HETEROTROPHIC FEEDING
We had constructed our two sexually mature colonies of Acropora valida (Dana, 1846) and Acropora prostrata
(Dana, 1846). Now we needed an appropriate feeding regime, as we knew this was important for providing
the nitrogen and phosphorus input required for healthy growth and gamete production.
Acropora species can feed on a huge range of prey types. They can absorb free amino acids from the water
column and consume bacteria and protists (picoplankton) and various species of phytoplankton that are
trapped in their mucus. Corals use this mucus layer as a feeding mechanism, employing cilia
located on the surface to move the mucus in and out of the mouth. Acroporacorals also have powerful stinging cells
that they use to capture large motile prey, such as copepods (zooplankton),and these form a large portion of their
heterotrophic feed in the wild. Having identified various prey groups, I had lengthy discussions
with Dr. Dirk Petersen, director of the SECORE Foundation. Dirk has a huge amount of knowledge about heterotrophic
feeding, which he gained during experimental work for his Ph.D. and after. He provided me with various dose rates
that he’d developed while he was at the Rotterdam Zoo, and I based our feeding regime (see images on page 88) on his
advice. The coral research system at the Horniman holds 530 gallons (2,000 L), split into four 130-gallon (500-L) components:
two 65-gallon (250-L) broodstock tanks and two 65-gallon (250-L) sumps that are connected. Each of our broodstock tanks was
isolated for 1–2 hours every morning to concentrate the food and allow uptake.We then added a yeast culture solution at 0.4 ounce
(12.4 ml) per 264 gallons (1,000 L) (picoplankton).We made this three times a week and cultured the yeast
with sugar at 75.2°F (24°C) for 24 hours before use. For the phytoplankton component of the diet we used Reed
Mariculture shellfish diet and aimed for an approximate cell concentration of 48,000 cells per liter. Each tank
also got two frozen cubes of rotifers (1,000–2,000 per liter) and 1.7 ounces (50 ml) of instar II Artemia nauplii
that had been enriched with New Era’s Live Food Enrichment to improve the fatty acid profile. This concentration
yields a prey density of approximately 1,136 naupliiper gallon (300 per L). This feeding regime places a lot of pressure on the
filtration, and we’ve had to increase the percentage of water changed to manage the correct water chemistry.
The feeding regime has also resulted in Derbesia marina and Aiptasia outbreaks, which we have managed both
biologically and chemically.During feeding there is a distinct change in the corals. At first the polyps retract, but over the course
of the isolation they greatly extend, often ejecting their mesenterial filaments (see facing page). The corals also
produce mucus webs to capture and ingest their prey.Each broodstock tank has high internal water flow to
ensure prey suspension and movement.
ANNUAL TEMPERATURE CYCLE
Traditional reefing maintains a nearly constant temperature, but in the wild there is often considerable change
throughout the year. This seasonal fluctuation is one important contributing factor in initiating gamete development.
We based the environmental parameters in our coral laboratory on a Fijian reef because we had a number
of large colonies shipped directly from Fiji, and through replicating the environmental parameters of that location
we intend to spawn these during 2014 (2x Acropora humilis, 4x Acropora valida, and 2x Acropora prostrata).
When planning this project I spent a great deal of time researching these environmental parameters to ensure
accurate replication of the environmental conditions experienced on the reef. We are using a U.S. NOAA sea
surface temperature (SST) data recording buoy to provide our annual temperature regime. This buoy, part of a
global climate change monitoring program coordinated by NOAA, has been uploading SST readings every three
days since 2000. One of the buoy’s roles is to collect data on the likelihood that a bleaching event will occur. We
wanted to create an annual temperature change, but we certainly didn’t want to cause a bleaching event in the
system, so I downloaded the previous 13 years of data and analyzed it. Based on the results, I chose 2011 as the best
model, one that wouldn’t create any bleaching issues. Using this data, Gary programmed the microprocessor to
control the heaters and chiller so that our system would replicate this regime. Overall the system performed well
during 2013, averaging 31.69 degrees (C) from a set point across the year. We plan to improve this by cooling
the space during the hot summer months.
