Identifying Environmental Determinants of Diurnal Distribution in Marine Birds and Mammals

Abstract

Marine birds and mammals move between various habitats during the day as they engage in behaviors related to resting, sleeping, preening, feeding, and breeding. The per capita rates of movement between these habitats, and hence the habitat occupancy dynamics, often are functions of environmental variables such as tide height, solar elevation, wind speed, and temperature. If the system recovers rapidly after disturbance, differential equation models of occupancy dynamics can be reduced to algebraic equations on two time scales. Identification of environmental factors that influence movement between habitats requires time series census data collected in both the absence and presence of disturbance.

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

“…Hourly tide heights and solar elevations, as well as wind speeds in the Strait (measured at Smith Island), were obtained from the National Oceanic and Atmospheric † The trapezoid {(C, P )|0 ≤ P ≤ C and K/2 ≤ C ≤ K} is forward invariant under model (6). This is because along the lines C = K/2, C = K, and P = 0, we have the inequalities dC/dt > 0, dC/dt ≤ 0, and dP /dt > 0, respectively.…”

“…Administration (NOAA). We nondimensionalized all environmental variables x so that 1 ≤ x ≤ 2 [1,3,4,6,7].…”

“…Model (6) contains 107 parameters: the two parameters K and α shown explicitly in equation (6), and three coefficients and 18 exponents as shown in equation (5) for each of five per capita flow rates shown in equation (6). The stochastic model (3)-(4) further contains the parameter ψ (as well as three variance-covariance parameters, but these can be computed from the fitted sum of squared residuals).…”

“…This created a very large number of possible models in the form of model (6). We chose a relatively small subset of these based on (a) the biologists' knowledge of what was likely to be most biologically reasonable for the system, (b) statistical investigations comparing the data and the environmental variables, and (c) the experience gained by weeks of trial and error searches for models that could fit the data.…”

“…Damania et al [4] and Moore et al [5] modeled the diurnal occupancy dynamics of a system of habitat patches in and around the Protection Island gull colony as a function of environmental factors with R 2 values up to 0.82. Henson et al [6] modeled the same system on two time scales, and studied recovery from disturbances as a function of environmental conditions. Henson et al [7] used a system of differential equations to predict the dynamics of colony occupancy and sleep in the Protection Island gull colony.…”