Correlated observer errors and their effects on survey estimates of needle-leaf loss