Hatchability of Nile crocodile (Crocodylus niloticus) eggs : association with bacteria and fungi in incubation boxes and in eggs that failed to hatch

Abstract

The South African crocodile industry has the potential to produce Nile crocodile skins, which are in high demand by the global fashion industry. The crocodile industry is an important economic resource to South Africa and Southern Africa. The main purpose of rearing Nile crocodiles at intensive production units is to produce viable hatchlings which will be grown to produce quality skins. The production of sufficient hatchlings depends on high hatching percentage of clutches. It has been observed that the hatching percentage varies among clutches. We do not know that whether the poor hatching percentages are associated with the microbial load from the incubation boxes or not. The aims of the study were to determine whether eggs in boxes that had more aerobic bacterial colonies and more fungal colonies before they received vermiculite and eggs are more prone to become lost from the pool potentially capable of yielding conceptuses developing to term than eggs in boxes that had fewer colonies before they received vermiculite and eggs, and whether there are species of aerobic bacteria, anaerobic bacteria and fungi that are present in unhatched eggs from clutches with low hatching percentage but not in clutches with high hatching percentage without the converse being true. At the time of hatching the farm personnel classify hatchlings or unhatched eggs into different categories e.g. hatchling survived, culled, foetus died in the shell, hatchling died, unfertilised egg, banded (the foetus died long before hatching), or rotten egg (putrefaction rendering the fertilisation status unknown). We assumed that banded or rotten eggs are more likely to have lost conceptuses due to the microbial load that was present in the incubation boxes before they were loaded with vermiculite and eggs than the other classes of eggs. A response variable Perhapsloststatus was therefore created. For each egg that was either classified as Banded or Rotten Perhapsloststatus was assigned the value of one. For each egg belonging to other categories Perhapsloststatus was assigned the value of zero. Eggs classified as unfertilised were excluded from Perhapsloststatus. The microbial load of aerobic bacteria and fungi from the incubation boxes before they were loaded with vermiculite and eggs was measured as colony forming units (CFU). The independent variable of interest was Count category (a categorical variable of aerobic bacteria and fungi colony counts), which was compiled as follows: 0 = Counts zero to below the 25th percentile, 1 = Counts from the 25th percentile to below the 50th percentile, 2 = Counts ranging the 50th percentile to below the 75th percentile, 3 = Counts from the 75th percentile to the maximum that was countable and 4 = Too numerous to count. Sterile swabs were used to collect samples from unhatched eggs for isolation and identification of aerobic bacteria, anaerobic bacteria and fungi from the 10 boxes with the lowest and highest hatching percentages, respectively. We used a population-averaged mixed-effect logistic regression model to determine the effect of Count category on Perhapsloststatus. The odds for eggs being banded or rotten (Perhapsloststatus = 1) in boxes with too numerous to count aerobic bacterial colonies (Count category 4) tended to be higher than for eggs in boxes with colony counts between zero and the 25% percentile (Count category 0) (P = 0.06). The odds of eggs being banded or rotten (Perhapsloststatus = 1) in boxes with fungal colony counts falling in any count categories 1 to 4 were the same as the odds of eggs being banded or rotten in boxes with fungal colony counts falling in Count category 0. The variety and frequency of species of aerobic bacteria, anaerobic bacteria and fungi from unhatched eggs of lowest hatching boxes and highest hatching boxes was almost the same. The species isolated from unhatched eggs are more likely pertaining to contamination from the cloaca of female crocodile or from the environment i.e. nesting material, personnel handling, vermiculite, incubation box or the incubator environment, as most of the isolated species present are normally present in soil, water, dust and also opportunistic pathogens in animals. This observational study shows that there is a trend of association between the aerobic bacterial load from the incubation boxes and the eggs in them to become either banded or rotten.