Year: 2016
Type: article
Abstract: We present a deterministic model for predicting daily energy expenditure of two raptors--femaie American Kestrels (Falco sparverius) and Whim-tailed Kites (Elanus leucurus)-wintering in coastal northwestern California.Inputs to the model include body mass, air temperature, photoperiod, energy expenditure of flight, and relative portions of the daytime spent in flight and nonflight activities.A simplified version of the model applicable to birds spending less than 20% of the day in flight and inflating daily energy expenditure by 6% or less is also presented.Inputs to the simplified version include body mass, air temperature, energy expenditure of flight, and relative portions of the 24-h day spent in flight and nonflight activities.Input data were estimated directly and indirectly.The validity of the model was tested by comparing predicted energy expenditure with energy expenditure estimated by observed food consumption of wild birds.The model predicted that individual female kestrels would expend 42.0-61.0kcai (2.04 to 2.96 W) daily and that individual kites would expend 105.6-118.3kcal (5.12 to 5.74 W) daily.Daily energy expenditure estimated by food consumption averaged 42.9 and 113.1 kcai (2.08 and 5.49 Vv') per individual kestrel and kite, respectively.The degree of correspondence between model prediction and field estimation of energy expenditure of kites was considered adequate for model validation.Even though the model predictions bracketed the field estimation of energy expenditure of kestrels, however, the model predictions were considered to be too high because of an erroneous temperature input, and the field estimation was considered to be too low because of an erroneous estimate of the biomass of an important group of prey.Correcting these errors indicated that the daily energy expenditure of kestrels should average 48.7 kcai (2.36 W) per individual.Using the corrected energy expenditure as a standard for female kestrels and the field estimate of energy expenditure as a standard for kites, the predictive accuracy of the versions of the model was evaluated relative to the predictive accuracy of 11 other models.Three of these models, including the two versions presented here, produced estimates that were within 5 % of the mean standard value.Eight of the models under-approximated the mean standard value by 10.3-49.5%; the other two over-approximated the mean standard value by 14.8 and 36.9%.
Language: English
Subfield: Ecology
Field: Environmental Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W150634703
