. Evolution 53, 326–35. coral bleaching event on reef fish assemblages. Brock, J.C., Wright, C.W., Clayton, T.D. and Marine Ecology-Progress Series 315, 237–47. Nayegandhi, A. 2004: LIDAR optical rugosity of Graham, N.A.J., Wilson, S.K., Jennings, S., coral reefs in Biscayne National Park, Florida. Coral Polunin, N.V.C., Bijoux, J.P. and Robinson, J. Reefs 23, 48–59. 2006: Dynamic fragility of oceanic coral reef Brock, J.C., Wright, C.W., Kuffner, I.B., ecosystems. Proceedings of the National Academy Hernandez, R. and Thompson, P. 2006: Airborne of Sciences of the United States of America 103, lidar sensing of massive stony coral colonies on patch 8425–29. reefs in the northern Florida reef tract. Remote Sensing Gratwicke, B. and Speight, M.R. 2005: Effects of Environment 104, 31–42. of habitat complexity on Caribbean marine fish Capolsini, P., Andréfouët, S., Rion, C. and Payri, C. assemblages. Marine Ecology-Progress Series 292, 2003: A comparison of Landsat ETM+, SPOT HRV, 301–10. Ikonos, ASTER, and airborne MASTER data for coral Green, E.P., Mumby, P.J., Edwards, A.J. and reef habitat mapping in south Pacifi c islands. Canadian Clark, C.D. 2000: Remote sensing handbook for Journal of Remote Sensing 29, 187–200. tropical coastal management. Paris: UNESCO. Chabanet, P., Ralambondrainy, H., Amanieu, Guenther, G.C. 2001: Airborne lidar bathymetry. In M., Faure, G. and Galzin, R. 1997: Relationships Maune, D.F., editor, Digital elevation model tech- between coral reef substrata and fish. Coral Reefs nologies and applications: the DEM users manual, 16, 93–102. Bethesda: The American Society for Photogrammetry Chapman, M.R. and Kramer, D.L. 1999: Gradients and Remote Sensing, 237–306. in coral reef fi sh density and size across the Barbados Hay, G.J., Marceau, D.J., Dube, P. and Bouchard, Marine Reserve boundary: effects of reserve pro- A. 2001: A multiscale framework for landscape tection and habitat characteristics. Marine Ecology- analysis: object-specific analysis and upscaling. Progress Series 181, 81–96. Landscape Ecology 16, 471–90. Coates, A.G. and Obando, J.A. 1996: The geologic Hedley, J.D. and Mumby, P.J. 2002: Biological and evolution of the Central American Isthmus. In remote sensing perspectives of pigmentation in Jackson, J.B.C., Budd, A.F. and Coates, A.G., coral reef organisms. Advances in Marine Biology 43, editor, Evolution and environment in tropical America, 277–317. Chicago: University of Chicago Press, 21–56. — 2003: A remote sensing method for resolving depth Connell, J.H. 1978: Diversity in tropical rain forests and subpixel composition of aquatic benthos. and coral reefs – high diversity of trees and corals is Limnology and Oceanography 48, 480–88. maintained only in a non-equilibrium state. Science Hedley, J.D., Harborne, A.R. and Mumby, P.J. 199, 1302–10. 2005: Simple and robust removal of sun glint for Connell, J.H., Hughes, T.P. and Wallace, C.C. 1997: mapping shallow-water benthos. International Journal A 30-year study of coral abundance, recruitment, of Remote Sensing 26, 2107–12. and disturbance at several scales in space and time. Hedley, J.D., Mumby, P.J., Joyce, K.E. and Phinn, Ecological Monographs 67, 461–88. S.R. 2004: Spectral unmixing of coral reef benthos Dubinsky, Z., editor 1990: Coral reefs. Amsterdam: under ideal conditions. Coral Reefs 23, 60–73. Elsevier. Hixon, M.A. and Beets, J.P. 1993: Predation, Fang, L.S., Liao, C.W. and Liu, M.C. 1995: Pigment prey refuges, and the structure of coral-reef fish composition in different-colored scleractinian corals assemblages. Ecological Monographs 63, 77–101. before and during the bleaching process. Zoological Hochberg, E.J. and Atkinson, M.J. 2003: Capabilities Studies 34, 10–17. of remote sensors to classify coral, algae, and sand as Foster-Smith, R.L. and Sotheran, I.S. 2003: Mapping pure and mixed spectra. Remote Sensing of Environment marine benthic biotopes using acoustic ground 85, 174–89. discrimination systems. International Journal of Remote Hochberg, E.J., Andréfouët, S. and Tyler, M.R. Sensing 24, 2761–84. 2003a: Sea surface correction of high spatial reso- Friedlander, A.M. and Parrish, J.D. 1998: Habitat lution Ikonos images to improve bottom mapping characteristics affecting fish assemblages on a in near-shore environments. IEEE Transactions on Hawaiian coral reef. Journal of Experimental Marine Geoscience and Remote Sensing 41, 1724–29. Biology and Ecology 224, 1–30.

