(Acari: Ixodidae) using rDNA - Semantic Scholar

Proc. Natl. Acad. Sci. USA Vol. 90, pp. 10221-10225, November 1993 Evolution

Investigation of the validity of species status of Ixodes dammini (Acari: Ixodidae) using rDNA DAWN M. WESSON*, DENSON KELLY MCLAINt, JAMES H. OLIVERt, JOSEPH PIESMANt, AND FRANK H. COLLINS*§ *Malaria Branch, Division of Parasitic Diseases, Mailstop F12, Centers for Disease Control, Atlanta, GA 30333; tDepartment of Biology, Georgia Southern University, Statesboro, GA 30460; and tMedical Entomology/Ecology Branch, Division of Vector-Borne Infectious Diseases, Centers for Disease Control, Ft. Collins, CO 80522

Communicated by George B. Craig, Jr., April 12, 1993

California and Arizona), L. scapularis (from Georgia and North Carolina), and . dammini (from Maryland, Massachusetts, New Jersey, New York, and Wisconsin); our goal is to further test the status of . dammini as a species distinct from I. scapularis.¶

ABSTRACT The two internal transcribed spacers (ITS1 and ITS2) of rDNA of three members of the Ixodes ricinus "complex" (Acari: Ixodidae) were sequenced. Sequence variation was assessed for the North American species 1. scapularis, 1. dammini, and 1. pacificus at three levels: within individual/population, between individuals of different geographic origin within a species, and between species. Both spacers are highly variable, particularly with regard to small deletions and additions which may arise via replication slippage. Homogenization of rDNA multigene arrays for particular sequence variants appears to occur at a relatively rapid rate, since I. pacificus sequences differ from the others at numerous invariant sites, facilitating the use of these sequences to assess sibling species relationships. Based on maximum parsimony and two distance methods (unweighted pair-group with arithmetic means and neighbor-joining), sequence variation in ITS1 and ITS2 suggests that 1. scapularis and 1. dammini are not distinct species and that even individuals from geographically isolated locations are very simila. Individuals from geographically separated populations of 1. pacificus appear to be relatively less closely related to each other but distinct from those of I. scapularis/dammini. In 1. scapularis/dammini, diversity within and between individuals from geographic populations contributed equally to total sequence diversity.

The status of Ixodes scapularis and Ixodes dammini as separate species (1) is a question of medical and biological importance. Like some other members of the Ixodes ricinus "complex" (2), both species are capable of transmitting Lyme disease spirochetes (3-5). The description of I. dammini as a (new) distinct species (1) raises questions as to whether or not the two species are equally competent as vectors of the spirochete that causes Lyme disease, especially because Lyme disease appears to be more prevalent in the northeastern and north central United States, the range of I. dammini, than in the southeastern United States, where I. scapularis occurs (6). Of biological interest are questions of gene flow and niche divergence in zones of overlap or range contiguity between sibling or incipient species and the genetic relationships between peripheral and central populations as !. dammini expands its range (7). The morphological basis for separating !. scapularis and !. dammini has not always provided for consistent species identification (8), and has recently been challenged. Indeed, Oliver et al. (9) synonomized !. dammini with !. scapularis based on hybridization/assortative mating experiments, morphometric analyses, and other studies. In the present study, we compare the nucleotide sequence of the internal transcribed spacer (ITS) of PCR-amplified rDNA of three members of the . ricinus "complex": Ixodes pacificus (from

