CoralASSIST: assessing the feasibility of Assisted Gene Flow as a large-scale and low-cost coral reef restoration approach

Assisted Gene Flow (AGF) involves the deliberate movement of individuals or gametes within their natural range to facilitate adaptation to environmental change. While AGF has the potential to facilitate adaptation there are also potential risks, for example costs to recipient population fitness due to resource trade-offs between traits.

Growth and contraction of ecosystem engineers, such as trees and corals, influence ecosystem structure and function. Dramatic changes in biodiversity are inevitable in the face of manmade global change. Although species have the capacity to adapt, it is unclear in many cases whether rates of adaptation are sufficient to cope with the current rate of climate change. While corals have the potential to adapt there are also potential risks, such as costs to recipient population fitness due to resource trade-offs between traits. For instance, are there resource trade-offs between increased thermo-tolerance and coral growth?

CoralASSIST is a project led by Dr. James Guest and Dr. Adriana Humanes at Newcastle University. My role in the project  is to use 3D models to reveal trade-offs between coral growth and thermo-tolerance.

This project uses 3D models from in situ coral colonies to measure coral growth over time without manipulating or damaging corals. We are following 300 colonies over time and comparing their growth rate, measured using 3D models, with their thermo-tolerance.

This is the first time that 3D models are used to investigate this trade-off in situ. These data will help parametrise models that predict the status of future coral reefs. Future reefs may adapt to projected ocean warming, yet their rate of growth could decrease, likely impacting a broad range of important ecosystem services. These results improve our capacity to measure the drivers underpinning ecosystem biodiversity, status and trajectory.

Fish smart seawalls

Our review on fish-smart seawalls finally was published in Frontiers in Ecology and the Environment! It summarises what are the key habitat features that fish are influenced by, and includes a decision tool for ecologists and managers to tailor the structural complexity of marine infrastructure for fish.

In a nutshell we found that:

  • Ecological engineering can mitigate the negative impacts of built infrastructure on marine assemblages.
  • Fishes are known to respond to certain habitat features on both natural and artificial reefs.
  • Knowledge hained from the association of fish with natural and artificial features can be applied in marine management to create “fish friendly” infrastructure.

To know more go to:

Screen Shot 2018-05-17 at 10.22.09 am

Ecosystem risk assessment

Ecosystem risk assessment is crucial for monitoring the impact of management interventions. The International Union for Conservation of Nature (IUCN) has a great team working on global, regional and local assessment of ecosystems: The Red List of Ecosystems team. I have collaborated with them in a couple of projects that were published in 2017 and 2018.

The paper led by Dr. Nick Murray is an excellent review on The role of remote sensing in structured assessments of ecosystem status. It highlights how:

  1. Unstructured use of remote sensing data for assessing ecosystem dynamics can introduce substantial error and uncertainty.
  2. It, then identifies case studies that have used satellite remote sensing to assess degradation of marine, aquatic and terrestrial ecosystem types.
  3. It provides guidance and a framework for integrating remote sensing data into ecosystem risk assessment.

Palau Echinopora sp. COTs Maygald (3)

Almost as useful is the paper led by Dr. Lucy Bland, which is the first red list of ecosystems assessment of a coral reef, and was published in 2017 in the Proceedings of the Royal Scoiety B. this work found that the Mesoamerican Barrier Reef System, the second longest barrier reef in the world, is Critically Endangered (plausibly Vulnerable to Critically Endangered), with notable differences among Red List criteria and data types in detecting the most severe symptoms of risk. This study provides a template for assessing risks to coral reefs and for further application of ecosystem models in risk assessment.



Marine Protected Area assessment

Not all my research is about 3D reefs. I also spend significant time thinking about the effectiveness of marine spatial management, such as marine protected areas. For example in 2016-2017 I supervised a study investigating the effectiveness of small no-take reserves in South Eastern Australia.

This work was published by Biodiversity and Conservation and identified the key drivers of fish, invertebrate and benthic communities. The study found that small reserves can enhance biodiversity and biomass on a local scale, but only if they have full no-take protection, are in sheltered locations with complex habitat, and have positive community involvement to engender support and stewardship.

Fishies_Lumes_LIRS - 1

I also led a much larger scale study looking at the effectiveness of no-take zones in MPAs on benthic communities. We used images from Autonomous Underwater Vehicles to quantify the benthic biota across three MPAs and seven degrees of latitude.

The study potential short-term effects of zoning (up to 10 years) on benthic communities, with important habitat-forming biota being more prevalent and abundant in no-take zones. This study advanced knowledge of marine benthic communities and their conservation in three ways.

