The Evolution of Recombination under Domestication: A Test of Two Hypotheses

Department of Genetics, Life Sciences Building, University of Georgia, Athens, Georgia, 30602, USA.
The American Naturalist (Impact Factor: 3.83). 02/2004; 163(1):105-12. DOI: 10.1086/380606
Source: PubMed
The successful domestication of wild plants has been one of the most important human accomplishments of the last 10,000 yr. Though our empirical knowledge of the genetic mechanisms of plant domestication is still relatively limited, there exists a large body of theory that offers a host of hypotheses on the genetics of domestication. Two of these that have not been addressed concern the role of recombination in the process of domestication. The first predicts an increase in recombination rate through domestication, while the second argues that recombination rate should serve as a preadaptation to domestication. This study makes use of data on chiasma frequencies available from almost a century of plant cytogenetical literature to test these two hypotheses. The results support the hypothesis that domestication selects for an increase in recombination, and in rejecting the preadaptation hypothesis, they suggest directions for future research into the possibility of preadaptation to domestication.


Available from: Jeffrey Ross-Ibarra, Feb 19, 2014
vol. 163, no. 1 the american naturalist january 2004
The Evolution of Recombination under Domestication:
A Test of Two Hypotheses
Jeffrey Ross-Ibarra
Department of Genetics, Life Sciences Building, University of
Georgia, Athens, Georgia 30602
Submitted September 18, 2002; Accepted May 13, 2003;
Electronically published January 28, 2004
Online enhancements: appendix tables.
abstract: The successful domestication of wild plants has been
one of the most important human accomplishments of the last 10,000
yr. Though our empirical knowledge of the genetic mechanisms of
plant domestication is still relatively limited, there exists a large body
of theory that offers a host of hypotheses on the genetics of do-
mestication. Two of these that have not been addressed concern the
role of recombination in the process of domestication. The first
predicts an increase in recombination rate through domestication,
while the second argues that recombination rate should serve as a
preadaptation to domestication. This study makes use of data on
chiasma frequencies available from almost a century of plant cyto-
genetical literature to test these two hypotheses. The results support
the hypothesis that domestication selects for an increase in recom-
bination, and in rejecting the preadaptation hypothesis, they suggest
directions for future research into the possibility of preadaptation to
Keywords: domestication, recombination, preadaptation, chiasma
The successful domestication of wild plants has been one
of the most important human accomplishments of the last
10,000 yr. Changes wrought by domestication enabled hu-
man populations to harness and control a food supply
tremendously greater than was previously possible. The
magnitude of these changes and the rapidity with which
they were effected are convincing evidence of the strong
directional selection to which these plants were subjected.
Though clearly each species is a product of its own unique
history, several patterns of morphological and genetic
* E-mail:
Am. Nat. 2004. Vol. 163, pp. 105–112. 2004 by The University of Chicago.
0003-0147/2004/16301-20342$15.00. All rights reserved.
change can nonetheless be discerned among many do-
mesticated plant species (Harlan 1992; Paterson 2002).
Morphological and physiological changes including gigan-
tism, loss of natural dispersal and defense mechanisms,
and loss of seed dormancy have been known for decades
and are well documented for a wide variety of crop plants
(Smartt and Simmonds 1995). Genetic evidence of strong
selective pressure (e.g., Wang et al. 1999), loss of genetic
diversity (Doebley 1989), and polyploidy (Hilu 1993) have
been shown for some of the more important crop plants,
but our empirical knowledge of the genetics behind the
domestication process is still rather limited.
Though the empirical literature on domestication is less
developed than could be desired, there exists a large body
of theory on the genetics of plant domestication. Both
quantitative and population genetics provide a host of hy-
potheses regarding domestication. Two of these that have
not been addressed concern the role of recombination in
the process of domestication.
The first, initially proposed by Rees and Dale (1974)
and later echoed by Burt and Bell (1987) and Otto and
Barton (2001), predicts an increase in recombination rate
through domestication. Both theory and simulations show
that selection generally favors an increased recombination
rate during periods of rapid evolutionary change (Otto
and Barton 1997). High recombination is of most value
when selection is strong and genetic variability is limited
by negative linkage disequilibrium (Feldman et al. 1997).