MATERIALS AND METHODS Tick Sources. Eggs of . scapularis from Georgia, . dammini from Massachusetts, and . pacificus from California were obtained from females in laboratory colonies maintained at Georgia Southern University. Nymphs of . scapularis from North Carolina and of . dammini from Maryland, Massachusetts, New Jersey, New York, and Wisconsin were provided by the Division of Vector-Borne Infectious Diseases, Centers for Disease Control and Prevention. These specimens originated from single females from each of the listed geographic locations. l. pacificus adults from Arizona were collected in the spring of 1991 in the Hualapai Mountains, southwest of the town of Kingman (10). Sample Preparation. Genomic DNA extraction (11), PCR of the ITS region with 18S and 28S primers, cloning, and sequencing procedures have been noted elsewhere (12). DNA was extracted either from single frozen nymphs or adults or from a mass of eggs (probably of mixed parentage). [Genomic DNA extracted from an egg mass should contain population level variation, since the average individual carries most of the population-wide rDNA diversity (13).] For PCR of the ITS1 region only, a primer was chosen in the 5.8S region based on sequence conserved between Drosophila melanogaster (14) and Anopheles gambiae (S. M. Paskewitz and F.H.C., unpublished data). The 5.8S primer, 5'GTGAATTCTTGCTGCGTTCTTCATCAC-3', was used in conjunction with the 18S primer used for PCR of the entire ITS region. Since a HindIll site was used as a cloning restriction site in the 5' end, the presence of an internal HindIII site in ITS1 interfered with the ability to clone that entire =550-bp segment. ITS1 was sequenced from 25 different clones (17 of which also contained ITS2): of 8 l. scapularis clones, 7 were from mass egg DNA (Georgia) and 1 was from a nymph (North Carolina); of 10!. dammini clones, 6 were from mass egg DNA (Maryland) and 4 were from single nymphs from four colonies differing in geographic origin (Massachusetts, New Jersey, New York, and Wisconsin); of 7 . pacificus clones, 6 were from mass egg DNA (California) and 1 was from an adult (Arizona). The 17 ITS2 sequences included the following: 5 . scapularis clones, 4 from mass egg DNA (Georgia) and 1 from a nymph (North Carolina); 8 . dammini clones, 3 from mass egg DNA (Maryland) and 5 from single nymphs Abbreviation: ITS, internal transcribed spacer. §To whom reprint requests should be addressed. $The sequences reported in this paper have been deposited in the GenBank database (accession nos. X63868 and L22265-L22289).

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

10221

10222

Proc. Natl. Acad. Sci. USA 90 (1993)

Evolution: Wesson et al.

from five colonies differing in geographic origin (Maryland, Massachusetts, New Jersey, New York, and Wisconsin); 4 I. pacificus clones, 3 from mass egg DNA (California) and 1 from an adult (Arizona). Phylogenetic Analysis of ITS Sequence. Sequences were aligned manually and the data were analyzed with programs in the PHYLIP package (J. Felsenstein, version 3.4, 1991). Relationships between populations and species were assessed by maximum parsimony, by the unweighted pairgroup method with arithmetic means (UPGMA), and by the neighbor-joining method. Distances were calculated by the Kimura two-parameter method (15), and input order of species and sequences was randomized. While various weighting schemes are possible, equal weight was given to each site within a sequence, the transition/transversion ratio was 2, and individual sites with deletions/additions were weighted like transitions. Statistical Analyses. Variation between DNA sequences, indexed by the magnitude of the distance measure, was partitioned into components arising from variation within a population, between populations, and between species by using C statistics (16). If the mean distance between species, populations, and sequences within a population are dmean.spp, dmean.pop, and dmean.in, then the proportion of sequence variation attributable to (i) variation between species is the mean of Cspp = (dspp - dmean.pop)/dmean.spp, (ii) variation between populations is the mean of Cpop = (dpop - dmean.in)/dmean.spp, and (iii) variation within a population is the mean of Cin = din/dmean.spp. Pairwise distance measures, dpp, dpop, and din, reflect sequence variation between, respectively, species, populations, and sequences within a population. Because the number of pairs increases with the number of sequences (n) according to (n2 - n)/2 and because each sequence will be represented in multiple pairwise comparisons when n > 2, degrees of freedom can be inflated and the independence of data points compromised when n > 2. Therefore, distances were measured between a sequence and a consensus sequence based on the other sequences at the appropriate level. For instance, dmean.pop is (di.con + d2.con + . . . + d,.c.n)/n, where di.con is the distance between population i and the consensus sequence. Likewise, distances between species are based on distances between their consensus sequences. This approach controls for inflated degrees of freedom and for nonindependence of data points. Also, consensus sequences at higher levels are based upon consensus sequences at lower levels, so that variation in the degree of sampling at lower levels is not reflected in relative contribution to the consensus sequence at a higher level. S: D: P:

A consensus sequence was derived by aligning the sequences either for a given species from a geographic region or laboratory colony or across the species. From the conspecific alignment, the most common nucleotide at a variable position was used in the consensus. For interspecific comparisons using the consensus sequences, as with other alignments, gaps were inserted where necessary to align the sequences.

