Skip to main content

Full text of "Proceedings of the California Academy of Sciences"

See other formats 


PROCEEDINGS 

OF  THE 


California  Academy  of  Sciences 


Volume  46 


SAN  FRANCISCO 

PUBLISHED  BY  THE  ACADEMY 

1988-1990 


PUBLICATIONS  COMMITTEE 
Tomio  Iwamoto,  Scientific  Editor 
Judy  Prokupek,  Managing  Editor 

Luis  F.  Baptista 

Thomas  F.  Daniel 

Terrence  M.  Gosliner 

Thomas  Moritz 
Wojciech  J.  Pulawski 


(US  ISSN  0068-547X) 

The  California  Academy  of  Sciences 

Golden  Gate  Park 
San  Francisco,  California  94 1 1 8 


PRINTED  IN  THE  UNITED  STATES  OF  AMERICA 
BY  ALLEN  PRESS,  INC.,  LAWRENCE,  KANSAS 


CONTENTS  OF  VOLUME  46 

Pages 

No.  1.  WELSH,  HARTWELL  H.,  JR.  An  ecogeographic  analysis  of  the  herpetofauna  of 
the  Sierra  San  Pedro  Martir  region,  Baja  California,  with  a  contribution  to  the 
biogeography  of  the  Baja  California  herpetofauna.  Published  December  7, 
1988  1-72 

No.     2.    DANIEL,  THOMAS  F.  Three  new  species  of  Hologr aphis  (Acanthaceae)  from 

Mexico.  Published  December  7,  1988   73-81 

No.  3.  ANDERSON,  M.  ERIC,  AND  ALEX  E.  PEDEN.  The  eelpout  genus  Pachycara  (Teleos- 
tei:  Zoarcidae)  in  the  northeastern  Pacific  Ocean,  with  descriptions  of  two  new 
species.  Published  December  7,  1988  83-94 

No.  4.  UBICK,  DARRELL,  AND  THOMAS  S.  BRIGGS.  The  harvestmen  family  Phalango- 
didae.  1 .  The  new  genus  Calicina,  with  notes  on  Sitalcina  (Opiliones:  Lama- 
tores).  Published  March  9,  1989  95-136 

No.     5.    ALMEDA,  FRANK.  Five  new  berry-fruited  species  of  tropical  American  Melas- 

tomataceae.  Published  March  9,  1989  137-150 

No.  6.  ROBERTS,  TYSON  R.  Systematic  revision  and  description  of  new  species  of 
suckermouth  catfishes  (Chiloglanis,  Mochokidae)  from  Cameroun.  Published 
August  24,  1 989  1 5 1-1 78 

No.     7.    ROGERS,  EDWARD.  American  dance  flies  of  the  Drapetis  assimilis  species  group 

(Diptera:  Empididae).  Published  December  20,  1989  179-192 

No.  8.  GREENE,  HARRY  W.  Ecological,  evolutionary,  and  conservation  implications 
of  feeding  biology  in  Old  World  cat  snakes,  genus  Boiga  (Colubridae).  Published 
December  20,  1 989  1 93-207 

No.     9.    ALMEDA,  FRANK.  New  species  and  taxonomic  notes  on  Mexican  and  Central 

American  Melastomataceae.  Published  December  20,  1989  209-220 

No.  10.  ANDERSON,  M.  ERIC.  Review  of  the  eelpout  genus  Pachycara  Zugmayer,  1911 
(Teleostei:  Zoarcidae),  with  descriptions  of  six  new  species.  Published  Decem- 
ber 20,  1989  221-242 

No.   1 1 .    PARENTI,  LYNNE  R.  A  phylogenetic  revision  of  the  phallostethid  fishes  (Ath- 

erinomorpha,  Phallostethidae).  Published  December  20,  1989  243-277 

No.  12.  DANIEL,  THOMAS  F.  New  and  reconsidered  Mexican  Acanthaceae.  IV.  Pub- 
lished September  1 1,  1990  279-287 

No.   1 3.    DANIEL,  THOMAS  F.,  AND  DIETER  C.  WASSHAUSEN.  Three  new  species  ofJusticia 

(Acanthaceae)  from  Panama.  Published  September  11,  1990  289-297 

No.  14.  ALMEDA,  FRANK.  New  species  and  new  combinations  in  Blakea  and  Topobea 
(Melastomataceae),  with  an  historical  perspective  on  generic  limits  in  the  tribe 
Blakeeae.  Published  September  11,  1990  299-326 

No.   15.    ALMEDA,  FRANK.  Three  new  species  ofConostegia  (Melastomataceae:  Micon- 

ieae)  from  southern  Central  America.  Published  September  11,  1990  327-335 

Index  to  Volume  46  ..  ...  337-346 


The  Scientific  Publications  Committee  acknowledges  with  thanks  the  individuals  listed  below 
who  served  as  manuscript  reviewers  for  this  volume. 

