REVIEWS
Efficient genetic markers
for population biology
Paul Sunnucks
I
laboratory (or, indeed, just to coln the past decade, several key
Population genetics has come of age. Three
laborate with a convenient labadvances in molecular genetics
important components have come together:
oratory), rather than choosing
have greatly increased the imefficient techniques to examine informative
the suite of genetic markers that
pact of population genetics on
segments of DNA, statistics to analyse DNA
biology. Most important have
would best answer the question
data and the availability of easy-to-use
been: (1) the development of
and then obtaining those data.
computer packages. Single-locus genetic
polymerase chain reaction (PCR),
We might consider three levels of
markers and those that produce gene
which amplifies specified stretches
molecular change that would progenealogies yield information that is truly
of DNA to useable concentrations;
vide information at different levcomparable among studies. These markers
(2) the application of evolutionels of population biology (Box 1).
answer biological questions most efficiently
arily conserved sets of PCR primers
First, the most sensitive genetic
and also contribute to much broader
(e.g. Ref. 1); (3) the advent of investigations of evolutionary, population and signals are genotypic arrays, most
hypervariable microsatellite loci2,3;
commonly encountered in the
conservation biology. For these reasons,
and (4) the advent of routine DNA
form of multiple microsatellite
single-locus and genealogical markers should
sequencing in biology laboratorloci scored in samples of individbe the focus of the intensive genetic data
ies. These innovations, coupled
uals. In sexual species, these
collection that has begun owing to the power
with the recent explosion of powerof genetics in population biology.
arrays are reshuffled at each generful analyses and relatively useration, and, therefore, are useful
friendly computer programs4, have
for the shortest- and finest-scale
Paul Sunnucks is at the Dept of Biological Sciences,
meant that much of the power
population processes, such as
Monash University, VIC 3168, Australia
inherent in molecular genetic data
individual identification and track([email protected]).
can be tapped, revealing otherwise
ing, parentage and relatedness of
unobtainable information at all
interacting individuals9–12. How3,5–7
levels of biotic hierarchy . The
ever, arrays can be recognizable
profound contribution of molecfor longer in organisms without
ular genetics is reflected in Molecular Ecology being among frequent genetic recombination13. Second, microsatellites,
the most cited primary ecological and/or evolutionary mitochondrial DNA (mtDNA) and other single-locus markjournals (Institute for Scientific Information).
ers (definitions and details in Boxes 2 and 3) can also be
analysed as individual genes with frequencies and geo‘Natural history of DNA’ results in a trail
graphic distributions (genic analyses). These properties
of information
change on larger spatial and temporal scales than genoIndividual organisms differ in the DNA sequences compris- typic arrays, and are effective markers of gene flow and
ing their genomes. This genetic variation can be considered population history, even in species with limited genetic
at the level of individual genes (genic) or of genotypes variation14,15. Third, slower again is the creation of new
(genotypic). The fate of a given genetic variant in time and alleles by mutation, thus the analysis of their evolutionary
in space will be influenced by the biology and circum- relationships (allele and/or gene genealogies, or phystances of the individuals through which it passes, includ- logenies) is informative about the longer-term processes
ing reproductive success, migration, population size, nat- of phylogeography, speciation and deeper taxonomic
ural selection and historical events. Population genetic phylogenetic reconstruction8.
models investigate the connection between these demographic features and the distribution of molecular genetic Basic properties of genetic markers in
variants7,8. By measuring genetic variation and by applying population biology
population genetic models, we can make inferences about There is an apparently bewildering array of genetic techthe biology of organisms. Processes that affect individuals niques available for population genetic analysis5,7,16. Howultimately accumulate into effects on populations, which, in ever, ignoring technical details and focusing on important
turn, influence speciation, and so on up the taxonomic hier- properties helps to make sense of the methods. Genetic
archy8. Thus, by examining genetic markers with appropri- markers are simply heritable characters with multiple
ate rates of change, and, therefore, suitable signals, infor- states at each character. Typically, in a diploid organism,
mation can be obtained about almost any population and each individual can have one or two different states
evolutionary process through the hierarchy of life.
(alleles) per character (locus). All genetic markers reflect
The rates of change of the distributions of different differences in DNA sequences, usually with a trade-off
genetic markers vary owing to the differential action of between precision and convenience. Separate loci can profundamental processes, including recombination, mutation vide independent tests of hypotheses, thus using many
and selective constraint (Box 1). Selecting appropriate together can yield extreme sensitivity. Genetic variation
genetic analysis is vital to the success of applying molecu- is usually organized hierarchically (for example, two allar genetics in population biology. It is a surprisingly com- leles within an individual, N individuals within a subpopumon mistake just to use techniques that are available in a lation and Y subpopulations within populations). Thus, data
TREE vol. 15, no. 5 May 2000
0169-5347/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0169-5347(00)01825-5
199
REVIEWS
Box 1. Rates of change of the major categories of genetic
variants in population biology
Genotypic
Genotypic arrays (composite genotypes of multiple loci): individual genotypes are labile – a single round of sexual recombination usually destroys a
genotype. They are quantified most rigorously using multiple single-locus
nuclear markers (typically microsatellites) (see Boxes 2–4 and Table 1 for
definitions concerning, and attributes of, genetic markers).
• Application: the individual level, including movements of individuals, noninvasive sampling, parentage and relatedness of interacting individuals7,9,11,12,17,28. Attributes of individual genotypes can address questions
such as the relationship between inbreeding and/or outbreeding and fitness4.