RESULTS ITS Sequences. The I. dammini, I. pacificus, and I. scapularis consensus sequences for ITS1 (Fig. 1) and ITS2 (Fig. 2) exhibited much more interspecific variation than did the intervening 5.8S rRNA gene. The 5.8S region was virtually identical among the three species; the I. scapularis 5.8S sequence and secondary structure are presented elsewhere (17). The ITS1 consensus sequences for I. dammini and I. scapularis differed at 5 of 353 sites (for ITS1, length range was 343-357 bp for I. scapularis and 342-354 bp for I. dammini), all S of which were variable within one or both species. In contrast, the consensus sequence for I. pacificus (ITS1 length range, 319-344 bp) differed from those of I. dammini and I. scapularis at 52 sites out of 365, 17 of which were not variable within any of the three species. The ITS2 consensus sequence of I. scapularis (ITS2 length range, 679-682 bp) differed from that of I. dammini (ITS2 length range, 678-686 bp) at only one of 693 sites, the single site of difference being variable in both species. The I. pacificus (ITS2 length range, 708-728 bp) consensus sequence differed from those of I. dammini and I. scapularis at 172 of 760 sites, 124 of which were not variable within any of the three species. Distribution of Sequence Variation. In ITS1, the fraction of intraspecifically variable sites was 23/353, 16/353, and 69/ 354 for I. scapularis, I. dammini, and I. pacificus, respectively. Intraspecifically variable sites were concentrated in six regions (Fig. 1 positions 30-40, 186-195, 217-227, 244261, 302-322, and 354-363). Ninety of the 108 variable sites were in these regions. These regions and the bases flanking them (±2) are A+T-rich (74%, 71%, and 69%, respectively, for I. scapularis, I. dammini, and I. pacificus) relative to the rest of ITS1 (51% A+T; for all species, x2 > 11.0, df = 1, P < 0.05). Thirty-one of 50 differences in the consensus sequences of I. scapularis, I. dammini, and I. pacificus were variable intraspecifically. Twenty of these 50 were in poly(A) [or poly(T)] stretches.

AAGCTTGGAT GCCGTCGGGA CGGCACAAAA AANMQAAA ACTGTGGTTC GGATGTGCCT ..........

C..........

..........

.........

=... --.. ..........

..........

.........

A

......................

.~-QC .....As -........A...

..........

60 56 58

S: TTTTCTTTGG GT-GGCCTCC GTTCCTTTCG ATGATGCGCC CAGCGGGAGC CCTCTGCTGC D: ........ ......... T. .........A......... P: ..

119 115

CTCGATT-GA GAGGAGATTT CAAGTGCACC GGGACAATCC GCGGGTTACA AAAGCCGCAT

178 174 177

..................

..........

..........

......... ...... ..........

L......

..........

S: D: P:

..........

S:

CGCCCCCTTT TTCGTCTCTT CTCGAGGATC GCCCTTTTGT TTTGATTAAC GAAGAGGCGT

.

.

.T.

.

..........

.-.

..

.

.

.......... ..........

......... ..........

T.

......... L.

.

.

...... .

....

. . . .

L..

. . .

D: P:

S: D: P:

S: D:

..

.

.

.

..

.-

. . . .

.

..

. . ..

. . . . . . . . . .

118

238 234 223

TTGAGAAGA_ G----C mT TTTTTGCGCT TTCGAGATGG CATGGGATGT GCCGGGCGC-

293

.......

288 283

....

-..........

G..GA._I...G. A .G C.A.

..

..............................

.......CT .......

..........

.......... ........

..........

A. ........_. A

cus (P). Dots signify that the se-

-GAMAAAAAA AAAA---=CA ACAACTTGTG TTTTGTCGTA CTGATTGAAA AATCCAAAA...................

