Local dive master Joel Pandino 
shows a coral reef formation in the 
Verde sea passage south of Manila 
on February 21, 2007

The ban on "Danish seine" fishing nets will take effect in mid-March, giving hundreds of fishing boats time to switch their equipment, said Asis Perez, head of the Bureau of Fisheries and Aquatic Resources.

"This is because of their adverse effect on the environment. They (the nets) have weights attached to them, it drags on the ocean floor, it hits the corals and damages the marine life," he told AFP.

Perez said the bureau took action after small fishermen and coastal communities made numerous complaints about big fishing boats using the nets, which are weighted more heavily than other forms of net used to trawl the seabed.

The bureau's licensing division said that at least 795 fishing boats were using this type of net and would be affected by the ban.

A Danish seine is a contraption consisting of a conical net connected to herding devices and heavy sinkers that are hauled typically using mechanical winches.



September 24th, 2013

Catlin Global Reef Record is a free online resource that will become invaluable to scientific researchers.
The Catlin Seaview Survey last year made one of the great wonders of the world – Australia's Great Barrier Reef – accessible to anyone with an internet connection.

Now the visual access to the reef that has been seen on Google Street View will go even further, with the launch of the Catlin Global Reef Record, a free online resource that will make the survey's imagery invaluable to scientific researchers.

"The Catlin Global Reef Record is a game-changing analytic tool that scientists and reef managers around the world now have at their fingertips," said UQ's Professor Ove Hoegh-Guldberg, who is also chief scientist of the Catlin Seaview Survey.

Visitors to the online library : will be able to explore approximately 180,000 panoramic underwater coral reef images with another 200,000 panoramas expected by late 2014.

"Through the Record we will be able to monitor change in marine environments now and in the future," Professor Hoegh-Guldberg said.

"It will be available to high schools, universities and scientists working on coral research."
An estimated 50 per cent of coral reefs worldwide have been lost in the past 50 years, with 75 per cent of coral reefs today threatened by local and global stressors, including overfishing, pollution, unsustainable coastal development, ocean acidification and ocean warming.

Professor Hoegh-Guldberg, the Director of UQ's Global Change Institute (GCI), said many countries did not have the resources required to regularly measure the health of their coral reef ecosystems.
"As a result, there is often limited baseline data available for identifying the drivers of change on coral reefs," he said.
"Without this information, understanding change and implementing coastal management strategies for arresting the downward trend in the condition of coral reefs can be extremely difficult."

The Catlin Global Reef Record is an initiative of the non-for-profit, Underwater Earth, and developed in collaboration with the GCI, the project's lead scientific partner.
Professor Hoegh-Guldberg said understanding change to coral reefs is important because almost 25 per cent of marine species live in and around coral reefs and one-eighth of the world's population expects to draw on marine resources, such as coral reefs, for their livelihood and wellbeing.

He said the images in the Record were scanned for coral species and automatically annotated using computer vision algorithms.
"The footage, meanwhile, is used to create 3D reconstructions of reef ecosystems, providing a visual way to assess reef populations," he said.

Additional environmental data from satellites, such as ocean temperature, as well as information on regional coral bleaching activity, is included to allow for advanced analysis of worldwide reef health.
The data is captured using a custom-designed underwater camera, which integrates three digital SLR cameras positioned at an angle that allows the Catlin Seaview Survey team to record 360-degree panoramas.

Divers navigate the camera for two-kilometre transects along the reef, capturing high-definition imagery every three seconds.
By the end of 2014, about 300 reef locations will have been recorded, including the Great Barrier Reef, 10 countries in the greater Caribbean region, and the Coral Triangle region.

While many of the stresses on coral reefs originate from local sources, such as fishing activity and pollution, there is now urgent concern over the impact of ocean warming and acidification on coral reefs.
Professor Hoegh-Guldberg said if current trends in greenhouse gas emissions continued, atmospheric CO2 was expected to increase to more than 80 per cent above pre-industrial levels by 2050 – a rate of increase which had few, if any, parallels in the past 50 million years.

