An Adaptive Radiation of Hawaiian Thomisidae: Biogeographic and Genetic Evidence

8 Pages·1999·8.2 MB·English

by  Jessica E Garb

1999. The Journal of Arachnology 27:71-78
AN ADAPTIVE RADIATION OF HAWAIIAN THOMISIDAE:
BIOGEOGRAPHIC AND GENETIC EVIDENCE
Jessica E. Garb: Zoology Department, University of Hawaii at Manoa, Honolulu,
Hawaii 96822
ABSTRACT. The Hawaiian Thomisidae are noted for being extremely species rich, as well as diverse
in morphology and ecology. This exceptional diversity led early systematists to place the species into
several genera with cosmopolitan distributions. It has been recently suggested that these species compose
a single large adaptive radiation. Species-area relationships for all thomisid species and for Misumenops
F.O. Pickard-Cambridge 1900 (Thomisidae) species for various island areas were generated. Further, a
phylogenetic hypothesis was constructed based on genetic distances between the Hawaiian thomisids and
various outgroups using a 450 bp region of the mitochondrial cytochrome oxidase I (COI) gene to test
for close genetic relationships. Despite the extraordinary isolation of the Hawaiian islands, the numbers
of Misumenops and total thomisid species were found to be significantly higher than predicted for an
island system ofits size. Phylogenetic analysis ofCOI suggests the Hawaiian thomisids are more closely
related to each otherthan to representatives ofgenera to which they have been previously assigned. These
results support the existence of a Hawaiian thomisid adaptive radiation, and merit further investigation
using comparative methods.
The biota of the Hawaiian archipelago is due to its extreme isolation (Howarth & Mull
characterized by a number oflarge species ra- 1992). However, the islands are exceptionally
diations. These radiations result from the ex- diverse at the species level, due to extensive
treme geographic isolation and topographical autochthonous speciation. The diversity of
diversity of the islands (Carson & Clague Hawaiian spiders closely follows this pattern.
1995). Natural colonization ofthe archipelago Of the 105 known spider families (Cod-
&
has been largely limited to organisms having dington Levi 1991), only 10 occurnaturally
exceptional dispersal capabilities, and those in the Hawaiian Islands, and many of these
species that were sucessful colonizers experi- are represented by a single genus (Simon
enced total genetic isolation from their source 1900; Suman 1964). Some of these genera
populations. Founders that made the long have undergone extensive speciation (Gilles-
journey frequently underwent rapid evolution pie et al. 1998). In particular, native species
due to drift and selection in extremely small of the genera Tetragnatha Latreille 1831 (Te-
populations (Carson 1994). The Hawaiian Is- tragnathidae), Argyrodes Simon 1864 and
lands themselves are volcanic, formed at a Theridon Walckenaer 1805 (Therididae), Mis-
“hot spot” connected to the Earth’s core. As umenops F.O. Pickard-Cambridge 1900
the Pacific tectonic plate continually rolls (Thomisidae), Sandalodes Keyserling 1883
northwestward over the hot spot, new islands (Salticidae) and Cyclosa Menge 1866 (Ara-
are formed. Consequently, they are arranged neidae) are known to be unusually diverse and
in chronological order. After initial coloniza- most, if not all, are endemic to the Hawaiian
tion of the archipelago, it appears that taxa Islands (Table 1). It is likely that many more
have frequently progressed down the island species within these groups remain to be dis-
chain in a step-wise manner, resulting in re- covered. However, because many species have
peated founder events as each new island is extremely small ranges of endemism, rapid
colonized (Wagner & Funk 1995). This degradation of natural areas along with in-
unique combination ofbiogeographical events creasing numbers of alien species, make these
may explain some of the unusual aspects of spiders exceptionally vulnerable to extinction
the composition of the Hawaiian biota. For (Gillespie & Reimer 1993).
example, Hawaii’s terrestrial biota is consid- The family Thomisidae attracted attention
ered depauperate at higher taxonomic levels early in the study ofHawaiian biology. R.C.L.
71
72 THE JOURNAL OF ARACHNOLOGY
—
Table 1. Species radiations ofHawaiian spiders.
