close

Вход

Забыли?

вход по аккаунту

?

Inbreeding in Finland.

код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 84~127-139 (1991)
Inbreeding in Finland
L.B. JORDE AND K.J. PITKANEN
Department of Human Genetics, University of Utah School of Medicine,
Salt Lake City, lltah 84132 (L.R.J.,;Samfundet Folkhalsans Genetiska
lnstitut, 00101 Helsinki, Finland (L.B.J.1;Department of Economic and
Social History, University of Helsinki, 00100 Helsinki, Finland (K.J.P.)
KEY WORDS
Consanguinity, First-cousin marriage
ABSTRACT
We have compiled data on the frequency of first-cousin
marriages in Finland using royal dispensation records for the time period
1810-1872 and national population statistics for the time period 1878-1920.
For the earlier period, 0.315%of Finland’s marriages were contracted between
first cousins (2,331 of 739,387). During the second time period, 0.174% of
Finland’s marriages took place between first cousins (1,325 of 761,976).These
figures, which yield average kinship coefficients of 0.00020 and 0.00011,
respectively, show that the level of inbreeding in Finland due to first-cousin
marriage has been quite low. An analysis of individual parishes shows that
first-cousin marriages are, on average, substantially less frequent than predicted by a random-mating model. In order to evaluate determinants of
first-cousin marriage, several predictive variables have been examined: parish ethnic composition (proportion of Swedish and Finnish speakers), husbands occupation (graded into 6 socioeconomic levels), geographic distance
between spouses’ premarital residences, population density, parish endogamy, and urban vs. rural residence. Various logistic and linear regression
models were analyzed in which consanguinity was the dependent variable.
The best predictors of consanguinity were ethnic composition and occupation.
The other variables were not in general significant predictors. These results
show that many of the “mate availability” factors that would be predicted
theoretically to account for consanguinity variation (population density, geographic isolation, urbanvs. rural residence) do not. Instead, the best predictors
of consanguinity at the first-cousin level are cultural factors such as ethnicity
and occupation. Evaluation of cultural variables can provide a greatly enriched interpretation of complex biosociai phenomena such as inbreeding.
Inbreeding has long been a subject of interest to human geneticists, reflectin its
importance both in evolutionary and me ical
genetics (see Cavalli-Sforza and Bodmer,
1971; Jorde, 1990; Lebel, 1983; Reid, 1973;
and Roberts, 1975 for reviews). Population
genetic models of inbreeding tend to treat
t,his phenomenon in a rather mechanistic
fashion, t pically incorporating only the effects of e ective population size and migration rates. Such treatments are particularly
inap ropriate for humans, since a large
num er of economic, demo aphic, and cult,ural factors can influence uman breeding
patterns. In order to attain a more realistic
assessment of inbreeding in human populations, a greater variety of possible causal
i
x
E
0 1991 WILEY-L.ISS, INC.
a
determinants needs to be evaluated. This is
usually difficult, since it is not often possible
to obtain a large sample of accurate data on
consanguinit and potential causal factors.
Because ofYits excellent archival records,
Finland’s population presents one opportunity in which to overcome these difficulties.
The quality and availability of these records,
as well as Finland’s distinctive distribution
of genetic diseases, has led to a number of
demographic (Mielke et at., 19871,epidemiologic (Mielke et al., 1984; Pitkanen et a].,
1989; Jorde et al., 1989, 1990), and population enetic studies (e.g., Jorde et a]., 1982;
Miel e et al., 1976; Nevanlinna, 1972; Norio,
a
- . ._. .- .
Received February 5. 1990;accepted August 3, 1990
128
L.B.JORDE AND K.J. PITKANEN
1981;OBrien et al., 1988a,b;Pitkanen et al.,
1988; Rogers and Jorde, 1987; Workman et
al., 1976).
Using dispnsation records and other population statistics, we have compiled data on
first-cousin marriages in Finland from 1810
through 1920. A set of nonconsanguineous
control marriages was also assembled for the
1810-72 time period. In addition, data were
obtained on occupation, geographic distance
between mates' residences, population size,
population density, urban vs. rural residence, and ethnic group (Swedish or Finnish
speaking). In this study, Finnish first-cousin
marriage rates are re rted for the first
time, and temporal an s atial variation in
consanguinity is assesse . Determinants of
consanguinity are evaluated using linear
and logistic regression models. We demonstrate that most of the traditionally measured variables are not good predictors of
consanguinity at the first-cousin level. Instead, cultural factors that are less often
considered, such as ethnicity and social
class, are the best predictors of first-cousin
marriage in this population.
$B
MATERIALS AND METHODS
Data collection
Prior to 1680, first-cousin marriages had
been prohibited b law in Finland for several
centuries. After t is date, it was possible to
marry a first cousin if a dis ensation was
obtained from the Crown. Tfis option was
exercised only b the Lutherans, who comprised 98% of t e population in the 19th
century. Because of strong religious proscriptions, first-cousin marria es did not occur among the Greek-Ortho ox minority.
Dispensations were also required for certain
other types of marria es (e.g., a man wishing
to marry his decease!c brother's wife).
Since Finland was part of Sweden until
1809, the Finnish dispensation records for
marriages occurring before 1810 are kept in
the Swedish National Archives in Stockholm. It was not economically feasible to
compile these records because they are scattered randomly in a large series of volumes.
Thus, the data collection in this study begins
in 1810, when dispensations were granted by
the Finnish Senate. These records are kept
in the Finnish National Archives in Helsinki. The requirement for dispensations for
first-cousin marriages was abolished by an
Imperial Proclamation on 4 November 1872.
From 1878 until 1920, ministers were asked
K
K
%
to record annually the total number of firstcousin marriages that took place in their
parish for inclusion in the Population Statistics. These data were collected using the
Population Change Tables (ouestonmuutostaulukot, statistical forms compiled by the
parish ministers) which are kept in the Central Statistical Office in Helsinki.
The dispensation records usually provide
data on the mates' names, civil status, occupation, and lace of residence (church parish). When p ace of residence was not mentioned, the actual dis nsation a lication
was located. This usua ly rovide t e missing information. In the 1 78-1920 time period, data were available at the aggregate
level only. Thus, it was not ssible to examine causal variables for t is time period.
Only spatial and temporal variation in these
first-cousin marriages is reported.