Downloaded from ppg.sagepub.com at SIMON FRASER LIBRARY on August 15, 2013 432 Progress in Physical Geography 31(4)

Hochberg, E.J., Atkinson, M.J. and Andréfouët, S. in tropical intertidal predatory gastropod assemblages. 2003b: Spectral reflectance of coral reef bottom- Oecologia 25, 199–210. types worldwide and implications for coral reef Kopilevich, Y.I., Feygels, V.I., Tuell, G.H. and remote sensing. Remote Sensing of Environment 85, Surkov, A. 2005: Measurement of ocean water 159–73. optical properties and seafloor reflectance with Hodgson, G., Kiene, W., Mihaly, J., Liebeler, J., Scanning Hydrographic Operational Airborne Shuman, C. and Maun, L. 2004: Reef check instruc- Lidar Survey (SHOALS): I. Theoretical background. tion manual: a guide to reef check coral reef monitoring. In Proceedings, Remote Sensing of the Coastal Los Angeles: University of California, 92. Oceanic Environment, San Diego, 31 July, SPIE Hoegh-Guldberg, O. 1999: Climate change, coral 5885. bleaching and the future of the world’s coral reefs. Kostylev, V., Erlandsson, J. and Johannesson, K. Marine and Freshwater Research 50, 839–66. 1997: Microdistribution of the polymorphic snail Holden, H. and LeDrew, E. 1999: Hyperspectral Littorina saxatilis (Olivi) in a patchy rocky shore identification of coral reef features. International habitat. Ophelia 47, 1–12. Journal of Remote Sensing 20, 2545–63. Kostylev, V.E., Erlandsson, J., Ming, M.Y. and Hubbard, D.K. 1997: Reefs as dynamic systems. In Williams, G.A. 2005: The relative importance Birkeland, C., editor, Life and death of coral reefs, New of habitat complexity and surface area in assess- York: Chapman and Hall, 43–67. ing biodiversity: Fractal application on rocky Huston, M.A. 1994: Biological diversity: the coexistence of shores. Ecological Complexity 2, 272–86. species on changing landscapes. Cambridge: Cambridge Kuffner, I.B., Brock, J.C., Grober-Dunsmore, R., University Press. Bonito, V.E., Hickey, T.D. and Wright, C.W. International Hydrographic Organization (IHO) 2007: Relationships between reef fi sh communities 1998: IHO standards for hydrographic surveys. Monaco: and remotely sensed rugosity measurements in International Hydrographic Organization, 1–26. Biscayne National Park, Florida, USA. Environmental Isoun, E., Fletcher, C., Frazer, N. and Gradie, J. Biology of Fishes 78, 71–82. 