P:

A ...GG

S: D: P:

---ACAC

...

.

.

351

.......

346

TT....

342

C

..

...%-.-

.......... ......... .......... .

..........

..........

FIG. 1. Consensus sequences, 5' to 3', for ITS1 of f. scapularis (S), I. dammini (D), and I. pacift-

..........

. . . . C....

...........-C

347 342 335

quence matches that of 1. scapularis, whereas letters correspond to single base species differences relative to I. scapularis; dashes signify absent sites. Underlined characters vary within a species.

Proc. Natl. Acad. Sci. USA 90 (1993)

Evolution: Wesson et al. S: a .. P: 5:

. ...... ... .

..........

..........

...........

..........

..........

...... .......

.......... ......

........... .......... ......... .......... .........1-.L....

GCAGCACGCA GTTT-TATGC TTTCTTGCGT TGCGTTTTCT T---------

----------

:

.D . .... . . . . ....... .G........... .............. .....

P:

..........

GT

.....

...

.........

.GTTTCTCGG AGCAAGTACG

A

D:

---TGAGCAA ATGCACGAGT GGTGCGATTG CACGCGTGCG CTTAACCA-G TCCTCCTCCT ....... ........

P:

GAG.T

S:

S:

D:

..........

.......... ......... ...........

...... ..........

G.G

...............

A.... T

-CCTACGAGT TTTTATCGAA CACTGCATGG GAAAACGAAA CTCGATGGAT ACCG-TTTGT ..........

..........

.CGGTTTG .G.CG

..........

........ ........ .... . A....

..............................

A

.G

.....

P:

A.

S: D: P:

GGAAAATCCC G-TACCAAAA AAATCTTTCG CACGTTGAAC GGCGCT--GT GACGTCGGTG

S: D: P:

CGTTAGAAA- -CGGAGATTT GAAACGGTTT CTTTTCGATC GATTCTGTTT T-CTTT--GG

S: D: P:

S: D:

..............

..........

...... .....

......

.

....

..........

..G

....

......... ..........

A A

=

........

.......... ......... ...........

........ ..........

..........

.........

.........

.................... .......... .A-.-----.........

.........GT.

......... ........... .......... ..........

.....

...............

...... ....TG.

...............

..... ........

T .....7

... TT

CGTGGATGTT GTTCGAGTGG ---AAAAA-C CGTGGGACGG A--GTTGGCG TTCCCGTT ..........

..........

......... ..........

C

......

AA.A

CCTTTGGGA

GAAACGCGGC GTCGGCAAAT GTAA-----...... ...... ...............

..........

P:

S: D:

......-.

........

JA..&TCTTGT

.CGGAGTTG.

----------

----ATCTCG TGGCGTTGAT

..........

..............................

GGCGTTGATT

TTGCA.QAAA AAAAA_GAAA AACGCTTCTG GGAGGAAAGT ...?......... CG.GT.---- -..T ..GATL. ..........

...............

.

..........

TTQ.QA.AA

CTT.-...C.

-----TTTTTT GCGTCGTAGC ...... ............ ............

GG.G.G.G..

S: D: P:

CTTCCGTCAG TCTAAG&CCT TCGCGTCCCC GATGAATACT GGAGCCATCC AGTAGGGGAA

S:

TGCCGTTGGA TTGCGCGCTT TCTTTTTTGT CGAATCGAAA GCGATGCAAA GCGTGTGTAC

S: D: P: S:

D:

P:

..........

.......... .......... ......

..........

..........

..........

...............

...............

...............

...............

..........

...............

..........

..............................

.......... .........T.

AACA_CTGTT AGTTCGCCAT CCTTTT-GTG TTTTTGACAT GTGGTAAGGA GGCGAAGTTT .

.

....... .. ..... TQ.T. TT........

.......

T M

..... .

.TT.......

.....----

T-----TGOT GGAAAGCGCA CAAAAT---- ----GAATTG ........

G......G....

A........ ..

.

..........

A.

........... _AA_GA TTCQ.-G .

.

.......... ..

214 214 239 271 271 289 326 326 347 380 380 407 420 420 465

474 474 519

534 534 ................. T. 579

....... .............