Parish ministers were under strict instructions to require dispensations for all
first-cousin marriages within their jurisdiction. Their employment could be terminated
if the failed to do so. Nearly everybody was a
mem er of the Lutheran Church at this time,
and ministers had close contact with their
parishioners. This makes it unlikely that
very many first-cousin marriages would remain unrecorded. Nevertheless, two additional procedures were employed to verify
the data: 1)An announcement was placed in
two publications, Gems and Oulun Sukututkija, asking enealogists to write to one of
the authors ( .J.P.)with information on any
first-cousin marriages they mi ht have discovered in genealogcal researc . Of 22 firstcousin marria es reported to us, 21 were
found in the atabase. Second, the firstcousin marriages in our database were
checked against published genealogies of the
Finnish nobility (Carpelan, 1954, 1958,
1965). Of the 120 first-cousin marriages ascertained from these sources, 97 were found
in our database. Of the 23 that were not
found, 12 represented marriages that took
place in a foreign country and were thus not
expected to be part of the database. Of the
remaining 11 marriages not found in the
database, 6 took place during the 1810-19
time eriod, when historical accounts indicate t at there was some underreportin of
first-cousin marriages. Based on these i g
ures, it appears that the ascertainment of
first-cousin marriages using dispensation
records is nearly complete.
In the case-control analysis described below, dispensations were used only if an ac-
P
r % ZR
R"
i
a
3
R
a
f
129
INBREEDING IN FINLAND
tual marriage was confirmed for the cou le.
Marriage entries were not found for f;73
(16%) first-cousin couples who had sought a
dis nsation. Near1 all of these couples did
in act get married: but there are several
reasons why it was not possible to locate
their marriage records. First, not all of the
marriage records are preserved, and some
parish records were not available in the National or Provincial Archives for the whole
period under investigation. Second, the marria e records seem to fail to register a portion
oft e marriages contracted in the parishes,
since in many cases we were able to find a
reference to the marriage in other records of
t.he couple’s parish. Third, we were not always able to locate the parish in which the
marriage was recorded. In fact, the place of
residence gwen by the dispensation records
did not always prove to be accurate. Man
communities consisted of a mother paris
and one or several subsidiary parishes
(chapelries).The name of the mother parish
(which also referred to the whole community) was often given as the place of residence, althou h the couple was actually livin in one oft e chapelries.
$he denominator data for this study,
which consisted of the total number of marriages for each parish, were given in the
Population Change Tables for the years 1812
to 1872. For the years 1810 and 1811 this
information was not available for individual
arishes. The occasional gaps during the
illowing period were filled by using the
arish marriage records (uihitty ‘en luette&t), or when these were missing, y employing straightforward interpolation
dures. For the period from 1878 to 19 0 the
marriage totals are published in the Po ulalion Statistics (Official Statistics of Fin and,
Series VI).
In order to perform a case-control analysis
0:’ consanguinity determinants, a control
sample of marriages was abstracted from the
parish marriage records. Control marriages
were matched with cousin marriages on the
basis of year of marriage and bride’s arish
of residence. Among the marriages t at fit
the matchin criteria, three control marriages were rawn randomly for each cousin
marriage. In a small number of cases, fewer
than three possible controls were available
within a matching category. When this happened, the same marriage could be used
more than once as a control.
r
a
E
a
I:
Three levels of geogra hic subdivision
were used in this study. he first level of
subdivision is the municipality, usual1
equivalent to one church parish, of whic
there were 500 in 1870. Because some arishes were not separate units during a I of
the study period, it was necessary to lump
them. The resulting number used in analyses was 457. The next level of subdivision is
the “deanery”(ecclesiastic units headed by a
Lutheran dean). There were 43 rural deaneries, each of which consisted of grou s of
parishes. The cities, which numberel 34,
were each treated as additional separate
subdivisions a t the deanery level. The largest unit of subdivision is the rovince, of
which there were 8 during the 1 th century.
Two of the provinces (Vaasa and Oulu) were
geographical1 very large and culturally hetero eneous. his led to the creation of two
adJtional administrative provinces in the
mid-20th century, Central Finland (KeskiSuorni) and Lapland. Because of this heteroeneity, we designated Central Finland and
fapland as separate provinces for the time
period covered by this study.
Geographic distances between mates’ residences were computed using the geographic
coordinates of the center of each municipality (parish). Since residences were often recorded only at the municipality level, a zero
distance would be assigned to mates who
originated from two different villages in the
same municipality. If the mates originated
from the same municipality, their marriage
was termed endogamous.
Population sizes for each municipality
were recorded in the Population (Census)
Tables, another statistical form completed
by the parish ministers. Prior to 1860 these
data were collected from the tables available
in the Central Statistical Office. During and
after 1860, the
ulation sizes were published as part o t e Population Statistics.
Population density was calculated b dividing the population size of each paris by its
land area.
Based on data from the 1880 census, it was
possible to classify each municipality into
one of four categories: 1) >95% Swedish
speakers; 2) 50-959b Swedish s akers; 3)
50-95% Finnish speakers; 4) >9 5% Finnish
s akers. This categorization is the basis of
t e “ethnicity” variable.
f
P
15
I
?YeP
K
dg
Data analysis
YK
i
?!
R“
K
130
L.B.JORDE AND K.J. PITKANEN
Since occupation was listed for most men
and women in both the dispensation and
marriage records, it was possible to group
occupations into seven broad occupational or
“social” classes: 1) nobilit and gentry; 2)
high-status farmers (rust Mare); 3) other
farmers; 4) crofters (tenant farmers); 5)
skilled workers and artisans; 6) servants
arid unskilled workers; 7) “other.”
The proportion of first-cousin marriages in
the population was estimated by dividing the
number of first-cousin marriages b the total
number of recorded marria es. T is figure
was in turn divided by 1 to obtain the
average kinship coefficient, F. For convenience and brevity, first-cousin marria e
and “consanguinity”are used synonymous y
in this report. The consequences of measuring consanguinity on the basis of first-cousin
marriages are discussed below.
A variation of Dahlberg‘s (1948) method
was used to compare the observed frequencies of first-cousin mating with the frequencies ex cted in a random-mating population. nder random mating, the expected
frequency of first-cousin marriages in a population is
K
f
K
7
g.
:-( ; ;)
-I--+--
where N is the population size and X and V
are the mean and variance of the number of
offspring per cou le (Jacquard, 1974). Assuming that the c istribution
r
of the number
of offspring is Poisson, the mean equals the
variance, and this formula reduces to 8/N. It
is clearly unrealistic to assume that all females would be eligible for marriage to a
given male. An approximate correction was
made by substituting N/3 for N as the “effective population size” (Cavalii-Sfona and
Bodmer, 1971;Jorde, 1980).Since the parish
(munici ality) was the most important sociopolitica unit in terms of regular personal
contacts, this analysis was performed using
parishes as the unit of subdivision.