PHOTOPERIOD AND LUNAR CYCLE
In 2011, Boch published a paper that investigated a number of the factors that affect gamete release and spawning
coordination, including sunset and twilight periods, Kelvin light temperatures during these periods, and the lunar
cycle. Through a series of controlled experiments, we determined that the period of total darkness post sunset,
moving through twilight, and prelunar rise in particular is an important factor in synchronizing spawning. Each
day past a full moon the period of absolute darkness extends by approximately 1 hour and, based on Boch’s
results, we have designed an artificial lunar cycle to replicate wild conditions. The research system at the museum
is located behind the scenes and has a long bank of high windows and security lighting through the middle of the
space (see image above). The last thing we wanted was external light levels influencing spawning times, so we
completely blacked out the system (see image on facing page). The front is raised during the day for maintenance
and we simply “put the system to bed” at the end of the day, giving us complete control of this critical parameter.
Our lunar LED units were built by Tropical Marine Centre, UK, specifically for this project and emit the correct
Kelvin temperature of the moon (4,100 K). We spent some time calibrating the LED power output to match
the annual cycle. Maximum lux output at full moon is just 0.1 lux, and our microprocessor controls the LED to
produce 0.1 lux at 100 percent output (full moon) and drops down to 0 percent (new moon) throughout the
month. We’re using the U.S. Naval Observation data for both the photoperiod and the lunar cycle, and this has
enabled us to precisely match conditions on the reef that the corals were collected from using GPS coordinates.
That covers the five parameters (colony size, heterotrophic feeding, temperature cycle, photoperiod, and lunar
cycle) we thought would be needed for successful spawning, and these were all put in place at the beginning
of January 2013. Then it was just a matter of running the system, ensuring water chemistry was closely
monitored, and waiting. Inevitably, there is some doubt when you’re running a new project and trying to break new
ground. You think you have covered each and every detail and meticulously planned. You think it will work because the theory is
sound, but I have to say that when I started to crack the colonies (a term used by researchers, referring to the
process of removing a branch and examining the crosssection for signs of gametes) in early August, I was blown
away on discovering that both the Acropora prostrate and the A. valida we had “built” 8 months previously
were gravid. The A. prostrata had small white oocytes (eggs). Acropora egg development is well documented,
so I could tell we still had 4–8 weeks before it would spawn. The Acropora valida was more developed, and
the oocytes already had orange/pink pigmentation (see image on page 86). These oocytes were in a later stage
of development, and that meant spawning would occur after the next artificial full moon.
Both colonies were cracked a number of times during the buildup to spawning to check development
and photograph the cross-sections. The fragments were preserved in 10 percent formalin and sent off for
histological sectioning. During this process the hard skeleton of the frag is dissolved in acid. The remaining
soft, preserved tissue is then embedded in wax, and 6–8-micron sections are sliced from the wax block,
mounted to microscope slides, and stained to highlight the tissue structures (the pink areas are the oocytes, the
dark purple is sperm). It’s a powerful tool because it has enabled us to document the stages of egg and sperm
development in the buildup to spawning. Microscopic slides also enable us to revisit the stages at any time, and
one of our future goals is to develop a comprehensive library of slides at various stages of development from
multiple species, under varying conditions. The image at the bottom of page 83 follows the oocyte development,
from white to pink, in the A. prostrata over 8 weeks, but I think image (d) best highlights the amazing potential
that captive spawning in a controlled system has for coral research. I preserved this fragment just 10 minutes
before it spawned, and you can actually see the egg sperm bundle inside the polyp mouth prior to release (see
image on page 81, bottom). Preserving fragments this close to spawning in the wild, in a stage called “setting”
(more on this in a bit), would be challenging, so to be able to do this in southeast London is very exciting!
Imaging at higher magnifications, 200x and above, has also enabled us to see sperm orientation along the
mesenterial filaments, within the preserved tissue, on the microscope slides.
View attachment 1038
View attachment 1039
View attachment 1040
View attachment 1041