  1. It generated the first baseline of biodiversity and abundance of benthic communities in the studied reefs.
  2. It identified the taxonomic resolution necessary to assess both short and long-term effects of MPAs, concluding that coarse taxonomic resolution is sufficient.
  3. It provided an example of statistical analyses and sampling design that is useful to detect changes of marine benthic communities across multiple spatial and temporal scales.

This slideshow requires JavaScript.

Measuring coral growth and contraction in 3D

3D models can be used to measure growth of corals and external erosion of skeletons.

I recently lead a paper presenting a novel method to measure coral growth and external erosion!!! There are no more excuses to use old school methods that harm corals (and potentially confound results) to measure their growth, now we can measure change in volume and surface area without even touching them!!!

Measuring growth of tabulate corals like this one

3D models of corals were used to measure the growth and erosion of coral colonies and skeletons in situ over a year. This has not been done before with the fine scale resolution approach that we have taken. We found that dead coral colonies eroded by 52%, while live colonies grew by 20% of their original volume and surface area. 3D models are a very accurate and precise method to quantify both volume and surface area of coral colonies in situ without manipulating them, avoiding negative effects on corals associated with more invasive methods. Our metrics were within the range of previously reported values, which we also briefly review.

This collection is part of the electronic supplementary material of a manuscript published in Scientific Reports (

The 3D models can be found here:

Please refer to it as: Ferrari et al. (2017) 3D photogrammetry quantifies growth and external erosion of individual coral colonies and skeletons. Scientific Reports 7 (16737) DOI:10.1038/s41598-017-16408-z

 Fish and 3D Reefs

3D Habitat Complexity metrics improve predictions of fish distribution and abundance

Our paper on how to use surface rugosity and other 3D metrics to predict fish distributions and abundance was published by Ecography recently!

Trumpet fish camouflages while hunting amongst the branches of a soft coral

We used underwater robots to map in 3D the ocean floor and then measured different aspects of the reef structural complexity. We also used bated underwater cameras to measure fish abundance and developed advanced statistical models to relate and predict the distribution and abundance of fish with 3D habitat complexity metrics!

Turns out not all reefs with high habitat complexity have more fish, neither more diversity! This was a surprise, but once I thought about it it made sense. Some fish, for example predators, do like complex reefs while others, like herbivores, prefer flatter reefs. Perhaps this is because herbivores like having a bit of space to detect predators? Or because they need more space to be able to flee a predator before they get caught?

Learn more in:

Fish hide amongst corals

Accuracy and Precision of 3D models

Habitat structural complexity is one of the most important factors in structuring biological communities. Recent advances in structure-from-motion and photogrammetry, like the ones used to make the 3D maps featured in this website, allow non-experts to reconstruct 3D digital representations of habitats and organisms from which structural complexity can be measured.

But little attention has been given to quantifying the errors in using these techniques (but see Figueira et al. 2015 in Remote Sensing) and variability in results under different surveying and environmental conditions, errors that are important when we want to examine changes in habitat complexity over space and time.

A super useful paper Characterizing measurement errors of habitat structural complexity from marine ecosystem 3D models was published by Ecology and Evolution!

Fig. 4 from Bryson et al. 2017 Imagery mosaics from multi-day surveys at Horseshoe Reef, Lizard Island, Great Barrier Reef, Australia.

In this paper we evaluated the accuracy, precision and bias in measurements of marine habitat structural complexity derived from 3D maps, using repeated surveys of artificial reefs (with known structure) and natural coral reefs. We quantified measurement errors as a function of survey image coverage, actual surface rugosity and the morphological community composition of corals.

We discovered habitat complexity measurements were between 8 -15% lower for in-water reconstructions when compared to the reference reconstruction, but highly repeatable with a standard deviation of 0.065 (4.9% of the true rugosity) for a resolution of 2.5 cm. During our in situ surveys, we discovered measurements could be biased by up to 7.5% because of varying environmental conditions across surveys. The errors were larger (~7.5%) at more complex sites dominated by coarse branching corals, compared to less complex sites dominated by massive or plating corals.

Figure 11 from Bryson et al. 2017 Relationships between 2 × 2m quadrat rugosity errors and quadrat rugosity modeled using OLS

The logistical advantages of these techniques in marine habitats mean they are likely to replace traditional techniques (such as chain-and-tape) in future ecologically-focused research. The quantitative relationships demonstrated in this study have important implications for data collection and the interpretation of measurements when examining changes in habitat complexity using 3D models.