The Dahlberg method has shortcomings
(Morton, 1955) which have been overcome
somewhat by incorporating the effects of
factors such as population growth, migration, and age correlation between mates
(Ha’nal, 1963; Cavalli-Sforza et al., 1966;
Leslie, 1983). Since information on age
structure, family size distribution, and overall migration rates is not readily available
for Finland’s 19th-century population, these
modifications could not be used here.
P
What are Ihe consequences of not including the effects of migration, population
growth, and age correlation? Migration between parishes would increase the available
mate pool and thus decreases the frequency
of first-cousin marriages. Fortunately for
this analysis, close consanguineous unions,
such as first-cousin marriages, are much less
sensitive to the effects of migration than are
more remote unions (Hajnal, 1963).Age correlation tends to increase the frequency of
consan ineous unions, but Leslie (1983)
showe that the effects of age correlation are
quite small under a wide variety of conditions. In addition, the use ofN/3 instead ofN
tends to compensate for the effects of age
correlation. Population growth, like migration, decreases the expected proportion of
first-cousin marriages in a population. Since
Finland’s population was growingat a rate of
approximately 1% per annum during the
19th century (Historiallinen tilasto, 19831,
the Dahlberg approach would underestimate the fre uency of first-cousin marriages
somewhat. owever, Hajnal (1963) showed
that, when the population owth rate is less
ect on first-cousin
than 3%per annum, its efF
marria e rates is “negligible.” Thus, it appears t at, even though the assumptions of
no population growth, no migration, and no
age correlation are violated in this population, their effects are likely to be minor.
Regression methods were used to assess
the contribution of each causal variable to
consanguinity. In one set of analyses, the
dependent variable consisted of the average
consanguinity level in each deane (n = 77)
or each parish (n = 457). The in ependent
variables were population size of the subdivision, ethnic categorization of the parish
(an average value was used at the deanery
level), population density, average geogra hic distance between mates’ residences,
an urban vs. rural status for the parish or
deanery. Multi le linear regression was
used to assess t e relationship between the
independent and de ndent variables. Stepwise and “all ossi le subsets” approaches
were used to c efine
r
the optimum subset of
independent variables. Nonparametric correlation techniques were also used to assess
the relationships between consanguinity
and some of the independent variables.
A stepwise multiple logistic regression approach was used to analyze the case-control
data. The dependent variable here was consan ineous vs. control marria e. The indepen ent variables consisted o geographic
(r
;t
a
7
cp
K
g.
r
7
131
INBREEDING IN FINLAND
distance between mates' residences, endogamous vs. exogamous marriage, and male's
occupation class. The solution to the logistic
regression model was obtained using a maximum likelihood technique.
RESULTS
Of a total of 2,342 first-cousin dispensation requests recorded from 1810 to 1872, all
but 11 were approved by the authorities.
There were 739,387 marriages in Finland
during this period. The percentage of firstcousin marriages is thus 0.315% and the
average kinship coefficient based on firstcousin marriages is 0.00020. During the
1878-1920 time period, 1,325 first-cousin
marriages were re orted. A total of 761,976
marriages took pace during this period,
yielding a substantially lower proportion of
first-cousin marriages (0.174%)and a kinshi coefficient of 0.00011.
Rgure 1 shows the average kinshi coefficient for each year of the study p e r i d There
is a gradual upward trend in consanguinity
during the 1810-72 time period. Consan-
P
many of the lar er urban parishes, the ministers often di not know whether or not
marriages had in fact taken place between
first cousins.
There is substantial spatial variation in
first-cousin marriage rates. Table 1, which
summarizes the average kinship coefficients
in each of Finland's 10 provinces, shows that
the highest consan inity levels are found in
Lapland, while the owest levels are found in
the provinces of Kuopio and Central Finland
(Keslti-Suorni).
The table shows that the patterns of provincial variation in consanguinit are roughly similar in the two time peris. To assess this quantitatively, the
correlation between kinship coefficients in
the two time nods was estimated using
deaneries as t e level of subdivision. Using
the 43 rural deaneries and 34 cities (hereafter also termed 'deaneries" for convenience), a product-moment correlation of
0.45 ( P < 0.0001) was obtained. However,
examination of the data showed that this
correlation was greatly influenced by one
outlier (a Lappish deanery). Removal of this
outlier reduced the correlation to 0.15
( P > 0.19). Since underreportin of firstcousin marriages was particular y common
in the urban deaneries, another analysis was
done in which on1 the 43 rural deaneries
were used. Here t e correlation coefficient
was 0.76 ( P <
and there were no serious outliers. Thus, for the rural portion of the
population, there is considerable constancy
of consanguinity rates in the two time
f
Y
J
r
B
..
i
periods.
Ap lication of Dahlberg's formula to each
paris showed that the average expected
proportion of first-cousin marriages was
0.0125. The average observed proportion of
first-cousin marria es was 0.0036, indicating substantially ewer first-cousin marriages than ex ected under random mating.
The observed re uency of first-cousin marriages exceeded t e expected fre uency in
only 49 of 457 parishes. The corre ation between the observed and expected roportions in each parish was somewhat ow but
statistically significant (Pearson's r = 0.253,
P < 0.001;Spearman's p = 0.099,P < 0.025).
In applying Dahlberg's formula, a Poisson
distribution of the number of progeny was
assumed. It is well known that this distribution does not fit human population data well,
since the variance of progeny size usually
exceeds the mean (Brass, 1958). For most
surveyed human populations, the variance is
in fact two to three times higher than the
R
Fig. 1. Average yearly kinship coefficient in Finland,
estimated according to the frequency of first-cousin marriages. Data were unavailable for years 1873-77.
f
p%
guinit levels are enerally lower during the
secon time perio and demonstrate a downward trend through time. Examination of
the data and historical accounts indicates
that there was considerable underreporting
of first-cousin marriages during the latter
period, particularly in the cities. Since firstcousin marriage no longer required a dis nsation, ministers were much more lax a out
counting and reporting these marriages. In
d
di
r
9
P
132
L.B.JORDE AND K.J. PITKANEN
mean (S uhler, 1976).Finland's growth rate
during t e 19th century corresponds rou hly
to an average number of offsprin of 3 Ke.,
offspring surviving to adulthood). f the variance is assumed to be twice the mean, then
the expected number of first-cousin marriages is 33%higher than under the Poisson
model. Using a variance of 6, a mean of 3, and
an effective population size of N/3, only 32 of
457 parishes exceeded the expected frequency of first-cousin marriages. Using a
variance of 9, this number falls to 20. Thus,
the use of a more realistic variancelmean
ratio further reinforces the conclusion that
there is avoidance of first-cousin marriage in
nearly all of Finland's parishes.