In domesticated plants this disequilibrium might have
been generated by population bottlenecks and genetic drift
(Felsenstein and Yokoyama 1976; Otto and Barton 1997,
2001) or by negative epistasis among beneficial alleles
(Charlesworth 1993; Barton 1995).
Gornall (1983) elaborated the second hypothesis, ar-
guing that high recombination rate should serve as a pre-
adaptation to domestication: because high recombination
rate increases response to strong selection, he argued,
plants without this advantage would be less likely to be
successfully domesticated.
The two hypotheses are not mutually exclusive but,
while based on very similar theoretical underpinnings,
Page 1
106 The American Naturalist
make distinct and readily testable predictions about pat-
terns of recombination rate in domesticated plants and
their wild relatives. Rees and Dale’s hypothesis predicts a
higher recombination rate in domesticated plants relative
to their wild progenitors, while Gornall’s hypothesis pre-
dicts a higher recombination rate in the wild progenitors
of domesticated plants relative to their other wild
This study makes use of data on chiasma frequencies
available from almost a century of plant cytogenetical lit-
erature to test these two hypotheses. The results not only
support the predictions of Rees and Dale and others but
also, in rejecting Gornall’s hypothesis, suggest directions
for future research into the possibility of preadaptation to
Recombination data were collected from published studies
citing chiasmata frequencies for species of vascular plants.
Though it has been suggested that chiasma frequency is
not always an accurate measure of recombination in plants
(Nilsson et al. 1993), recent studies have shown the direct
one-to-one relation between chiasmata and recombination
events (King et al. 2002; Knox and Ellis 2002). Data on
other characteristics were obtained from electronic data-
bases, floras, and a variety of other sources. A text file of
the recombination data used is available as table B1 in the
online edition of the American Naturalist, and a complete
list of the sources consulted is available on request to the
Five parameters were recorded for the recombination
data: haploid chromosome number (N), total number of
chiasmata per nucleus (XMA), excess chiasmata per nu-
cleus ( ), chiasmata per bivalent (XS p XMA NIIp
), and recombination index ( ).XMA/N RI p XMA N
Missing data were calculated using the above relationships
or, in the case of haploid number, taken from the Index
to Plant Chromosome Numbers (IPCN) database (2002)
or from Darlington and Wylie (1956). Data from each
study were entered into the database as a species average
weighted by the number of plants or cells of each species
examined in the study. Species data from multiple
publications were averaged to give a single value in the
final data set. If entries for the same species differed in
one of the other characteristics recorded, however, those
entries were not averaged but instead kept separate in the
data set. Data were recorded separately for male and female
meioses; only data from male meiosis were included in
this study. Whenever possible, chiasma frequencies at
metaphase were used. Likewise, data from plants with su-
pernumerary chromosomes were excluded whenever pos-
sible. Nonetheless, sources that included only data from
other phases of meiosis or from plants with supernumerary
chromosomes were included in the database.
Each species was classified for five additional charac-
teristics: ploidy level, life form (annual or perennial), mat-
ing system (selfing, mixed, or outcrossing), weediness
(weed or not), and domestication status (wild, cultivated,
or domesticated). Data on ploidy level, life form, and mat-
ing system were occasionally available from the original
source but were otherwise gleaned from a wide range of
sources. Plants were considered weeds if named as such
by three or more sources as cited in the Global Compen-
dium of Weeds (2002). The domestication status provided
by the original source was used whenever available. If not
available, various sources were consulted to determine the
appropriate status; the majority of the determinations,
however, were made using Smartt and Simmonds (1995).
Plants grown horticulturally and plants for which no clear
evidence of domestication was available (mostly forage
grasses and relatives of domesticated taxa) were considered
cultivated but not domesticated for purposes of this study.
Statistical analyses took the form of an across-species
ANOVA of the entire data set followed by pairwise com-
parison of specific groups. The raw data for haploid num-
ber and chiasma frequency were significantly nonnormally
distributed and were transformed using a Box-Cox trans-
formation for use in the ANOVA and pairwise tests.