%
Native Considered
Family Genus species endemic Reference
Tetragnathidae Tetragnatha >52 100 Gillespie et at (1998)
Therididae Theridion >13 100 Simon (1900)
Therididae Argyrodes >30 100 Gillespie et al (1998)
Thomisidae Misumenops 17 100 Suman (1970)
Araenidae Cyclosa >7 100 Simon (1900)
Salticidae Sandalodes >9 88 Simon (1900)
Perkins, a pioneering Hawaiian naturalist and one in which extensive adaptive morphologi-
collector for the Fauna Hawaiiensis spent lit- cal evolution has played a significantrole. Ex-
tle time in the pursuit of spiders. However, he amples of explosive adaptive radiations in is-
recognized that “the Thomisidae are probably land archipelagos illustrate how closely
the most interesting and important group in related species often display great morpholog-
&
the Hawaiian spiders” (Perkins 1913). His ical diversity (Grant Grant 1989). This phe-
collections were examined by Eugene Simon nomenon is well documented in other Hawai-
(1900), who described 14 new species and a ian spiders (Gillespie 1994; Gillespie et at
new genus of Thomisidae. These 14 species 1997), and other Hawaiian taxa (Roderick &
were placed in the cosmopolitan genera Mis- Gillespie 1998). Consequently, if the Hawai-
umena Latreille 1804, Diaea Thorell 1869 and ian thomisids constitute an adaptive radiation,
Synaema Fabricius 1775 as well as the new they may represent another exceptional op-
endemic genus Mecaphesa Simon 1900. In a portunity for understanding evolutionary and
subsequent revision of Hawaiian Thomisidae, ecological mechanisms that generate rapid di-
Suman (1970) noted that the genitalia of the versification.
Hawaiian Diaea was extremely similar to that In this study I examine the species richness
ofthe HawaiianMisumena and these two gen- of Hawaiian thomisids in a biogeographical
era were likely the same. Further, examination framework. Classical biogeographical theory
of the setation and eyes of Hawaiian spiders (MacArthur & Wilson 1967) states that island
previously assigned to Diaea and Misumena species diversity is a balance between immi-
revealed that they were actually more similar gration and extinction. What determines the
to representatives of Misumenops, a cosmo- rate of immigration and extinction depends
politan genus, than to other representatives of largely on the size ofan island and its distance
Diaea and Misumena. Accordingly, represen- from a source. In this context one would pre-
tatives of Diaea and Misumena were placed dict that the isolation and small size ofHawaii
inMisumenops. Suman consideredthe endem- would result in a species poor fauna. Here I
ic Mecaphesa to be related to the genus Ozyp- examine the species-area relationship for Ha-
tila Simon 1964, and further suggested that waiian thomisids at the family and generic
that all of the Hawaiian thomisids were likely levels in comparison to other island areas.
derived from three separate colonizers, one Also, I conduct a phylogenetic analysis ofHa-
that gave rise to the Hawaiian Misumenops, waiian thomisid species and representatives of
one that gave rise to the endemic genus Me- the genera to which they have been assigned
caphesa, and one that gave rise to the Ha- historically, in order to provide corroborative
waiian Synaema (Suman 1970). Suman's re- evidence for a within-archipelago radiation.