Multiple linear regression was used to assess the effects of predictor variables on the
observed s atial variation in consanguinity
rates in t e 1810-72 time period. Using
deaneries as units of analysis, five independent variables were evaluated: deanery
population size, ethnicity (averaged across
parishes), population density, avera e geogra hic distance between mates' resi ences,
an urban vs. rural status of the deane .
The "best possible subsets" technique inxcated that ethnicity and population size were
the best predictive subset of independent
variables, with a multiple correlation value
R
f
R
%
a
TABLE 1.
of 0.458 ( P < 0.001). A stepwise multiple
regression produced the same result, entering ethnicit first, followed by population
size. When t e 43 rural deaneries were analyzed separately, three variables entered the
regression equation: ethnicit population
size, and population density. owever, the
effect of population density was due primarily to two La pish deaneries, which have
very high F v a ues and low po ulation densities. The non arametric corre ation between
density and (Spearman's 1is very low and
nonsignificant: p = -0.01. he nonparametric correlation between population size and
F,on the other hand, is significant and in the
expected direction: p = -0.33 ( P < 0.025).
Using parishes as the unit of analysis, both
re ession techniques obtained the same resu ts as the deanery-level analysis: ethnicity
and DoDulation size were sienificant medictors bf'consanguinity (multyple R = 6.3201,
P < 0.0003).
. - .- - ,.
Ethnicity had the highest univariate correlation with consanguinity level, both at the
deanery and parish levels (r = -0.413 and
-0.287, respectively). Table 2 shows that
consanguinity levels are substantially
higher for marriages occurring in parishes
with Swedish-speaking majorities. Among
the 78 parishes with Swedish-speaking ma-
K
ii
P
P
P
f
P
Average kinship cuefficients in each of Finland's 10 provinces'
1878-1920
1810-1872
Province
Uusimaa
Turku and Pori
I&me
Viipuri
Mikkeli
Kuopio
Central Finland
Vaaua
Oulu
lapland
No. of
mamaees
Kinship
coefticien t
No. of
marriages
Kinship
coefficient
77,599
131,346
67.225
121.693
69,297
87,459
21,aw
94.948
56,736
11,225
3.82
1.61
1.39
1.61
2.04
1.39
1.20
1.92
2.04
6.74
95,432
124.569
71.002
129.949
57,214
89,821
28,650
88.499
0.65
0.97
0.73
0.98
0.60
0.63
1.01
1.95
1.56
4.70
59,684
17,156
'Kinship cuefficienh are multiplied by LO'
1'ABLE 2. Distribution of consanminitv bv oarish ethnic classification (1810-72 time oeriod onlvl'
Ethnic classification
Swedish mnjority > 95%
Swcdish majority < 95%
Finnish majority < 95%
Finnish maioritv > 95%
'Kinship cwfticients are multiplird by 1 0 '
No. of mnmageu---53,109
48.167
45,349
675.6fi
Kinship coefficient
3.41
4.42
2.51
1.61
133
INBREEDING IN FINLAND
jorit,ies, 18 (23%)had a higher proportion of
first-cousin marriages than predicted by the
Dahlbcrg formula. The corresponding percentage was much lower (8%)among the 379
parishes with Finnish-speaking majorities.
There is also a si ificant difference in
consanguinity levels etween the urban and
rural deaneries, with the urban average being significantly higher than the rural average !0.00033 and 0.00022, respectively;
P < 0.003 b nonparametric Mann-Whitney
U-test). In t e multiple re ession analysis,
however, the effect of ur an versus rural
residence vanishes after controlling for ethnicity. This indicates that the apparent urban-rural difference is primarily the result of
a conibunding effect with ethnicity.
The case-control approach permits a somewhat finer-grained investigation of the determinants of first-cousin marriage in this
population. However, because the matching
criteria included bride’s parish of residence,
some of the variables used in the multiple
regression analysis could not be le ‘timately
used in the case-control analysis. ree variables were evaluated: husband’s social class,
arish endogamy, and geographic distance
Eelw een mates’ residences.
Table 3 shows that there are substantial
differences in the social class distribution
I?
x
r
lf
between first-cousin and control marriages.
The frequencies for gentry, high-status
farmers, and other farmers are considerably
higher in the first-cousin marriage category
than among the control marriages. The reverse is true for crofters, skilled workers,
and servants and unskilled workers. The
difference between these two distributions is
highly significant ( x 2 = 753.7, P < 0.0001).
Thus, first-cousin marriage is much more
likely to occur among the wealthy and landowning part of the populace. For comparison,
the social-class distribution of “other” types
of marital dispensations (e.g., the marriage
of a man with his deceased brother’s wife) is
also shown. The striking excess of u per
social classes that was observed for i r s t cousin dispensations is not seen among marital dis nsations of other types.
The Ktribution of endogamous versus exogamous marriages in cousin and control
marriages is shown in Table 4. Endogamous
marriage was significantly more common
among the control marriages than among
the cousin marriages (91% versus 81%,respectively; x2 = 131.0, P < 0.0001). This result is consistent with the fact that average
geographic distance between the residences
of cousin mates is significantly greater than
that of control mates (15.8km versus 4.5 km;
TABLE 3. Social class distrihution in cousin and control marriage samples
Cousin marriages
No.
Percentage
Suciiil
- .__
c1a.s
Ndility and gentry
Hiutt.status farmers
Other farmers
*743
86
Crofters
Skilled workers
%wants and unskilled workers
Other and unknown
n
-hi
--_
---
294
5
208
105
2.136
907
445
2.060
49
3.5
1.8
36.1
15.4
7.5
34.9
0.8
168
67
1,401
454
129
340
3
6.6
2.6
54.7
17.7
5.0
13.3
0.1
25x52
5,910
1,969
Other dispensations
No.
Percentage
TABLE 4. Endogamy in cousin and control marriages
- -________
Observed
Cousinfi
Controls
--_
9M
214
73
17.4
4.4
48.4
10.9
3.7
14.9
0.3
Control mamagesNo.
