Artificial sea salts are the basic
ingredients of any marine aquarium whose owner does not have
access to natural seawater. Myriad brands and formulations
exist, and opinions based largely on personal preference and
a few limited studies (see Borneman, 2006 for a review
of these studies) continue to fuel debates about which
brand is best. Much discussion has occurred regarding the
elemental composition of salts, and although impressive longevity
for some marine species has been accomplished by using these
artificial salts, we have little idea how artificial sea salts
compare to each other and to natural seawater. It is a question
that should be answered, if possible.
The original idea to study the practical effects of artificial
sea salts in a reef aquarium was initiated by the Marine and
Reef Society of Houston (M.A.R.S.H.) who enlisted the voluntarily
services of Eric Borneman as principle investigator and to
design an appropriate experiment. The generous M.A.R.S.H.
membership and several other contributors (see acknowledgements)
deemed the project worthy, and we embarked on what has turned
into an extensive study of the differences between various
commonly available artificial sea salt mixes' effects on organisms
common to reef aquaria. Our hypotheses were that there would
be differences in the survival, growth and reproduction of
species in equivalent aquaria utilizing different sea salt
mixes with a null hypothesis that there would be no differences.
Natural seawater controls were provided consisting of samples
from three oceanic sources.
Proposed Study: Artificial
Salts' Effects on Growth and Reproduction in Reef Aquaria
Ten 10-gallon tanks are set up on
a long bench and illuminated with two sets of 6' long high
output T5 (one white and one blue) fluorescent bulbs (Innovative
Lighting). A single new Maxi-Jet 1200 powerhead (Marineland)
is mounted in the center of each tank's short end, facing
lengthwise just below the water's surface. An automatic feeder
is the only other equipment used on the tanks, and it is filled
with 0.1g of a homogenized mix of multiple dry foods (Formula
One, Formula Two (Ocean Nutrition), Green 2000,
Labs) and beta glucan (SeaVive)).
Each tank houses clonally derived ramets
of various coral species, including stony corals, soft corals,
zoanthids and corallimorphs. We chose Montipora digitata,
Cladacora caespitosa and Porites cylindrica
as representative scleractinians. In addition, a corallimorpharian,
Sinularia grayi, Zoanthus sociatus and Plexaura
flexuosa of the same genotype were used as non-scleractinian
coral species. We also included two algae of the same genotype,
Chaetomorpha sp. and Gracilaria sp. Other organisms
included Astraea tecta, Nassarius sp., Cerith
sp. snails, Clibanarius tricolor hermit crabs, Asterina
sea stars, Leucetta sp. sponges and Amphiprion percula
clownfish from the same brood to minimize genetic variability.
Our idea was to provide a representative sample of the numerous
phyla encountered in reef aquaria including corals, gastropods,
echinoderms, crustaceans, algae, sponges and vertebrates to
each of the tanks.
No live rock was used in the tanks because of potential variation
in the diversity of life introduced into the tanks by live
rock, but we added a layer of autoclaved
crushed coral as a substrate for benthic flora and
We acquired 11 batches of each artificial salt brand separated
as far in time and space as was practical from various vendors
across the country to ensure both variation in batches and
that the products were those normally available to aquarists.
In no case were any samples provided directly by their manufacturer,
in order to ensure that samples were representative of those
available commercially. Each bag was new and unopened, and
approximately 1500g of salt was removed directly from the
bag or container into new, clean Ziploc bags and sealed immediately.
We recorded the donor, the place and date of purchase and
the size of the container. Each group of salt samples were
then marked as Samples A-J, rather than by manufacturer, in
order to ensure the study was conducted blindly.
Sea salt mixes are prepared twice. In the first case, 35g
of each salt are added to 1 liter of double-distilled water
and stirred for 30 minutes on a stir plate fitted with a stir
bar. At the end of this period, we qualify the resultant solution
in terms of its clarity and amount of undissolved and suspended
material. We also take immediate readings of pH and salinity
using lab quality pH meters and refractometers calibrated
before each reading. These data are recorded to calculate
the volume of salt needed to fill each tank water container
and to determine variation within salt brands. Then, deionized
water is used to fill triple-washed containers with water
and enough salt to bring the solution to 35 psu salinity,
and the solution is stirred using a clean powerhead for at
least one hour or until all salts are dissolved. This water
is prepared once each month for a total of 10 months, accounting
for 10 separate batches of each salt mix to be added with
a nearly 100% water change to each tank each month.