Paul  H.  Arnaud,  Jr.,  William  C.  Burger,  Daniel  M.  Cohen,  James  C.  Cokendolpher,  Thomas  Croat, 
W.  G.  D'Arcy,  L.  H.  Durkee,  Barry  E.  Hammel,  Robert  W.  Henderson,  James  Henrickson,  Walter 
Ivantsoff,  Tomio  Iwamoto,  Vincent  F.  Lee,  Douglas  F.  Markle,  Thomas  Morley,  Lucinda  McDade, 
Theodore  J.  Papenfuss,  Walter  J.  Rainboth,  Richard  H.  Rosenblatt,  Jay  M.  Savage,  W.  O.  Stevens, 
Carol  Todzia,  William  J.  Turner,  Harold  K.  Voris,  Dieter  C.  Wasshausen,  Robert  L.  Wilbur,  Richard 
W.  Winterbottom,  John  J.  Wurdack. 


PROCEEDINGS 

OF  THE 

CALIFORNIA  ACADEMY  OF  SCIENCES 


Vol.  46,  No.  1,  pp.  1-72,  21  figs.,  6  tables.  December  7,  1988 


AN  ECOGEOGRAPHIC  ANALYSIS  OF  THE  HERPETOFAUNA  OF 

THE  SIERRA  SAN  PEDRO  MARTIR  REGION,  BAJA  CALIFORNIA, 

WITH  A  CONTRIBUTION  TO  THE  BIOGEOGRAPHY  OF  THE 

BAJA  CALIFORNIA  HERPETOFAUNA 

By 

Hartwell  H.  Welsh,  Jr. 

Pacific  Southwest  Forest  and  Range  Experiment  Station,  Forest  Service, 

U.S.  Department  of  Agriculture,  1700  Bayview  Drive, 

Arcata,  California  95 52 11 


ABSTRACT:  An  ecogeographic  analysis  of  the  distributions  of  the  herpetofauna  of  the  Sierra  San  Pedro 
Martir  Region  of  Baja  California  was  undertaken.  The  Martir  Region  is  part  of  northern  Baja  California 
from  approximately  latitude  30°N  to  31°N.  Over  3,000  locality  records  from  the  literature,  museums,  and  field 
work  were  analyzed  across  a  matrix  of  ecogeographic  formations  based  on  regional  climates,  physiography, 
and  extant  vegetation.  Numerical  and  heuristic  methods  of  biogeographic  analysis  indicated  the  65  species 
present  occur  in  seven  distinct  patterns.  Examination  of  these  contemporary  patterns  within  and  beyond  the 
Martir  Region  led  to  a  hypothesis  of  five  historical  patterns  of  evolutionary  development  among  the  Baja 
California  herpetofauna.  This  hypothesis  indicates  that  vicariance,  dispersal  (diffusion),  and  in  situ  evolution 
are  interacting  phenomena  contributing  to  contemporary  herpetofaunal  distributions  in  peninsular  and  insular 
Baja  California  and  the  Pacific  Southwest. 

Received  Sept.  30,  1987.  Accepted  March  21,  1988. 

ences,  and  concluded  that  "...  continued  pro- 

INTRODUCTION  .  .  .          c  .,  f 

gress  in  solving  the  questions  of  the  origins  of 

There  is  a  growing  consensus  that  ecological  biogeographic  areas  will  require  biogeographers 

factors  limit  the  geographic  range  of  all  organ-  to  clearly  distinguish  the  origins  of  species  from 

isms  (e.g.,  Udvardy  1969;  MacArthur  1972).  Ro-  the  origins  of  areas  and  then  develop  new  meth- 

tramel  (1973:229)  expressed  it  thusly:  ".  .  .  bio-  ods  of  eco-geographic  analysis"  (Rotramel  1973: 

geographic   areas   originate   from   and   are  230). 