Percentage
371
552
Exogamous
Endogamous
EXpecred
Observed
Expected
230
1,575
5,318
1,716
5,177
693
134
L.B.JORDE AND K.J. PITKANEN
P < 0.0001 by both Student's t- and MannWhitney U-tests). Figure 2, which depicts
the average geographic distances in cousin
and control marria es, shows that the zero
class is substantial y larger among controls
than among cousin marriages. Cousin marriages are seen more frequently in the larger
distance classes. In particular, very few control marriages take place among mates removed more than 150 km.,while several
percent of the cousin marria es do.
The fact that geographic Astance and exogamy are both greater for cousin marriages
than controls may seem counter-intuitive at
first. In the context of isolation-by-distance
models, consanguinity (in terms of random
f
inbreeding) is predicted to be inversely related to exogamy and geographic distance
between mates. Tablc 5, which shows the
case-control difference in geo
tances after stratifying by socia
to ex lain the pattern seen in
tion. he geographic distances among cousins are significantly greater than those
among controls for the gentry, high-status
farmers, and skilled workers. The differences are not significantly different for the
other three major social classes. A highly
similar pattern is seen for the distribution of
endogamy by social class (Table 6 ) . These
results indicate that the geographic distance
and endogamy effects are confined to the
\
-_
Percentage
I Cousin marriage8
C 3
0
50
100
150
200
Control marriages
250
300
,300
Distance in kilometers
Fig. 2. Distribution of geographic distances between
mates for cotisin and control marriages. The "0" class
includesonlydistancesofO km. The"5O"classrepresents
distances 1 0 and s 50: the "100" class represents distances > 50 and 100; and so on.
TABI, E 5. Aueragc geographic distances between mates' residences
in cousin and control marriages, stratified by social class'
Ceographis distance (km>-..~
.
Social class
Nobility and gentry
High-status farmers
Other farmers
Crofters
Skilled workers
Servants and unskilled workers
Total
X
_
Cousin
~
75.1
7.4
3.8
2.9
14.9
-
~
2.1
15.8
'Significance levels were determined hy the nonpemmetrir Mann.Whilney ll+xt.
c 2- x
Significance level
39.0
3.3
O.OOO1
0.008
~
9.6
3.5
4.1
0.76
0.82
2.4
0.22
4.5
---
0.W3
o.ooo1
135
INBREEDING IN FINLAND
cousin marriage in Norwa from 1870 to
1891 was 1.7% (Saugstad: 1977a). This
higher figure may reflect the fact that the
Percent
dispensation re uirement in Norwa
endogamous
matings
was
abolished in 180 , which was earlier t an
l'n
Sweden
(1844)
and
Finland
(1872).
Cousin
Control
Social class
A factor that would have contributed
Nobility and gentry
41.2
70.7
somewhat to Finland's lower first-cousin
Highdatus farmers
70.9
86.5
marriage rate is a cultural prohibition
90.4
91.0
Other fnrmers
90.7
91.1
Crofters
-..
..-.
against marriages between the offsprin of
71.2
84.8
Skilled workers
brothers in parts of eastern Finland. t h e
Servants and unskilled workers
91.8
93.6
effect of this prohibition can be gauged by
examinin the frequency of isonymous marriages in inlands eastern provinces (this is
the only part of Finland in which family
wealthier and more mobile elements of soci- surnames were used consistently during the
ety, where there is a certain degree of non- 19th century). Oral tradition indicates that
random inbreedin . Many members of these the ban was observed most closely in the
classes contracte marriages amon geo- province of Kuopio (Lukkarinen, 1933). Of
graphically removed first cousins, whi e geo- the 187 first-cousin marriages that took
graphic distance had little effect on firstlace bet.ween couples in Kuopio who had
cousin marriage among the poorer classes.
family names, only 29 (16%) were isonyThe combined effects of social class, endog- mous. This is substantial1 lower than the
amy, and geographic distance upon consan- 25% that would be expecteJin the absence of
guineous marriage were assessed using a prohibition. In the province of Viipuri the
stepwise multiple logistic regression. The ban was not observed, and 83 of 297 firstfirst variable to enter the logistic regression cousin marriages (28%)were ison
equation was social class. Endogamy en- Mikkeli, another province in whic the ban
In
tered the equation next. Because endogamy was not observed, 48 of 211 marriages (23%)
and eographic distance are quite highly were isonymous. While isonymy would have
corre ated, the latter variable did not enter provided an accurate estimate of consanthe e uation. Both social class and endog- guinity at the first-cousin level in the latter
amy gad highly significant X2 values for two provinces, it would have produced a
entry into the e uation. The overall x2 value serious underestimate in the province of
for goodness of Zt of the logistic model had a Kuopio. These findings underscore the need
significance level of 0.054, indicating a mar- to exercise caution when estimating consanguinity from isonymy data.
ginal fit of the logistic model to these data.
What proportion of a
TABLE 6. Parish endogamy rates in cousin and control
marriages. stratified by social class
a
K
d
I
P
r.
P
DISCUSSION
In a general survey of first-cousin marriage in Euro an populations, Nee1 et al.
(1949) conch ed that approximately 1%of
European marriages were contracted between first cousins during the 19th and early
20th centuries. Finland's first-cousin marriage rate of 0.32 is lower than this figure,
but it is well within the range of the PO ulations reviewed by Nee1 et al., as we 1 as
others reviewed more recently by Lebel
(1983). Alstrom (1958), in a survey of firstcousin marriages in Sweden (1750-1844),
found that the first-cousin marriage rate
during the 19th century varied between 1%
and 1.5%. This is consistent with a series of
studies of individual Swedish parishes, in
which most first-cousin marria e rates
ranged from 1 to 2% (Book, 1948; ook and
Mawe, 1955; Larson, 1956).The rate of first-
F
f
i
marriages?
population size, age structure, mating preferences, the number of generations since the
founding of the population, and other factors. Nee1 et al. (1949) concluded that first72% of the
there was a 1:1ratio of first-cousin to secondcousin marriages in Norway, indicating
again that a large proportion of inbreeding
might be included in the first-cousin marriage figure. Since dispensations were
anted only for first-cousin marriages in
gnland, it is not possible to assess the
amount of inbreeding due to consanguinity
at more remote levels. However, more extensive enealogic evaluations of inbreedin
have een done in a few individual Finnis
t
a
136
L.B.JORDE AND K.J. PITKANEN
parishes. Nevanlinna (1972)com iled an extensive genealogy for the paris of Sawijarvi and obtained an average inbreeding
coefficient of 0.00867 (based on individuals
whose grandfathers were born before 1850).
