I think most field observers would agree that the identification of gulls is a special challenge.  The process of putting the right name to the right gull takes considerable practice and patience.  It also depends on the state of the information base available for making reasoned identifications.  Over the last twenty or so years the information base regarding gull species has changed.
     I extracted the numbers of four-year, pinked-footed gulls seen on Oregon Christmas between 1974 and 1993 (Table 1) from the CBC database (Shipman, 1997).  I then looked at absolute number (Figure 1) and relative numbers (Figure 2) of gulls seen.  Absolute numbers give a picture of general fluctuation in winter gull populations.  Changes in the number of counts or observer availability may influence absolute counts.  Relative numbers allow for the examination of changes in species composition over time without variations in total numbers getting in the way.  Absolute number should fluctate more substatially than relative numbers.
     Over time the numbers of gull species counted has changed as a function of changes in the information base, demonstrating what the most often cited weakness in Christmas Bird Count data: observer bias.

Thayer's Gull

     Prior to 1973, Thayer's Gull was considered a subspecies of Herring Gull.  Figure 7 shows the relative number of Thayer's Gulls reported over the twenty years immediately after the split.  As observers got used to the idea of Thayer's Gull, more and more were seen, finally peaking in about 1981.  It seems unlikely that the peak represents the actual state of Thayer's Gull population.  Rather, over zealousness on the part of observers with a new piece of information, but no real field experience regarding the identification of this species, probably lead to larger numbers of birds being misidentified.  Clicking on any individual graph will take you to a full sized copy.
 

Herring Gull

     Herring Gull reports show a steady decrease over the past 20 years (Figure 6).  This closely correlates with the increase in reports of Glaucous-winged x Western Gull hybrids (Figure 5).  Hybrids were only rarely reported before 1984, but drops in Herring Gull, Thayer's Gull and even Glaucous-winged Gull counts coincide with the increased reporting of hybrids.  The presence of high concentrations of GW x W hybrids was documented in the literature as early as 1971 (Scott,1971) and hybrid distribution along the Pacific Coast was well defined by 1974 (Hoffman, 1978), but observers skilled enough to sort out the problem on Christmas counts did not appear for nearly 10 years.

     Herring and Thayer's Gulls reported on Christmas bird counts prior to about 1986 should be considered highly suspect.  The majority of these reports are most probably misidentified Gw x W hybrids.

     Observer bias may also be seen in the band wagon effect that goes along with changes in the identification information base. It is best demonstrated in the spike of Thayer's Gulls reports in 1982-3.  Field observers were seeing Thayer's Gulls where there were (most probably) none to be seen.

References

Bell, Douglas A. 1996. Genetic Differentiation, Geographic Variation and Hybridization in Gulls of the Larus Glaucescens-occidentalisComplex. Condor 98:527-546.

Bell, Douglas A. 1997. Hybridization and Reproductive performance in Gulls of the Larus Glaucescens-occidentalisComplex. Condor 99:585-594.

Hoffman, W., J.A. Wiens, & J.M. Scott. 1978. Hybridiation between gulls (Larus glaucescens and L. occidentalis) in the Pacific Northwest. Auk 95:441-458.

Scott, J.M. 1971 Interbreeding of the Glaucous-winged Gull and Western Gull in the Pacific Northwest. Calif. Birds 2:129-133.

Shipman, John W. 1998. Christmas Bird Count Database Project.
[On-line] Available WWW: http://www.nmt.edu/~shipman/z/cbc/ homepage.html