“…Although all methods for estimating numbers of birds include caveats about sampling during periods with extreme weather conditions, surprisingly little research has addressed the magnitude of bias introduced by weather. To summarize, these biases can be reasonably controlled by avoiding periods of (1) strong wind (> 11 km/hr: Dawson, 1981a;Emlen and Dejong, 1981;Robbins, 1981b; but see Anderson and Ohmart, 1977, for a recommended threshold of 20 km/hr), (2) light to heavy precipitation (O'Connor and Hicks, 1980;Dawson, 1981a;Robbins, 1981b), (3) extreme temperatures « 7°C or > 24°C) (Robbins, 1981b), and (4) fog (Robbins, 1981b). Some of these recommended levels were not specifically identified in the literature cited, but represent my interpretation of results therein.…”