Modeling territory attendance and preening behavior in a seabird colony as functions of environmental conditions

Abstract

In previous studies we developed a general compartmental methodology for modeling animal behavior and applied the methodology to marine birds and mammals. In this study we used the methodology to construct a system of two differential equations to model the dynamics of territory attendance and preening in a gull colony on Protection Island, Strait of Juan de Fuca, Washington. We found that colony occupancy was driven primarily by abiotic environmental conditions, including tide height, time of day, solar elevation, and wind speed over open water. For birds in the colony, preening behavior was driven to some extent by abiotic environmental conditions (including time of day, solar elevation, humidity, and wind speed on the colony), but apparently was driven primarily by local and/or biotic effects not included in the model. In terms of R(2) values, the model explained 65% and 37% of the variability in colony occupancy and preening data, respectively, as a function of these six abiotic environmental factors.

Feather mite abundance varies but symbiotic nature of mite‐host relationship does not differ between two ecologically dissimilar warblers

“…These microhabitats differ by many abiotic factors, including temperature and humidity, which could directly affect the ability of mites to survive or reproduce on their hosts (Meléndez et al., ; Wiles et al., ). It is also possible that because canopy species may be exposed to harsher environmental elements (e.g., rain, wind, and fluctuating ambient temperatures) than understory species, they may need to preen more often to maintain feather condition, thus providing more uropygial oil for mites to consume (Henson, Galusha, Hayward, & Cushing, ). Differences associated with the geographic locations of our study areas did not appear to be influential in driving the species differences, as S. cerulea from both regions were infested with much greater numbers of mites than P. citrea in either region.…”

Predicting Gull/Human Conflicts With Mathematical Models: A Tool for Management

“…By contrast, Appledore Island occurs in the open ocean and is subject to almost continuous winds. Wind clearly alters the behavior and movement patterns of gulls (Henson et al [2007a, b]). Although wind direction was included as a variable in the Appledore model, its highly variable nature and relative unpredictability seemed responsible for much higher levels of stochasticity for the loafing patterns there.…”

“… The number of gulls in the loafing site can be described with a two‐compartment model consisting of the loafing site and a remote location (places other than the loafing site). Fluctuations in numbers of gulls at the loafing site occur in direct response to environmental variables that vary in time t. In particular, gulls arrive at the site at a per capita rate proportional to a function of environmental variables E 1 ( t ), and leave at a per capita rate proportional to a function of environmental variables E 2 ( t ). The rate functions E 1 ( t ) and E 2 ( t ) are multiplicative functions of powers of environmental variables (Damania et al [2005], Henson et al [2007a, b], Moore et al [2008]). This is equivalent to the assumption of log‐linear, or Poisson, regression of rates on environmental factors (McCullagh and Nelder [1989]).…”

“…The rate functions E 1 ( t ) and E 2 ( t ) are multiplicative functions of powers of environmental variables (Damania et al [2005], Henson et al [2007a, b], Moore et al [2008]). This is equivalent to the assumption of log‐linear, or Poisson, regression of rates on environmental factors (McCullagh and Nelder [1989]).…”

“…We nondimensionalized and scaled variables S , T , W , B , H so that their values always lay between one and two (Damania et al [2005], Henson et al [2007a, b], Moore et al [2008]). This is accomplished by subtracting off the minimum value, dividing by the new maximum value, and adding one.…”