|dc.description.abstract||Over 85% of Australia’s population live in urban areas and many turtle populations occur on Australia’s east coast where urban development is particularly concentrated. In the state of NSW, over half of the freshwater coastal wetlands have been highly modified or completely destroyed, and urban freshwater creeks often have only a narrow strip of weedy bushland left along their banks. Even though habitat degradation may result in declines in density and distribution of turtle populations, there are few data on Australian freshwater turtles in urban areas.
In addition to extreme habitat alteration, urban waterways are innundated with anthropogenic contaminants from sources including wet weather surface runoff and industrial and sewage discharges. Pollutants can impact all systems of the body with potentially severe effects on reproduction and survival that can result in deterioration of animal populations. Turtles are particularly susceptible to anthropogenic contaminants due to their intimate contact with the aquatic environment, an often high trophic level, their ability to accumulate toxins, and their longevity. For almost all contaminants, the degree of accumulation in and effect on reptile species is unknown. Sublethal effects in field situations are particularly poorly studied and have never been documented in pleurodiran turtles.
As a pioneering work in Australian reptile ecotoxicology, this thesis takes a broad approach, but focuses primarily on immunotoxicity and reproductive toxicity – two areas that greatly impact the size and continuance of animal populations. The aim of the thesis is to provide baseline data on haematology, cellular immunology and tissue metal concentrations for freshwater turtles in Sydney – data which were lacking for all Australian turtle species prior to this study. After initial assessment of the distribution and density of freshwater turtles in Sydney, the study examines the potential for Sydney’s turtles as sentinel species for measuring the effects of pollution on haematology, cellular immunity, and parasite loads; and considers the relationships between urban metal pollution and reproductive variables. The relative suitability of non-lethally sampled tissues (blood, carapace, egg) for use in biomonitoring is also assessed.
Three species of Australian freshwater turtles were found in the Sydney region, with Chelodina longicollis occurring naturally in the area, and populations of Emydura macquarii and Elseya latisternum likely to have originated from translocated individuals. The North American turtle Trachemys scripta elegans was not encountered during this study despite concerns that it was establishing in the Sydney area. Chelodina longicollis populations were widespread, although poor recruitmment was indicated by low capture rates and comparatively low percentage of juveniles at some sites. Not so widespread, Emydura macquarii was present in much larger numbers than C. longicollis and with a high juvenile component in some areas of southeastern Sydney.
I provide information on erythrocyte and leucocyte parameters in C. longicollis over a range of sites, pollution conditions, and seasons. In C. longicollis, numbers of lymphocytes, heterophils and eosinophils varied over sites, but not due to pollution from sewage treatment plant outfalls. There was significant temporal variation in erythrocyte, lymphocyte, eosinophil, heterophil, and basophil number, the heterophil:lymphocyte ratio, and haematocrit, but not consistently among sites. Future studies should ensure simultaneous sampling across sites for comparative purposes. Similarly, turtle populations downstream of sewage treatment plant outfalls showed no consistent difference in number, body condition, blood haemogregarine load, or leech (haemogregarine vector) load from upstream populations. Leech (Helobdella papillornata, with some Placobdella sp.) load and haemogregarine numbers increase dramatically once C. longicollis reach a carapace of 110 mm. The number of leeches on turtles varied across season, year, and site. Turtles with large numbers of leeches had reduced haematocrit, but the presence of leeches had no other correlations with haematological parameters. Haemogregarine numbers did not change across season or year, and were not correlated with haematological variables. The hypothesis that pollutants lead to an increase in normal blood protozoa due to reduced immunity thus was not supported.
The concentration of metals in C. longicollis and E. macquarii carapace and in lagoon sediments varied significantly over four urban and four national park sites, but not based on this split. Pollution in periurban areas, such as illegal dumping of toxic wastes and atmospheric deposition of pollutants, means that each site must be classified separately as to degree of metal pollution. There was little or no affect of species, size, sex, or gravidity on metal concentrations in the carapace of adult turtles.