vision also resulted in the recognition of METHODS
several new species, for a total of21 described —
species, 17 of which are in the genus Misu- Biogeographical analysis* A species-area
menops. curve was generated for both total thomisid
Recently, Lehtinen (1993) has suggested species and for species in the genus Misumen-
that all Hawaiian thomisids comprise an adap- ops. Data for the distribution ofthomisids and
tive radiation generated from a single founder. Misumenops species have been assessed in
—
garb—EVIDENCE FOR A HAWAIIAN THOMISIDAE RADIATION 73
—
Table 2. Species used in sampling of a 450 bp region of mitochondrial cytochrome oxidase L
Species Collection locality
Misumenops anguliventris Simon 1900 Kauai, Oahu, Maui, Hawaii Island
Misumenops cavatus Suman 1970 Hawaii
Misumenops discretus Suman 1970 Kauai
Misumenops editus Suman 1970 Oahu
Misumenopsfacundus Suman 1970 Hawaii Island
Misumenops insulanus Keyserling 1890 Oahu
Misumenops imbricatus Suman 1970 Oahu, Maui
Misumenopsjunctus Suman 1970 Molokai
Misumenops nigrofrenatus Simon 1900 Kauai, Oahu, Hawaii Island
Misumenops vitellinus Simon 1900 Kauai, Oahu, Maui, Hawaii Island
Mecaphesaperkinsi Simon 1900 Oahu
Mecaphesa semispinosa Simon 1900 Oahu
Synaema naevigerum Simon 1900 Maui
Synaema globulosum Fabricius 1775 Mt. Carmel, Israel
Xysticus sp. Westchester, New York
Ozyptila georgiana Keyserling 1880 Westchester, New York
Misumena vatia Thorell 1870 Maryland
Misumenops asperatus Hentz 1870 Maryland
Misumenoides sp. Indiana
Diaea sp. New Zealand
various island systems worldwide by a num- collected from North America and Israel and
ber ofresearchers: Andaman and Nicobar (Ti- others were kindly supplied by Dr. J. Robin-
kader 1977), Japan (Ono 1988), Britain and son, Dr. G. Dodson, Cor Vink and Dr. W.
Ireland (Roberts 1985), Puerto Rico (Petrunk- Shipley.
evitch 1930), Cuba (Bryant 1940), Tonga Collection and analysis ofgenetic data.
(Marples 1959), Fiji (Marples 1957), Pitcairn DNA was extracted from the 13 Hawaiian
&
Islands (Benton Lehtinen 1995), Philip- species and 7 non-Hawaiian species listed in
pines (Barrion & Litsinger 1995), Samoa Table 2, using a phenol-chloroform prepara-
(Marples 1957), Rapa (Berland 1924), Maur- tion followed by ethanol precipitation (Pal-
itius (Simon 18975), Galapagos Islands umbi et al. 1991). Voucher specimens were
(Banks 1902), Tahiti (Berland 1934), and retained and will be deposited in the Bishop
Marquesas Islands (Berland 1927). In the gen- Museum, Honolulu, Hawaii. A 450 bp region
eration ofthe Misumenops species area curve, ofthe mitochondrial gene cytochrome oxidase
Diaea species from Tonga, Fiji and Samoa I (COI) was amplified for at least two indi-
were substituted for Misumenops, as these viduals per species using a thermal cycler,
were likely incorrectly diagnosed and are with universal primers Cl-J-1718 and Cl-N-
closely related to the Hawaiian thomisids (Su- 2191 (Simon et al. 1994). The COI gene was
man 1970). Predictions of the total expected selected because it evolves rapidly in the
number of thomisid species and Misumenops third-codon position (Simon et al. 1994), pro-
species in Hawaii were calculated based on viding sufficient variation to determine rela-
the species-area relationships generated. tionships between closely related species. This
Collection a—nd identification ofHawaiian gene has been useful for assessing genetic di-
Thomisidae. Hawaiian thomisid species versity and phylogenetic relationships in other
were collected from native ecosystems on Hawaiian arthropods (Roderick & Gillespie
Kauai, Oahu, Maui, Molokai and Hawaii Is- 1998). PCR products were sequenced using an
land (Table 2). Specimens were collected by ABI 377 automatic sequencer and were
beating vegetation. Sexually mature spiders aligned using Sequencher 3.0. Pairwise ge-
were identified to species using the Bishop netic distance (percent nucleotide difference
Museum reference collection and Suman’s between sequences) was calculated for all spe-
(1970) key. Additional outgroup species were cies using the Kimura 2-parameter correction
74 THE JOURNAL OF ARACHNOLOGY
Area (km^)
—
Figure 1. Species-area relationship for total species in the family Thomisidae. Linear regression and
a 95% Cl on log-transformed data are shown. Equation ofline is log(species) = 0.398 log(area) - 0.917,
= 0.70, P < 0.0001. Expected number of species in the family Thomisidae is 7, while 21 species are
observed.