maintained  by  the  interactions  of  organisms  with  The  ecogeographic  characteristics  of  a  species' 

physical  and  biotic  factors  in  their  environ-  distribution  can  be  viewed  as  the  present  sum- 

ments."  He  noted  a  lack  of  placing  proper  em-  mation  of  the  continuous  evolutionary  process 

phasis  on  ecological  factors  in  biogeographic  of  interaction  and  compromise  that  occurs  be- 

analysis,  which  he  attributed  to  historical  influ-  tween  that  species  and  its  environment.  A  new 

adaptation  to  the  environment  that  promotes 

greater  survivorship  of  offspring  within  an  area, 

1  Formerly  with  the  U.S.  Fish  and  Wildlife  Service,  National       & 

Ecology  Center,  Dept.  of  interior,  1 300  Blue  Spruce  Drive,  Ft.      or  population  expansion  into  new  areas,  is  a  new 
Collins,  Colorado  80524.  characteristic  of  the  ecological  strategy  of  that 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


species,  and  a  new  chapter  in  its  evolutionary 
development.  While  congruent  areas  for  different 
species  are  theoretically  impossible  if  the  areas 
are  denned  strictly  in  terms  of  the  organisms' 
ecological  niches,  groups  of  species— both  plant 
and  animal— do  coexist  and  coevolve  in  loose 
congruency  in  biotic  communities.  Given  that 
the  ecological  parameters  of  a  species  are  indic- 
ative of  the  historical  process  through  which  that 
species  has  evolved,  it  is  logical  to  assume  that 
a  similarity  of  ecogeographic  constraints  among 
populations  of  different  species  in  the  same  area 
may  indicate  a  common  history  for  those  species. 
This  common  history  may  not  have  entailed  di- 
rect ecological  interactions  but  nonetheless  con- 
stitutes a  coevolution,  where  these  species'  dis- 
tributions are  similar  even  when  examined  on  a 
scale  approaching  that  of  microhabitat.  The  elu- 
cidation of  such  similar  ecogeographic  patterns 
is  the  fundamental  methodology  of  historical 
biogeography,  and  their  elucidation  and  inter- 
pretation comprises  the  essence  of  the  discipline. 
The  remote,  mostly  inhospitable,  and  sparsely 
populated  peninsula  of  Baja  California,  pre- 
served in  a  relatively  pristine  state,  is  an  ideal 
natural  laboratory  for  the  study  of  phenomena 
that  influence  and  regulate  the  distributions  of 
terrestrial  vertebrates.  The  biotic  diversity  and 
general  community  structure  of  the  flora  and  fau- 
na of  the  Baja  California  peninsula  was  reported 
by  Nelson  (1921).  Schmidt  (1922)  made  the  first 
comprehensive  study  of  the  diverse  herpeto- 
fauna  of  the  Peninsular  Region  (peninsula  and 
related  islands),  describing  peninsular  distribu- 
tions on  the  basis  of  Nelson's  Faunal  Districts. 
Savage  (1960)  presented  a  classic  historical  anal- 
ysis of  the  distributions  of  the  peninsular  her- 
petofauna,  adding  significantly  to  our  under- 
standing of  the  origins  of  these  distributions  and 
those  of  the  entire  North  American  herpeto- 
fauna.  However,  Savage's  analysis  had  three 
shortcomings:  (1)  he  assumed  that  the  peninsula 
had  existed  as  a  more  or  less  stable  physiographic 
unit  since  the  Eocene;  (2)  he  chose  not  to  incor- 
porate vital  evidence  provided  by  the  regional 
insular  distributions;  and  (3)  despite  noting  1 3 
different  distribution  patterns  among  the  pen- 
insular herpetofauna,  he  based  his  numerical 
comparison  and  subsequent  distributional  anal- 
ysis on  only  four  Herpetofaunal  Areas,  appar- 
ently modeled  after  Nelson's  Faunal  Districts. 
This  resulted  in  an  oversimplification  of  distri- 


bution patterns  and  relationships,  obscuring  some 
and  entirely  concealing  others.  Nelson's  (1921) 
Faunal  Districts  and  Savage's  slightly  more  re- 
fined Herpetofaunal  Areas  give  a  general  picture 
of  major  herpetofaunal  distribution  trends  on 
peninsular  Baja  California,  but  under  close  scru- 
tiny, they  fail  to  delineate  ecological  or  geograph- 
ic distribution  barriers  for  much  of  the  herpe- 
tofauna. 

