Genotyping of B. licheniformisbased on a novel multi-locus sequence typing (MLST) scheme
© Madslien et al.; licensee BioMed Central Ltd. 2012
Received: 22 February 2012
Accepted: 14 September 2012
Published: 10 October 2012
Bacillus licheniformis has for many years been used in the industrial production of enzymes, antibiotics and detergents. However, as a producer of dormant heat-resistant endospores B. licheniformis might contaminate semi-preserved foods. The aim of this study was to establish a robust and novel genotyping scheme for B. licheniformis in order to reveal the evolutionary history of 53 strains of this species. Furthermore, the genotyping scheme was also investigated for its use to detect food-contaminating strains.
A multi-locus sequence typing (MLST) scheme, based on the sequence of six house-keeping genes (adk, ccpA, recF, rpoB, spo0A and sucC) of 53 B. licheniformis strains from different sources was established. The result of the MLST analysis supported previous findings of two different subgroups (lineages) within this species, named “A” and “B” Statistical analysis of the MLST data indicated a higher rate of recombination within group “A”. Food isolates were widely dispersed in the MLST tree and could not be distinguished from the other strains. However, the food contaminating strain B. licheniformis NVH1032, represented by a unique sequence type (ST8), was distantly related to all other strains.
In this study, a novel and robust genotyping scheme for B. licheniformis was established, separating the species into two subgroups. This scheme could be used for further studies of evolution and population genetics in B. licheniformis.
Bacillus licheniformis is a Gram positive, thermophilic spore forming soil bacterium closely related to B. subtilis. It is widely used in the fermentation industry for production of enzymes, antibiotics and other chemicals and is generally regarded as a non-pathogen [1, 2]. However, there are several reports of B. licheniformis- associated human infections such as bacteremia and enocarditis, bovine abortions and food borne diseases which raise the question of its pathogenic potential [3–9]. More commonly, representatives of this species have caused spoilage of milk, bread and canned foods leading to severe economic losses to the food industry [10–13].
B. licheniformis is ubiquitous in the environment and able to grow under a wide range of temperatures (15–55°C) in both anaerobic and aerobic conditions making this species a highly potent food contaminant [14–16]. During starvation, the cells may form thermo-stabile endospores in a process known as sporulation . These spores are resistant against many decontamination and preservation steps applied by the food industry such as pasteurization, pressure, freezing, extreme pH, radiation and desiccation . In the presence of nutrients (germinants) spores may germinate and grow out into vegetative cells which can multiply in the absence of competing microflora [18, 19]. Germination can be further accelerated by external stress such as a short, sublethal heat step (usually at 65–95°C) [20–22]. This phenomenon, known as “activation”, is utilized in the “double heat treatment” (a modified tyndallisation), a decontamination strategy where spores that are activated in the primary heat step can be inactivated or killed as germs in the secondary heat treatment . Recent publications have provided new insight into the complexity of spore germination [20, 24, 25]. The observed diversity in germination between and within populations makes spore behavior prediction challenging  and might explain why spore decontamination strategies sometimes fail. Detecting strains with increased potential of causing food spoilage would therefore be of great value to the food industry.
Several molecular typing methods have been applied in order to characterize the population structure within B. licheniformis[27–30]. Multi-locus sequence typing (MLST) has the advantage to other molecular typing methods of being unambiguous and easily portable between laboratories . It has been applied to numerous species including members of the B. cereus family and Clostridium spp. [32–36] and has been used for epidemiological purposes identifying strains that could cause human infections [37, 38]. Basically, it relies on the sequence of several (usually six to eight) conserved house-keeping genes which are independently distributed in the genome. The method is therefore considered to be robust, discriminatory and capable of revealing the deeper evolutionary relation of populations that are studied [39, 40]. No MLST scheme has so far been developed for B. licheniformis.
The purpose of this study was to establish a MLST scheme for B. licheniformis in order to reveal the evolutionary relationship of 53 strains of this species and to see whether food-contaminating strains were restricted to certain lineages.
