Reef monitoring sampling methods
A map of Queensland and the GBR showing sectors or regions where data is collected. Sectors labelled in blue have permanent survey sites. Those labelled in black are sampled using broadscale surveys only.
The AIMS Long-term Monitoring Program (LTMP) is designed to detect changes in reef communities at a subregional scale. In this context, a subregion encompasses inshore, mid-shelf and outer shelf reefs across the continental shelf within one band of latitude (a sector).
Reef surveys involve three approaches:
- broadscale manta tow surveys of crown-of-thorns starfish populations and reef-wide coral cover
- intensive photographic surveys of stationary seafloor (benthic) organisms on fixed transects
- intensive visual counts of reef fish, juvenile corals, crown-of-thorns starfish, coral-eating snails and coral disease and bleaching.
The Inshore Marine Monitoring Program (MMP) is designed to detect changes in reef communities adjacent to the Wet Tropics, Burdekin, Mackay, Whitsunday and Fitzroy Natural Resource Management Regions.
The core survey reef sample comprises two components:
- the perimeter of each reef is surveyed using manta tows
- fish and seafloor organisms are surveyed intensively at three sites, in a habitat that is standardised across reefs.
The intensive survey sites are located in the first stretch of continuous reef (excluding vertical drop-offs) to be encountered when following the perimeter from the back reef zone towards the front reef in a clockwise direction, usually on the north-east flank of the reef.
Sites are at least 250 m apart where possible. Within each site, five 50 m transects (paths along which seafloor organisms will be studied) are laid along the reef slope parallel to the reef crest at about 6-9 m depth. Transects are marked with a star picket at each end and with lengths of reinforcing rod at 10 m intervals.
A schematic diagram of the sampling design on a single reef.
In inshore waters, higher turbidity means that neither the manta tow or fish surveys are included in the Marine Monitoring Program (MMP). This high turbidity means that coral communities can be very different at different depths and reefs are also less extensive. For these reasons the sampling design of the MMP includes sampling at two depths, 2 m and 5 m at only two sites. The length of the five transects at each site is also reduced to 20 m.
Broadscale (manta tow) surveys
AIMS began broadscale surveys of the Reef in the early 1980s. These surveys were incorporated into the Long-term Monitoring Program in 1992. Broadscale surveys cover reefs in 11 sectors. Reefs in six of the sectors are surveyed intensively.
The primary objective of the broadscale surveys is to detect and monitor populations of crown-of-thorns starfish. Manta tow surveys also provide estimates of percentage cover of living hard corals, living soft coral and recently dead hard coral, allowing assessment of the impact of starfish outbreaks and other large-scale disturbances.
Manta tow sampling
AIMS diver using a manta board
The percentage of the perimeter of each reef that is covered with living hard and soft coral and dead hard coral is calculated from the manta tow results.
The manta tow technique is used to provide a general description of large areas of reef and to gauge broad changes in abundance and distribution of organisms on coral reefs. The advantage of manta tow over other survey techniques is that it enables large areas of reefs to be surveyed quickly and with minimal equipment.
The technique involves towing a snorkel diver (observer) at a constant speed behind a boat. The observer holds on to a manta board attached to the boat by a 17-metre length of rope. The observer makes a visual assessment of specific variables during each two-minute manta tow, and records these data on a data sheet attached to the manta board when the boat stops.
Results from Manta tows
Coral cover, the number of crown-of-thorns starfish per reef, and the average number of the starfish per tow are used to assess the outbreak status of each reef. There are four outbreak categories: active outbreak, incipient outbreak, recovering, and no recent outbreak.
An ‘active outbreak’ is defined as at least 1.0 crown-of-thorns starfish per tow. This represents a density that is highly likely to cause a net decline in corals. A density of 0.22 starfish per tow is referred to as ‘incipient outbreak’ level.
Manta tow estimates were compared with intensive scuba surveys undertaken by the Great Barrier Reef Marine Park Authority and with live coral estimates obtained from video footage (Reefwatch Australia). The level of agreement found suggest that broadscale surveys methods are effective in monitoring starfish populations and for estimating coral cover with enough accuracy to detect moderate- to large-scale disturbance and recovery.
For more detailed information see Standard operational procedure number 9.