.........

156 156 179

FIG. 2. Consensus sequences, 5' to 3', for ITS2 of I. scapularis (S), I. dammini (D), and l. pacdifcus (P). Dots signify that the underlying sequence matches that of I. scapularis, whereas underlying letters correspond to single base differences relative to I. scapularis; dashes signify absent sites. Underlined sites vary within a

.......... .... .......... .....

...... .........TG.

..............................

100 100 119

......... .......... ........... .....

..GAT..T..

P:

D: P:

60 60 60

ATCATATATC AAGAGAGGAG AATTTGTTTT CTACCTCGTT TTGACTGTGT CGGATCGTGG ..........

10223

680 680 723

594 594 639 653 653 689

species.

In ITS2, the fraction of intraspecifically variable sites was 22/692, 31/692, and 94/745 for I. scapularis, L. dammini, and l. pacificus, respectively. Almost two-thirds (96) of the 147 intraspecifically variable sites were clustered in 13 small regions of combined length 106 bases (Fig. 2 positions 48-53, 75-79, 155-157, 231-238, 252-262, 385-392, 438-444, 494503, 523-527, 669-678, 687-693, 717-731, and 745-755). As with ITS1, these and the immediately flanking (±2) sequences were significantly richer in A+T (66%, 65%, and 61%, respectively, for l. scapularis, L. dammini, and I. pacificus) than ITS2 as a whole (53%; for all species, x2 > 3.9,

simonious trees. The 25 sequences used in the analysis represented variation within and between populations of all three species. All maximum parsimony (minimum evolution) trees shared certain features: (i) I. pacificus sequences (n = 7) clustered together, the most divergent members being more closely related to each other than to any sequence of the other two species; (ii) at least some I. scapularis sequences (n = 8) were more closely related to I. dammini sequences (n = 10) than to conspecific sequences, and a similar pattern held for some I. dammini sequences; and (iii) sequences

dfn =

Table 1. Mean distances between ITS1 or ITS2 sequences within populations, between populations, and between species Mean distance ± SD Within Between Between Species population populations species ITS1 S 0.018 ± 0.009 0.029 0.028 (vs. D) 0.159 (vs. P) D 0.013 ± 0.006 0.008 ± 0.004 0.028 (vs. S)

1,

P < 0.05).

Intraspecific variation in ITS1 was equally attributable to deletions/additions (n = 46 among 25 subclones) and transitions/transversions (n = 49), while among ITS2 subclones (n = 17) there were 50%6 more transitions and transversions (n = 83) than deletions/additions (n = 55). For ITS1 and ITS2, the transition/transversion ratio was 1.49. Apportioning Sequence Diversity. In general, distances between ITS1 and ITS2 sequences within a population were only slightly, if any, smaller than those between individuals from different geographic populations within the same species (Table 1). Between-species distances were high relative to within-species distances only for species pairs containing I. pacificus. The distance between I. scapularis and I. dammini was of the same magnitude as intraspecific distances. When sequence diversity was partitioned into its components for I. scapularis and I. dammini, the betweenspecies contribution was relatively small or even negative (suggesting minimal contribution) (Table 2). Most of the sequence variation for I. scapularis and I. dammini was intraspecific, and most of this represented intrapopulation variation. (Table 3). Parsimony Analysis. The DNA parsimony algorithm, when applied to ITS1 sequence diversity, yielded 36 equally par-

P

0.041 ± 0.073

0.155

ITS2 S

0.007 ± 0.001

0.007

0.168 (vs. P) 0.159 (vs. S) 0.168 (vs. D)

0.003 (vs. D) 0.228 (vs. P) D 0.012 ± 0.004 0.005 ± 0.003 0.003 (vs. S) 0.227 (vs. P) P 0.040 ± 0.061 0.055 0.117 (vs. S) 0.228 (vs. D) Except for interspecific comparisons, distances are calculated relative to a consensus sequence at the appropriate level for each species (given that n > 2). S, I. scapularis; D, I. dammini; P, I.

pacificus.

10224

Evolution: Wesson et al.