To test the importance of domestication in determining
recombination rate, a stepwise regression analysis was per-
formed on the entire data set. Life form, mating system,
weediness, domestication, and all two-way interactions
were included as variables in the initial model, but all
three- and four-way interactions were ignored. More direct
tests of the effect of domestication were carried out by
making pairwise comparisons of domesticated species to
their wild progenitors and of the progenitors to their near-
est congeners. These were compared using both a standard
paired t-test as well as a Wilcoxon paired-samples test of
the untransformed data. To test for possible correlations
between these pairwise differences and the other species
characteristics, the sets of (transformed) pairwise differ-
ences were then analyzed by two stepwise regressions,
again ignoring all three- and four-way interaction terms.
The first regression analyzed the effect of the other species
characteristics on the size of the difference in recombi-
nation rate between paired taxa (e.g., whether selfing taxa
have smaller differences than outcrossing taxa). The sec-
ond looked for correlations between a change in these
species characteristics as a result of domestication and the
difference in recombination rate (e.g., whether a change
from selfing to outcrossing is correlated with larger dif-
ferences in recombination than no change in mating sys-
tem). Statistical calculations were carried out using Sta-
tistica 5.5 (Statsoft 2000).
Page 2
Recombination under Domestication 107
Table 1: Regression analysis of entire data set for chiasmata per
Intercept 1 28.347 28.347 322.955 .000
Domestication 2 .552 .276 3.14 .045
Mating 2 .637 .319 3.63 .028
Weediness 1 .028 .028 .318 .573
Life form 1 .056 .056 .634 .427
Weediness # life form 1 .434 .434 4.941 .027
Error 191 16.765 .088
Note: The following factors showing no data were removed from the model
because their corresponding P values were
1.1: ,domestication # weediness
form, , ,domestication # life domestication # mating weediness # mating
.life form # mating
Figure 1: Observed means and 95% confidence intervals of the three categories of domestication; transformed data shown
Data were collected for 601 species of vascular plants from
124 genera and 37 families. After elimination of incom-
plete cases, the across-species analyses of chiasma fre-
quency were limited to a sample size of 196, including 46
domesticated species. Species used for pairwise compari-
sons are listed in table A1 in the online edition of the
American Naturalist.
In order to separate the effect of chromosome number
in determining chiasma frequency, it was desirable to use
a measure of chiasma frequency independent of haploid
number and ploidy level. In spite of claims that excess
chiasmata is independent of chromosome number (Burt
and Bell 1987; Koella 1993), all measures of chiasma fre-
quency in this data set are significantly correlated with
haploid number (Pearson product-moment correlation of
10.5 for all measures besides chiasmata per bivalent, sig-
nificant at the level), and all but chiasmata per
! .05
bivalent are significantly influenced by ploidy level (
in all cases). Chiasmata per bivalent is least correlated
with haploid number (Pearson product-moment corre-
lation of 0.20) and is independent of ploidy level (
). Though only data using chiasmata per bivalent are
reported here, analyses performed using the other mea-
sures do not differ qualitatively from those presented.
Results from an initial across-species regression analysis
are shown in table 1. Though domestication is not the
sole determinant of recombination rate in the final model,
it is clearly significant. A post hoc analysis reveals that
although cultivated and wild plants are not discernibly
different from each other, domesticated plants have a sig-
nificantly higher recombination rate than either of the
former (Scheffe´ test, for wild and for cul-
! .03 P ! .01
tivated), providing support for the hypothesis that do-
mestication increases recombination rate. It is worth not-
ing that none of the other characteristics (mating system,
life form, etc.) interact significantly with domestication,
suggesting that the role these have played in the effect of
domestication on recombination rate is relatively insig-
nificant. Including the progenitors of crop plants in the
regression model as a category distinct from other wild
plants adds no meaningful information; the relative sample
size of the category is small and the standard error such
that it cannot be distinguished from either other wild or
domesticated taxa.