Fixed transects at intensive survey sites
Seafloor organisms and reef fish are surveyed on the five marked transects within each site. LTMP core reefs are currently surveyed every second year (in odd-numbered years) while a second set of reefs chosen to assess the effect of the 2004 Great Barrier Reef Marine Park re-zoning plan is surveyed in the alternate (even-numbered) years.
Inshore reefs surveyed as part of the Marine Monitoring Plan (MMP) are also surveyed every second year, although are sometimes visited annually to document the effects of large disturbances such as coral bleaching or cyclones.
Seafloor photographic surveys on fixed transects
Holding digital still cameras approximately 40 cm above the substrate, the scientists take 50 photographs at approximately 1m intervals along each transect. Percentage cover of corals and other benthic categories are estimated from five points on each image, so approximately 200 systematically dispersed points are sampled from each transect.
Christie and Mapstone (unpublished) sampled images to investigate how the distribution of sampling points affects estimates of percentage cover. They found that estimates based on a face-centred cubic pattern of five points per screen did not differ from estimates based on single random points, so using fixed points represents an efficient sampling strategy for estimating the percentage cover when using basic equipment configurations.
They also showed that 750 random points per transect represents the ‘best’ possible sampling intensity for transects filmed at this focal distance. A comparison of mean and precision estimates for fewer than 750 points indicated that at 200–250 points there was no significant difference for any benthic group or lifeform.
The MMP uses the same sample methods. However, because of the smaller area of many inshore reefs, sampling is compressed so that 40 photographs are taken tat 50cm intervals along each transect.
Accuracy and observer variability
Hard and soft corals are typically identified to genus level. Algae and other organisms are placed into functional groups. A comparison of identifications made in the field with the video footage found that ‘benthic groups’ and ‘families of hard coral’ could be identified most accurately from video images (both with an accuracy around 90 per cent). Life forms of hard corals proved to be the least accurate and most variable level of classification (74 per cent). Observer error was small relative to estimates of cover.
For more detailed information see Standard operational procedure number 10.
Counts of reef fishes on fixed transects
Fish species, representing 10 families, are counted on the five 50 m transects in each of three sites at each reef. All species are largely non-cryptic (‘cryptic’ fish are fish that are often hard to see, for example because they hide) and are easily identified underwater. They include commercial and non-commercial fish.
Because the surveys span the annual recruitment season, fish less than one year old are excluded from counts. These are distinguished by their small size and often distinctive colouration. Large, mobile fish are counted using a 5 m wide belt transect, while damselfishes are counted separately on transects 1 m wide. The lengths of all large predatory fishes are estimated visually.
The general survey procedure at each site involves an experienced observer swimming along the transect line counting large mobile fishes to a distance of 2.5 m either side of the mid-line. An assistant swims approximately 10 m behind the observer and lays a surveyor's tape over the substratum along the centre line of the transect. When large, mobile fish have been counted on the five transects of a site, an observer returns along the same transects (now marked with a tape along the centre line) and counts damselfishes (Pomacentridae) on a transect 1 m wide.
For more detailed information see Standard operational procedure number 3.
SCUBA searches on fixed transects
Scuba searches are designed to provide a more detailed picture of the causes and relative scale of coral mortality than is possible with either the manta tow technique or the video technique. Scuba searches are made on the fixed transects. A 2 m belt (1 m either side of the central tape measure) is visually searched along each 50m transect (or 20m for the MMP), and numbers are recorded for the following: crown-of-thorns starfish, scars caused by that starfish, Drupella species (coral-feeding snails), Drupella species scars, white syndrome coral disease, black band coral disease and unknown scars. An estimate of coral bleaching as a percentage of live coral cover is also recorded.
For more detailed information see Standard operational procedure number 9.
Counts of juvenile hard corals
Recruitment is a basic demographic process, and counts of juvenile hard corals are helpful for predicting how coral communities may change in following years.
Juvenile hard corals that are defined as colonies that measure less than 5cm diameter. Since 2007, these juveniles have been counted in a belt 34 cm wide (length of dive slate) along the first 5 m of each transect. The MMP counts juvenile corals in a 34 cm wide belt along the full 20 m of each of their transects.