Greater  ecological  and  geographical  resolution 
is  now  possible  with  the  extensive  taxonomic  and 
distributional  data  available  from  more  recent 
studies  of  the  regional  herpetofauna.  Loomis  et 
al.  (1974)  compiled  an  updated  checklist  and  re- 
ported the  general  distributions  of  the  Baja  Ca- 
lifornia herpetofauna.  Linsdale  (1932),  Tevis 
(1944),  Murray  (1955),  Bostic  (1971),  Murphy 
(1975,  1976,  1 983a),  Welsh  (1976a),Seib(  1980), 
and  Welsh  and  Bury  (1984)  have  investigated 
community  distributions  in  various  parts  of  the 
peninsula.  Numerous  other  authors  have  con- 
tributed to  knowledge  of  distributions  and  ecol- 
ogy for  single  species  of  the  regional  herpeto- 
fauna (see  species  accounts  below).  However,  a 
comprehensive  ecogeographic  study  of  the  entire 
Baja  California  herpetofauna  has  yet  to  be  con- 
ducted. Murphy's  (1983a)  study  is  the  best  at- 
tempt to  date;  using  new  distributional  data  from 
remote  areas  of  southern  Baja  California  and  the 
Gulf  islands,  he  defined  more  accurate  southern 
peninsular  biotic  subdivisions  for  analysis. 

Murphy  lacked  complete  and  accurate  distri- 
butional data  for  some  species  in  the  northern 
areas;  his  analysis  emphasized  the  southern  pen- 
insular patterns.  He  lumped  several  northern  dis- 
tribution patterns  together  in  a  "marginal  track" 
and,  I  submit,  underplayed  their  overall  signif- 
icance as  evidence  of  important  evolutionary 
trends  complementing  the  tectonic  events  that 
shaped  the  distributions  of  much  of  the  Baja  Ca- 
lifornia herpetofauna.  The  northern  peninsular 
patterns  are  important  because  the  north  is  the 
proposed  "gate"  for  species  migrating  down  from 
the  continent  (Savage  1 960),  and  the  majority  of 
the  lineages  present  in  Baja  California  are  rep- 
resented there.  Its  diversity  of  habitat  types, 
stemming  from  its  marked  elevational  gradient, 
and  from  its  situation  between  tropical  and  tem- 
perate latitudes,  is  the  greatest  on  the  peninsula. 

This  study  focuses  on  the  northern  area  of  the 
peninsula,  specifically  on  the  Sierra  San  Pedro 
Martir  Region  of  north  central  Baja  California 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


Sierra   San   Pedro 
Martir    Region 


Museum  of  Vertebrate  Zoology 

University  of  California 


FIGURE  1 .    The  location  of  the  Sierra  San  Pedro  Martir  Region,  Baja  California  Norte,  Mexico. 


(hereafter  referred  to  as  the  Martir  Region,  or 
simply  the  Region).  The  Martir  Region  consists 
of  that  area  of  northern  Baja  California  from 
approximately  30°N  to  31°N  latitude  (Fig.  1).  It 
contains  65  species  of  reptiles  and  amphibians. 
Four  of  Nelson's  five  Faunal  Districts  and  three 
of  Savage's  four  Herpetofaunal  Areas  intersect 


within  the  Region  (Fig.  2a).  Hastings  and  Turner 
(1965)  investigated  the  climates  of  Baja  Califor- 
nia. They  correlated  their  data  with  the  work  of 
Shreve  and  Wiggins  (1964)  on  plant  distributions 
and  described  six  phytogeographic  divisions  in 
Baja  California.  Three  of  these  vegetation  areas 
intersect  in  the  Martir  Region  (Fig.  2b). 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


FIGURE  2.  Biotic  and  climatic  subdivisions  of  the  Sierra  San  Pedro  Martir  Region,  Baja  California,  Mexico,  a)  Herpetofaunal 
Areas  (Savage  1960):  (1)  Californian;  (2)  Colorado  Desert;  (3)  Peninsular  Desert,  (b)  Phytogeographic  Provinces  (after  Shreve 
and  Wiggins  [1964]  and  Hastings  and  Turner  [1965]):  (1)  San  Pedro  Martir;  (2)  Lower  Colorado  Valley;  (3)  Vizcaino,  (c) 
Geographic  relationships  of  the  major  climatic  areas:  (1)  Pacific  Climate;  (2)  Montane  Climate;  (3)  Gulf  Climate;  (4)  Central 
Desert  Climate. 