MLST analysis of B. licheniformis
Bacterial growth and biochemical identification
All strains were stored at −70°C, plated on sheep blood agar (Columbia blood agar, Oxoid, UK) and grown at 30°C overnight. Biochemical characterization was performed on pure cultures by using API 50 CH cassettes (bioMÃ©rieux, Marcy l’Etoile, France) according to the instructions given by the manufacturer . Color changes were examined after 24 and 48 h at 30°C and compared to the Bacillus identification profile database, API Lab1 (version 4.0). The reaction profiles of these tests were compared with the ApiwebTM database provided by the manufacturer.
Bacteria were grown on sheep blood agar at 30°C overnight. Single colony material was inoculated in 20 ml Luria broth (LB). The bacterial culture was grown overnight at 30°C and centrifuged at 3000 × g for 10 min. The supernatant was discarded and the pellet resuspended in 1 ml enzymatic lysis buffer (20 mM Tris·Cl, pH 8.0, 20 mM Tris·Cl, pH 8.0, 1.2% Triton® X-100, 20 mg/ml lysozyme). Further DNA extraction was performed according to the protocol provided by DNeasy Blood and Tissue Kit (Qiagen, USA). The final DNA concentration ranged from 8–72 ng/ul with a mean 260/280 absorbance ratio of 1, 89 (Nanodrop ND-1000 Spectrophotometer, Thermo Fisher Scientific, USA).
Amplicon size (bp)
Target gene (s)
Target (from-to bp)in NC_00627
GGT AAA GGG ACA CAG GCT GA
BL01030: adk ;adenylate kinase
TCG AGT AAA GGC TGG GTT TG
TAT GAT GTA GCA CGC GAA GC
BL00444: ccpA; transcriptional regulator (compl)
TAT CCC CAA GCG CTC TTT TA
ACG GTT CTG TTC CCA TTC AG
BL00079: recF; recombination protein F
CAT CAC GGC CAT TGA CAT AG
GGG TCC CGA CGG CCA ACA AA
BL01285: sucC; succinyl-CoA synthetase, subunit beta
GGC CGG TTC CCC TCC GTA GT
AGG TCA ACT AGT TCA GTA TGG ACG
BL02798: rpoB; DNA-directed RNA polymerase, subunit beta
AAG AAC CGT AAC CGG CAA CTT
GAA GTG CTT GGT GTC GCA TA
BL01518: spo0A; response regulator
TGT GTA GCC GAA AAG TGA CG
AAA TCA AAG CGG TTT TCC TG
BL02280: pyrE; orotate phosphoribosyltransferase
AGG ATC CGC TTT CCA TTC TT
CTT ACG GGC TGA GCA AGT TC
BL00185: glpT; glycerol-3-phosphate permease
CAC GAA AAT GTT GGC AAG TG
ATC GTT GAG GGT GAC TCT GC
BL00081:gyrB DNA gyrase subunit B
AAA TTT CTT CGA GCT GCT GGT
Real-time PCR and sequencing
The nine primer sets were applied on a subset of 20 strains to see which combination of loci that gave the highest level of discrimination and still being congruent (visual evaluation). The amplification reactions were performed in 20 μl using 2 μl DNA extract (approximately 20 ng of DNA) as a template. Real-time PCR reactions were performed in a LightCycler® 480 System using LightCycler® 480 SYBR Green I Master (Roche Diagnostics GmbH, Germany) according to recommendations given by the manufacturer of the kit. The temperature program was as follows: 5 min initial denaturation at 95°C followed by 35 cycles of denaturation at 95°C for 10 s, annealing at 56°C for 10 s and primer extension at 72°C for 30 s. The amplifications were terminated after a final elongation step of 7 min at 72°C. The PCR fragments were verified by electrophoresis using Bioanalyzer (Agilent Technologies, USA). PCR products were purified and sequenced by Eurofins MWG Operon (Ebersberg, Germany) using the dideoxy chain termination method on a ABI 3730XL sequencing instrument (Applied Biosystems, USA).
