2025 NZES National Virtual Event

Mortality of birds associated with the construction and operation of photovoltaic solar farms (PVSFs) is a recognised problem globally. Given the dramatic increase in the size and number of planned PVSFs in Aotearoa, especially near sensitive bird habitats, it is important that the potential impacts on birds are assessed and accounted for. However, we have a conundrum. There has been no monitoring of effects of PVSFs in Aotearoa to date to base risk assessments on. In addition, published studies largely come from arid areas in California, so the level of inference achievable for situations in this country is unknown. Thus, the ability of the solar industry to assess and mitigate risks is uncertain. Bird mortality overseas is largely associated with collisions with solar infrastructure (panels, transmission lines, fences, buildings, vehicles). Hypotheses explaining collisions include birds mistaking PV panels as water or birds becoming disorientated when detecting polarised light from panels, although the causal mechanisms by which fatalities occur are unknown. Some overseas studies have estimated cumulative quanta of tens or hundreds of thousands of bird deaths/yr associated with PVSFs. There are suggestions that disproportionately large numbers of waterbirds are recorded and proximity to water may be a contributing factor. Collision risk has yet to be studied in Aotearoa, so the quantum of risk posed by PVSFs for birds is unknown. While some authors have discounted bird mortality as impacting birds at a population level, the situation in Aotearoa is different because many threatened species are at risk, some of which have small populations where the loss of individual breeding birds could have disproportionate negative effects on species recovery. There are numerous examples of bird species being found dead overseas that are ecologically like threatened species of concern in this country. But there appears to be considerable variability in the numbers of birds found dead at PVSFs likely reflecting both ecological and site-based factors such as bird behaviour, habitat type, geographic location and proximity to water, and PVSF age, size, configuration and association with other solar power generation types. The development of PVSFs has been rapid and recent in Aotearoa and developments run the risk of outpacing our understanding of risks. The uncertainties about potential impacts and effective mitigation techniques point to urgent knowledge gaps in the context of this emerging industry in Aotearoa. Until meaningful monitoring is conducted, it would be prudent to apply the precautionary principle to consenting PVSFs by not building PVSFs on or next to significant bird sites, applying infrastructure cautiously with appropriate experimentation, and developing and testing a suite of mitigation and adaptive management options to avoid or minimise risks.

Pitfall and funnel traps are widely used methods for monitoring lizards in Aotearoa New Zealand. Positive results of trapping are easy to interpret; if you catch a lizard, it is certainly present at the trapping site. However, negative results are difficult to confidently relate to species absence. A lizard may be present at the site, but not be captured by a trap, because it: 1) does not encounter the device or 2) encounters the device but does not enter. The latter outcome is a significant determinant of the efficiency of trapping and how likely a negative result will represent an occupancy state. To investigate the interaction rates of lizards with pitfall and funnel traps, we paired devices with downward facing trail cameras set to take a still image every 10 seconds. We generated 709 days of images at six different sites in the Mackenzie Basin, targeting McCann’s (Oligosoma maccanni), grass (Oligosoma aff. polychroma Clade 5), Mackenzie (Oligosoma prasinum), Lakes (Oligosoma aff. chloronoton ‘West Otago’), and southern long-toed (Oligosoma aff. longipes ‘southern’) skinks. We manually processed 2,255,895 unique images to determine lizard presence and the time at trap and related these metrics to the number of lizards caught in traps. Comparisons indicated that trapping was highly inefficient. Lizards entered traps on less than 5 % of the occasions that they were in the cameras’ focal range (~1.5 m2). Low interaction rates are problematic for applications that require certain thresholds of captures e.g., mark-recapture, and when trying to confidently determine absence for rare species or populations at low densities. We suggest that trail camera photographs provide a more efficient means to detect lizards, but significant improvements to image processing are needed if this methodology is to be widely applied.