By performing an across-taxa regression for genera and
families, the data make possible a test for preadaptation
Page 3
Figure 2: Pairwise comparison of domesticated taxa to their nearest congener. In each case the chiasmata per bivalent of the domesticate is plotted
against the chiasmata per bivalent of the progenitor. Numbers refer to species in table A1 in the online edition of the American Naturalist.
Figure 3: Pairwise comparison of progenitor taxa to their nearest congener. In each case the chiasmata per bivalent of the progenitor is plotted
against the chiasmata per bivalent of the congener. Solid line, regression when all data points are included. Dotted line, regression when the two
outlier taxa (filled circles) are removed from the analysis. Numbers refer to species in table A1 in the online edition of the American Naturalist.
Page 4
Recombination under Domestication 109
Table 2: Number of species pairs, taxa excluded, average difference in chiasma per bivalent and
its SE, and P values associated with domesticate-progenitor (D-P) and progenitor-congener (P-
C) pairwise comparisons
Species pairs 26 22 20 22 20
Taxa excluded 7,10,17,18 7,10,14
7,10,17,18 7,10,14
Average difference (SE) .095 (.044) .097 (.046) .117 (.045) .048 (.083) .031 (.037)
Paired t .02 .023 .009 .283 .208
Wilcoxon .028 .029 .017 .19 .161
Note: A paired t-test was performed on the transformed data, and a Wilcoxon paired-samples test was calculated
using nontransformed data. The numbers of taxa excluded in each comparison refer to table A1 in the online
edition of the American Naturalist, and the rationalization for each comparison is explained in the text. The mean
SD of replicate measurements of chiasma per bivalent within a species was .140.
at these taxonomic levels. Were preadaptation acting at
these higher levels, genera or families that gave rise to
domesticated taxa would be expected to have higher re-
combination rates than genera or families that did not.
Such an analysis reveals no significant effect ( forP
1 .7
both families and genera; results do not differ if data from
domesticated species are included in the analysis), sug-
gesting that preadaptation is not an important factor at
the genus or family level.
A more direct test of the role of recombination is a
pairwise test of the actual taxa involved. These are shown
graphically in figures 2 and 3, and P values for these tests
are reported in table 2. Though the overall data conform
reasonably well to normality after transformation, the set
of transformed pairwise comparisons (both the raw data
and the set of differences themselves) is still markedly
nonnormal. While the t-test is generally robust to non-
normality, a nonparametric test is perhaps more appro-
priate for these analyses; results for both a standard paired
t-test and the Wilcoxon paired-samples test are reported
in table 2.
The comparison of domesticates to their wild progen-
itors is shown in figure 2. Domesticated taxa in general
show a significant increase in recombination when com-
pared to their wild relatives (table 2). Hexaploid domes-
ticated oats (Avena sativa) differ in ploidy from their most
direct wild progenitor, the tetraploid Avena insularis. Given
that chiasmata per bivalent is independent of ploidy level,
this difference should not affect the results; however, even
if Avena is removed from the list, the results do not change
Of the 26 species pairs from table A1 compared above,
recombination data for both a progenitor and a congener
were available in only 22 cases. For three of the progenitor
taxa (Cucumis, Pennisetum, and Triticum), the closest avail-
able congener was of different ploidy level; removal of
these taxa once again does not change the results. Two
cases (Oryza and Vicia) were found to be statistical outliers
(Grubb’s test for outliers, ). In both cases, the con-
! .05
geners used, though the closest relative of the progenitor,
are dramatically different from other species in their genera
and probably should be excluded as misrepresentative.
Nonetheless, results from both comparisons (the original
22 and the 20 nonoutliers) are shown in table 2 and are
plotted on figure 3. The comparison does not detect a
significant difference between progenitors and their con-
geners, thus failing to support Gornall’s hypothesis of
preadaptation. Though the number of taxa analyzed is
small, it is unlikely that any large difference was missed
by this test. Using the same reduced set of 22 or 20 taxa
to make the paired domesticate-progenitor comparison
still produces a statistically significant result (table 2), in-
dicating that even if recombination rate is of preadaptive
value, its effect is less than that of domestication in in-
creasing recombination rate via selection. Moreover, a post
hoc power analysis (Hintze 2001) based on the 20 pro-
genitor-congener pairs does not reveal any real lack of
power (80% power to detect a difference of 0.061 in the
transformed data, a difference less than that found in any
of the comparisons of domesticates and progenitors).