Proc. Natl. Acad. Sci. USA 90 (1993)

Table 2. Contribution of intra- and interspecies diversity to total sequence diversity in ITS1 and ITS2 for I. scapularis and I. dammini Source of diversity Within species Between species

% contribution ITS1 72.4 27.6

ITS2 114.5 -14.5

derived from the same population were never all more closely related to each other than to conspecific sequences derived from other geographic localities. Also, the five localities from which I. dammini sequences originated did not cluster according to geographic proximity. Seventeen ITS2 sequences were employed in a maximumparsimony analysis that yielded the same qualitative results as for ITS1. Here, there were 69 equally parsimonious trees. Again, I. pacificus sequences (n = 4) always clustered, whereas l. scapularis (n = 5) and I. dammini (n = 8) sequences always failed to assort along currently recognized species designations. When ITS1 and ITS2 were combined into a single sequence and the analysis was repeated, four equally parsimonious trees were obtained, from which a consensus tree was generated (Fig. 3). The I. pacificus (n = 4) sequences clustered as a closely related group in every instance, whereas I. scapularis and I. dammini always failed to cluster according to geographic proximity of sites of origin or according to current species designations. Distance Methods. As with maximum parsimony, both UPGMA and the neighbor-joining distance methods failed to separate I. scapularis sequences from those of l. dammini whether the analysis focused on ITS1, ITS2, or ITS1 plus ITS2. Again, l. pacificus sequences always formed a cohesive, separate assemblage.

DISCUSSION Relationships Among Species. The results do not suggest that I. scapularis and I. dammini are members of separate species, because neither parsimony nor distance methods resolved the rapidly evolving sequence, ITS1 and ITS2 into assemblages corresponding to current species designations and because there was not even a single consistent base difference, out of over 1100 bases sequenced, between the two. Yet, I. pacificus differed from I. scapularis/dammini at numerous sites observed not to vary within any of the species. Both parsimony and distance methods based on ITS1 and/or ITS2 sequences recognized L. pacificus sequences as a subset distinct from the others. The inference, based on sequence data, that l. scapularis and I. dammini represent a single species, is consistent with other data. For instance, ethological isolation tests in which females of a single "species" are offered a choice between males of their own and another species, and reversed tests in which males are presented with two females, fail to reveal ethological isolation between l. scapularis and L. dammini but do reveal isolation between I. scapularis/dammini and I. pacificus (9). Postmating isolation (e.g., hybrid inviability) is Table 3. Contribution of within-population, between-population, and between-species diversity to total sequence diversity in ITS1 and ITS2 for L. scapularis and L. dammini % contribution Source of diversity ITS1 ITS2 Within population 81.6 107.9 -13.2 9.2 Between populations 31.6 -17.2 Between species

dam-MD scap-NC scap-GA

dam-WI

FIG. 3. Consensus minimum-evolution tree based on sequence variation in ITS1 and ITS2 among 16 ITS subclones. Clones are identified by species (pac, I. pacificus; scap, I. scapularis; dam, I. dammini) and geographically (two-letter designations for states). not detected between crosses of I. scapularis and I. dammini but is for crosses between either of these and I. pacificus. Laboratory studies of host preference reflect little or no difference among the three species (18) and all three are competent laboratory vectors for Borrelia burgdorferi (4, 5). Conversely, biological differences seem to exist between the more northern populations identified as I. dammini and the more southern populations identified as I. scapularis. Field-collected northern populations are most often found on rodents as immature organisms (3, 19) and are easily collected on flags, while the southern populations are often found on lizards (20) and are seldom found when flagging. In addition, B. burgdorferi infection is common (25% in nymphs, 50% in adults) in I. dammini (21), while it is much less common (0-0.4%) in I. scapularis (5, 22). These differences could reflect the range of available vertebrate hosts rather than variation inherent in the tick population. In the laboratory, larval and nymphal I. scapularis and I. dammini did not differ in host preference with regard to mice, lizards, and chickens. They preferred mice, lizards, and chickens, in that order (18). So why the apparent difference in nature, not only in host preference and questing strategy, but in tick infection rate? In two regions other than the eastern United States where B. burgdorferi transmission occurs (in other Ixodes tick species), the presence of lizards as part of the known vertebrate host assemblage has been linked to lowered spirochete transmission (23-25). This is true in Europe, where I. ricinus is the tick vector (26-28), and in the western United States where I. pacificus and I. neotomae are the tick vectors (6, 29). In both cases, although lizards are excellent hosts for immature ticks, they do not act as efficient reservoirs of B. burgdorferi. Thus, lizard populations, if present, may be heavily infested by immature ticks; in addition, ticks parasitizing lizards are more likely to be