for nucleotide bias (Kimura 1980), and a phy- pared to other island systems examined. For
logenetic tree was constructed from these dis- example, Hawaii has the highest observed di-
tances using a neighbor-joining algorithm in versity of Misumenops species (17), in con-
PAUP* 4.0 (Swofford 1998). The Xysticus sp. trast to the 1-4 species observed in other is-
sequence was used to root the resulting tree. land systems. The number of Misumenops
RESULTS species on the 15 non-Hawaiian islands ap-
— peared to be normally distributed, and assum-
Biogeographical analysis. For the spe- ing normality, Hawaii falls far above the
cies-area in the family Thomisidae, a highly 99.999th percentile of the distribution. These
significant linear increase was found between
results suggest that although thomisid diver-
the number of species and area of the island
s=ystem (log-transformations, slope = 0.39, R- MsiitsyuimnetnheopHsawacoinisatnitiustleasndsaidsihsipgrhopionrgteinoenraatle,
0.70, see Fig. 1). Perhaps not unexpectedly, amount of this diversity. Further, the restric-
Hawaii has an unusually high diversity for its
size, fallingjust above the 95% confidence in- tion of 17 ofHawaii’s 21 thomisids to a single
terval for the regression. The slope ofthe line genus suggests that biodiversity in this island
indicates that an archipelago the area of Ha- archipelago is a function of autochthonous
waii (28,314 sq. km), should have a total of speciation, rather than exter—nal colonization.
Analysis ofgenetic data. Construction of
7 thomisid species, but 21 occur in Hawaii.
The relationship for representatives of the ge- a distance based phylogeny from 450 bp of
nus Misumenops is quite different (Fig. 2). COI revealed that the smallest genetic dis-
Whether or not Hawaii is included in the re- tances were foundbetween Hawaiianthomisid
gression, there is no significant relationship species (Fig. 3). Thus the Hawaiian thomisids
between island size and number ofMisumen- appearto be more closely related to each other
ops species = 0.49, P = 0.43, for 15 is- than they are to representatives ofMisumena,
lands not including Hawaii). By any measure, Misumenoides, Diaea, Synaema, and Ozyptila.
Hawaii appears to be an extreme outlier com- The Hawaiian Synaema and non-Hawaiian S.
garb—EVIDENCE FOR A HAWAIIAN THOMISIDAE RADIATION 75
—
Figure 2. Species-area relationship for species in the genus Misumenops, along with a simple linear
regression on log-transformed data. Equation ofline is log(species) = 0.054 log(area) — 0.067, = 0.49,
P = 0.43. Expected number of Hawaiian species in the genus Misumenops is 0.6, while 17 species are
observed.
globulosum are extremely distant while the isid andMisumenops species, theMisumenops
Hawaiian Synaema is comparatively much species represent nearly all ofthe thomisid di-
more similar to the Hawaiian Misumenops, versity. It appears, therefore, that the number
supporting Lehtinen’s (1993) claim that the of successful colonizers is consistent with the
Hawaiian 'Synaema' is not a true Synaema, island biogeography model. However, the
Small genetic distances between non-Hawai- model does not account for subsequent pro-
ian Misumenops asperatus and all Hawaiian cesses that occur over evolutionary time.
Misumenops species sampled suggests thatthe Thus, despite few colonization events, autoch-
Hawaiian species are likely descended from a thonous speciation raises the species diversity
Misumenops ancestor. However, because Mis- to levels higher than predicted by island area.
umenops asperatus falls within the clade con- Such a pattern might suggest that most of the
taining all Hawaiian thomisids, more than one available niche space for the family is being
colonization event is implied for the Hawaiian occupiedby species in the genusMisumenops.
thomisid fauna. Sampling of many more Mis- Colonizing species may have undergone rapid
umenops species from other geographic re- ecological divergence in the absence of com-
gions must be conducted in order to firmly
petition from other genera that exploit a va-
establish what proportion of Hawaii's present
riety of ecological strategies elsewhere.