Finally, data on mating system (outcrossing, mixed, in-
breeding), weediness, and life form (annual vs. perennial)
were used to determine whether these characters, or
changes in these characters, are correlated with differences
found in the pairwise tests. None of the ANOVAs of pair-
wise differences in recombination rate with regard to these
species characteristics or changes in these characteristics
was significant ( in all cases) for either of the
1 .10
comparisons (domesticate-progenitor or progenitor-
congener). The increase in recombination rate due to do-
mestication does not seem to be affected by any of the
other traits studied or changes in these traits, nor does the
lack of preadaptation seem to be explained by these other
Page 5
110 The American Naturalist
The results from both an across-species analysis and a
paired comparison of domesticates to their progenitor taxa
strongly suggest that the domestication process generally
increases the recombination rate of a species. Domesti-
cated taxa show a higher overall recombination rate than
nondomesticated taxa (fig. 1), and pairwise comparison
to their progenitors reaffirms the result. Though chiasmata
per bivalent is undoubtedly the most appropriate measure
of recombination for this analysis, it is comforting to note
that all measures tested showed similar results.
In spite of the data available, the literature includes few
observations that domestication might affect chiasma fre-
quencies. Though several studies have published data on
both a domesticate and its progenitor, very few authors
have noted the difference in recombination rate (but see
Koul et al. 1989), and none have offered a plausible ex-
planation. In addition to these few empirical papers, a
study by Burt and Bell (1987) on recombination rate and
life span in mammals has occasionally been cited as evi-
dence of the effect of domestication on recombination rate
(Koul et al. 1989; Otto and Barton 2001). While their data
do show a statistically significant increase in excess chiasma
among several domesticated mammal species, their study
does not include any of the progenitors of these species,
and the effect disappears if the analysis is done using mea-
sures of recombination shown to be less strongly correlated
with chromosome number (data not shown). With the
exception of the limited work mentioned above, then, this
study is the first to conclusively show the general effect of
domestication on recombination rate.
In contrast to the lack of empirical work, there is a large
body of theory that would predict a correlation between
domestication and a change in recombination rate. Much
effort has been devoted to determining the potential
sources of the negative linkage disequilibria and the con-
ditions under which recombination is favored. It is widely
agreed, however, that strong directional selection (Feldman
et al. 1997), especially at multiple loci (Otto and Barton
1997) or in concert with genetic drift (Felsenstein and
Yokoyama 1976; Otto and Barton 2001), can generate neg-
ative disequilibria sufficient to select for increased recom-
bination. And while it has been suggested that introgres-
sion from wild relatives could potentially select for a
decreased recombination rate in domesticates (Lenormand
and Otto 2000), the majority of the conditions provided
by the process of domestication (new environment, strong
directional selection at multiple loci, and, at least in some
cases, small population sizes) concur with those thought
to select for recombination.
Theoretical analysis offers many scenarios for the evo-
lution of recombination; it also cautions of the negative
consequences of recombination rate. Not only does re-
combination break down negative disequilibria among
beneficial alleles, but it also disrupts positive disequilibria
among them as well, potentially splitting up adaptive gene
complexes (Barton 1995). This “recombination load” is
thought to limit selection for increased recombination and
effectively set a threshold above which higher recombi-
nation is selected against. Selection against excessive re-
combination would predict a regression slope of
!1 be-
tween the recombination rate of a domesticate and that
of its progenitor. Such a relation is clearly seen in figure
2 ( ). Interestingly, the regression line crosses theP
! .05
1 : 1 line of equality at just above 2 chiasmata/bivalent, a
number suggested by Kondrashov (1988) as a potential
limit beyond which the effects of recombination become
detrimental. Perhaps most significantly, a slope of
not in keeping with an alternative explanation for the as-
sociation between domestication and high recombination
rate: the idea that increased homozygosity, and not selec-
tion, is the primary cause of increased recombination.
