Ecological Archives E092-019-A2

Per Fauchald, Henrik Skov, Mette Skern-Mauritzen, Vera Helene Hausner, David Johns, and Torkild Tveraa. 2011. Scale-dependent response diversity of seabirds to prey in the North Sea. Ecology 92:228–239.

Appendix B. Two-stage models relating the count of each species group to year and geographically fixed variables.

TABLE B1. Summary of two-stage models. Independent variables are: Year, bottom Depth, distance to coast (Dist), geographical position (g(X,Y)) and fishing Gear (fish only). Stage (1) Logistic: Generalized Additive Models (GAMs) of the presence of individuals using a logit link with a binomial distribution. Stage (2) Gamma: GAMs of the counts conditional on presence, using a loge link with a Gamma distribution. Number of observations (N), percent deviance explained (%dev) for each model, and estimated degrees of freedom (edf) for each variable are given.

        Parametric terms   Smooth terms
      % Year Gear   g(X,Y) s(dist) s(depth)
    N dev df P df P   edf P edf P edf P
Little auk 1) 9609 33 18 ***       19.5 *** 2.1 *** 1.0 NS
  2) 764 47 18 ***       19.1 *** 1.0 * 1.0 NS
Puffin 1) 9609 33 18 ***       18.5 *** 3.7 *** 1.9 NS
  2) 725 36 18 ***       14.9 *** 3.6 NS 8.8 **
Common murre 1) 9609 9 18 ***       20.6 *** 3.4 *** 3.4 ***
  2) 5960 24 18 ***       21.8 *** 2.8 ** 3.7 ***
Razorbill 1) 9609 14 18 ***       19.9 *** 3.3 ** 3.3 ***
  2) 2017 59 18 ***       22.6 *** 3.9 ** 3.7 ***
Gannet 1) 9609 18 18 ***       19.9 *** 3.1 *** 3.9 ***
  2) 2149 26 18 ***       21.4 *** 2.7 NS 3.5 ***
Fulmar 1) 9609 32 18 ***       20.1 *** 2.4 *** 3.5 ***
  2) 3793 32 18 ***       23.4 *** 3.6 ** 4.0 ***
Kittiwake 1) 9609 5 18 ***       15.6 *** 2.6 *** 3.8 ***
  2) 4534 21 18 ***       20.8 *** 3.9 *** 3.9 ***
G. black-b. gull 1) 9609 9 18 ***       21.2 *** 3.5 *** 3.9 ***
  2) 3548 20 18 ***       19.8 *** 3.9 * 2.0 ***
Herring gull 1) 9609 12 18 ***       21.7 *** 3.9 *** 4.0 ***
  2) 3844 22 18 ***       20.2 *** 3.9 *** 3.6 ***
Common gull 1) 9609 32 18 ***       23.7 *** 3.3 *** 1.0 NS
  2) 2256 23 18 ***       20.3 *** 2.6 ** 3.8 **
Herring 1) 7518 21 18 *** 2 ***   22.5 *** 3.9 *** 3.6 ***
  2) 6506 18 18 *** 2 ***   23.1 *** 3.8 *** 3.9 ***
Sprat 1) 7518 48 18 *** 2 ***   21.8 *** 3.9 *** 3.1 ***
  2) 4566 33 18 *** 2 *   23.8 *** 3.9 *** 2.0 NS
Krill 1) 8037 18 18 ***       17.5 *** 1 NS 2.8 NS
  2) 2860 23 18 ***       21.0 *** 3.0 NS 1.0 **
C. finmarchicus 1) 8037 20 18 ***       22.4 *** 3.1 NS 1.0 NS
  2) 2299 27 18 ***       23.0 *** 1.0 NS 3.5 ***
C. helgolandicus 1) 8037 7 18 ***       20.5 *** 1.0 NS 1.1 NS
  2) 4024 11 18 ***       22.6 *** 3.2 * 2.5 NS
Para/pseudoc. 1) 8037 13 18 ***       23.5 *** 1.0 * 1.0 *
  2) 3010 16 18 ***       20.0 *** 1.0 NS 3.7 ***

*** P < 0.001, ** 0.001 < P < 0.01, * 0.01 < P < 0.05, NS P > 0.05


FIG. B1. Time series of the winter abundance of seabirds (upper), pelagic fish (middle) and zooplankton (bottom) in the North Sea as predicted from the two-stage models. All data are log10 transformed. Seabirds are predicted number of birds in the study area. Fishes and zooplankton are catch per unit effort (CPUE) summed over the study area (see text). Fishes: sprat and herring are number of individuals caught per hour of trawling. Zooplankton: krill, C. finmarchicus, H. helgolandicus and Para/pseudocalanus are number of individuals counted per Continuous Plankton Recorder (CPR) sample.

FIG. B2. Average winter distribution of seabirds in the North Sea as predicted from the two-stage models.

FIG. B3. Average winter distribution of herring and sprat (upper) and zooplankton (lower) in the North Sea as predicted from the two-stage models.
[Back to E092-019]