2003: Multi-spectral mapping of reef bathymetry Kutser, T., Dekker, A.G. and Skirving, W. 2003: and coral cover; Kailua Bay, Hawaii. Coral Reefs 22, Modeling spectral discrimination of Great Barrier Reef 68–82. benthic communities by remote sensing instruments. Jones, G.P. and Syms, C. 1998: Disturbance, habitat Limnology and Oceanography 48, 497–510. structure and the ecology of fishes on coral reefs. Kutser, T., Miller, I. and Jupp, D.L.B. 2006: Mapping Australian Journal of Ecology 23, 287–97. coral reef benthic substrates using hyperspectral Jones, G.P., McCormick, M.I., Srinivasan, M. space-borne images and spectral libraries. Estuarine and Eagle, J.V. 2004: Coral decline threatens fi sh Coastal and Shelf Science 70, 449–60. biodiversity in marine reserves. Proceedings of the Lara, E.N. and Gonzalez, E.A. 1998: The relationship National Academy of Sciences of the United States of between reef fi sh community structure and environ- America 101, 8251–53. mental variables in the southern Mexican Caribbean. Joyce, K. 2004: A method for mapping live coral cover Journal of Fish Biology 53, 209–21. using remote sensing. Brisbane: School of Geography, LaRocque, P.E., Banic, J.R. and Cunningham, University of Queensland, 137. A.G. 2004: Design description and fi eld testing of the Joyce, K.E. and Phinn, S.R. 2002: Bi-directional SHOALS-1000T airborne bathymeter. In Kamerman, refl ectance of corals. International Journal of Remote G.W., editor, Proceedings, Laser Radar Technology Sensing 23, 389–94. and Applications IX, SPIE 5412, 162–84. Karlson, R.H. 1999: Dynamics of coral communities. Luckhurst, B.E. and Luckhurst, K. 1978: Analysis Dordrecht: Kluwer Academic. of infl uence of substrate variables on coral-reef fi sh Kleypas, J.A., Buddemeier, R.W., Archer, D., communities. Marine Biology 49, 317–23. Gattuso, J.P., Langdon, C. and Opdyke, B.N. Lyzenga, D.R. 1978: Passive remote-sensing techniques 1999: Geochemical consequences of increased for mapping water depth and bottom features. Applied atmospheric carbon dioxide on coral reefs. Science Optics 17, 379–83. 284, 118–20. Marceau, D.J. 1999: The scale issue in the social and Knudby, A. and LeDrew, E. 2007: Measuring structural natural sciences. Canadian Journal of Remote Sensing complexity on coral reefs. In Proceedings, AAUS 24, 347–55. Annual Symposium, Miami, USA, 9–10 March. McClanahan, T.R. 1988: Coexistence in a sea-urchin Kohn, A.J. 1968: Microhabitats abundance and food of guild and its implications to coral-reef diversity and Conus on atoll reefs in Maldive and Chagos Islands. degradation. Oecologia 77, 210–18. Ecology 49, 1046–62. — 1999: Predation and the control of the sea urchin Kohn, A.J. and Leviten, P.J. 1976: Effect of habitat Echinometra viridis and fl eshy algae in the patch reefs complexity on population-density and species richness of Glovers Reef, Belize. Ecosystems 2, 511–23.