Interlocus allelic correlations (linkage disequilibrium, LD): associations of alleles at pairs of nuclear loci will persist according to their recombination rate,
the effective population size (Ne) and natural selection. Textbook population
genetics holds that newly generated correlations of unlinked loci in ideal
populations decay to zero in approximately ten generations, but associations
among linked loci might persist for hundreds or thousands of generations.
• Application: population level including Ne estimation, understanding
metapopulation dynamics, and recognizing recent colonizations and introductions7,29.
Genic
Allele and/or haplotype frequencies are population statistics that can be
changed by genetic drift, founder effect, gene flow and selection. They are
estimated most accurately from multiple, separate, nuclear loci and mitochondrial DNA (mtDNA).
• Application: includes estimating gene flow and population subdivision8.
Gene genealogies
Sequence of mtDNA or nuclear genetic regions [microsatellites and single
copy nuclear (scn) markers] will evolve as determined by the mutation rate,
selection and population parameters, such as Ne and changes in Ne.
• Application: among populations or species, including intraspecific
phylogeography and population history, systematics and biodiversity
assessment1,5,8,19,30.
Box 2. Favourable attributes of genetic markers in
population biology
Assayable by polymerase chain reaction (PCR): PCR can use low quantities of even degraded DNA and can target specific DNA regions.
Comparability (‘connectibility’): PCR primers that amplify homologous
regions over a wide taxonomic range generate directly comparable data,
thus facilitating profound insights by meta-analysis (deriving conclusions
from the results of many studies)8,22,23,25.
DNA rather than protein: unlike allozymes (Box 3), DNA can be extracted
from old material, is more convenient for collection and storage of samples,
is more variable, is PCR-assayable and can yield gene genealogies.
Gene genealogies and frequency data: molecular genealogies uniquely
can untangle current and historical processes, and can yield information on
demographic trends19. Markers giving both allele and/or haplotype frequency
and sequence data are informative over a range of scales5,16,19.
Many separate loci available: the use of multiple markers has two main
advantages: (1) overcoming stochastic biological sampling that can cause
different patterns in loci with similar histories; and (2) detection of nonconcordant characters that are likely to be biologically interesting (e.g. affected
by natural selection).
Rapid development and screening: ideal markers would be useable in new
taxa with little further development and amenable to rapid screening.
Single-locus as opposed to multilocus markers: multilocus approaches
(i.e. they can visualize many anonymous genes simultaneously, for example,
RAPD and AFLP) are technically convenient but imprecise, and have many
major technical and/or analytical drawbacks, such as dominance (only one
allele identified)31–33. Data are of limited comparability among studies. By
contrast, single-locus, codominant (both alleles identified) or haploid organellar markers supply robust data for input into precise analyses. Data are
comparable among studies, thus contribute globally.
are amenable to analysis by approaches closely related to
mainstream statistics familiar to ecologists, including
analyses of variance, spatial autocorrelation, Mantel testing and diverse multivariate analyses.
200
Recent major advances in analysis in molecular
population biology
After some considerable lags behind theory (mostly because suitable data were not readily available before the
advent of microsatellites), population genetic analysis is
making great use of maximum likelihood, Bayesian statistics and Markov chain Monte Carlo simulation techniques
to extract more information from genetic data4,17,18. A fundamental advance is the development of analyses of gene
genealogies focusing on evolutionary relationships among
individual alleles sampled from populations within
species3,19 (Box 4). These techniques (notably ‘coalescence
approaches’3,4,8,19 and ‘nested clade analysis’19) allow clear
testing of historical and spatial hypotheses, such as distinguishing current restricted gene flow from past gene flow,
and investigating the direction and relative timing of
events, such as range expansions. Finally, models of molecular evolutionary change are becoming sophisticated and
can greatly improve genealogical inference16. These developments have resulted in fundamental advances in what
can be learnt about populations. Luikart and England4 cite
26 different analyses in four categories of population biology [relatedness and parentage, dispersal and migration,
inbreeding, and effective population size (Ne )] that can provide previously unattainable information, and this counts
only those primarily involving microsatellites. The new
approaches are making great strides in analysing nonequilibrium situations that predominate in conservation biology
and in studies of invading organisms15,19,20.
Choosing genetic markers for a given question
A population genetic survey must start with a decision
regarding appropriate genetic markers. The main issues
are outlined below, and attributes of specific markers are
summarized in Boxes 2 and 3, and Table 1.
Sensitivity
A marker must have the correct sensitivity for the question
(Box 1 and Table 1). It is possible to have too much information (if entities are too different there is nothing to link
them) or too little information (no signal). Accumulated
data have given us a good idea about what sorts of markers
are probably informative at a given level in the biotic hierarchy, but pilot studies are usually a good idea. Among
markers with suitable resolution, choices can be made on
more pragmatic bases (Box 2).
Multilocus or single-locus?
Usually, there is a trade-off between practicality and accuracy of genetic markers. One manifestation of this is the
dichotomy between multilocus DNA techniques [usually
RAPDs (randomly amplified polymorphic DNA) and AFLP
(amplified fragment length polymorphic DNA)] and singlelocus techniques [commonly microsatellites and single
copy nuclear (scn)DNA regions] (Boxes 2 and 3). Multilocus
approaches are technically convenient, but have some
marked weaknesses and limitations, including that a substantial proportion of the variation they detect can be nonheritable or not even derived from the target organism.