Evolution: Wessm et-aL collected than those on corresponding homeotherms, because feeding ticks remain on lizards 3 times longer (9). In the absence of lizard populations-in the northeastern United States, for example-immature ticks will be found predominately on rodents and birds. Lizard populations in the southern United States may in fact serve to inhibit transmission of the Lyme spirochete by an otherwise highly competent vector, while in the Northeast this zooprophylactic effect is absent (7). Species Evolution. As the time since divergence of populations increases, the geneology of DNA sequences converges on the phylogeny of the populations, and there is little gain in information by sequencing multiple genes or DNA from multiple individuals (30). However, in the current analysis, evidence suggests that I. scapularis and I. dammini are, in fact, the same species. Thus, the problem of "lineage sorting" (DNA geneology is set before populations split) may arise. Also, with gene flow between populations, geneologies are not tree-like but contain loops (30). Our data set addresses these potential problems in two ways: (i) multiple sequences were examined per "species," and (ii) the length of the sequences examined was large (>1000 bp for ITS1 and ITS2 combined). Theoretical simulation studies indicate that the probability of obtaining the correct tree with the neighborjoining method is high for sequences over 1000 bases in length and high for shorter sequences (500-1000 bp) where the rate of evolution is high (31). Thus, our sample sizes and the lengths of the sequences examined should be sufficient to address the relationships between populations of I. scapularis and I. dammini. While parsimony and distance methods produced slightly different consensus phenograms, they agreed in failing to separate I. dammini populations/ sequences from those of I. scapularis. Yet I. pacificus populations and sequences were always recognized as a closely interrelated assemblage. All three "species" shared the same hypervariable regions and similar relative contribution to intraspecific variability from base substitutions versus deletions/additions (50%o for each). ITS domains varying intraspecifically also contributed preponderately to interspecific variation, suggesting that interspecific differences reflect stochastic processes in regions experiencing little if any selective constraint. Thus, failure of parsimony and distance methods to distinguish between I. scapularis and I. dammini, while distinguishing them from I. pacificus, is indicative of a shared gene pool. Inferred relationships between I. scapularis/dammini populations did not correspond to their respective geographic proximities. This suggests that the populations shared a common ancestor in the recent past or that they are bound by gene flow. Condusions. DNA sequencing of ITS1 and ITS2 of rDNA reveals that these sequences are highly variable in ticks, particularly with regard to small deletions and additions which may arise via replication slippage. Homogenization of rDNA arrays for particular sequence variants appears to occur at a relatively rapid rate, facilitating the use of these sequences to assess species relationships. Based on maximum parsimony and two distance methods, sequence variation in ITS1 and ITS2 provides strong evidence that I. scapularis and I. dammini are in fact a single species in which even geographically isolated populations are very similar. Similarly, isolated populations of I. pacificus are closely related to each other but are distinct from those of I. scapularis/dammini. Our results are in agreement with those

Proc. Natl. Acad. Sci. USA 90 (1993)