day thomisid diversity has arisen thi'ough au-
Although Hawaiian Misumenops species
tochthonous species radiation.
are extremely diverse (Fig. 2), it is possible
DISCUSSION that this result may be biased by data points
The high diversity of Misumenops species representing inaccurate estimates of diversity.
in Hawaii, coupled with the short genetic dis- Many of the data points used to generate the
tances between species, is a pattern similar to relationships between area and diversity were
that found in other Hawaiian terrestrial arthro- extracted from surveys of unequal sampling
pods that have undergone species radiation effort. Further, current revisions of these is-
&
(Roderick Gillespie 1998). While the is- lands’ spider fauna ought result in different
lands are exceptionally diverse in both thom- generic diagnosis of species. Consequently,
76 THE JOURNAL OF ARACHNOLOGY
Xysticus sp., Westchester, NewYork
Ozyptila georgiana, Westchester, New York
Synaema globulosum, Mt, Carmel. Israel
Diaea sp, New Zealand
,
Misumenoides sp. Maryland
,
Misumena vatia, Maryland
Misumenops cavatus, Hawaii, Hawaii
Synaema naevigerum, Maui, Hawaii
Misumenops asperatus, Maryland
Mecaphesaperkinsi, Oahu, Hawaii
Misumenops discretus, Kauai, Hawaii
Misumenopsjunctus, Molokai, Hawaii
Misumenops vitellinus, Oahu, Hawaii
Misumenops nigrofrenatus, Kauai, Hawaii
Misumenops imbricatus, Oahu, Hawaii
Misumenopsfacundus, Hawaii, Hawaii
Misumenops editus, Oahu, Hawaii
Misumenops insulanus, Oahu, Hawaii
Misumenops anguliventris, Kauai, Hawaii
Mecaphesa semispinosa, Oahu, Hawaii
—
Figure 3. A distance based phytogeny constracted from 450 bp of COL Genetic distances were caL
culated using Kimura’s (1980) 2-parameter corrected distance measure.
many of these points may represent over or tribe Misumenini Simon 1895, including spi-
under estimates oftrue diversity. It is possible ders in the genera Misumena, Misumenoides,
that a more complete data set would reveal and Misumenops (Ono 1988), are commonly
that Hawaii is not overly diverse in Misumen- known as flower spiders because they mimic
ops or total thomisid species. However, Japan the coloration of flowers on which they sit in
(the most thoroughly studied island group), order to capture pollinating insects (Gertsch
has an area 10 times the size of Hawaii yet 1979). Recent research reveals that certain
only twice the number of thomisid species species of this tribe have ecological roles as
(Ono 1988). Moreover, only three of these nectivores and possible pollinators (Pollard et
species are Misumenops. The relatively small al. 1995). While genetic data suggest that the
genetic distances among Hawaiian species are Hawaiian thomisids are descendants offlower
consistent with the hypothesis that the high spiders, Hawaiian thomisids are extremely di-
morphological and ecological diversity ofHa- verse in their substrate affinities. Forexample,
waiian thomisids is primarily the result of au- the green and brown speckled Misumenops
tochthonous radiation. editus, endemic to the summit of Mt. Kaala,
The Hawaiian Misumenops possess excep- Oahu, is perfectly camouflaged amongst its
tional ecological diversity when compared to primary microhabitat of moss patches. Like-
their continental congeners. Members of the wise, M. aridus and M. nigrofrenatus are well
family Thomisidae are well known for their hidden on their substrate of white filamentous
employment of mimicry to ambush prey. The lichen. Other species are more specific to
A
garb—EVIDENCE FOR A HAWAIIAN THOMISIDAE RADIATION 77
green foliage. Many Hawaiian Misumenops des lies Juan Fernandez. The Natural History of
that are ecologically separated as a result of Juan Fernandez and Easter Island, 3:419-437.
differential substrate affinity are sympatric Berland, L. 1927. Notes sur les Araignees recueil-
and may be very close relatives (Sumae lies aux eles Marquises par le R.P. Simen Del-
mas. Bull. Du Museum National d’Histoire Na-
1970
). turelle, Paris, 38:366-368.