These drawbacks have been shown in organisms including
plants, nematodes, flies, birds and mammals (Box 2). A fundamental limitation is dominant inheritance – DNA fragments can be scored only as present or absent, in contrast
to codominant inheritance where each of the two alleles at a
locus in an individual can be identified and thus analysed
more precisely. As a consequence of simultaneous visualization of many dominant markers, multilocus data typically
TREE vol. 15, no. 5 May 2000
REVIEWS
are analysed as pairwise comparisons of complex patterns
that only have meaning relative to others in the same
study. Thus, multilocus data can be compared only superficially among studies. Allele frequencies are rarely available, making many of the powerful analyses discussed here
impossible. By contrast, single-locus markers are far more
flexible, informative and connectible, because they can be
analysed as genotypic arrays, as alleles with frequencies
and as gene genealogies. It is sometimes asserted that multilocus techniques are more economical, but this is doubtful, especially per unit information21. In that study, singlelocus microsatellites detected much that RAPD markers
had overlooked: a highly divergent cryptic species, interspecific gene flow and high levels of genetic recombination, as well as some more complex aspects of biology.
Single-locus markers are even more economical when the
value of comparing data sets is considered22,23.
It is no coincidence that most of the important recent
advances in population genetic analysis cannot use multilocus data (Box 4); they depend on comparisons of attributes
of alleles within and among loci, which are not provided by
multilocus techniques. Notwithstanding, multilocus techniques can generate many variable bands, and, consequently, can be powerful in applications such as gene mapping and analysis of quantitative traits. In one elegant
example, a variable RAPD band was cloned and converted to
a single-locus codominant marker of asexuality in an aphid24.
Gene genealogies and frequencies
Markers capable of yielding gene genealogies present
some enormous benefits over those that do not (Box 4).
Analysis of the relationships between demographics and
genealogies are a major growth area leading to some previously unimaginable advances in what can be deduced
about population processes and history3,4,7,8,15,19. For example, molecular phylogenies can help untangle current
structure from the effects of historical events19 and might
be the only way of obtaining long-term demographic information for conservation planning6,8. Nested clade analysis19 is formulated in a particularly clear statistical and
hypothesis-testing framework, and can be implemented
by the computer program GEODIS (D. Posada, http://
bioag.byu.edu/zoology/crandall_lab/geodis.htm). Worked examples, including inference of spatial and temporal details of
postglacial range expansion in gophers (Geomys bursarius),
can be found in Ref. 19. As well as contributing to such developments, another global advantage of the acquisition of
genealogies is to facilitate the increasingly profound inferences becoming possible using data sets from comparable
(‘connectible’ sensu5) genetic markers. These include bringing together data from diverse organisms to uncover the
major evolutionary impacts of phenomena such as glaciation, and use of standard measures of genetic divergence to
integrate biological and phylogenetic species concepts22,23,25.
Organelle and nuclear DNA
Cells from most eukaryotes contain biparentally inherited
nuclear DNA, as well as DNA in organelles (mitochondria;
and chloroplasts in plants) that is usually inherited uniparentally. This difference in transmission, and some major
differences in patterns of evolution, causes organellar DNA
and nuclear DNA gene genealogies to reflect different
aspects of population biology and history. Mitochondrial
DNA has a lower Ne than nuclear markers, and, consequently (under most demographic scenarios), mtDNA variants become diagnostic of taxa more rapidly. Comparison
of nuclear and mitochondrial genotypes can help recognize
TREE vol. 15, no. 5 May 2000
Box 3. Attributes of specific genetic markers most useful
in population biology
Anonymous single copy nuclear (scn)DNA: several methods have been
used to develop PCR-assayable scnDNA regions34. These techniques can be
technically convenient and can yield sensitive data, but the actual markers
are transferable only among similar taxa. The technique is too recent for its
full potential to be known.
Conserved primers for amplifying specific variable nuclear regions (e.g.
introns): primers in evolutionarily conserved DNA regions that flank more
variable regions can be technically convenient, and can yield connectible
frequency and genealogical data15,35. Loci tend to be too invariant for the
shortest-term population processes. The technique is too recent for its full
potential to be known.
Protein electrophoresis (allozymes): modest numbers of variable, codominant, nuclear allozymes are available with minimal development. The technique is inexpensive, but requires high-quality samples, often reveals little
variation and gives limited genealogical information. It is used extensively in
studies of gene flow.
Sequences from known mtDNA and nuclear regions amplified by conserved or specific primers: is technically convenient for most organisms,
and is the mainstay of intraspecific phylogeography and of systematics.
Single-locus microsatellites: numerous hypervariable, codominant nuclear
markers are available, assayed via PCR to reveal length variation among alleles. These provide sensitive, connectible data from individual identification
through to shallow phylogeny. Loci have a wide range of evolutionary rates,
thus examine different timescales. Development of loci is laborious, but
once developed they can have moderate taxonomic breadth of application,
thus suitable markers might be available in the literature36 (unpublished information can be obtained via e-mail: [email protected]). Microsatellites are the
mainstay of modern population genetics other than systematics.
Box 4. Important recent developments in genetic
population biology analysis
Using low probabilities of individual genotypes (multiple single loci)
Parentage and/or relatedness analysis: these find probable parents from pools
of candidates, test individual parentage and/or relatedness hypotheses and estimate kinship, with adjustable assumptions and giving reliability estimates4,10,17.
Migration and nonequilibium populations – assignment tests and mixed stocks
analysis: assignment tests estimate the population of origin of individuals, thus
identifying individual migrants (and even their descendants)11,20 and uncovering attribute-biased dispersal. Related tests (e.g. mixed stocks analysis) examine population admixture4,15,28. These analyses estimate the same phenomena
as trapping and/or marking, but have the enormous advantage that ‘recapture’
is not required: ‘…very few birds have bands, but all have genotypes’28.