The Staden Package  was used for alignment, editation and construction of consensus sequences based on the ABI sequence chromatograms. Consensus sequences were entered into the MEGA4  software and aligned by CLUSTALW . Sequences were trimmed to be in frame and encode an exact number of amino acids. Dendograms for each locus (Additional file 1) were constructed in MEGA4 using the Neighbor-Joining method (NJ) with branch lengths estimated by the Maximum Composite Likelihood method [45, 47]. Branch quality was assessed by the bootstrap test using 500 replicates. A subset of six loci including adk, ccpA, recF, sucC, rpoB and spo0A, which gave the highest tree resolution and still being congruent (visual evaluation, Additional file 1), was selected for the final MLST scheme (highlighted in Table 1). The trimmed sequences were entered into BioNumerics software v. 6.6, (Applied Maths NV) as fasta files and used to generate allelic profiles for each isolate based on the six loci. Each unique allelic profile defined a sequence type (ST). A cluster analysis was performed using the allelic profiles as categorical coefficients and a dendogram was constructed based on the UPGMA method. The tree branch quality was estimated by calculating the cophenetic correlation coefficients. Sequence analysis was performed using the START2 software package  where the number of nucleotide differences and ratio of nonsynonymous to synonymous substitutions (dN /dS ) were calculated. MEGA5 was used to construct a phylogenetic tree based on the concatenated sequences (adk;ccpA;recF;rpoB;spo0A;sucC) by the NJ-method with branch lengths estimated by the Maximum Composite Likelihood method [47, 49]. Minimum spanning tree (MST) was generated in BioNumerics v.6.6 (Applied Maths NV) using the categorical coefficient.
Index of associaton (IA)
To test the null hypothesis of linkage equilibrium (alleles are independent) between the alleles of the six MSLT loci, IA values were calculated in START2 by the classical (Maynard Smith) and the standardized (Haubold) method . The test was repeated on a dataset containing only one isolate per ST in order to avoid the risk of a bias toward a clonal population for strains with the same epidemiological history (e.g. the abortifacient strains) .
Results and discussion
Characteristics of B. licheniformis MLST loci
Length of sequenced fragment (bp)
No. of variable sites
% of variable sites
Mean % GcpC
In cases were recombination is rare it is generally recommended to concatenate the sequences before calculating dendograms . This concatenated dendogram corresponded well with the allel-based dendogram and is presented in Additional file 3. A small difference between the allel-based and the concatenated dendogram was observed. NVH1032 (ST8) was positioned slightly closer to group A isolates in the latter. When examining individual loci, NVH1032 (ST8) clustered together with group A for all loci apart from adk. It is therefore reasonable to assume that NVH1032 (ST8) could be regarded as a group A member. However, none of the MLST allels of NVH1032 was shared by any other strains in our collection (Additional file 2) underpinning the genetic distinction of NVH1032 (ST8) from the other strains.
A robust and portable typing scheme for B. licheniformis was established. This method, based on six house-keeping genes separated the species into two distinct lineages. These two lineages seem to have evolved differently. The food spoilage strain NVH1032 was distantly related to all other strains evaluated. The MLST scheme developed in the present study could be used for further studying of evolution and population genetics of B. licheniformis.
We thank Ingjerd Thrane for valuable technical assistance in order to complete this work. The work was supported by grants from the Norwegian Research Council (grant 178299/I10) and the Norwegian Defence Research Establishment (FFI).