While the results obtained strongly support the hy-
pothesis that domestication selects for increased recom-
bination rate, they offer no support for the hypothesis that
recombination rate is an important preadaptation to do-
mestication. Pairwise comparisons reveal that not only do
the wild progenitors of crop plants generally possess lower
recombination rates than their domesticated descendants
but also that they are not discernibly different from their
wild congeners. Furthermore, if recombination functions
as a preadaptation to domestication, constraints due to
recombination load should be evident as recombination
increases in the congeners of domesticated species. As re-
combination increases in congeneric taxa, selection in fa-
vor of progenitors with higher recombination should de-
crease, resulting in a slope of
!1 as seen in figure 2. Though
this relation is in fact seen in figure 3, the removal of the
two outlier taxa reveals that it is probably artifactual; the
remaining tightly-clustered comparisons do not reveal any
such relation (the regression line is not different from 1).
Likewise, across-taxa comparisons at the genus and family
levels similarly find no effect of preadaptation, and there
are no correlations with mating system, life form, or weed-
iness that explain the lack of an effect.
Though we are beginning to form an idea of the effect
that domestication has had on the genetics of plant species,
we are still far from understanding the role genetics has
played in determining which species were successfully do-
mesticated. It seems implausible that ecological factors
alone (mating system, life form) could explain the im-
pressive discrepancy between the 250,000 species of flow-
ering plants and the few hundred species of domesticates
(Hawkes 1983; Diamond 2002). Yet the very extent to
which humans have modified or utilized the botanical di-
Page 6
Recombination under Domestication 111
versity of their environments almost requires a mechanistic
explanation for this discrepancy, and it seems quite plau-
sible that genetic factors could have played a central role
in the success or failure of the domestication process.
While the genetic bases of many important domestication
traits have been identified (Paterson 2002), we will not
know how these genes have influenced the success of do-
mestication until they have been studied in many wild
species as well. Other hypothesized genetic preadaptations,
such as polyploidy (Hilu 1993), have been shown to be
unimportant in determining the successful domestication
of plant species, and the present analysis suggests that
recombination rate is likewise of little importance. Clearly,
much work is still needed before we can approach a more
complete understanding of the genetic mechanisms in-
volved in the domestication process.
I would like to thank the patience, advice, and guidance
of J. Hamrick, S. Otto, and D. Promislow, as well as the
helpful comments of many reviewers, including R. S.
Cornman, M. Eubanks, M. J. Godt, K. Parker, A. Paterson,
M. C. Q. Provance, P. Ross, S. Ross, and two anonymous
reviewers. Thanks are also due to A. Burt for kindly sharing
his data. This work was funded in part by a fellowship
from the University of Georgia.
Literature Cited
Barton, N. H. 1995. A general model for the evolution
of recombination. Genetical Research 65:123–144.
Burt, A., and G. Bell. 1987. Mammalian chiasma fre-
quencies as a test of two theories of recombination.
Nature 326:803–805.
Charlesworth, B. 1993. Directional selection and the evo-
lution of sex and recombination. Genetical Research
Darlington, C. D., and A. P. Wylie. 1956. Chromosome
atlas of flowering plants. Macmillan, New York.
Diamond, J. 2002. Evolution, consequences, and future
of plant and animal domestication. Nature 418:700–
Doebley, J. F. 1989. Isozymic evidence and the evolution
of crop plants. Pages 165–191 in D. E. Soltis and P. S.
Soltis, eds. Isozymes in plant biology. Dioscorides,
Portland, Oreg.
Feldman, M. W., S. P. Otto, and F. B. Christiansen. 1997.
Population genetic perspectives on the evolution of
recombination. Annual Review of Genetics 30:261
Felsenstein, J., and S. Yokoyama. 1976. The evolutionary
advantage of recombination. II. Individual selection
for recombination. Genetics 83:845–859.
A global compendium of weeds. 2002. http://
Gornall, R. J. 1983. Recombination systems and plant
domestication. Biological Journal of the Linnean So-
ciety 20:375–383.