Emydura macquarii had higher concentrations of blood Fe than C. longicollis from a different site, but this is possibly due to an increase in haemoglobin resulting from the site’s low aquatic oxygen concentration rather than any increased environmental exposure.
Chelid turtles in Sydney do not show much promise as a biomonitoring tool. Carapace analysis is largely discounted as a potential tool for metal biomonitoring due to poor correlations between potentially toxic metals in non-lethally samplable tissues (carapace, claw) and internal organs (liver, kidney) or bone (femur). However, carapace metal concentrations still potentially reflect long-term metal presence or different dietary exposures as evidenced by the significant variation in concentrations over sites. A rare correlation was found for concentrations of aquatic Pb and carapace Pb, and a correlation was also found for concentrations of blood Pb and carapace Pb in E. macquarii. Thus any potential for tissue biomonitoring seems to lie with this highly ecotoxicologically relevant metal. Although two other ecotoxicologically relevent metals, Cu and Se, were significantly higher in egg contents of C. longicollis compared to E. macquarii, these elements are also essential and a lack of baseline values means it is not known if this simply reflects natural taxonomic variation. Ni, a metal of toxicological concern in sea turtles, was not present in egg contents, and only variably present in eggshell. The absence of Pb from eggs, despite its presence in many maternal tissues, suggests that selective metal uptake into eggs may be protective of toxic elements, rather than eggs serving as a maternal method of toxic metal elimination as has been previously suggested. The paucity of toxic metal detection in eggs renders them unlikely tissues for biomonitoring.
The maternal tissue or tissues or environmental source from which egg metals originate remains obscure, although a significant negative effect of maternal carapace concentrations of Ca and Mg on eggshell thickness in E. macquarii indicates that there may be mobilisation of Ca and Mg from the carapace for eggshell formation. The only metal whose eggshell concentration correlated with eggshell thickness was Mg, indicating that ecotoxic metals previously associated with eggshell thinning are not problematic in the Sydney chelids. As with North American turtles living at polluted sites, none of the chelid hatchlings were found to have any overt abnormalities. Hatching success was poor and hatching mass low for eggs of both C. longicollis and E. macquarii, although results from natural nests are required to determine whether or not this was an outcome of hormonally-induced oviposition and artificial incubation.
It is difficult to interpret metal concentrations found in the soft tissues, calcified tissues, and eggs of chelonians due to the paucity of comparative data, and much more research is required on tissue metal concentrations before patterns will emerge. This especially applies to pleurodires for which no previous information is available. From comparisons with the limited data available for other freshwater turtles, marine turtles, and other aquatic reptiles, it does not appear that Sydney’s turtle populations have unusually high metal concentrations in tissues. Exclusion of toxic metals such as Pb from the egg may also be protective to the developing embryo. An ability to live in polluted habitats, while limiting the accumulation of toxic contaminants, may be one key to their persistence in urban waterways from which other freshwater fauna have disappeared. Reproductive impacts such as low embryo survival and small hatchling weights require more rigorous examination, but may have less effect on these animals which have such naturally high egg and hatchling mortality.
Although it was generally hard to demonstrate biochemical, physiological or population impacts of contaminants, C. longicollis from a site with severe sewage pollution did display unusual alterations in a number of haematological variables, body condition, and carapace bone structure. Despite this, the population was large and had a comparatively high ratio of juveniles. Additionally, the adverse haematological alterations appeared reversible. Thus, successful populations in Sydney probably are more dependent on basic ecological needs being met, than on low levels of environmental contaminants. The ongoing persistence of chelid populations in Sydney is likely to be dependent to some extent on their opportunistic diets, which generally make animals less vulnerable to habitat modification and the reduction in prey item diversity following pollution (Mason 1996, Allanson & Georges 1999), with a further benefit possibly bestowed at some sites on E. macquarii by its omnivory.||en|