190
ShortCommunications
[Auk, Vol. 101
A Re-analysisof Hybridization between Mallards and Grey Ducks in New Zealand
MALCOLM
I-•ADDON 1
Departmentof Zoology,Universityof Otago,P.O. Box56, Dunedin,New Zealand
In the literature there is some disagreementconcerning the risk, in countriessuchas New Zealand,
Australia, and South Africa, of losing pure stocksof
indigenous duck speciesbecauseof hybridization between those stocks and populations of introduced
Mallards (Anas platyrhynchos).Dorst (1970: 252)
or embryo deaths." Greig acceptedthese conclusions
but neverthelesspointed out that someintrogressive
differencesin fertility and hatchingsuccess
resulting
from hybridization would affectthe Mallard stockto
a greater degree than that of the Grey Duck. Williams
(1981: 64) repeated this claim when he stated"At this
any pairing, such as G x M or G/M x M, etc., the
hybridization could occur.
The work by Williams and Roderick (1973) thus
has attractedsomeattention. Unfortunately, they did
not analyze their data fully, and, when this is done,
claimed that "... in New Zealand, for example, pure
different conclusionscan be reached.The purposeof
stock of the wild duck (Anassuperciliosa
superciliosa) this paper is to reanalyze their data and to discuss
is disappearing becauseit has hybridized with the the significanceof conclusionsreached from this reintroduced mallard duck." (Weller 1969, 1980) also analysis.
has expressedconcern that pure stocksof both the
The following notationswill be used throughout
New Zealand Grey Duck (Anas superciliosa
superci- this paper and were copied from Williams and Rodliosa)and the Australian BlackDuck (Anassuperciliosa erick. G/M meansan F.1. hybrid derived from a Grey
rogersi)could be eliminated at least partly as a result male x Mallard female pairing. Similarly, M/G deof hybridization with the Mallard. On the other hand, notes an F.1. hybrid from a Mallard male x Grey febased on breeding and hybridization experiments male pairing. The referencesmade to • hybrids conwith the New Zealand Grey Duck and the introduced cern the progenyobtainedfrom pairing an F.1. hybrid
Mallard, Williams and Roderick (1973) suggestedthat with a pure Grey or pure Mallard. In reference to
time there is no evidence at all to support the rather
wild notion that Mallards are interbreeding so freely
with Grey Ducks that eventually no pure Grey Ducks
will exist.What little evidence there is, suggestsGrey
Ducks are retaining their purity while Grey blood
has more strongly infiltrated the Mallard."
Williams and Roderick's (1973) study has been cited by a number of workers who have been consid-
ering similar problems.Braithwaiteand Miller (1975),
for example, did not consider that the potential for
hybridization between the Mallard and the native
Black Duck of Australia was particularly great. When
discussingthe barriers to hybridization between the
Black Duck and the Mallard they stated (p. 60) that
"Apart from the restrictionto gene flow enforcedby
speciesrecognition and pairing behaviour, there is
also evidence from studiesof mallard-grey duck hybridization (Williams and Roderick 1973) of reduced
hybrid fertility and, most important, of a high rate
of infertility of eggs and a substantially increased
rate of embryonic death from matings between the
two species."Greig (1980), in a later paper, was trying
to draw
attention
to the
threat
that
he considered
hybridization to pose to the genetic integrity of native dabbling duckssuchas thosefound in New Zealand and in South Africa. He stated(p. 88) that "Williams and Roderick (1973) also studied the breeding
success
of the hybrids and found that, depending on
the cross,there may be a proportion of infertile eggs
Presentaddress:13 Fairmile, Aylesbury, Buckinghamshire,HP21
7JT, England.
speciesof the drake is mentioned first and that of the
hen second.
Williams and Roderick(1973) concludedthat pure
populations of Grey Duck were not at risk from introgressivehybridization with the Mallard. This conclusion was reached and supported by pointing out
the differencesin fertility and the levels of embryo
death that were observedin their experiments.They
argued that (p. 69)
TABLE1. Fertility and hatching success(from Williams and Roderick 1973).