Natural seawater was used as a control. For the first three
months water was utilized from the National Cephalopod Research
Station at the University of Texas Medical Branch in Galveston,
Texas. Its water is pulled from approximately 30 miles offshore
to the facility. The water was then sequentially filtered
to 0.22 microns through Millipore filters for use as control
water. During the next four months seawater controls were
obtained from surface waters above the Flower Garden Banks
coral reefs, approximately 100 miles off the Texas coast.
For the final three months water was provided from Stetson
Bank, a mixed coral and hardbottom community 70 miles off
the Texas coast. In all cases, the seawater was filtered to
The salts used in the study were as follows:
Sea Salt (Kent Marine)
Seas Bioassay Formula (Marine Enterprises)
Sea (Red Sea Fish Pharm)
Marinemix (HW Weigandt)
Marin® (Tropic Marin
(Energy Savers Unlimited)
Initially, all tanks were maintained in a climate-controlled
room with HEPA filtration and were covered with acrylic to
reduce evaporation and limit the input of airborne contaminants.
After filling each tank with the seawater, the tanks were
cycled with the addition of a known amount of homogenized
flake food (once) and the addition of reagent grade ammonium
chloride (Sigma). The cycling period lasted for six weeks
before the autoclaved substratum was added to the tanks. The
following week, the species were added to each tank with positions
determined randomly but with each species in the same position
within each tank to minimize microhabitat variation. Recently,
the tanks have been moved to the main lab to ensure security
because the climate controlled room could not be locked and
some concern arose over potential incidents that could occur
without the added security.
All tanks have the following water
quality parameters measured at the start of the experiment,
weekly (within batches), monthly (old and new between batches)
and at the end of the experiment:
Temperature: measured using a thermometer (Fisher Scientific)
Salinity: measured using a calibrated lab quality refractometer
pH: measured using a calibrated pH pen (Milwaukee Instruments)
Oxygen: measured using an oxygen probe (YSI, Inc.)
PAR: measured using a lab quality PAR meter (LiCor Industries)
Ammonia: measured using colorimetric test kit (Salifert)
Nitrite: measured using colorimetric test kit (Salifert)
Nitrate: measured using colorimetric test kit (Salifert)
Phosphate: measured using colorimetric test kit (Salifert)
Calcium: measured using titration based test kit (Salifert)
Alkalinity: measured using titration based test kit (Salifert)
Each volunteer responsible for water quality testing must
read the tanks for the entire duration of a test period (=
one month) to ensure reproducibility of results. All instruments
were calibrated prior to each test.
Following testing, the following actions may be taken. Salinity
is corrected through the addition of the same batch of salt
if hyposaline, or through the addition of Class I deionized
water to maintain 35psu. Calcium will not be maintained but
will be allowed to decline over the course of the month-long
periods because calcium is generally not limiting to calcification.
Alkalinity will be maintained weekly using a mixture of reagent-grade
sodium bicarbonate (VWR) and sodium carbonate (Sigma) brought
to a pH of 8.2 to maintain an alkalinity of 3.0meq/l and any
additions are recorded as to the volume of carbonates added.
pH will be not be adjusted during each month-long period.
Every other day, 0.1g of homogenized food is added to each
tank through the automatic feeders. Initially, food was added
manually through 0.1g aliquots
prepared and stored in closed microcentrifuge tubes.
All tanks receive nearly a 100% water change (siphoning
to the gravel level) each month for 10 months, with a new
batch of the same salt brand being used for each water change.
Between batches, species are counted, photographed, measured
and weighed as described above. Any mortalities that occur
are recorded and replaced immediately or as soon as possible
after losses are recorded. Soft corals, because of the water
contained in their tissues, are recorded qualitatively in
terms of their appearance and assigned a ranking. The gorgonian,
being rigid, is measured by length, width and height, using
and qualitatively by description. Cladocora and zoanthids
are measured by counting mature and incipient polyps. Montipora
and Porites are weighed, measured with calipers or
described by percent coral tissue alive according to the Atlantic
and Gulf Rapid Reef Assessment (AGRRA) protocol. Algae are
cleaned of detrital material or other algae, patted dry and
weighed. Fish are weighed using the buoyant weight technique.