Harlan, J. R. 1992. Crops and man. American Society of
Agronomy/Crop Science Society of America, Madison,
Hawkes, J. G. 1983. The diversity of crop plants. Harvard
University Press, Cambridge, Mass.
Hilu, K. W. 1993. Polyploidy and the evolution of do-
mesticated plants. American Journal of Botany 80:
Hintze, J. 2001. NCSS and PASS. Number Cruncher Sta-
tistical Systems, Kaysville, Utah.
Index to plant chromosome numbers. 2002. http://
King, J., L. A. Roberts, M. J. Kearsey, H. M. Thomas, R.
N. Jones, L. Huang, I. P. Armstead, W.G. Morgan, and
I.P King. 2002. A demonstration of a 1 : 1 correspon-
dence between chiasma frequency and recombination
using a Lolium perenne/Festuca pratensis substitution.
Genetics 161:307–314.
Knox, M. R., and T. H. N. Ellis. 2002. Excess heterozy-
gosity contributes to genetic map expansion in pea
recombinant inbred populations. Genetics 162:861–
Koella, J. C. 1993. Ecological correlates of chiasma fre-
quency and recombination index of plants. Biological
Journal of the Linnean Society 48:227–238.
Kondrashov, A. S. 1988. Deleterious mutations and the
evolution of sexual reproduction. Nature 336:435–440.
Koul, P., A. K. Koul, and I. A. Hamal. 1989. Reproductive
biology of wild and cultivated carrot Daucus carota L.
New Phytologist 112:437–444.
Lenormand, T., and S. P. Otto. 2000. The evolution of
recombination in a heterogeneous environment. Ge-
netics 156:423–428.
Nilsson, N. O., T. Sall, and B. O. Bengtsson. 1993. Chi-
asma and recombination data in plants: are they com-
patible? Trends in Genetics 9:344–348.
Otto, S. P., and N. H. Barton. 1997. The evolution of
recombination: removing the limits to natural selec-
tion. Genetics 147:879–906.
———. 2001. Selection for recombination in small pop-
ulations. Evolution 55:1921–1931.
Paterson, A. 2002. What has QTL mapping taught us
about plant domestication? New Phytologist 154:591
Page 7
112 The American Naturalist
Rees, H., and P. J. Dale. 1974. Chiasmata and variability in
Lolium and Festuca populations. Chromosoma 47:335.
Smartt, J., and N. W. Simmonds. 1995. Evolution of crop
plants. Longman, Harlow, Essex.
Stasoft. 2000. STATISTICA for Windows. Statsoft, Tulsa,
Wang, R.-L., A. Stec, J. Hey, L. Lukens, and J. Doebley.
1999. The limits of selection during maize domestica-
tion. Nature 398:236–239.
Associate Editor: John Willis
Page 8
  • Source
    • "A further possibility is that local directional selection on a quantitative trait selects for the most extreme phenotype and, by proxy, the most highly recombining individuals. This is, for example, one way to explain why domestication (commonly a form of strong directional selection) is sometimes associated with increased recombination rates [72,73]. Alternatively, there may be direct selection for variance between workers in their behavior and selection for locally high recombination rates might achieve this. "
    [Show abstract] [Hide abstract] ABSTRACT: Social hymenoptera, the honey bee (Apis mellifera) in particular, have ultra-high crossover rates and a large degree of intra-genomic variation in crossover rates. Aligned with haploid genomics of males, this makes them a potential model for examining the causes and consequences of crossing over. To address why social insects have such high crossing-over rates and the consequences of this, we constructed a high-resolution recombination atlas by sequencing 55 individuals from three colonies with an average marker density of 314 bp/marker. We find crossing over to be especially high in proximity to genes upregulated in worker brains, but see no evidence for a coupling with immune-related functioning. We detect only a low rate of non-crossover gene conversion, contrary to current evidence. This is in striking contrast to the ultrahigh crossing-over rate, almost double that previously estimated from lower resolution data. We robustly recover the predicted intragenomic correlations between crossing over and both population level diversity and GC content, which could be best explained as indirect and direct consequences of crossing over, respectively. Our data are consistent with the view that diversification of worker behavior, but not immune function, is a driver of the high crossing-over rate in bees. While we see both high diversity and high GC content associated with high crossing-over rates, our estimate of the low non-crossover rate demonstrates that high non-crossover rates are not a necessary consequence of high recombination rates.