My  approach  was  to  formulate  a  base  map  of 
regional  habitats  with  which  to  compare  distri- 
butions and  seek  evidence  of  patterns  across 
species.  Such  evidence  was  then  used  to  define 
potential  historical  groupings.  Independent  evi- 
dence of  congruency  among  overall  distributions 
was  then  sought  as  corroboration  for  the  group- 
ings derived  from  the  regional  analysis.  Such  cor- 
roboration I  considered  as  evidence  that  the  re- 
gional groupings  are  each  the  result  of  a  common 
evolutionary  history  among  its  members,  and 
not  the  result  of  random  processes. 

ACKNOWLEDGMENTS 

I  want  to  dedicate  this  work  to  the  memory  of 
James  R.  Koplin,  a  friend  and  mentor  who 
showed  me  the  real  value  of  a  sense  of  humor. 
The  Direction  General  de  la  Fauna  Silvestre  de 
Mexico  is  gratefully  acknowledged  for  providing 
me  the  permits  to  do  this  research.  My  family, 
Aida  Meling,  and  the  Observatorio  Nacional  de 
Mexico  generously  provided  field  logistical  sup- 
port. C.  Olen,  A.  Szurak,  R.  Bebout,  A.  Chakos, 
L.  Harlow,  F.  Adams,  A.  Spolsky,  R.  Stebbins, 
D.  Wake,  J.  Koplin,  D.  Kitchen,  J.  Houck,  K. 
Tollestrup,  A.  Lind,  R.  Drewes,  and  R.  B.  Bury 
provided  support,  assistance,  and  encourage- 
ment along  the  way.  I  am  grateful  to  M.  Bogen, 
T.  Fritts,  R.  B.  Bury,  H.  M.  Smith,  R.  Etheridge, 


T.  Papenfuss,  K.  Tollestrup,  H.  Greene,  M.  Ha- 
ney,  A.  Lind,  D.  Waters,  and  three  anonymous 
reviewers  for  helpful  comments  on  earlier  drafts. 
I  thank  C.  Houle  and  R.  Nelson  for  help  with 
tables  and  graphics,  and  L.  Mandell-Lewis  for 
the  summary  in  Spanish.  T.  Fritts  of  the  San 
Diego  Natural  History  Museum,  J.  Wright  of  the 
Los  Angeles  County  Museum  of  Natural  His- 
tory, A.  Leviton,  R.  Drewes,  and  J.  Vindim  of 
the  California  Academy  of  Sciences,  and  R.  Steb- 
bins of  the  Museum  of  Vertebrate  Zoology,  Uni- 
versity of  California,  Berkeley,  kindly  provided 
locality  records  from  their  respective  institu- 
tions. I  wish  to  acknowledge  the  National  Ecol- 
ogy Center  of  the  U.S.  Fish  and  Wildlife  Service, 
the  Mattole  Institute,  Humboldt  State  Univer- 
sity, the  Museum  of  Vertebrate  Zoology  of  the 
University  of  California,  and  the  Theodore  Roo- 
sevelt Memorial  Fund  for  providing  support  or 
facilities  that  helped  bring  this  project  to  fruition. 

MATERIALS  AND  METHODS 

The  following  procedures  were  followed  in  the 
investigations  of  the  distributional  patterns  of 
the  herpetofauna  of  Baja  California:  (1)  deri- 
vation of  a  base  map  of  the  ecogeographic  for- 
mations of  the  Sierra  San  Pedro  Martir  Region 
on  the  basis  of  climatic  data,  physiographic  fea- 
tures, and  extant  vegetation  associations;  (2) 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


compilation  of  locality  and  ecological  data  for 
the  65  herpetofaunal  species  occurring  within  the 
Martir  Region  from  field  investigations,  museum 
records,  and  available  literature;  (3)  plotting  of 
ecogeographic  distributions  of  regional  species, 
and  grouping  of  species  whose  distributions  are 
similar  and  thus  indicative  of  a  possible  histor- 
ical relationship;  (4)  assessment  of  the  validity 
of  such  proposed  historical  groupings  by  com- 
parison of  intra-Regional  patterns  with  broader 
distribution  patterns  (in  the  rest  of  Baja  Califor- 
nia and  beyond),  to  seek  corroborative  evidence 
of  a  common  evolutionary  history. 