The dispensation data show that 17 of 3,596
marriages in this parish were contracted
between first cousins, givin an average kinship coefficient of O.oO03, w ich is nearly 30
times less than Nevanlinna’s value. OBrien
et a]. (1988a) obtained an average inbreeding coeff icient of 0.002 for individuals born in
the parish of Sottun a from 1825 to 1900
(corresponding roughy to the offspring of
marria es occurring between 1810 and
1872). $he average kinship coefficient for
this parish usin first-cousin marriages is
0.0005, a fourfol8 difference. In contrast to
ven by Nee1 et al. and Geddethe figures
these igures indicate that, for these 2
Dahl,
parishes at least, only a fraction of the total
inbreeding coefficient is detected by evaluating first-cousin marriages. Thus, the kinship
coefficients re orted here represent only a
lower limit to t e amount of consanguinity in
Finland.
A number of potential causal variables
were evaluated in terms of their effects on
spatial variation in first-cousin marria e
rates. Some of the findings were initial y
unexpected. For example, opulation density has often been found to ave a ne ative
association with consanguinity (see kbel
1983 and Roberts, 1975 for reviews).
strong negative correlation (r = -0.86) between population density and first- and second-cousin marriages was observed, for example, in Norway (Saugstad, 1977b). In the
present study, there was very little association between population density and firstcousin marriage. Alstrom (1958),who examined patterns of first-cousin marriage in
Sweden, also failed to find an association
between these two variables. On the other
did correlate negain Finland, as prereflect the fact
that most members of a parish did come into
contact with one another through churchrelated activities, a fact which has been
shown to be important in e idemic transmission (Jorde et al., 1989). .fphus, these activities facilitated contact of a certain number of
peo le, regardless of opulation density.
&ile population ensity was not an important determinant of consanguinity in
most of Finland’s opulation, Lapland is a
notable exception. #he Lappish parishes had
R
%
5
P
R
P
K
A:
0
the highest rates of consanguinity (1-3%
first-cousin marriages) and the lowest average population density in Finland (0.28 persons per km2). It is also noteworth that, in
recognition of the difficulty of fin& eligible mates, couples living in Laplan were
exey t from payin the usual dis ensation
fee. n! both the wedish and !r orwegian
studies, the isolated northern portions of
these countries also had the highest rates of
consanguinity (Alstrom, 1958; Gedde-Dahl,
1973).
The fact that higher consanguinity levels
are seen in the urban portion of Finland’s
population than in the rural portion is also
inconsistent with the findings of many other
studies (e.g., Freire-Maia et a]., 1983;GeddeDahl, 1973; Gomez, 1989; Rao et al., 1972).
This apparent anomaly is largely ex lained
by the fact that there is an excess of Jwedish
speakers, who had elevated consanguinity
rates, in the urban areas. Also, the gentry
class, which had the highest first-cousin
marriage rate of any social class, was distributed preferentially in the urban areas (both
of these associations yielded highly significant x2 values).
Geographic distance between mates’ residences and endogamy were two other variables that failed to conform to expectations.
Geographic distances were larger, and endogamy was less frequent, in cousin marriages than in the control marriages, contrary to the findings of many other studies
(Roberts, 1975; Tchen et a]., 1977) and the
predictions of isolation-b -distance models.
Again, an examination o social class differences he1 s to clarif this issue. The geographic istance di erence was
among the entry and high-status armers,
who were t e wealthiest and most mobile
elements of society. A similar social class
effect is seen in Sweden, where over 90% of
the marriages amon the nobility were exo amous at the palish eve1 (Alstrom, 1958). n
addition, Swedlund and Boyce (3.983)found
higher i s o n y y and exogamy rates among
the social elite” of Deerfield, Massachusetts. The members of Finland’s upper
classes may have had an interest in arranging first-cousin marriages, for reasons outlined below.
By far the best redictors of first-cousin
marriage rates in inland are two cultural
factors that generally receive less attention
in studies of consanguinity: social class and
%
8
0
x
a
f
F
r
rtest
P
137
INBREEDING IN FINLAND
ethnicity. Some previous studies have assessed the relationship between socioeconomic status and consanguinity and have
generally found a negative relationship
(Imaizumi, 1986; Khlat, 1988; Rao et al.,
1972; Schull and Neel,1972).In Finland the
relationship is positive, with the gentry and
wealthy farmin classes having strikingly
higher rates ofg first-cousin marriages. A
study of first-cousin marria e among the
Swedish nobility (Alstrom, 1 58; Fraccaro,
1958) revealed the same pattern. In fact,
3 4 % of the Swedish nobility married first
cousins during the 19th century. This proportion was even higher among Finland‘s
small noble class: approximately 10%.
Among the Finnish gentry, and especially
amon the nobility, first-cousin marriage
was o en socially and economically advantageous.
It has often been suggested that firstcousin marriage is a way in which land and
other economic interests can be retained in
the same family (Calderon, 1989; Guz et al.,
1989; Pettener, 1985). This is thought to
have contributed to the increase in firstcousin marriage seen in many European
opulations during the 19th century, when
k n d began to be divided among offspring
(Cavalli-Sforza and Bodmer, 1971; McCullough and O’Rourke, 1986). The fact that
first-cousin marria e in Finland is hi hest
among the land-ho ding classes woul apear to be consistent with this h othesis.
f;owever, Finnish farms were usua 1 inherited in total by only one child in t i e 19th
century, so first-cousin marriage would not
tend to reunite divided parcels of land. The
eographic distance effect suggests that motilit and assortative mating among the
wea Iythier classes were more important factors: these families would be more likely to
become se arated as they migrated, but the
would stil enjoy social contact. Such mobil
ity was not characteristic of Finland’s lower
classes.
In Sweden, the financial cost of obtaining
a dispensation, which included an administrative fee, could be quite high in some cases.
Alstrom (1958)indicates that it was sometimes equivalent to several months’ cash
wages for a laborer, although the cost appears to have been typically much lower than
this. After this fee was abolished in Sweden
(18281, the fre uency of first-cousin marriages increase from 1% to about 1.5%.
Alstrom (1958) concluded that this financial
burden often prevented couples of modest
8
a
P
YP
P
8
d
a
means from obtainin dispensations and
thus helped to explain t e high percentage of
dispensations among the wealthier classes.