- Boer AS, Priest F, Diderichsen B: On the industrial use of Bacillus licheniformis: a review. Appl Microbiol Biotechnol. 1994, 40: 595-598.View ArticleGoogle Scholar
- Eveleigh DE: The microbiological production of industrial chemicals. Sci Am. 1981, 245: 120-130.View ArticleGoogle Scholar
- Salkinoja-Salonen MS, Vuorio R, Andersson MA, Kampfer P, Andersson MC, Honkanen-Buzalski T, Scoging AC: Toxigenic strains of Bacillus licheniformis related to food poisoning. Appl Environ Microbiol. 1999, 65: 4637-4645.PubMedPubMed CentralGoogle Scholar
- Agerholm JS, Krogh HV, Jensen HE: A retrospective study of bovine abortions associated with Bacillus licheniformis. J Vet Med, Series B. 1995, 42: 225-234.View ArticleGoogle Scholar
- Blue SR, Singh VR, Saubolle MA: Bacillus licheniformis bacteremia: five cases associated with indwelling central venous catheters. Clin Infect Dis. 1995, 20: 629-PubMedView ArticleGoogle Scholar
- Santini F, Borghetti V, Amalfitano G, Mazzucco A: Bacillus licheniformis prosthetic aortic valve endocarditis. J Clin Microbiol. 1995, 33: 3070-3073.PubMedPubMed CentralGoogle Scholar
- Sugar AM, McCloskey RV: Bacillus licheniformis sepsis. JAMA. 1977, 238: 1180-PubMedView ArticleGoogle Scholar
- Tabbara KF, Tarabay N: Bacillus licheniformis corneal ulcer. Am J Ophthalmol. 1979, 87: 717-719.PubMedView ArticleGoogle Scholar
- Haydushka I, Markova N, Vesselina K, Atanassova M: Recurrent sepsis due to Bacillus licheniformis. J Global Infectious Dis. 2012, 4: 82-83.View ArticleGoogle Scholar
- Thompson JM, Dodd CER, Waites WM: Spoilage of bread by Bacillus. Int Biodeter Biodegr. 1993, 32: 55-66.View ArticleGoogle Scholar
- Pirttijärvi TSM, Graeffe TH, Salkinoja-Salonen MS: Bacterial contaminants in liquid packaging boards: assessment of potential for food spoilage. J Appl Microbiol. 1996, 81: 445-458.View ArticleGoogle Scholar
- Heyndrickx M, Scheldeman P: Bacilli Associated with Spoilage in Dairy Products and Other Food. Applications and systematics of Bacillus and relatives. Edited by: Berkeley R. 2002, Oxford: Blackwell, 64-82.View ArticleGoogle Scholar
- Sorokulova IB, Reva ON, Smirnov VV, Pinchuk IV, Lapa SV, Urdaci MC: Genetic diversity and involvement in bread spoilage of Bacillus strains isolated from flour and ropy bread. Lett Appl Microbiol. 2003, 37: 169-173.PubMedView ArticleGoogle Scholar
- Shariati P, Mitchell WJ, Boyd A, Priest FG: Anaerobic metabolism in Bacillus licheniformis NCIB 6346. Microbiology. 1995, 141: 1117-1124.View ArticleGoogle Scholar
- Clements LD, Miller BS, Streips UN: Comparative growth analysis of the facultative anaerobes Bacillus subtilis, Bacillus licheniformis, and Escherichia coli. Syst Appl Microbiol. 2002, 25: 284-286.PubMedView ArticleGoogle Scholar
- Fields ML, Zamora AF, Bradsher M: Microbial analysis of home-canned tomatoes and green beans. J Food Sci. 1977, 42: 931-934.View ArticleGoogle Scholar
- Piggot PJ, Hilbert DW: Sporulation of Bacillus subtilis. Curr Opin Microbiol. 2004, 7: 579-586.PubMedView ArticleGoogle Scholar
- Setlow P: Spores of Bacillus subtilis: their resistance to and killing by radiation, heat and chemicals. J Appl Microbiol. 2006, 101: 514-525.PubMedView ArticleGoogle Scholar
- Moir A: How do spores germinate?. J Appl Microbiol. 2006, 101: 526-530.PubMedView ArticleGoogle Scholar
- Lovdal I: PhD thesis. Germination of Bacillus species related to food spoilage and safety. 2012Google Scholar
- Paidhungat M, Setlow P: Spore germination and outgrowth. Bacillus subtilis and its closest relatives: from genes to cells. Edited by: Sonenshein AL, Hoch JA, Losick R. 2002, Washington, D.C: ASM, 537-548.View ArticleGoogle Scholar
- Keynan A, Evenchik Z: Activation. The bacterial spore. Edited by: Gould GW, Hurst A. 1969, New York: Academic Press, 359-396.Google Scholar
