718 
THE WILSON JOURNAL OF ORNITHOLOGY • Vo l 124. No. 4. December 2012 
TABLE 5. Observed and expected frequencies of pairings in nesting Aechmophorus grebes and grebes with chicks 
among Western (n = 69). Clark's (/i - 14). and intermediate grebes (// = 7). and in all pairs assessed among Western (n = 
82). Clark's (n = 31). and intermediate grebes (n = II). 
Pair composition 
Pairs nesling/tending young 
All pairs assessed 
Observed frequency Expected frequency Percent deviation 
Observed frequency 
Expected frequency 
Percent deviation 
WGXWG 
31 
26.36 +17.6 
36 
27.00 
+33.3 
WGXCG 
0 
10.85 -100 
0 
20.67 
-100 
WGXIG 
7 
5.43 +29.0 
10 
7.33 
+36.4 
CGXCG 
7 
1.02 +584.6 
15 
3.78 
+296.8 
CGXIC. 
0 
1.10 -100 
1 
2.77 
-63.9 
IGXIC. 
0 
0.24 -100 
0 
0.45 
-100 
X Goodness of fit test y 2 
= 48.41, df = 5. P < 0.0001 
T = 59.5 
1. df = 5. P < 0.0001 
on intermediary plumages between species to 
identify hybrids. DNA analysis could provide 
more reliable results. Genetic data may not solve 
the problem in the case ol Aechmophorus grebes. 
Studies ot the genetic dissimilarity between 
Clark's and Western grebes have had inconsistent 
results (Guerra and Speed 19%, Nuechterlein and 
Buitron 1998. Kerr ct al. 2007). The relaliv. 
frequency of mixed species pairs of up to 5% 
depending on populations (Nuechterlein anc 
Buitron 1998), and the observations of mixec 
pairs with viable offspring (Ratti 1979; Storer am 
Nuechterlein 1985; Nuechterlein and Buitror 
1998; pers. obs.) leave little doubt that hybrid) 
exist. Observations of intermediates with yotin^ 
were sufficient proof for the fertility of mixec 
species pairs' descendants for Nuechterlein ami 
Buitron (1998). They considered that hybridiza¬ 
tion for Aechmophorus grebes may reflect an 
adaptive mate choice rather than a mistake. Thus, 
debatable individuals or intermediates can be 
considered suspected (Eichhorst and Parkin 1991 ) 
or presumed (Nuechterlein 1981a) hybrids 
Randier (2008), in his meta-analysis of avian 
hybrid zones, concluded assortative mating was 
-stronger m narrow hybrid zones compared to 
wider ones. Th.s supports the presumed hybrids 
"m n ~ S r ,udy . and , for Califomia-Oregon (Konter 
-Oil) f° r largely sympatric Aechmophorus 
populations. Lower percentages were recorded 
n P'aine Canada, a region of relative allopatry 
where Clark s Grebes represent only I - 4 % c p 
the Aechmophorus population. Nuechterlein and 
Buitron U 998) re P orted intermediates represent- 
ed <4,o of the population in 1980 and in the 
early iqqos at Delta Marsh. Manitoba. Konter 
(2009), ln a survey of different lakes in 
Manitoba and Saskatchewan in 2008. found 
3% intermediates. 
Reinforcement of incompatibilities by selection 
in sympatry (Servedio and Noor 2003) should lead 
to a taster reduction of interbreeding in sympatric 
than in largely allopatric populations (Randier 
2006). Increased percentages of intermediates in 
Utah and Cali torn ia-Oregon a priori contradict 
growing reinforcement of incompatibilities be¬ 
tween both Aechmophorus grebes. Hybridization 
becomes an ongoing process acting against 
narrowing of hybrid zones (Randier 2008) if. in 
situations with limited choice. Aechmophorus 
grebes do not pair assortatively. 
Gene flow can prevent speciation even at low 
rates (Winker 2010) and development of postzy- 
gotic incompatibilities possibly originating ~- 
million years after divergence (Price and Bouvier 
2002). We may question whether the existing 
prezygotic barriers to random mating in Aechmo¬ 
phorus species are not being lost as numbers of 
intermediate Aechmophorus grebes generally in¬ 
crease (Konter 2011). Repeated backcrossing may 
give rise to individuals of mixed ancestry that can 
^ UlOUIIgUIJIICU ..v*... ,,u,w.. 
(Howard et al. 2003. Randier 2004) and they are 
absorbed into one parental species. The process of 
introgressive hybridization (Anderson and Hu- 
bricht 1938). besides introducing new variation 
into the parental population, may reduce differ* 
ences in advertising calls between Western and 
Clark s grebes. This is a possible reason for the 
near total absence ot Clark's Grebes at Panguitch 
Lake in 2010; only two of 43 grebes assessed 
were Clark’s and six displayed intermediate traits. 
Clark’s Grebes were a specialty at this lake in the 
late 20th century (Mclvor 1998). 
