2.1 Study site and data collection
Data collection took place between 2013 and 2020 at two African penguin
colonies in the Western Cape Province, South Africa: Robben Island
(33°48’ S, 18°22’ E) and Stony Point (34˚22’S, 18˚53’E) (Figure 1). From
2013 onwards, penguins were captured in each colony, and injected with
passive integrated transponders (PITs). For 2013 and 2014 these were
Half Duplex (HDX), 134.2 kHz, ISO 11784/11785 compliant, 32mm glass PITs
(31.2 [l] × 3.85 [d] mm, weight 0.8 g), injected subcutaneously
into the back of the neck. From 2015, Full Duplex (FDX-B), 134.2 KHz,
ISO compliant, 12 mm PITs (Biomark, Boise ID, USA) were injected
subcutaneously into the skin flap posterior to the left leg. Subsequent
encounter data of tagged penguins were then collected from 2014 onwards.
As part of routine nest monitoring between March and October, the
presence and breeding status (breeding or nonbreeding) of tagged birds
was identified using a hand-held transponder reader (Datamars model
GES3SEU with external stick antenna from 2014–2015, Allflex model RS420
from 2016 onwards). Each year captured untagged birds were tagged under
the same protocol (Table A1, Appendix). Supplementing this, ground
reader systems (Biomark IS1001 with loop antenna) were installed across
commonly used highways to/from the sea; one reader was installed at
Robben Island in 2015, and two at Stony Point in 2016 and 2017
respectively. Although not completely curtailed in either season, data
collection using the hand-held transponder readers was negatively
impacted by an Avian Influenza outbreak in 2018 (Molini et al .
2020), which limited close approaches by researchers to penguin nests,
and by the COVID-19 pandemic in 2020, which limited the number of
person-days spent in the field (relative to other breeding seasons),
particularly at Robben Island.