Using gene genealogies and distributions of other allelic properties
Microsatellite allele length data (in the light of molecular genetic models) and
DNA sequences yield gene genealogies3,4,7,8. Genealogies illuminate population
processes, phylogeographic events and speciation, because they add the
dimension of evolutionary (thus temporal) relationships among alleles, and these
can be related to spatial organization3,4,8,19. Other allele properties (lengths, frequencies and distribution shapes) yield sensitive information4. These procedures
are becoming well validated using large data sets, simulation and experiments3,4,7,8. Gene genealogies and marker attributes can be used to estimate
current, past, changes in, timing of changes in, and even rate of changes in Ne,
thus uncovering population history and overcoming previously intractable problems3,4,15,28. For example, BOTTLENECK (Ref. 4) infers recent bottlenecks without prebottleneck data, by examining observed relationships between allelic
diversity and heterozygosity compared with molecular genetic models.
hybrid individuals, asymmetrical mating preferences and
stochastic effects on variants for which ancestral taxa were
polymorphic. These phenomena can cause phylogenetic
trees of some genes to not match those of the taxa that
carry them5. Accordingly, it is generally preferable to use a
suite of markers capable of detecting such phenomena.
Rapid development and screening
Major gains in efficiency in new research can arise if genetic
markers have already been developed or can be transferred
from earlier work, and the possibility of rapid screening,
such as single-stranded conformation polymorphism (SSCP)
201
REVIEWS
Table 1. Attributes of markers commonly used in molecular population biologya
PCR
assay
Single
locus
Mitochondrial (and chloroplast)
Sequence
Yes
Yes
RFLP
No, large
Yes
Multilocus nuclear
Mini- and/or microNo, large
No
satellite ’fingerprints‘
Yes
No
RAPDb
Yes
No
AFLPb
Yes
No
rDNAc
Single-locus nuclear (single copy nuclear, scn)
Allozymes
No, protein
Yes
Minisatellites
Few
Yes
Microsatellites
Yes
Yes
Anonymous scn
Yes
Yes
Specific scn
Yes
Yes
rDNAc
Yes
In effect
Codominant
Allele
genealogy
feasible
No. loci
readily
available
Connectibility Rapid
of data
transfer to
among studies new taxa
Overall
variabilityg
Yesd
Yesd
Yes
Yes
Single
Single
Direct
Direct
Yes
Yes
Low–high
Low–moderate
No
No
Many
Limited
Yes
High
No
No
No
No
No
No
Many
Many
Few
Limited
Limited
Limited
Yes
Yes
Yes
High
High
Moderate–high
Yes
Yes
Yes
Yes
Yes
Yes
Rarely
Rarely
Yes
Yes
Yes
Yes
Moderate
Moderate
Many
Many
Moderate
Few
Direct
Indirecte
Indirecte
Indirecte
Direct
Direct
Yes
Few
Some
No?f
Yes?f
Yes
Low–moderate
High
High
Moderate?f
Moderate?f
Low–moderate
aMore
details in Boxes 2 and 3.
RAPD (randomly amplified polymorphic DNA) and AFLP (amplified fragment length polymorphic DNA) bands can be converted to single-locus markers,
in which case they behave like ‘anonymous scn’or ‘specific scn’ categories.
crDNA consists of tandem arrays of a few regions. In some taxa the arrays are effectively identical and regions act as single loci, but in some taxa there can
be many different sequences within individuals, in which case rDNA acts more like a multilocus system.
dmtDNA and chloroplast DNA are haploid and show one of a range of alternative positive states, in contrast to dominant markers that are either present or absent.
eData from these markers are indirectly, but meaningfully, connectible given adequate models of molecular evolution.
fInsufficient research effort has been put into these markers.
gVariability depends on variation per marker and number of markers obtained readily. The assessment here approximates the outcome of a typical marker system.
bSome
(Refs 7,8,13,16), can yield important savings in resources.
These issues of technical convenience should be included
in assessments of relative suitability of candidate marker
systems for each project.
Prospects
The advances outlined here depend on the application of
sophisticated statistics to the sensitive data that can be
obtained from microsatellites and from other single-locus
markers, and the development of methods for extracting
demographic information from gene genealogies. Many
computer programs and refinements or new applications
of statistics can be found on the Internet (Ref. 4). Although
many sorts of patterning in genetic data (e.g. allele frequency distributions and evolutionary divergences among
alleles) are coming into common use, others are underutilized, a good example being allelic correlations among loci
(linkage disequilibium). Although genotypic approaches
have made in-roads into modelling nonequilibrium demographic scenarios, much progress remains to be made15.
For example, assignment tests assume linkage and Hardy–
Weinberg equilibria, but these are unlikely to hold during
colonizations or in recently perturbed populations, such
as those of conservation concern. Violations of these
assumptions might have important effects on the accuracy
of current tests (B. Rannala, pers. commun.). On a technical note, approaches that detect high levels of sequence
variation in nuclear genome regions, such as introns and
variable regions of ribosomal RNA (rRNA) genes, have not
yet supplied sufficient markers for many taxa. Impressive
efficiencies in marker development and screening have
been made in biomedicine and agriculture, particularly
codominant AFLPs and microchip-based screening of
single nucleotide polymorphisms (SNPs). However, these
202
require substantial input of resources and it remains to be
seen what impact they will have in studies of species that
are not of economic interest.