Gastropods, crustaceans and echinoderms are counted as absent
or present. Sponges are accounted for by their presence or
absence and the number of incipient sponges being produced
on their surface or found within the tank. Each month, any
filamentous algae or cyanobacteria are removed from the species
to limit the observed effects on species to salts, rather
than algal overgrowth, as responsible for any partial or total
mortality. The tanks' front pane was cleaned with a razor
weekly so that photographic documentation of each tank could
be recorded, and the tanks' sides are wiped clear of nuisance
algae before siphoning old water prior to the water change.
Coralline algae and other sessile invertebrates, such as bryozoans,
were left intact where present. The powerhead filters were
cleaned and strands of attached Chaetomorpha and Gracilaria
were removed, dried and weighed. Also, qualitative descriptions
of each tank were performed monthly and occasionally weekly,
as needed. This process was repeated for a total of 10 salt
batches (= 10 months).
Our argument and contention is that all salt mixes are supposed
to be reasonable substitutions for seawater, and that a 100%
water change shouldn't stress the organisms unless some sort
of significant acclimatization to large shifts in water chemistry
occurs over the interim period. In these tanks, the test period
is only one month. While we concur that 100% water changes
do not seem well-tolerated in established tanks, theoretically,
the total water change should be, if anything, beneficial.
It will also highlight any results of salts that had particularly
notable deleterious effects. A 100% change will likely show
salt-based effects more strongly than would smaller water
At the end of the experiment, all species will be reweighed
as described above. All reproductive events will be counted.
All mortalities will be counted. All algae will be scraped
from the glass and weighed. The calcareous species will be
soaked in bleach to dissolve tissues, then the remaining calcareous
crusts will be weighed. All other organisms, such as amphipods,
polychaetes and other flora and fauna will be removed, filtered
and counted or weighed.
Inter- and intra-tank variation of
all data will be determined using repeated measures (ANOVA).
Additional pairwise comparisons may be required using Student's
t-tests. Nonparametric tests will be performed on ranked
microbiota will play similar roles in all systems and will
exert relatively small effects on the species' growth and
other species introductions will occur through the use of
living organisms that will affect the species' growth and
species are healthy and free of pre-existing conditions
that would affect their growth and survival.
salt tested is representative of a "batch" of
salt. Each salt mix (n=10) will be purchased from different
sources at different times. Each purchase of salt results
from a separate "batch."
in each of the aquaria are sufficiently similar that they
will not affect the species' growth and survival.
additions to tanks in terms of buffer, water and food are
sufficiently similar or of such small effects that they
do not affect the species' growth and survival.
populations are assumed to be normally distributed in terms
of growth and survival in their original systems; in the
ocean or with aged water using many salts. An assumption
exists that species are alive and growing, some reproduction
is occurring and that mortality and recession exist.
data collected may not be valid for comparison with wild
reefs (true control).
The data we have collected to date have been extensive and
surprising. Not only will the targeted results be interesting,
but so are many of the unexpected results we have encountered
over the course of the experiment to date. This experiment
ends on the weekend of MACNA (Sept. 23/24, 2006). The preliminary
results and highlights will be presented at that meeting,
where we hope to see a strong attendance. The data analysis,
however, due to the overwhelmingly large amount of data, will
take time to perform and present in a usable and digestible
manner. These results will first be presented for consideration
in a peer-reviewed journal and then written as a series of
articles for aquarium magazines around the world. We very
much look forward to presenting the results of our work to
the aquarium hobby and hope this is a step toward understanding
more about the intrinsic properties of various artificial
seawater mixes available to the trade. Furthermore, thousands
of articles have been written in peer-reviewed journals using
synthetic sea salts predominantly from two brands - and under
the assumption that they perform equivalently to natural seawater.
The results of our work will further determine if such assumptions
in the literature can be considered valid.
Foremost, we would like to thank our spouses, Brandee and
Mark, for tolerating our time away to complete this project,
as well as helping us out in our time of need. This work was
supported by various dedicated members of the Marine Aquarium
and Reef Society of Houston (M.A.R.S.H.), the Dallas-Fort
Worth Marine Aquarium Society (D.F.W.M.A.S.) and the Marine
Aquarist Association of South Texas (M.A.A.S.T.). Donations
of equipment and livestock were provided by The Atlanta Reef
Club, the Orlando Reef Caretakers Association, Exotic Aquatic
and Pets, Innovative Lighting, Marine Depot, Maroon Lagoon,
Premium Aquatics, Salifert, Village Tropical and various members
of M.A.R.S.H. We would also like to thank all others not mentioned
above who advised and supported us in this project.