- Brown JV, Wiles R, Prentice G: The effect of a modified tyndallization process upon the spore forming bacteria of milk and cream. J Soc Dairy Technol. 1979, 32: 109-112.View ArticleGoogle Scholar
- Hornstra LM, Ter Beek A, Smelt JP, Kallemeijn WW, Brul S: On the origin of heterogenity in (preservation) resistance of Bacillus spores: input for a ‘systems’ analysis approach of bacterial spore outgrowth. Int J Food Microbiol. 2009, 134: 9-15.PubMedView ArticleGoogle Scholar
- Ghosh S, Setlow P: Isolation and characterization of superdormant spores of Bacillus species. J Bacteriol. 2008, 191: 1787-1797.View ArticleGoogle Scholar
- Brul S, van Beilen J, Caspers M, O Brien A, de Koster C, Oomes S, Smelt J, Kort R, Ter Beek A: Challenges and advances in systems biology analysis of Bacillus spore physiology; molecular differences between an extreme heat resistant spore forming Bacillus subtilis food isolate and a laboratory strain. Food Microbiol. 2011, 28: 221-227.PubMedView ArticleGoogle Scholar
- Duncan KE, Ferguson N, Kimura K, Zhou X, Istock CA: Fine-scale genetic and phenotypic structure in natural populations of Bacillus subtilis and Bacillus licheniformis. Implications for bacterial evolution and speciation. Evolution. 1994, 48: 2002-2025.View ArticleGoogle Scholar
- De Clerck E, De Vos P: Genotypic diversity among Bacillus licheniformis strains from various sources. FEMS Microbiol Lett. 2004, 231: 91-98.PubMedView ArticleGoogle Scholar
- Palmisano MM, Nakamura LK, Duncan KE, Istock CA, Cohan FM: Bacillus sonorensis sp nov., a close relative of Bacillus licheniformis, isolated from soil in the sonoran desert, arizona. Int J Syst Evol Microbiol. 2001, 51: 1671-1679.PubMedView ArticleGoogle Scholar
- Daffonchio D, Borin S, Frova G, Manachini PL, Sorlini C: PCR fingerprinting of whole genomes: the spacers between the 16S and 23S rRNA genes and of intergenic tRNA gene regions reveal a different intraspecific genomic variability of Bacillus cereus and Bacillus licheniformis. Int J Syst Bacteriol. 1998, 48: 107-116.PubMedView ArticleGoogle Scholar
- Maiden MCJ, Bygraves JA, Feil E, Morelli G, Russell JE, Urwin R, Zhang Q, Zhou J, Zurth K, Caugant DA: Multilocus sequence typing: a portable approach to the identification of clones within populations of pathogenic microorganisms. Proc Natl Acad Sci USA. 1998, 95: 3140-3145.PubMedPubMed CentralView ArticleGoogle Scholar
- Helgason E, Tourasse NJ, Meisal R, Caugant DA, Kolsto AB: Multilocus sequence typing scheme for bacteria of the Bacillus cereus group. Appl Environ Microbiol. 2004, 70: 191-201.PubMedPubMed CentralView ArticleGoogle Scholar
- Jost BH, Trinh HT, Songer JG: Clonal relationships among Clostridium perfringens of porcine origin as determined by multilocus sequence typing. Vet Microbiol. 2006, 116: 158-165.PubMedView ArticleGoogle Scholar
- Lemee L, Bourgeois I, Ruffin E, Collignon A, Lemeland JF, Pons JL: Multilocus sequence analysis and comparative evolution of virulence-associated genes and housekeeping genes of Clostridium difficile. Microbiology-Sgm. 2005, 151: 3171-3180.View ArticleGoogle Scholar
- Neumann AP, Rehberger TG: MLST analysis reveals a highly conserved core genome among poultry isolates of Clostridium septicum. Anaerobe. 2009, 15: 99-106.PubMedView ArticleGoogle Scholar
- Olsen JS, Skogan G, Fykse EM, Rawlinson EL, Tomaso H, Granum PE, Blatny JM: Genetic distribution of 295 Bacillus cereus group members based on adk screening in combination with MLST (Multilocus Sequence Typing) used for validating a primer targeting a chromosomal locus in B.anthracis. J Microbiol Methods. 2007, 71: 265-274.PubMedView ArticleGoogle Scholar
- Urwin R, Maiden MCJ: Multi-locus sequence typing: a tool for global epidemiology. Trends Microbiol. 2003, 11: 479-487.PubMedView ArticleGoogle Scholar
- Sullivan CB, Diggle MA, Clarke SC: Multilocus sequence typing - data analysis in clinical microbiology and public health. Mol Biotechnol. 2005, 29: 245-254.PubMedView ArticleGoogle Scholar