Much of the current boom in molecular ecology has
resulted from user-friendly computer packages such as
GENEPOP (Ref. 26) (version 3.1d can be obtained from
http://www.cefe.cnrs-mop.fr/). However, although the principle of interconnection among packages embodied in
GENEPOP greatly facilitates development of the field, it still
has a long way to go. Hours can be spent converting data
from one format to another, and trying to find elusive typographical errors that cause programs to crash. Finally,
greater understanding of molecular evolution is central
to increased precision in inferences about population
processes. Two important examples are refining the assumptions and approaches to absolute dating of population
events using molecular clocks, and a clearer understanding
of the occurrence and impact of convergent evolution of
microsatellite allele lengths (length homoplasy)3,7,8,23,24.
By providing solutions to difficult problems, molecular
ecology is an invaluable part of modern population biology;
it is most powerful when integrated with whole animal biology7,27. Technical and analytical developments have facilitated each other, and we are rapidly achieving great resolution for interesting biological questions. Primarily this will
be via the intelligent application of multiple, single-locus,
codominant, genealogy-yielding genetic markers.
Acknowledgements
I apologize to the many excellent researchers whose work I
have not been able to cite directly in this review. Tristan
Marshall, Chris Simon, Gordon Luikart, Andrea Taylor, Alex
Wilson, Luciano Beheregaray, Steve Jordan, Lou Rodgerson
and an anonymous reviewer made valuable comments on this
article. Dave Briscoe coined ‘the natural history of DNA’.
TREE vol. 15, no. 5 May 2000
REVIEWS
Thanks to Bruce Rannala for useful discussions. I thank Mike
Bruford, Bob Wayne and Andrea Taylor for their lasting effects
on my way of thinking about molecular population biology.
This is publication 308 of the Key Centre for Biodiversity and
Bioresources at Macquarie University, Sydney.
References
1 Simon, C. et al. (1994) Evolution, weighting, and phylogenetic utility of
mitochondrial gene sequences and a compilation of conserved
polymerase chain reaction primers. Ann. Entomol. Soc. Am. 87, 651–701
2 Jarne, P. and Lagoda, J.L. (1996) Microsatellites, from molecules to
populations and back. Trends Ecol. Evol. 11, 424–429
3 Goldstein, D.B. and Schlötterer, C., eds (1999) Microsatellites:
Evolution and Applications, Oxford University Press
4 Luikart, G. and England, P.E. (1999) Statistical analysis of microsatellite
data. Trends Ecol. Evol. 14, 253–256
5 Avise, J.C. (1994) Molecular Markers, Natural History and Evolution,
Chapman & Hall
6 Schierwater, B. et al., eds (1994) Molecular Ecology and Evolution:
Approaches and Applications, Birkhäuser
7 Smith, T.B. and Wayne, R.K., eds (1996) Molecular Genetic
Approaches in Conservation, Oxford University Press
8 Burke, T. et al., eds (1998) Special Issue – Phylogeography. Mol. Ecol.
7, 367–545
9 Taylor, A.C. et al. (1997) Unusual relatedness structure detected by
microsatellite analysis, and parentage analysis in an endangered marsupial,
the Northern Hairy-nosed wombat, Lasiorhinus krefftii. Mol. Ecol. 6, 9–20
10 Marshall, T.C. et al. (1998) Statistical confidence for likelihood-based
paternity in natural populations. Mol. Ecol. 7, 639–656
11 Paetkau, D. et al. (1998) Gene flow between insular, coastal and
interior populations of brown bears in Alaska. Mol. Ecol. 7, 1283–1292
12 Taberlet, P. et al. (1999) Non-invasive genetic sampling: look before
you leap. Trends Ecol. Evol. 14, 323–327
13 Wilson, A.C.C. et al. (1999) Microevolution, low clonal diversity and
genetic affinities of parthenogenetic Sitobion aphids in New Zealand.
Mol. Ecol. 8, 1655–1666
14 Taylor, A.C. et al. (1994) Genetic variation of microsatellite loci in a
bottlenecked species: the northern hairy-nosed wombat Lasiorhinus
krefftii. Mol. Ecol. 3, 277–290
15 Davies, N. et al. (1999) Determining the source of individuals:
multilocus genotyping in non-equilibrium population genetics.
Trends Ecol. Evol. 14, 17–21
16 Hillis, D.M. et al., eds (1996) Molecular Systematics, Sinauer
17 Goodnight, K.F. and Queller, D.C. (1999) Computer software for
performing likelihood tests of pedigree relationship using genetic
markers. Mol. Ecol. 8, 1231–1234
18 Shoemaker, J.S. et al. (1999) Bayesian statistics in genetics – a guide
for the uninitiated. Trends Genet. 15, 354–358
Organizing a meeting?
If you would like your workshop, conference or symposium to have a free
entry in TREE ’s Meetings Diary,
please send the details to:
The Editor, Trends in Ecology &
Evolution, 84 Theobald’s Road,
London, UK WC1X 8RR
(or [email protected]).
TREE vol. 15, no. 5 May 2000
19 Templeton, A.R. (1998) Nested clade analysis of phylogenetic data:
testing hypotheses about gene flow and population history. Mol. Ecol.
7, 381–397
20 Rannala, B. and Mountain, J.L. (1997) Detecting immigration by using
multilocus genotypes. Proc. Natl. Acad. Sci. U. S. A. 94, 9197–9201
21 Sunnucks, P. et al. (1997) Genetic structure of an aphid studied using
microsatellites: cyclic parthenogenesis, differentiated lineages, and
host specialization. Mol. Ecol. 6, 1059–1073
22 Johns, G.C. and Avise, J.C. (1998) A comparative summary of genetic
distances in the vertebrates from the mitochondrial cytochrome
b gene. Mol. Biol. Evol. 15, 1481–1490
23 Avise, J.C. et al. (1998) Speciation durations and pleistocene effects on
vertebrate phylogeography. Proc. R. Soc. London Ser. B 265, 1707–1712
24 Simon, J-C. et al. (1999) Molecular markers linked to breeding system
differences in segregating and natural populations of the cereal aphid
Rhopalosiphum padi L. Mol. Ecol. 8, 965–973
25 Avise, J.C. and Johns, G.C. (1999) Proposal for a standardized
temporal scheme of biological classification for extant species.
