Cyanobacteria or blue green algae are known for their extensive and highly visible blooms in rivers or
dams. One of the most important cyanobacteria is Microcystis aeruginosa which can synthesize various
microcystins that can affect the health of terrestrial and aquatic animals. Commercial Nile crocodile
(Crocodylus niloticus) farming in South Africa is based on keeping breeders (adult males and females) in
big dams on farms (captive-bred approach). Unfortunately, cyanobacterial blooms in the breeder dams
are a concern to farm owners, managers and veterinarians. This research project focussed on the
monitoring of microcystins in the Hartbeespoort Dam and a crocodile breeding dam over a period of nine
months. A commercial, but expensive, Abraxis ELISA kit was compared to a much cheaper and robust
Norwegian-developed ELISA to detect microcystins in fresh water. Another objective was to determine if
microcystins were present in the contents of crocodile eggs and dead hatchlings.
Water samples were collected monthly from August 2014 to April 2015 at two sites, the Hartbeespoort
Dam (control site) and the breeding dam of a commercial Nile crocodile (Crocodylus niloticus) farm. In
addition, various water quality parameters including nitrate, phosphorous, chlorophyll a, oxygen
saturation, pH and total dissolved solids (TDS) were determined to assess eutrophication. During the
crocodile hatching season microcystin concentrations in unfertilized eggs, egg-shell membranes and in
the yolk and liver of dead hatchlings were determined using liquid chromatography-mass spectrometry
Water quality parameters showed that there was no significant difference between the two dams' (the
Hartbeespoort and the breeding dam) eutrophic state i.e. phosphates, TKN and nitrates; they both
seemed to be becoming more eutrophic as the nutrient supply to the dam was increasing. Furthermore,
microcystin concentrations during peak summer months were generally higher at the Hartbeespoort Dam
compared to the crocodile breeding dam. The two ELISAs as performed on water samples "as is" and
following an adsorbent disk/methanol extraction method were positively correlated; however, the correlation between the two assays was much stronger when using the adsorbent disk/methanol
extraction as compared to using water "as is". Besides dissolved oxygen all the other water quality
parameters were not significantly different (p > 0.05) between the two sites.
Microcystin concentrations (MC-LR, MC-RR, MC-YR) in the crocodile egg and hatchling samples
collected from batches with a good hatching rate (? 90%) ranged between 0 - 1.76 ng/g, with the highest
concentration in the eggshell membranes. Microcystin concentrations in samples collected from batches
with a bad hatching rate (? 10%) ranged from 0 - 1.63 ng/g with the highest concentration detected in
the hatchling yolk. Although the "tissue" concentration levels were probably underestimated with the
extraction method employed for LC-MS as the percentage recovery from spiked samples were very low.
Bayesian analysis suggests that the liver, yolk and unfertilized egg all have similar microcystin
concentrations, while the membranes have (with moderate to high certainty) higher microcystin
In conclusion, when using the Norwegian ELISA it seems as though the use of a resin-containing
adsorbent disk followed by methanol extraction is more reliable than analysing water "as is". Following
methanol extraction the results of the two ELISAs were strongly correlated, which suggests that the two
ELISAs provide comparable results. There appears to be no difference in microcystin concentrations
among good and bad clutches across all tissue types or within a specific tissue type. Vertical transmission
of microcystins to the Nile crocodile egg does occur, but due to the small sample size, final conclusion
cannot be made if microcystin affects Nile crocodile hatchling mortality and/or hatching of eggs.
Future studies will include a longitudinal study to be done since a single season of breeding is insufficient
to conclude that microcystins do not contribute to the low hatching rate in Nile crocodiles.