The impact of sublethal pesticide exposure on behaviour, physiology, and molecular markers in the Central Bearded Dragon (Pogona vitticeps): a trigger for inclusion in conservation, government, and risk assessment policy

Assessment of non-target impacts of pesticides used widely in agriculture and pest management rarely consider reptiles. Despite their integral role in all ecosystems, particularly arid ecosystems, reptiles are not included in risk assessments. Two pesticides used in agricultural pest management are fipronil and fenitrothion. Across two field-based before-after-control-impact (BACI) design experiments in semi-arid and arid Australia the sublethal exposure of these pesticides was assessed in a widespread, arid-zone agamid species, the Central Bearded Dragon (Pogona vitticeps) using relevant sampling intervals. Firstly, to investigate the impact of these pesticides on basic physiological and behavioural parameters of P. vitticeps, fipronil and fenitrothion were applied in a single ecologically relevant dose via oral gavage at study site 1 (Nombinnie Nature Reserve) and applied via ultra-low volume (ULV) aerial spray at study site 2 (Fowlers Gap Arid Zone Research Station). Exposure scenarios were designed to replicate the use of these two pesticides in locust control in Australia. At study site 1,I assessed blood, movement activity, and body condition parameters (SBMI) before and after dosing (chapter 2). At study site 2 ,I assessed blood, movement activity, SBMI, ectoparasite count and measured environmental conditions (chapter 3). To build our knowledge further and incorporate a holistic approach, based at study site 1,I also examined the impact of sublethal pesticide exposure on molecular biomarkers before and after exposure, using oxidative stress biomarkers, protein carbonyl and DNA damage (8-OHdG) (chapter 4) as well as relative telomere length (rTL) (chapter 5).

I found that at study site 1 (chapter 2), where lizards were exposed to pesticides via oral gavage, that temperature and fipronil sulfone levels significantly influenced lizard activity in the morning period of movement, where fipronil-treated individuals moved at least 49% less than fenitrothion-treated and control lizards. This was detected from 7 days after dosing through to the end of the experiment at post 28 days. Physiological measures did not change significantly before or after exposure to both pesticides, however, other indicators showed evidence of exposure which remained for the entirety of our monitoring period. On average, cholinesterase inhibition was still > 30% in fenitrothion-treated compared to control lizards at the end of 4 weeks, and fipronil sulfone blood residues remained at 0.219 pg/mL. Comparing this to lizards exposed to ULV aerial spray at study site 2 (chapter 3), fipronil-treated lizards again showed significant changes in activity measures. Interestingly, I found that higher levels of fipronil sulfone in the blood of fipronil-treated individuals correlated to significantly higher activity levels. Fenitrothion and control groups both consistently saw a decrease of up to 30 - 45% in activity over the experiment, where the fipronil group saw a 51% increase in activity between post 7 d and post 14 d. Results also showed that SBMI was not impacted by pesticide treatment but significantly influenced by sex, as well as the number to ticks present on P. vitticeps. Despite persistent depression in ChE (< 25%) in fenitrothion-treated lizards there was no significant relationship between activity or body condition measures. My findings from chapter 2 and 3 clearly demonstrate that an ecologically relevant sublethal dose of fipronil alters the behaviour of P. vitticeps, which has the potential to modify feeding, reproductive, and predator evasion behaviours. Persistence of both pesticides in the blood of all treatment lizards throughout both field experiments indicate lizards are unable to clear these toxins within 14 days - 28 days post exposure. This may be significant for compounding exposure and latent toxicity.

My findings surrounding molecular-based impacts of exposure complemented the behaviour and physiology results to form a comprehensive view on how these pesticides are influencing the biology and ecology of P. vitticeps. In my use of oxidative stress biomarkers of exposure (chapter 4) results suggested no statistically significant treatment effect of either pesticide on parameters measured, however, 8-OHdG levels decreased by >45% for both pesticide treatment groups and not controls. Protein carbonyl levels showed a high degree of individual variation that proved more influential than pesticide exposure itself. My findings on relative telomere length (rTL) changes build upon my oxidative stress results indicating that impacts of sublethal fenitrothion exposure are only detected when utilising molecular-based biomarkers of exposure. Results demonstrate a significant relationship between haemoglobin (Hb) levels and rTL aside from treatment effect. There was no significant treatment effect on rTL or 8-OHdG levels over time, however effect size was large with indication of impact. Fenitrothion-treated lizards saw an average decrease of 46% in 8-OHdG levels and increase of 42% in rTL over the experiment which is of potential biological importance. Fipronil-treated lizards on the other hand saw an average decrease of 25% in 8-OHdG levels and a 6% increase in rTL. Both markers showed negligible change before and after exposure for control lizards. Highlighted in the collective findings of this thesis is the importance of using a specific set of biomarkers of exposure dependent on the type of pesticide you are investigating. Fipronil impacts were only detected using behavioural markers whereas fenitrothion impacts were only detected using molecular markers, combined with blood biomarkers of exposure and physiology across both pesticide treatments. Overall, my thesis provides novel insights into the impacts that widely used

pesticides have on a widespread lizard species in a field-based and ecologically relevant setting.

Further, I highlight the susceptibility that reptiles have to a selection of common pesticides and

the inherent need for higher prominence in wildlife ecotoxicological research to inform risk

assessments formulated by regulatory bodies and conservation policy.