Proc. Natl. Acad. Sci. U. S. A. 96, 7358–7363
26 Raymond, M. and Rousset, F. (1995) GENEPOP (version 1.2):
a population genetics software for exact tests and ecumenicism.
J. Hered. 86, 248–249
27 Moritz, C. and Lavery, S. (1996) Molecular ecology: contributions from
molecular genetics to population ecology. In Frontiers of Population
Ecology (Floyd, R.B. et al., eds), pp. 433–450, CSIRO Publishing
28 Waser, P.M. and Strobeck, C. (1998) Genetic signatures of
interpopulation dispersal. Trends Ecol. Evol. 13, 43–44
29 Tishkoff, S.A. et al. (1996) Global patterns of linkage disequilibrium at
the CD4 locus and modern human origins. Science 271, 1380–1387
30 Crozier, R.C. et al. (1999) Towards complete biodiversity assessment:
an evaluation of the subterranean bacterial communities in the Oklo
region of the sole surviving natural nuclear reactor. FEMS Microbiol.
Ecol. 28, 325–334
31 Hurme, P. and Savolainen, O. (1999) Comparison of homology and linkage
of random amplified polymorphic DNA (RAPD) markers between
individual trees of Scots pine (Pinus sylvestris L.). Mol. Ecol. 8, 15–22
32 Pérez, T. et al. (1998) An evaluation of RAPD reproducibility and
nature. Mol. Ecol. 7, 1347–1358
33 Rabouam, C. et al. (1999) Features of DNA fragments obtained by
random amplified polymorphic DNA (RAPD) assays. Mol. Ecol. 8, 493–504
34 Burt, A. et al. (1996) Molecular markers reveal cryptic sex in the human
pathogen Coccidioides immitis. Proc. Natl. Acad. Sci. U. S. A. 93, 770–773
35 Congdon, B.C. et al. Rapid population expansion and peripheral
isolation in marbled murrelets: historical vs contemporary
evolutionary processes. Evolution (in press)
36 Slate, J. et al. (1998) Bovine microsatellite loci are highly conserved in
red deer (Cervus elaphus), sika deer (Cervus nippon) and Soay sheep
(Ovis aries). Anim. Genet. 29, 307–315
Letters to TREE
Correspondence in TREE may address topics raised
in very recent issues of TREE, or (occasionally)
other matters of general current interest to ecologists and evolutionary biologists. Letters should be
no more than 500 words long with a maximum of
12 references and one small figure; original
results, new data or new models are not allowed.
Letters should be sent by e-mail to [email protected]. The decision to publish rests with the
Editor, and the author(s) of any TREE article criticized in a Letter will normally be invited to reply.
Full-length manuscripts in response to previous
TREE articles will not be considered.
203
CORRESPONDENCE
References
Table 1. The number and percentage of species in different vascular plant
groups on Barro Colorado Island, Panamaa
Plant group
Shade tolerant trees
Pioneer trees
Lianas (woody vines)
Shrubs
Forest herbs
Herbaceous vines
aData
No. of species
267
89
171
93
75
83
34
11
22
12
10
11
% of woody species
43
14
28
15
–
–
taken from Ref. 9.
of forest diversity11 might be premature. The
focus of previous research on the ability of
tree species to partition resources in gaps
might have caused us to overlook the
importance of gaps for many other groups of
vascular plants (Table 1). Future research is
necessary to quantify further the proportion of
species in these groups (and others, such as
epiphytes) that require gaps for persistence in
the community.
Stefan A. Schnitzer
Walter P. Carson
Dept of Biological Sciences,
University of Pittsburgh, Pittsburgh,
PA 15260, USA
([email protected]; [email protected])
References
1 Brokaw, N. and Busing, R.T. (2000) Niche
versus chance and tree diversity in forest gaps.
Trends Ecol. Evol. 15, 183–188
2 Schnitzer, S.A. and Carson, W.P. Treefall gaps
and the maintenance of species diversity in a
tropical forest. Ecology (in press)
3 Schnitzer, S.A. et al. (2000) The impact of lianas
on tree regeneration in tropical forest canopy
gaps: evidence for an alternative pathway of
gap-phase regeneration. J. Ecol. 88, 655–666
4 Gentry, A.H. and Dodson, C. (1987)
Contribution of nontrees to species richness
of a tropical rain forest. Biotropica
19, 149–156
5 Dirzo, R. et al. (1992) The effects of gap size
and age on the understorey herb community
of a tropical Mexican rain forest. J. Ecol.
80, 809–822
6 Goldblum, D. (1997) The effects of treefall gaps
on understory vegetation in New York State.
J. Veg. Sci. 8, 125–132
7 Levey, D.J. (1988) Tropical wet forest treefall
gaps and distributions of understory birds and
plants. Ecology 69, 1076–1089
8 Denslow, J.S. et al. (1990) Growth responses of
tropical shrubs to treefall gap environments.
Ecology 71, 165–179
9 Croat, T.B. (1978) The Flora of Barro Colorado
Island, Stanford University Press
10 Denslow, J.S. (1995) Disturbance and diversity
in tropical rain forests: the density effect.