- Coffey TJ, Pullinger GD, Urwin R, Jolley KA, Wilson SM, Maiden MC, Leigh JA: First insights into the evolution of streptococcus uberis: a multilocus sequence typing scheme that enables investigation of its population biology. Appl Environ Microbiol. 2006, 72: 1420-1428.PubMedPubMed CentralView ArticleGoogle Scholar
- Feil EJ, Cooper JE, Grundmann H, Robinson DA, Enright MC, Berendt T, Peacock SJ, Smith JM, Murphy M, Spratt BG, et al: How clonal is Staphylococcus aureus?. J Bacteriol. 2003, 185: 3307-3316.PubMedPubMed CentralView ArticleGoogle Scholar
- Logan NA, Berkeley RCW: Identification of Bacillus strains using the API system. J Gen Microbiol. 1984, 130: 1871-1882.PubMedGoogle Scholar
- Maiden MCJ: Multilocus sequence typing of bacteria. Annu Rev Microbiol. 2006, 60: 588-Google Scholar
- Rozen S, Skaletsky H: Primer3 on the WWW for general users and for biologis programmers. Methods Mol Biol. 2000, 132: 365-386.PubMedGoogle Scholar
- Staden R: The Staden sequence analysis package. Mol Biotechnol. 1996, 5: 233-241.PubMedView ArticleGoogle Scholar
- Tamura K, Dudley J, Nei M, Kumar S: MEGA4: molecular evolutionary genetics analysis (MEGA) software version 4.0. Mol Biol Evol. 2007, 24: 1596-1599.PubMedView ArticleGoogle Scholar
- Higgins D, Thompson J, Gibson T: CLUSTALW: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 1994, 22: 4673-4680.PubMedPubMed CentralView ArticleGoogle Scholar
- Tamura K, Noi M, Kumar S: Prospects for inferring very large phylogenies by using the neighbour-joining method. Proc Natl Acad Sci. 2004, 101: 11030-11035.PubMedPubMed CentralView ArticleGoogle Scholar
- Jolley KA, Feil EJ, Chan MS, Maiden MC: Sequence type analysis and recombinational tests (START). Bioinformatics. 2001, 17: 1230-1231.PubMedView ArticleGoogle Scholar
- Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S: MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol Biol Evol. 2011, 28: 2731-2739.PubMedPubMed CentralView ArticleGoogle Scholar
- Kotetishvili M, Kreger A, Auters G, Orris JG, Ulakvelidze A, Tine OC: Multilocus sequence typing for studying genetic relationships among Yersinia species. J Clin Microbiol. 2005, 43: 2674-2684.PubMedPubMed CentralView ArticleGoogle Scholar
- Lovdal IS, Hovda MB, Granum PE, Rosnes JT: Promoting Bacillus cereus spore germination for subsequent inactivation by mild heat treatment. J Food Prot. 2011, 74: 2079-2089.PubMedView ArticleGoogle Scholar
- Ghosh S, Zhang P, Li Y, Setlow P: Superdormant spores of Bacillus species have elevated wet-heat resistance and temperature requirements for heat activation. J Bacteriol. 2009, 191: 5584-5591.PubMedPubMed CentralView ArticleGoogle Scholar
- Hoffmann K, Wollherr A, Larsen M, Rachinger M, Liesegang H, Ehrenreich A, Meinhardt F: Facilitation of direct conditional knockout of essential genes in Bacillus licheniformis DSM13 by comparative genetic analysis and manipulation of genetic competence. Appl Environ Microbiol. 2010, 76: 5046-5057.PubMedPubMed CentralView ArticleGoogle Scholar
- Waschkau B, Waldeck J, Wieland S, Eichstädt R, Meinhardt F: Generation of readily transformable Bacillus licheniformis mutants. Appl Microbiol Biotechnol. 2008, 78: 181-188.PubMedView ArticleGoogle Scholar
- Thorne CB, Stull HB: Factors affecting transformation in Bacillus licheniformis. J Bacteriol. 1966, 91: 1012-1020.PubMedPubMed CentralGoogle Scholar
- Maiden MC: Multilocus sequence typing of bacteria. Annu Rev Microbiol. 2006, 60: 561-588.PubMedView ArticleGoogle Scholar
- Saitou N, Nei M: The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol. 1987, 4: 406-425.PubMedGoogle Scholar
- Felsenstein J: Confidence limits on phylogenies: an approach using the bootstrap. Evolution. 1985, 39: 783-791.View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.