The administrative fee in Finland was equivalent to that of Sweden, and it was raised in
1818. Cash was in chronically short supply
in Finland at this time, laborers often receiving most of their income as goods rather than
cash. Obtaining sufficient cash for a dispensation mi ht have been difficult for common
people. dso, the procedure for obtaining a
dispensation, which involved writing a petition, could have been dauntin for commoners, many of whom were i literate. The
Lutheran ministers were probably of little
help to these couples, since they t i d l y discouraged first-cousin marria e! f the combination of the dispensation ee and the procedure itself were the primary cause of social
class differences in first-cousin marriages,
one would expect to see highly similar socialclass distributions of first cousin and “other”
types of dispensations (the fee and rocedure
being the same for the latter). Ta le 3 indicates that this was not the case. Furthermore, the distribution of other types of dispensations showed almost no variation
across ethnic groups, and the provincial distribution of these dispensations showed
much less variation than did the first-cousin
marriage distribution. Thus, while the dispensation fee and procedure may have contributed slightly to social class differences in
first-cousin marriages, it a pears that the
other previously discussed actors played a
more im rtant role.
The et nic differences in first-cousin marriages can be understood at least in part by
considering the economic sition of the
Swedish-speakin minority. uring the 19th
century, much of inland’s land and many of
its economic interests were controlled by
Swedish speakers, which consisted of only
15%.of the population (it is important to
int out, however, that even among the
Kedish speakers, a large portion were
farmers of very modest means). It would
have been quite natural for this economically
powerful minority to preserve its domination
through a variety of means, including firstcousin marriage.
A primary conclusion emer ‘ng from this
study is that, with the possib e exce tion of
Lapland, first-cousin marriage in finland
was not usually the result of a lack of avail-
P
f y
g
F
R“
f
!i?
f
’According to recordsoldiscussionsin the Finnish Dlet rVullru.
prcaasiakirjcilL 11172.
138
L.R. JORDE AND K.J. PITKANEN
able mates. Rather, it was often a desired
marriage pattern among certain advantaged
portions of the po ulation. A similar conclusion was reache by Calderon (19891, who
found that first-cousin marriages in Spain
were often desirable for economic reasons,
while second-cousin marriages were more
the result of a lack of othcr available mates.
These findings suggest that very different
causal factors may be operating at different
levels of consan uinity. As a result, investigators of causal eterminants of consanguinity should stratify their samples according to
de eeofrelationshi .
K i s study also emonstrated that, at
least at the level of first-cousin marriage,
cultural factors such as ethnicit and social
class are much more important actors than
many of the more commonly studied variables such as
ulation size, geogra hic
distance, popu ation density, and ur a d
rural residence. In order to understand complex biosocial phenomena such as inbreeding, these types of variables should be
measured and evaluated whenever possible.
B
f
B
r
rp
1
ACKNOWLEDGMENTS
gg:!
This research was su ported by
from NSF (BNS-8703!411,
the
Juselius Foundation, and the Academy of
Finland. We are ateful for aid and comdur Eriksson, Johan Fellments from Drs. A Y
man James Mielke, Kenneth Morgan, Elizabeth OBrien, Derek Roberts, and Alan
Rogers. We also wish to acknowled e the
invaluable technical assistance of anna
Erkkila, Kaisa Kauranen, and Tarja Raisanen-Toumi. Our ap reciation is extended
to Mr. Rauno Selin rom the Finnish National Archives and Mr. Kari De erstedt
from the Central Statistical Office or their
aid regarding the archival materials, and to
Dr. Aarno Strommer and Mr. Pentti Voipio,
who helped in obtaining information from
the Finnish genealogists.
fl
P
f
LITERATURE CITED
Alstrom CH (1958) First-cousin marriages in Sweden
1750-1844 and a study ofthe population movement in
some Swedish subpopulations from the genetic-statistical viewpoint. Acta Genet. Med. Gemellol. (Roma)
8:295-369.
Biiok J A (1948) The frequency of cousin marriages in
three North Swedish parishes. Hereditas 34:252-255.
Book JA, and MAwe CE (1955) The incidence of cousin
marria es in a West-Swedish rural community. Am. J.
Hum. t e n e t . 7:4263-429.
Brass W (1958) Models of birth distributions in human
populations. Bull. Inst. Int. Statist. 36:16.5-178.
Calderon R (1989)Consanguinity in the archbishopric of
Toledo, Spain, 1900-79. I. Types of consanguineous
mating in relation to premarital mi ation and i t s
23:253-266,
Carpelan T(1954)Attartavlorfor de B Finlands riddarhus inskrivna atterna. 1 (A-G).H e h n k i .
CarpelanT(1958)Attarvlorfor de pi4 Finlands riddarhus inskrivna atterna. I1 (H-RI.Helsinki.
Carpelan T(1965jAttart;ivlor for de pa Finlands riddarbus inskrivna atterna. I11 (S-0).Helsinki.
Cavalli-Sforza LL, and Bodmer WF (1971 )The Genetics
of Human Populations. San Francisco: Freeman.
Cavalli-Sforza LL. Kimura M. and Rarrai 111966)The
probability of consanguineous marriages. Ckmetics
54:37-60
Dahlberg G (1948) Mathematical Methods for Population Genetics. Basel: s. Karger.
Fraccaro M (1958) The incidence of consanguineous
marriages in Sweden. Hereditas 44:65-74.
Freire-Maia N,Chautard-Freire-Maia EA, De AguiarWolter IP. Azevedo-Fialho M. Rarros de Azevedo M,
Krieger H, and Barbosa. CAA (1933) Inbrerding studies in Brasilian schoolchildren. Am. J. Med. Genet.
16:331-355.
Cedde-Dahl T (1973) Population structure in Norway.
Hereditas 73211-231.
Gomez PGrl989jConsanguinity: gengraphical variation
and temporal evolution in the north of the Iberian
peninsula. 191&-1968(Lebn,Spain).Int. J.Anthropol.
4:119-124.
GUz K. Dedeoglu N. and Luleci C: 1989' Thi, frequi!ncv
and medical effects of consanguineous marriages in
Antalya. Turkey. Hereditas I 1 I :79.-83.
Hajnal J 19631 Concepts of random mating and the
frequencv of consanguineous marriages. Proc. H.Soc.
Lond. IBio1.l I59:125-174.
Historiallinen tilasto (Historical Statistics) (1983) In K
Vattula (ed.): Suomen Taloushistoria. Helsinki.
Tanini.
lmaizumi Y 11986)A recent survey of consanguineous
marriages in Japan: religion and socioeconomic class
effects. Ann. Hum. Biol. 13:317-330.
Jacquard A (1974)TheGenetic Structure ofPopulations.
New York: Springer-Verlag.
Jorde LB (1980) The genetic structure of subdivided
human populations: a review. In JH Mielke and MH
Crawford (eds.):Current Develo ments in A n t h r o p
ligcal Genetics, Volume 1. New {ork: Plenum Press,
pp. 135-208.