In order to appreciate fully the extent of
Berland, L. 1934. Les Araignees de Tahiti. Bull.
species radiation in the Hawaiian Misumen- Bernice P. Bishop Museum, 113:97-107.
ops, a complete phylogenetic construction is Bryant, E.B. 1940. Cuban spiders in the Museum
warranted. Only with a well-supported phy- of Comparative Zoology, Bull. Mus. Comp.
logenetic hypothesis can the idea of species ZooL, 86(7):247-554.
radiation among the Hawaiian thomisids be Carson, H.L, 1994. Genetical processes of evolu-
tested rigorously. From such ahypothesis, one tion of high oceanic islands. Pp. 259-262. In A
can determine the number of colonization Natural History of the Hawaiian Islands. (E.A.
events and the ecological diversity of this CarKsaoyn,,eHd..)L. U&nivD..AH.awCaliaiguPere.ss,19H9o5n.olGuelou,loHgaywaaini.d
group. It is clear from datapresented here that biogeography of the Hawaiian Islands. Pp. 14-
the Hawaiian Misumenops do not fit the clas- 29. In Hawaiian Biogeography, Evolution on a
sical model for island biogeography as they Hot Spot Archipelago. (W.L. Wagner & V.A.
are species rich in an isolated and small island Funk, eds.), SmithsonianInstitutionPress,Wash-
area. ington, D.C.
ACKNOWLEDGMENTS Coddington, J.A. & H.W. Levi. 1991. Systematics
and evolution of spiders (Araneae). Ann. Rev.
I thank Rosemary Gillespie, George Rod- EcoL Syst., 22:565-592.
erick, A. Bohonak, and two anonymous re- Gertsch, W.J. 1979. American Spiders. 2Eded. Van
viewers for help in preparation of this manu- Nostrand Reinhold Co., New York.
script, B. Thorsby, M. Rivera, and J. Gubili Gillespie, R.G. 1994. Hawaiian spiders of the ge-
for laboratory assistance, M. Heddle, C. Ew- nus Tetragnatha: III. Tetragnatha acuta clade. J.
ArachnoL, 22:161-168.
ing, B. Thorsby, M. Rivera, A. Vandergast, D. Gillespie, R.G., H.B. Croom & L. Hasty. 1997.
Preston, A. Asquith, A. Medieros and E. Lar-
Phylogenetic relationships and adaptive shifts
son for assistance in collecting Hawaiian spi- among major clades of Tetragnatha spiders (Ar-
ders, W. Shipley, C. Vink, J. Robinson and G. aneae: TetragnatMdae) in Hawaii. Pacific Sci.,
Dodson for shipment of non-Hawaiian thom- 51(4):380-394.
isids, S. Swift at the Bishop Museum for ac- Gillespie, R.G. & N. Reimer. 1993. The effects of
cess to collections, the managers ofVolcanoes alien predation of ants (Hymenoptera: Formici-
and Haleakala national parks, Hawaii State dae) on the Hawaiian endemic spiders (Araneae:
Dept, of Fish and Wildlife reserves and The TetragnatMdae). Pacific Sci., 47(l):21-33.
Nature Conservancy ofHawaii’s preserves for Gillespie, R.G., M.A. Rivera & J.E. Garb. 1998.
Taxonomy and phylogeography ofHawaiian Ar-
permission to collect thomisids. This work is
aneae. Bull. British ArachnoL Soc., 11:41-51.
partly supported by a student research grant
Grant, B.R,& P.R. Grant. 1989. Evolutionary Dy-
from the University of Hawaii Ecology, Evo-
namics of Natural Populations. Univ. Chicago
lutionary and Conservation Biology Program
Press, Chicago.
and from AAS, as well as NSF grants to R. Howarth, EG. & W.P. Mull. 1992. Hawaiian In-
Gillespie. sects and Their Kin. Univ. Hawaii Press, Hono-
lulu.
LITERATURE CITED Kimura, M. 1980. A simple method for estimating
Banks, N. 1902. Papers fromthe Hopkins Stanford evolutionary rate of base substitution through
Galapagos Expedition. Proc. Washington Acad. comparative studies of nucleotide sequences. J.