10225

from assortative mating and hybridization experiments, from morphometric and isozyme analysis, from chromosome karyotyping, and from vector competence and host preference studies. We thank J. M. C. Ribeiro for supplying ticks from Arizona, and B. Holloway and the staff of the National Center for Infectious Diseases Biotechnology Core Facility for synthesizing oligonucleotides. D.M.W. is a National Research Council Associate. 1. Spielman, A., Clifford, C. M., Piesman, J. & Corwin, M. D. (1979) J. Med. Entomol. 15, 218-234. 2. Hoogstraal, H. (1981) Annu. Rev. Entomol. 26, 75-99. 3. Levine, J. F., Wilson, M. L. & Spielman, A. (1985) Am. J. Trop. Med. Hyg. 34, 355-360. 4. Burgdorfer, W. & Gage, K. L. (1986) Zentralbl. Bakteriol. Mikrobiol. Hyg. Ser. A 263, 15-20. 5. Piesman, J. & Sinsky, R. J. (1988) J. Med. Entomol. 25, 336-339. 6. Lane, R. S., Piesman, J. & Burgdorfer, W. (1991) Annu. Rev. Entomol. 36, 587-609. 7. Spielman, A., Wilson, M. L., Levine, J. F. & Piesman, J. (1985) Annu. Rev. Entomol. 30, 439-460. 8. Coan, M. E. & Stiller, D. (1986) J. Med. Entomol. 23, 446-453. 9. Oliver, J. H., Owsley, M. R., Hutcheson, H. J., James, A. M., Chen, C., Irby, W. S., Dotson, E. M. & McLain, D. K. (1993) J. Med. Entomol. 30, 54-63. 10. Olson, C. A., Cupp, E. W., Luckhart, S., Ribeiro, J. M. C. & Levy, C. (1992) J. Med. Entomol. 29, 1060-1062. 11. Collins, F. H., Porter, C. H. & Cope, S. E. (1990) Am. J. Trop. Med. Hyg. 42, 417-423. 12. Wesson, D. M., Porter, C. H. & Coffins, F. H. (1992) Mol. Phylogenet. Evol. 1, 253-269. 13. Learn, G. H., Jr., & Schaal, B. A. (1987) Evolution 41, 433438. 14. Tautz, D., Hancock, J. M., Webb, D. A., Tautz, C. & Dover, G. A. (1988) Mol. Biol. Evol. 5, 366-376. 15. Kimura, M. (1980) J. Mol. Evol. 16, 111-120. 16. Wilkinson, G. S. & Chapman, A. M. (1991) Genetics 128, 607-617. 17. Wesson, D. M. & Coffins, F. H. (1992) Nucleic Acids Res. 20, 1139. 18. James, A. M. & Oliver, J. H. (1990) J. Med. Entomol. 27, 324-330. 19. Main, A. J., Carey, A. B., Carey, M. G. & Goodwin, R. H. (1982) J. Med. Entomol. 19, 655-664. 20. Rogers, A. J. (1953) Dissertation (Univ. Maryland, College Park). 21. Piesman, J., Mather, T. N., Donahue, J. G., Levine, J., Campbell, J. D., Karakashian, S. J. & Speilman, A. (1986) Acta Trop. 43, 263-270. 22. Magnarelli, L. A., Anderson, J. F., Apperson, C. S., Fish, D., Johnson, R. C. & Chappell, W. (1986) J. Wildl. Dis. 22, 178-188. 23. Lane, R. S. (1990) Am. J. Trop. Med. Hyg. 42, 75-82. 24. Manweiler, S. A., Lane, R. S. & Constantine, H. T. (1992) Am. J. Trop. Med. Hyg. 47, 328-336. 25. Matuschka, F., Fischer, P., Musgrave, K., Richter, D. & Spielman, A. (1991) Am. J. Trop. Med. Hyg. 44, 100-107. 26. Aeshlimann, A., Chamot, E., Gigon, F., Jeanneret, J. P., Kesseler, D. & Walther, C. (1986) Zentralbl. Bakteriol. Hyg. 263, 450-458. 27. Stanek, G., Pletschette, M., Flamm, H., Hirschl, A. M., Aberer, E., Kristoferitsch, W. & Schmutzhard, E. (1988) Ann. N. Y. Acad. Sci. 539, 274-282. 28. Landbo, A. S. & Flong, P. T. (1992) Med. Vet. Entomol. 6, 165-167. 29. Brown, R. N. & Lane, R. S. (1992) Science 256, 1439-1442. 30. Felsenstein, J. (1988) Annu. Rev. Genet. 22, 521-565. 31. Jin, L. & Nei, M. (1991) Mol. Biol. Evol. 8, 356-365.