Ecol. Appl. 5, 962–968
11 Hubbell, S.P. et al. (1999) Light gap
disturbances, recruitment limitation, and tree
diversity in a neotropical forest. Science
283, 554–557
376
% of species
1 Schnitzer, S.A. and Carson, W.P. (2000) Have
we forgotten the forest because of the trees?
Trends Ecol. Evol. 15, 375–376
2 Brokaw, N. and Busing, R.T. (2000) Niche
versus chance and tree diversity in forest gaps.
Trends Ecol. Evol. 15, 183–188
3 Fraver, S. et al. (1998) Delimiting the gap phase
in the growth cycle of a Panamanian forest.
J. Trop. Ecol. 14, 673–681
4 Jennions, M.D. (1997) Stability in coral
communities: a natural experiment.
Trends Ecol. Evol. 12, 3–4
Reply from N. Brokaw and
R.T. Busing
Microsatellite frequencies
in different taxa
Schnitzer and Carson’s valuable letter1 helps
define and extend the ideas in our recent
TREE Review2. They clarify that it is pioneer
tree species, rather than shade-tolerant
tree species, that account for the degree of
niche partitioning observed within treefall
gaps. This makes sense. Pioneer species
depend more on gap conditions than do
shade-tolerant species, in terms of
germination, establishment, growth and
survival. With all these behaviors more fully
responding to gap conditions, there is more
potential for gap partitioning among
pioneers. Also, the gap phase in the forest
growth cycle can be short3, and we would
expect species with rapid life cycles, such as
many pioneers, to be the species most closely
adapted to it.
The apparent gap partitioning among
pioneer species and among lianas1, as
opposed to shade-tolerant species,
suggests that the extent to which niche
versus chance controls community structure
can depend on life-history strategy and
growth form of the group considered. It will
also depend on such environmental features
as disturbance regime and seasonality, and
on the spatial and temporal scales
considered4.
Schnitzer and Carson1 point out that the
density effect is transient on a site-by-site
basis. By definition, once trees in a gap site
thin to background forest density the site’s
tree species richness should equal
background richness on an area basis.
However, over the forest as a whole, gaps are
continually appearing, thus the density effect
is permanent at that scale. The density effect
is simply a manifestation of the high
immigration rate to the seedling–sapling pool,
which is promoted by treefall disturbance and
helps sustain stand-scale tree diversity.
Nicholas Brokaw
A recent TREE Review by Sunnucks1 on
molecular markers for population biology took
for granted that, in principle, all markers could
be used for all taxa. He made the point that the
choice of markers depends primarily on the
goals of the study, discussing the availability
of markers in any particular taxon as a
practical consideration. However, it seems that
different taxa are not equivalent for the
different possible methods of investigation.
Microsatellites are five times less abundant in
the genomes of plants than in mammals2.
Furthermore, within a given class, abundance
and distribution of microsatellites vary greatly,
such as between Lepidoptera and
Hymenoptera. For example, there have only
been five studies published on Lepidoptera3–7,
whereas 47 were published on Hymenoptera
during the same period (1997–1999).
The bias towards certain taxa in
microsatellite studies seems to lie, at least in
part, in their frequency within the genome of
the respective study species, and in the
structure of the microsatellites and their
flanking regions. Both factors are probably
important to explain why about 75
microsatellite loci are available for population
studies for the honey-bee Apis mellifera 8,
although a similar study (in the same
laboratory and using the same method) on
Parnassius mnemosyne (the clouded Apollo)
yielded only three loci3. Apart from the
silkworm moth (Bombyx mori ), for which 15
microsatellites have been studied, no more
than four scorable microsatellite loci have
ever been used for population studies in
Lepidoptera species, in contrast with high
numbers found in the Hymenoptera (Table 1).
The more we know about the organization
of the genome in different organisms, the
more complex their differences are. This
should lead to different molecular markers
being appropriate in different taxa to tackle a
given biological question.
Manomet Center for Conservation
Sciences, PO Box 1770, Manomet,
MA 02345, USA ([email protected])
Gabriel Nève
Emese Meglécz
Richard T. Busing
Forestry Sciences Laboratory, 3200 SW
Jefferson Way, Corvallis, OR 97331, USA
([email protected])
0169-5347/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved.
Laboratoire de Systématique Evolutive,
Case 5, Université de Provence, 3 Place
Victor Hugo, F-13331 Marseille cedex 3,
France ([email protected];
[email protected])
TREE vol. 15, no. 9 September 2000
CORRESPONDENCE
Table 1. Number of scorable loci in various Hymenoptera
and Lepidoptera species
Species
No. of scorable loci
Refs
Hymenoptera
Vespula rufa
Apis mellifera
Bombus terrestris
47
75
26
9
8
8
3
4
2
4
15
3
4
5
6
7
Lepidoptera
Parnassius mnemosyne
Parnassius smitheus
Melitaea cinxia
Lymantria dispar
Bombyx mori
References
1 Sunnucks, P. (2000) Efficient genetic markers
for population biology. Trends Ecol. Evol.
15, 199–203
2 Lagercrantz, U. et al. (1993) The abundance of
various polymorphic microsatellite motifs
differs between plants and vertebrates.
Nucleic Acids Res. 21, 1111–1115
3 Meglécz, E. and Solignac, M. (1998)
Microsatellite loci for Parnassius mnemosyne
(Lepidoptera). Hereditas 128, 179–180
4 Keyghobadi, N. et al. (1999). Influence of
landscape on the population genetic structure
of the alpine butterfly Parnassius smintheus
(Papilionidae). Mol. Ecol. 8, 1481–1495
5 Palo, J. et al. (1995) Developing microsatellite
markers for insect population structure:
complex variation in a checkerspot butterfly.