Sci., 4:49-86. Mol. Evol. 2:87-90. —
Barrion, A.T. & J.A. Litsinger. 1995. Riceland spi- Lehtinen, P.T. 1993. Polynesian Thomisidae
ders ofSouth and SoutheastAsia. CenterforAg- meeting of old and new world groups. Mem.
riculture and Bioscience, Wallingford, UK, Queensland Mus., 33(2):585-591.
Benton, TG, & P.T Lehtinen. 1995. Biodiversity MacArthur, R.H. & E.O. Wilson. 1967. The The-
and originofthenon-flyingterrestrialarthropods ory of Island Biogeography. Monog. Population
ofHenderson Island. Bio. J. Linn. Soc., 56(1-2): Biology L Princeton Univ. Press, Princeton.
261-272. Marples, B.J. 1957. Spiders from some Pacific is-
Berland, L, 1924. Araignees de File de Paques et lands II. Pacific Sci., 11:386-395.
78 THE JOURNAL OF ARACHNOLOGY
Marples, B.J. 1959. Spiders from some Pacific is- quences and a compilation of conserved poly-
lands III. The Kingdom of Tonga. Pacific Sci., merase chain reaction primers. Ann. Entomol.
13:362-367. Soc. America., 87:651-701.
Ono, H. 1988. A revisional study of the spider Simon, E. 1897. Arachnological records par Al-
family Thomisidae of Japan. National Science luaud a File Maurice. Ann. Soc. Entomol.
Museum, Tokyo, Japan. France, 66:271-276.
Palumbi, S.R., A. Martin, S. Romano, W.O. Simon, E. 1900. Arachnida. Pp. 443-519. InFauna
McMillian, L. Stice & G. Grabowski. 1991. The Hawaiiensis, Vol. 2. (D. Sharp, ed.), Cambridge
simple fool’s guide to PCR. Dept, of Zoology, Univ. Press, Cambridge, UK.
University of Hawaii, Honolulu. Suman, TW. 1964. Spiders of the Hawaiian Is-
Perkins, R.C.L. 1913. Introduction. Pp. 15-228. In lands: catalog and bibliography. Pacific Insects,
Fauna Hawaiiensis, Vol. 1. (D. Sharp, ed.) Cam- 6:665-687.
bridge Univ. Press, Cambridge. Suman, TW. 1970. Spiders of the Family Thom-
Petrunkevitch, A. 1930. The spiders of Puerto isidae in Hawaii. Pacific Insects, 12(4):773-864.
Rico. Trans. Connecticut Acad. Sci., 30:1-355; Swofford, D.L. 1998. PAUP*. Phylogenetic Anal-
31:1-191. ysis Using Parsimony (*andothermethods). Ver-
Pollard, S.D., W.M. Beck & G.N. Dodson. 1995. sion 4. Sinauer Associates, Sunderland, Massa-
Why do male crab spiders drink nectar? Anim. chusetts.
Behav., 49:1443-1448. Tikader, B.K. 1977. Studies ofthe spider fauna of
Roberts, M.J. 1985. The Spiders of Great Britain Andaman and Nicobar Islands, Indian Ocean.
and Ireland. Vol. 1. Atypidae-Theridiosomati- Rec. Zool. Surv. India, 72:153-212.
dae, 229 pp. Harley Books, Colchester, UK. Wagner, WL. & V.A. Funk. 1995. Hawaiian Bio-
Roderick, G.K. & R.G. Gillespie. 1998. Speciation geography. Evolution On a Hot Spot Archipela-
and phylogeography of Hawaiian terrestrial ar- go. Smithsonian Institution Press, Washington,
thropods. Mol. Ecol., 7:309-322. D.C.
Simon, C., E Fratti, A. Beckenbach, B. Crespi, H.
Liu & P. Flook. 1994. Evolution, weighting, and Manuscript received 1 May 1998, revised4Novem-
phylogenetic utility of mitochondrial gene se- ber 1998.