influence du climat et de la prédation sur l'utilisation de l'habitat et la ...

59
Third, we evaluated the behavioural response of porcupines to variations In Tc. We
recorded time spent outside of the den, number of activity bouts, and a nocturnality index
(equation 2 below) for each day (from 00:00 am to 24:00 pm) and compared them to the
mean Tc of the day (recalculated from continuous measures of Ta, wind speed and net
radiation made at our weather station, equation 3 below). We also compared the duration of
activity bouts to the mean Tc during the activity bout. Finally, we determined which
microhabitats were used when porcupines were outside of their den, according to Tc
measured in the stand at the time of observation. Because porcupines are sexually dimorphic,
we tested for sexual differences in behaviour.
Meteorological conditions
Our automated weather station measured mr temperature In the shade (Ta) using a
temperature sensor (8-bit Temperature Smart Sensor, Hobo, Onset) placed 1 m above ground,
wind speed using a wind and direction sensor (Wind speed/Direction Smart Sensor, Hobo)
placed 2 m above ground, and net radiation (0-100 !lm) using a net radiometer sensor (model
NR-Lite Net Radiometer, Kipp & Zonen) placed 1.5 m above ground. Ail sensors were
connected to a datalogger (four channels Micro-station, Hobo) that collected data every
minute.
We placed temperature loggers (SmartButton Temperature Loggers, ACR systems lnc.)
inside nine dens that were regularly used by porcupines during our long-term study, but that
were not occupied during the sampling period. Loggers were placed on the ground as far as
possible from the entrance of the den (0.5 to 6.5 m deep, depending on den structure).
Loggers recorded temperature every 15 minutes.
Thermal mapping
We worked during two winters to establish the thermal map: 29 January to 30 March
2004 and 13 December 2004 to 8 February 2005. Following Bakken et al.(1983), we built
two heated taxidennic mounts in the fonn of North American porcupines. To coyer the
copper core we used pelts from animaIs that died naturally in our study area during winter
2003. We built the mounts in a standing posture with the head facing down. We lined the
inside surface of the hollow copper mould with 24 gauge Teflon-coated iron wire that was
connected to an external control circuit and to a Campbell CR-21X data logger (Campbell
Scientific). Electrical power was provided by four 6 V deep discharge batteries in series.