The effect of this rapid increase on marine ecosystems, in particular the Great Barrier Reef, would be devastating.
"The magnitude and rate of increases in sea surface temperatures and ocean acidity – both caused by the increasing concentrations of greenhouse gases, including CO2, in the atmosphere – is likely to exceed the ability of many marine species to adapt and survive," he said.
"The Catlin Global Reef Record is one way we can help turn this around."

Explore further: 'Street-view' comes to the world's coral reefs

More information:

Provided by University of Queensland



Coral ecologists are seeking to document changes to reefs, like this one near Yanuca, Fiji, by taking underwater photos and having a computer evaluate the images automatically. Photo courtesy of David Kline.

Laura Petersen, E&E reporter 
Greenwire: Tuesday, September 3, 2013

LA JOLLA, Calif. -- Coral ecologists are taking a page out of the national security playbook: facial recognition technology.

Techniques that can pick a terrorist out of a crowd are being adapted for underwater images, producing data on reefs in record time.

"It's a technology domain whose time has arrived," said Greg Mitchell, an optical ecologist at Scripps Institution of Oceanography. "There is going to be a huge expansion of our ability to understand what's going on in these systems."

Mitchell and his colleagues teamed up with University of California, San Diego, computer vision scientists, who aim to automatically extract information from photos or videos. Over the past several years, the group has developed software that analyzes a coral reef picture in seconds.

The software not only saves hours of "mind-numbing" work, but also opens the door to much larger and more frequent surveys. If a camera attached to an underwater vehicle patrolling protected areas takes thousands of pictures, the information can be processed in a matter of days and given to resource managers.

"The more data you have, the better ecological insight you have into success or failure of a management plan," said David Kline, a Scripps coral ecologist working on the project.

Monitoring on that scale wouldn't be possible without advances in digital cameras, more efficient batteries, faster computers and ever-cheaper data storage, Mitchell noted.

"It's big data applied to ecology," he said.

The team has made the software available for free online, and researchers from around the world laud it as a powerful new tool.

Rusty Brainard, chief of the Coral Reef Ecosystem Division at the National Oceanic and Atmospheric Administration, said the program will help his team analyze images more efficiently. However, he noted that faster analysis is not necessarily required by reef managers because "most management decisions aren't being made tomorrow."

But there are other potential benefits to automation, Brainard added. Sometimes human-collected survey data are questioned by fishermen or developers whose activities may be restricted as part of reef management, he said. If the automatic analysis can prove it's more accurate than humans, he said, it might assuage those concerns and smooth the management process.

'The algorithm does better'

To document changes to coral reefs, researchers and managers first need base-line data about the quantity and type of corals in each area.

Historically, divers swam along predetermined lines and counted what they saw, taking notes on an underwater clipboard. These days, researchers swim or are towed behind boats along those same lines, but take pictures and later count corals in the lab.

Digital cameras allow them to cover more ground, but it could take up to 20 minutes to identify 200 points in one image. If a researcher took 1,000 pictures on just one dive, much data could remain unprocessed simply because there wasn't enough time.

The computer vision program helps solve that problem. But can managers trust the technology to be as good as a human expert?

The computer correctly identified corals versus other materials like rock, sand or algae between 92 and 95 percent of the time, the researchers said. Within corals, it correctly assigned the coral's genus -- the biological classification level above species -- 97 percent of the time.

"The algorithm does better [than humans] distinguishing genus level of corals," said Oscar Beijbom, a computer vision science Ph.D. student working on the project.

But the computer had some trouble specifically classifying the full spectrum of reef components: rock, algae, microalgae, sand and several classes of corals. It correctly identified randomly selected points in the picture between 74 and 83 percent of the time.

While the computer makes mistakes, so do humans. Even coral experts disagree over classification, or make mistakes because of the repetitive nature of the task, the researchers noted. The computer "turns out to be mostly within the error bars of what people do," said David Kriegman, a computer science professor at UCSD developing the software.