Total
number
Parents
Drake
Hen
Number
ber
of
of
Em-
of
fertile
fertile
bryos
eggs
eggs
eggs
living
G x G
M x G
10
74
10
74
10
74
10
70
G/M x G
M/G x G
33
43
33
33
33
33
31
29
3AxG
G x M
M x M
9
84
22
9
47
22
9
104
22
9
85
21
G/M x
M/G x
3/4x
G x
M
M
M
G/M
130
50
29
89
98
47
18
70
95
47
18
71
85
44
18
48
M
G/M
G
M
G
M
G/M
G/M
M/G
M/G
64
9
43
68
64
9
43
63
64
9
41
63
54
0
37
59
47
26
36
26
36
26
34
25
x
x
x
x
x •
x 3/•
January
1984]
Short
Communications
191
"Whichever pairing occurs,if breeding synchrony
either to pure Mallard or Grey Ducks or to other
can be established,the eggs laid will usually be
fertile. However, fertility can act as a barrier to
further hybridizationdependingon the way the
F.1. hybrids pair. The F.1. generationphenotypically resemblesthe paternalspeciesand might be
expectedto pair with it. If this occursas a Grey x
G/M pairingßsomeeggsin the clutch will be in-
hybrids; (2) embryo deathscannotbe said to lower
the reproductivesuccessof any of the combinations
of pairs significantly;(3) it may be that a Mallard x
Grey Duck pairing is more likely to occurthan a Grey
Duck x Mallard combination (although complete
overlap of the laying period between the Mallard
and the hybrid forms was observedby Williams and
fertile [Table 1]. Less likely pairingsße.g. M/G x
Roderick), but it is not the case that one combination
Grey, may alsoproduceinfertile eggs.
Embryodeathsalsolower the success
of someof
the pairings.In general the pairingswhich had a
high percentage
of infertileeggsalsosufferedmost
embryo deaths.Although a maternal effectcannot
will be more successfulthan the other at producing
progeny;and (4) there is no evidenceof a sex-linked
geneticincompatibility between the Mallard and the
Grey Duck.
The analysiscarried out in this paper implies that,
if the pairing of disparateforms is successfulin the
wild, then reproductive successpresentsno significant barriers to successfulintrogressivehybridization between the Mallard and the Grey Duck. This
meansthat Williams and Roderick's(1973) breeding
experimentscannot be usedßin a valid way, to support the idea that there is little risk of indigenous
speciesof dabbling duckslosing their specificintegrity through hybridization with introduced Mal-
be excluded [bantam hens were used for incuba-
tion], the high percentageinfertility and embryo
deaths recorded for Grey x G/M, and to a lesser
extent for M/G x Grey, is strongly suggestiveof
sex-linkedgenetic incompatibility....
Thus, data from this captive study are corroborated by some from the wild and together they
suggestthat the mallard x Grey pairing is more
likely to occur and produce progeny than the
Grey x mallard combinationand that the F.1. will
breed back to the mallard more successfully(at the
egg stage)than to the Grey."
These conclusionsdepend on the existenceof significant differencesbetween the breeding successof
the various combinations
of mated ducks. When
these
data are statisticallyanalyzedßhoweverßno significant differences
are found.
The columns
in Table
lards.
This was written while I was the recipient of a
University of Otago ResearchCommittee Grant. I
thank Grant Gillespie for his critical reading of the
manuscript and both Vernon Squire and Pat Langhorne for the loan of their cottagewhile I was completing this paper.
1
(column 1 against 2 and column 3 against 4) were
comparedusing 2 x 17 Chi-squaredcontingencytables. This enabled all combinationsto be compared
at once and would avoid artificially stressingany particular pairing.
The first comparisonmade was to answer the question of whether or not there was a significantdifference in the proportionof fertile eggsproducedbetween any of the 17 combinationsconsideredby
Williams and Roderick (1973). The contingency table
showed there to be no significant differences(X2=
12.24, df = 16).
LITERATURE CITED
BRAITHWAITE,L. W., & B. MILLER. 1975. The Mallardß
Anas platyrhynchos,and Mallard--Black Duck,
Anassuperciliosa
rogersi,hybridization. Australian
Wildl.
Res. 2: 47-61.
DORST,J. 1970. Before nature dies (French transla-
tion by C. O. Sherman).BostonßHoughton Mifflin.
GREIG,J. C. 1980. Duck hybridization:a threat to
speciesintegrity. Bokmakierie 32: 88-89.
WELLER,
M.W. 1969. Potentialdangersof exoticwa-
terfowl introductions.
Wildfowl
20: 55-58.
The second comparison made was to decide
1980. The island waterfowl. Amesß Iowa,
whether or not there was a significantdifferencein
Iowa State Univ.
Press.
the proportion of embryossurviving (or converselyß
proportion of embryo deaths) between the various WILLIAMS,g. 1981. The duckshooter'sbag. Wellington, New Zealand, The Wetland Press.
combinations.Again, the respective contingency taß& C. RODERICK.1973. The breeding perforble showed no significant differences (X2 = 11.26,
manceof Grey duckAriassuperciliosa,
mallardArias
df = 16).
platyrhynchos
and their hybrids in captivity. InThus, in contradiction to the conclusions of Wil-
liams and Roderick (1973), the evidence from the
breeding experimentsimplies that: (1) fertility differencesdo not act as a barrier to further hybridization deriving from the F.1. hybrids breeding back
tern.
Zool.
Yearbook
13: 62-69.
Received27 April 1983, accepted6 September
1983.