Hereditas 123, 295–300
6 Bogdanowicz, S.M. et al. (1997) Microsatellite
DNA variation among Asian and North
American gypsy moths (Lepidoptera:
Lymantriidae). Ann. Entomol. Soc. Am.
90, 768–775
7 Reddy, K.D. et al. (1999) Microsatellites in the
silkworm, Bombyx mori: abundance,
polymorphism, and strain characterization.
Genome 42, 1057–1065
8 Estoup, A. et al. (1993) Characterization of
(GT)n and (CT)n microsatellites in two insect
species: Apis mellifera and Bombus terrestris.
Nucleic Acids Res. 21, 1427–1431
9 Thorén, P.A. et al. (1995) Unusually high
frequency of (CT)n and (GT)n microsatellite
loci in a yellowjacket wasp, Vespula rufa (L.)
(Hymenoptera: Vespidae). Insect Mol. Biol.
4, 141–148
microsatellites from the invertebrate Phylum
Onychophora2,3, when libraries from other
invertebrates and vertebrates cloned in
parallel were unproblematic. Similar
comments would apply to most other genetic
marker systems. Even the usually highly
reliable mtDNA can throw up surprises; for
example, its use in a restricted clade of
Sitobion aphids is severely limited by the
presence of multiple nuclear copies4.
The undoubted taxonomical variation in
density, physical properties, mutation and
evolutionary behaviour of microsatellites are
important topics of research5. Accordingly, I
highlighted increased knowledge of molecular
evolution as one of the most important
research areas in molecular population
genetics. However, none of this undermines
my main premise that single locus
codominant markers capable of yielding allele
phylogenies are worthy of the bulk of research
effort, because they provide connectible data
that inform us about ecology and evolution at
a variety of levels in the hierarchy of life. Such
markers and the data they generate
(particularly as compared with multilocus
Dept of Genetics, La Trobe University,
Bundoora, Melbourne, VIC 3083, Australia
([email protected])
References
1 Nève, G. and Meglécz, E. (2000) Microsatellite
frequencies in different taxa. Trends Ecol. Evol.
15, 376–377
2 Sunnucks, P. and Wilson, A.C.C. (1999)
Microsatellite markers for the onychophoran
Euperipatoides rowelli. Mol. Ecol. 8, 899–900
3 Curach, N. and Sunnucks, P. (1999) ‘Molecular
anatomy’ of an onychophoran:
compartmentalized sperm storage and
heterogeneous paternity. Mol. Ecol. 8, 1375–1386
4 Sunnucks, P. and Hales, D.F. (1996) Numerous
transposed sequences of mitochondrial
cytochrome oxidase I–II in aphids of the genus
Sitobion (Hemiptera: Aphididae). Mol. Biol. Evol.
13, 510–524
5 Goldstein, D.B. and Schlötterer, C., eds (1999)
Microsatellites: Evolution and Applications,
Oxford University Press
• You will have advance, full-text access to the entire contents of each issue of TREE
•
•
Reply from P. Sunnucks
•
Meglécz1
TREE vol. 15, no. 9 September 2000
Paul Sunnucks
Why TREE readers should access BioMedNet
•
Nève and
are quite right that
performance and availability of genetic
markers might vary greatly among taxonomic
groups, even over and above strong biases in
research effort and technical factors. In
particular, they highlighted the taxonomically
heterogeneous nature of microsatellites.
Anyone who has cloned microsatellites from
even a modest diversity of species is likely to
have encountered one sort of challenge or
another. Indeed, my own research has run
into unusual difficulties in obtaining usable
dominant markers) are of ongoing, as well as
current, use in population biology and
molecular evolution.
When certain taxa or questions throw up
technical barriers, we need to be imaginative
and resourceful in finding markers that
fulfil the most important criterion – generating
the type of data we have decided we need. If
microsatellites are rare in a genome, we might
decide to work harder at obtaining them: Nève
and Meglécz point out that microsatellites are
five times less abundant in the genomes of
plants than mammals, but that still leaves
plenty of loci for everything, except perhaps
gene mapping. We can also investigate the
availability of other single locus, codominant,
phylogeny-yielding markers, such as
codominant AFLPs, SNPs or rDNA spacers.
•
•
within the BioMedNet Journal Collection. Access to the Journal Collection is free
until 1 September 2000.
You will be able to access an archive of past issues of TREE, backdated to
January 1998.
You will be able to access news articles on evolution and ecology from all Trends and
Current Opinion journals, including Trends in Genetics, Current Opinions in Genetics
and Development, Trends in Microbiology and Parasitology Today.
With an institutional subscription to BioMedNet Reviews, you gain access to over
5000 life science review articles in an expanding database. Free trial access to
BioMedNet Reviews is available until 1 September 2000.
You can create your own ‘virtual journals’ within BioMedNet Reviews, tailored exactly to
your research needs, using a mix of subject areas and key words to select content from
the database.
You can set up personalised e-mail alerts within Reviews to stay informed whenever
new, relevant content is posted online.
You will get up-to-date coverage on a wide variety of life-science issues and
information, including policy, funding, on-the-spot conference reports and daily
commentaries in the News & Comment section.
BioMedNet is continually evolving in response to your feedback. Registering with
BioMedNet is FREE! Visit BioMedNet today at www.bmn.com and use the feedback
form to tell us what you think!
0169-5347/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved.
377