How it works

Once found only in science fiction movies, facial recognition technology is everywhere -- it's how a camera phone picks out people to focus on, and how Facebook suggests friends to tag in a picture.

But while faces are easier to differentiate, it's difficult to pick out any one object in a coral reef because the materials are jumbled together. So Beijbom and Kriegman designed software to decode images using texture and color.

Coral diver
A diver photographs a reef around Totoya island in Fiji using fluorescent light. Corals glow in bright colors under fluorescent light, making it easier for the computer to identify them in the pictures. Photo by Keith Ellenbogen.

Examined pixel by pixel, corals, algae, rocks and even sand have unique patterns -- a bit like fingerprints, Beijbom said. The researchers decided to identify the patterns associated with each element and build a reference database with thousands of examples.

They started with 2,000 images from a National Science Foundation long-term research site, a coral reef by Moorea Island in French Polynesia in the South Pacific. Coral ecologists had manually identified 400,000 points within the images, providing a "gold mine" from which to build the reference database.

Now, when presented with a new coral reef image, the computer extracts information about color and texture and compares it to the reference database. It labels hundreds of points within each picture based on what they most closely match. The whole process takes all of 20 seconds, compared to a person's typical five to 20 minutes.

While ecologists would love to simply upload images into the computer and have results spit out in a half hour, it's unlikely they will achieve complete automation, Beijbom said.

The same species of corals have different shapes at different depths and in different parts of the world. Even in the same location, different water and light conditions could skew the analysis, so researchers have to manually label a small part of their new data sets to calibrate the computer, before letting it loose on the rest of their images.

"I don't think it's a matter of being fully automated," Beijbom said. "It's needing less and less annotations."

Fewer annotations would be welcomed by Vincent Moriarty, who works on the Moorea Coral Reef Long Term Ecological Research site, conducting coral reef surveys and annotating the thousand images taken each year.

"I am looking forward to handing this part of my job over to a computer at some point," he said.

Improved accuracy

Accuracy improves as the database grows with each new processed image; this is how the computer "learns." That's in part why the team freely provided the software online at CoralNet. Researchers from around the globe can upload their images for automatic analysis, which helps the software continue to learn and improve with more examples from different locations.

One CoralNet user is the Catlin Seaview Survey, which is on a mission to systematically document the world's coral reefs in 360-degree, high-resolution images. The group has taken more than 100,000 images of the Great Barrier Reef and is now working in the Caribbean, and praised CoralNet for extracting useful information from its images 50 times faster than a human expert.

"CoralNet is a very powerful tool for automated image annotation," said Manuel González-Rivero, lead shallow-reef survey scientist for the Catlin Seaview Survey. "It is flexible to our needs and user-friendly."

The Scripps and UCSD team continues to tweak the program to improve accuracy. For example, they are investigating fluorescent imaging of coral reefs.

Under fluorescent blue light, corals glow in bright reds, greens and yellows. "It's an underwater acid trip," Kline said.

The bright colors make it much easier for humans -- and the computer -- to identify corals in images. It could also help improve the accuracy of the automatic analysis by recognizing dead corals that have the same texture pattern as live corals, but should not be logged as such.

Another benefit of the fluorescent images: Baby corals that are almost impossible to see in regular photographs become bright beacons.

"We are trying to improve it," Kline said, "so we can quantify how many baby corals are being produced, which is really important for the recovery of a reef."


Over 11 million reef fish and millions of other reef dwellers are taken from the world’s reefs each year to meet the demand of 700,000 U.S. aquarium hobbyists. Those figures are doubled for reef wildlife lost to hobbyist demand globally.
Over 1 Million of those reef animals are taken from Hawaii's coral reefs, making Hawaii the world’s third largest supplier of reef wildlife for U.S. saltwater aquarium hobbyists.

The unsustainable demand for reef wildlife is driven by the astronomical mortality rates associated with trying to keep wildlife captive in tiny glass boxes. Nine fish may die for every one that makes it to a hobby tank.