    Full-text · Article · Dec 2015 · Genome Biology
    • "This can explain the elevated CO rates in Caprinae, as an evolutionary younger taxon. Regarding Caprinae, more CO events were present in goats and sheep than in Barbary sheep, which could be explained by the process of domestication that favors higher recombination rates in these 2 species[Ross-Ibarra, 2004]. However, only 1 specimen of Barbary sheep was available for this study and, considering the individual variability within species, the evidence for the effect of domestication is weak and demands further research. "
    [Show abstract] [Hide abstract] ABSTRACT: Despite similar genome sizes, a great variability in recombination rates is observed in mammals. We used antibodies against SYCP3, MLH1 and centromeres to compare crossover frequency, position along chromosome arms and the effect of crossover interference in spermatocytes of 4 species from the family Bovidae (Bos taurus, 2n = 60, tribe Bovini; Ovis aries, 2n = 54, Capra hircus, 2n = 60 and Ammotragus lervia, 2n = 58, tribe Caprini). Despite significant individual variability, our results also show significant differences in both recombination rates and the total length of autosomal synaptonemal complexes (SC) between cattle (47.53 MLH1 foci/cell, 244.59 µm) and members of the tribe Caprini (61.83 MLH1 foci, 296.19 µm) which can be explained by the length of time that has passed since their evolutionary divergence. Sheep displayed the highest number of MLH1 foci per cell and recombination density, although they have a lower diploid chromosome number caused by centric fusions corresponding to cattle chromosomes 1;3, 2;8 and 5;11. However, the proportion of MLH1 foci observed on the fused chromosomes in sheep (26.14%) was significantly lower than on the orthologous acrocentrics in cattle (27.6%) and goats (28.2%), and their distribution along the SC arms differed significantly. The reduced recombination rate in metacentrics is probably caused by interference acting across the centromere.
    No preview · Article · Sep 2015 · Cytogenetic and Genome Research
  • Source
    • "In support of this relationship, both the Peking duck (46) and turkey (46), the remaining domesticated species among the birds, have an above average number of β-keratins. Given that domestication may increase recombination rate [58,59], the extreme variation in β-keratin copy numbers among birds may be partially linked to higher recombination rates on β-keratin loci and the domestication of these species. The differential expression of feather β-keratins is related to their genomic locus [57], signifying that expansion of feather β-keratins, through unequal crossing over events on specific loci, may be induced by artificial selection. "
    [Show abstract] [Hide abstract] ABSTRACT: Vertebrate skin appendages are constructed of keratins produced by multigene families. Alpha (α) keratins are found in all vertebrates, while beta (β) keratins are found exclusively in reptiles and birds. We have studied the molecular evolution of these gene families in the genomes of 48 phylogenetically diverse birds and their expression in the scales and feathers of the chicken. We found that the total number of α-keratins is lower in birds than mammals and non-avian reptiles, yet two α-keratin genes (KRT42 and KRT75) have expanded in birds. The β-keratins, however, demonstrate a dynamic evolution associated with avian lifestyle. The avian specific feather β-keratins comprise a large majority of the total number of β-keratins, but independently derived lineages of aquatic and predatory birds have smaller proportions of feather β-keratin genes and larger proportions of keratinocyte β-keratin genes. Additionally, birds of prey have a larger proportion of claw β-keratins. Analysis of α- and β-keratin expression during development of chicken scales and feathers demonstrates that while α-keratins are expressed in these tissues, the number and magnitude of expressed β-keratin genes far exceeds that of α-keratins. These results support the view that the number of α- and β-keratin genes expressed, the proportion of the β-keratin subfamily genes expressed and the diversification of the β-keratin genes have been important for the evolution of the feather and the adaptation of birds into multiple ecological niches.
    Full-text · Article · Dec 2014 · BMC Evolutionary Biology
Show more