Downloaded from ppg.sagepub.com at SIMON FRASER LIBRARY on August 15, 2013 Anders Knudby et al.: Remote sensing for coral reef biodiversity studies 433

McCormick, M.I. 1994: Comparison of fi eld methods Paulay, G. 1997: Diversity and distribution of reef for measuring surface-topography and their associ- organisms. In Birkeland, C., editor, Life and death of ations with a tropical reef fi sh assemblage. Marine coral reefs, New York: Chapman and Hall, 298–353. Ecology-Progress Series 112, 87–96. Philpot, W.D. 1989: Bathymetric mapping with Minghelli-Roman, A., Chisholm, J.R.M., passive multispectral imagery. Applied Optics 28, Marchioretti, M. and Jaubert, J.M. 2002: 1569–78. Discrimination of coral reflectance spectra in the Pittock, A.B. 1999: Coral reefs and environmental Red Sea. Coral Reefs 21, 307–14. change: adaptation to what? American Zoologist 39, Mitchell, N.C. and Clarke, J.E.H. 1994: Classifi cation 10–29. of sea-floor geology using multibeam sonar data Preston, J.M. 2004: Resampling sonar echo time series from the Scotian Shelf. Marine Geology 121, 143–60. primarily for seabed sediment classification. USA: Moberg, F. and Folke, C. 1999: Ecological goods and Quester Tangent Corporation, 12. services of coral reef ecosystems. Ecological Economics Purkis, S.J., Myint, S.W. and Riegl, B.M. 2006: 29, 215–33. Enhanced detection of the coral Acropora cervicornis Mobley, C.D. 1994: Light and water: radiative transfer in from satellite imagery using a textural operator. natural waters. San Diego: Academic Press. Remote Sensing of Environment 101, 82–94. Molles, M.C. 1978: Fish species-diversity on model Reaka-Kudla, M.L. 1997: The global biodiversity and natural reef patches – experimental insular of coral reefs: a comparison with rain forests. In biogeography. Ecological Monographs 48, 289–305. Reaka-Kudla, M.L., Wilson, D.E. and Wilson, E.O., Moran, D.P. and Reaka, M.L. 1988: Bioerosion and editors, Biodiversity II, Washington, DC: Joseph availability of shelter for benthic reef organisms. Henry Press, 83–107. Marine Ecology-Progress Series 44, 249–63. Riegl, B.M. and Purkis, S.J. 2005: Detection of shallow Moyer, R.P., Riegl, B., Banks, K. and Dodge, R.E. subtidal corals from IKONOS satellite and QTC 2005: Assessing the accuracy of acoustic seabed View (50, 200 kHz) single-beam sonar data (Arabian classification for mapping coral reef environments Gulf; Dubai, UAE). Remote Sensing of Environment in South Florida (Broward County, USA). Revista 95, 96–114. De Biologia Tropical 53, 175–84. Risk, M.J. 1972: Fish diversity on a coral reef in the Mumby, P.J. and Edwards, A.J. 2002: Mapping Virgin Islands. Atoll Research Bulletin 153, 1–6. marine environments with IKONOS imagery: Rogers, C.S., McLain, L.N. and Tobias, C.R. 1991: enhanced spatial resolution can deliver greater Effects of Hurricane Hugo (1989) on a coral-reef in thematic accuracy. Remote Sensing of Environment St-John, USVI. Marine Ecology-Progress Series 78, 82, 248–57. 189–99. Mumby, P.J. and Harborne, A.R. 1999: Develop- Sale, P.F. 1991: Habitat structure and recruitment ment of a systematic classification scheme of in coral reef fi shes. In Bell, S.S., McCoy, E.D. and marine habitats to facilitate regional management Mushinsky, H.R., editors, Habitat structure – the and mapping of Caribbean coral reefs. Biological physical arrangement of objects in space, London: Conservation 88, 155–63. Chapman and Hall, 438 pp. Mumby, P.J., Green, E.P., Edwards, A.J. and Sale, P.F. and Dybdahl, R. 1975: Determinants of Clark, C.D. 1997: Coral reef habitat-mapping: how community structure for coral-reef fishes in an much detail can remote sensing provide? Marine experimental habitat. Ecology 56, 1343–55. Biology 130, 193–202. Smith, V.E., Rogers, R.H. and Reed, L.E. 1975: Mumby, P.J., Hedley, J.D., Chisholm, J.R.M., Automated mapping and inventory of Great Barrier Clark, C.D., Ripley, H. and Jaubert, J. 2004a: Reef zonation with Landsat. OCEANS 7, 775–80. The cover of living and dead corals from airborne Stumpf, R.P., Holderied, K. and Sinclair, M. remote sensing. Coral Reefs 23, 171–83. 