In hobby tanks, most wildlife will die within weeks or months from stress related disease; from cramped or failed environments; from improper food; and, generally because the vast majority of those attempting to keep them are too inexperienced.

Unlike freshwater aquariums which contain nearly 100% captive-bred and raised animals, nearly all the wildlife kept in saltwater aquariums is captured from coral reefs, with fewer than 5% bred in captivity. Greed drives the endless plundering of the world's reefs and inflicts needless suffering and death on tens of millions of wild animals annually.

Worldwide, the aquarium trade is decimating wildlife populations, causing local disappearance of species that once were common and upsetting the delicate balance of fragile coral reef ecosystems. Over millennia coral reefs evolved to optimal balance to ensure survival and continuing function as nursery. Left unmonitored, the aquarium trade will cause species extinctions and contribute to ecosystem collapse.

Hawaii reefs, which are quite narrow in scope, are visibly suffering from aquarium trade exploitation. Aquarium catch reports, fish surveys and monitoring reports, and both resident and visitor “eyes on the reef” confirm that local populations and individual species are crashing.

Coral reefs are left unbalanced as natural populations are altered and key species depleted. Algae suffocation now threatens many Hawaii reefs.

Many fish species collected in Hawaii are endemic, occurring nowhere else in the world. Most of the animals taken are herbivores, algae eaters essential to coral reef health. All species collected are vital to reef balance.

Coral and live rock extraction is 100% illegal on Hawaii reefs, yet for a $50 annual license and permit, commercial collectors have access to every living animal on every reef in Hawaii that isn't protected - virtually 99% of Hawaii's coral reefs are impacted by this industry.

The most heavily oppressed fish in Hawaii is the yellow tang. Hundreds of thousands of their young are taken by the trade annually and they are among the top 10 fish most popular fish in the U.S. market. In the areas where they’re heavily collected, their populations are down by 70 – 90%.

Yellow Tangs will live an average age of 11 years on a protected reef. Many will survive for 20 – 30 years and their full potential lifespan is over 40 years.

In captivity, some yellow tangs may survive a year or more, but the vast majority will likely be killed off within the first month or two of capture from shipping and other stressors, starvation and hobbyist mistakes.

An estimated 1% of wild yellow tangs will live to be at least 5 to 7 years old, and as an adult fish, will spawn for the first time and contribute to the survival of their species. The other 99% will have played key roles in the ecosystem, like maintaining the proper algae balance, and, ultimately contributing to the food web.

Hawaii’s butterflyfishes are another example of species decimated by the aquarium trade because of their beauty and fragility. Many species of butterflyfish feed primarily on corals and starve to death in captivity. They are also especially fragile in transport and many are shipped out with no guarantee to arrive alive. Their wild populations have plummeted.

New aquariums need to be cycled, an essential process requiring organic materials and 2 - 6 weeks to develop the chemistry and proper levels of beneficial bacteria for nutrient cycling. Live rock, sand and fish are used to accomplish this. When fish are used, the word "torture" frequently occurs in online marine aquarium sites describing the process, but that doesn't stop the so-called experts and fish sellers from recommending the practice.

Over 260 species of Hawaii’s reef fish and creatures are taken by the aquarium trade with no concern for ecosystem impacts or animal welfare. Reef wildlife is treated as inert products with “shelf lives” rather than as fragile living creatures, and the trade is managed by the state as if they are a limitless resource with no consequences to the coral reef ecosystems from which they are harvested. Also ignored are Hawaii's communities and residents who rely upon healthy and beautiful coral reefs for their socio-economic well being.

Reporting fraud in the trade is rampant and widely acknowledged. Experts estimate the true number of fish and creatures taken by the trade to be at least double what they officially report.

Throughout the Indo-pacific, the aquarium trade is also having disastrous effects. At least 25 - 30 million animals, representing over 2,000 species, are traded each year, over-collecting is rampant and many fish are caught using cyanide. 90% of the reefs show missing groups of fish and rare species.

Turquoise fish in stag horn coral  aliceiswonderful.tumblr~dot~com

No comments:

Post a Comment