Jorde LB (1991) Inbreeding in human populations. In R
Dulbecco (ed.): Encyclopedia of Hunian Biology. New
York: Academic Press (in press).
Jorde LB. Workman PL, and Eriksson AW 11982) Genetic microevolution in the &and Islands, Finland. In
MH Crawford and J H Mielke (eds.):Current Develo
rnents in Anthropological Genetics, Vol. 2, New YorL
Plenum Press, pp. 333-365.
Jorde LR, Pitkanen KJ, and Mielke JH (198Y)Predicting
smallpox epidemics: a statistical analysis of two Finnish populations. Am. J. Hum. Biol. 1:621-629.
Jorde LB, Yitkanen KJ, Mielke JH, Fellman JO, and
Eriksson AW (1990)Historical epidemiol of smallpox in Kitee, Finland. In AC Swedlund an% Armela0s (eds.): The Health and Disease of Populations in
!'ransition. New York: S. Prager (in press).
Khlat M (1988)Social correlates of consan ineous marHum. Biol.
riage in Beirut: a population based
60t541-548.
Larson CA (1956, The fre uency of first cousin marriages in a south Swedi& rural community. Am. J.
Hum. Genet. 8:151-153.
Label HR (1983)Consanguinity studies in Wisconsin I:
e f f i q on inbreeding levels. J. Biosoc.
&.
SLUE
INBREEDING IN FINLAND
Secular trends in c o n s a n ~ i n e o u smarriage, 18431981. Am. J . Med. Genet. 15:543-560.
Ifislie PW (1983) Age correlation between mates and
average consanguinity in age-structured human p o p
ulations. Am. J. Hum. Genet. .’3.5:962-977.
hkkarinen J (1933) Suomalaisia naimatapoja. Aineksia suomalaisten kansojen avioliiton historiaan. Helsinki: Suomalaisen Kirjallisuuden Seura.
MrChllough J M , and ORourke DH (1986) Geogra hic
distribution of Consanguinity in Europe. Ann. &m.
!3iol. 13359-367.
M d k e JH, Workman PL, Fellman J and Eriksson AW
(!976) Po ulation structure ofthe k a n d Islands, Finland. In !
I Harris and K Hirschhorn (eds.):Advances
in Human Genetics, Vol. 6. New York: Plenum Press,
pp. 241-321.
Mielke.JH, Jorde LB,Tra PG, Anderton DI,, Pitkanen
K. and Eriks n AW (1!%44) Historical epidemiology of
smallpox in E a n d , Finland: 1751-1890. Demography
21:27 1-295.
Mielke JH, Pitkanen K. Jorde LB, Fellman JO, and
I<riksson AW (1987) Demogra hic atterns in the
Aland Islands, Finland, 1750-1&0. A b k . Popul. Res.
Finland 2557-74.
Morton NE (1955t Non-randomness in consanguineous
marriage. Ann. Hum. Genet. 20:116-124.
Nee1 JV,Kodani M,Brewer R, and Anderson RC (1949)
The incidence of consan ineous matings in Japan.
Am. J . Hum. Genet. 1 : 1 g 1 7 8 .
Ntvanlinna HR (1972) The Finnish population structure. Hereditas 71:195-235.
Norio R (1981) Diseases of Finland and Scandinavia. In
H Rothschild (ed.): Biocultural Aspects of Disease.
New York: Academic Press, pp. 359-415.
O’Brien E.Jorde LB, Ronnlof R, Fellman JO. and Eriksson AW (1988aj Inbreedin and genetic disease in
h t t u n g a . Finland. Am. J. bhys. Anthropol. 7,5:477186.
OBrien E.Jorde LB, Ronnlof R, Fellman JO. Eriksson
AW (1988b) Founder effect and genetic disease in
Sottunga, Finland. Am. J. Phys. Anthropol. 77:335346.
Pc!tener D (1985) Consanguineous marriages in the
Upper Bologna Appennine (1565-1980): Microgeo-
139
graphic variations. pedigree structure and correlation
ofinbreedin secular trend with changes in population
size. Hum. Ifiol. 57267-288.
Pitkanen K. Jorde LB, Mielke JH. Fellman JO. and
Eriksson AW (1988) Marital mi ation and genetic
structure in Kitee, Finland. Ann. E m . Riol. 15:23--34.
and Jorde LB (1989)Smallpox
Pitkanen K, Mielke <JH,
and its eradication in Finland: implications for disease
control. Popul. Studies 43:95-111.
Rao PSSS,Inbaraj SG. and Jesudian G (197’Lt Ruralurban dimerentials in consanguinity. J. Med. &net.
9:174-178.
Reid RM (1973) Inbreeding in human po ulations. In
MH Crawford and PL Workman (eds.):bethods and
Theories of Anthro logical Genetics. Albuquerque:
University of New g x i c o Press, pp. 83-1 16.
Roberts DF (19751 Genetic studies of isolates. In AEH
Emery (ed.): Modem Trends in Human Genetics,
Vol. 2. London: Buttenvorths, pp. 221-269.
Rogers AR, and Jorde LR (1987)Theeffect of noiirandom
migration on genetic differences between populations.
Ann. Hum. Genet. 51:169-176.
Saugstad LF f 1977a)Inbreeding in Norway. Ann. Hum.
Genet. 40:481491.
Saugstad LF ( 1977b)The relationship between inbreeding, mi ation and population density in Norway. Ann.
Hum. &net. 40331-341.
Schull WJ, and Nee1 JV (1972) The effects of parental
consanguinity and inbreeding in Hirado, Japan. V.
Summary and interpretation. Am. J. Hum. Genet.
24:425-453.
Spuhler J N (1976)The maximum opportunity for natural selection in some human populations. In EBW
Zubrow (ed.): Demo aphic Anthropology: Quantitative Ap roaches. f i u q u e r q u e : University of New
Mexico L e s s . pp. 185-226.
Swedlund AC, and Boyce A J (1983) Mating structure in
historical populations: estimation by analysis of surnames. Hum. Biol. ,55252-262.
Tchen P, h i s E, Fein old J, et al. (1977) Inbreeding in
recessive diseases. j u m . Genet. 38:163-167.
Workman PL. Mielke JH. and Nevanlinna HR (1976)
The geneticstructiire of Finland. Am. .J. Phys. Anthropol. 44:341-368.
Документ
Категория
Без категории
Просмотров
2
Размер файла
1 177 Кб
Теги
finland, inbreeding
1/--страниц
Пожаловаться на содержимое документа