2003: Determination of water depth with high- Mumby, P.J., Skirving, W., Strong, A.E., Hardy, resolution satellite imagery over variable bottom types. J.T., LeDrew, E.F., Hochberg, E.J., Stumpf, Limnology and Oceanography 48, 547–56. R.P. and David, L.T. 2004b: Remote sensing of coral Suzuki, H., Matsakis, P., Andréfouët, S. and reefs and their physical environment. Marine Pollution Desachy, J. 2001: Satellite image classification Bulletin 48, 219–28. using expert structural knowledge: a method based Newell, N.D. 1972: Evolution of reefs. Scientific on fuzzy partition computation and simulated American 226, 54–65. annealing. In Proceedings, Annual Conference of the Ormond, R.F.G., Roberts, J.M. and Jan, R.Q. International Association for Mathematical Geology, 1996: Behavioural differences in microhabitat use IAMG, Cancun, Mexico. by damselfishes (Pomacentridae): Implications for Syms, C. and Jones, G.P. 2000: Disturbance, habitat reef fi sh biodiversity. Journal of Experimental Marine structure, and the dynamics of a coral-reef fish Biology and Ecology 202, 85–95. community. Ecology 81, 2714–29.

Downloaded from ppg.sagepub.com at SIMON FRASER LIBRARY on August 15, 2013 434 Progress in Physical Geography 31(4)

Tenix 2005: Laser Airborne Depth Sounder, LADS Mk Ward, T.J., Vanderklift, M.A., Nicholls, A.O. and II. Retrieved 14 June 2007 from http://www.tenix. Kenchington, R.A. 1999: Selecting marine reserves com/Main.asp?ID=116 using habitats and species assemblages as surrogates Tuell, G.H., Feygels, V., Kopilevich, Y., for biological diversity. Ecological Applications 9, Weidemann, A.D., Cunningham, A.G., Mani, 691–98. R., Podoba, V., Ramnath, V., Park, J.Y. and Webb, G.E. 1998: Earliest known Carboniferous Aitken, J. 2005: Measurement of ocean water shallow-water reefs, Gudman Formation (Tn1b), optical properties and seafloor reflectance with Queensland, Australia: implications for late Devonian Scanning Hydrographic Operational Airborne reef collapse and recovery. Geology 26, 951–54. Lidar Survey (SHOALS): II. Practical results and Weber, T. 1996: A new generation of multi-beam comparison with independent data. Proceedings, sweeping echosounder for surveying – ATLAS Remote Sensing of the Coastal Oceanic Environ- FANSWEEP 20. In Proceedings, OCEANS ’96, ment, San Diego, USA, 31 July. Prospects for the 21st Century, Fort Lauderdale, USA, Turner, J. and Klaus, R. 2005: Coral reefs of the 23–26 September. Mascarenes, Western Indian Ocean. Philosophical Wiens, J.A. 1989: Spatial scaling in ecology. Functional Transactions of the Royal Society of London Series A Ecology 3, 385–97. – Mathematical Physical and Engineering Sciences Wilkinson, C. 2004: Status of coral reefs of the world. 363, 229–50. Townsville: Australian Institute of Marine Science. US Geological Survey (USGS) 2006: EAARL – Wood, R. 1998: The ecological evolution of reefs. Annual Airborne lidar system for high-resolution submerged Review of Ecology and Systematics 29, 179–206. and sub-aerial topography. Retrieved 14 June 2007 Yonge, C.M. 1940: The biology of reef building corals. from http://coastal.er.usgs.gov/remote-sensing/ London: British Museum, 353–91. advancedmethods/eaarl.html Veron, J.E.N. 1995: Corals in space and time: the biogeography and evolution of the Scleratinia. Ithaca, NY: Cornell University Press.

Downloaded from ppg.sagepub.com at SIMON FRASER LIBRARY on August 15, 2013