Derivation  of  Regional  Ecogeographic 
Formations 

Heyer  (1967:259)  described  a  life  zone  as  "an 
ecological  altitudinal  or  latitudinal  zone,  char- 
acterized by  specific  climate  parameters  and  sec- 
ondarily by  vegetation."  Miller's  (1951)  "eco- 
logic  formations"  reflected  geomorphological 
components  that  sometimes  play  a  greater  role 
in  the  distribution  of  species  than  do  altitudinal 
or  vegetative  parameters.  I  have  incorporated 
both  these  concepts  in  my  division  of  the  Region 
into  "ecogeographic  formations."  These  forma- 
tions were  delineated  on  the  basis  of  both  field 
data  and  literature  pertaining  to  physiographic 
features,  climate,  and  dominant  vegetational  as- 
sociations (using  perennial  plant  species).  Bio- 
geographic  assemblages  generally  represent  gra- 
dients with  regard  to  space,  time,  and  biotic  form, 
and  boundaries  between  assemblages  are  there- 
fore at  best  approximate  and  variable,  changing 
continuously,  and  varying  somewhat  for  each 
biotic  form  (Udvardy  1969).  Therefore,  I  have 
used  ecotones  to  establish  approximate  bound- 
aries between  ecogeographic  formations. 

Compilation  of  Species  Accounts 

The  herpetofauna  throughout  the  Martir  Re- 
gion was  surveyed  from  22  June  to  13  August 
1973,  from  9  April  to  25  June  1974  (except  6 
days  in  late  May  1974),  for  2  weeks  in  June  1976, 
and  2  weeks  in  May  1979.  Much  of  the  Region 
is  accessible  by  road  and  many  areas  were  reached 
by  vehicle.  In  the  more  remote  and  inaccessible 
central  and  south  central  parts  of  the  Region  travel 
was  by  foot.  A  total  of  2,187  km  by  road  and 
766  km  by  trail  were  logged  during  these  inves- 
tigations. Approximately  80-90%  of  the  areal  ex- 


tent of  the  Martir  Region  was  observed.  Eleva- 
tion (with  a  barometric  altimeter),  physiography, 
climate,  vegetation,  and  microenvironmental 
conditions  were  recorded  wherever  animals 
were  observed  or  collected.  Distributional,  eco- 
logical, and  behavioral  data  were  recorded  on 
the  65  species  that  comprise  the  herpetofauna 
of  the  Martir  Region.  Two  hundred  thirty-one 
voucher  specimens  were  collected;  now  depos- 
ited in  the  Museum  of  Vertebrate  Zoology,  Uni- 
versity of  California,  Berkeley.  Additional  lo- 
cality records  of  captured  and  released  or 
observed  animals  totaled  523.  Museum  and  lit- 
erature records  used  in  the  study  totaled  2,618. 
A  total  of  3,372  locality  records  were  obtained 
from  65  collecting  sites.  The  collecting  sites  are 
listed,  north  to  south,  in  Appendix  A,  and  are 
shown  in  Figure  3.  Distributional  data  from  the 
following  museums  were  used:  California  Acad- 
emy of  Sciences  (CAS  and  SU),  San  Diego  Nat- 
ural History  Museum  (SDNHM),  Los  Angeles 
County  Museum  of  Natural  History  (LACM), 
and  the  Museum  of  Vertebrate  Zoology,  Uni- 
versity of  California,  Berkeley  (MVZ).  Sources 
for  literature  records  are  noted  in  the  species  ac- 
counts. All  known  localities  of  museum  and  lit- 
erature records  were  visited  to  determine  existing 
vegetational  and  physiographic  features,  and  to 
note  available  habitat  types.  Phylogenetic  orga- 
nization of  families  in  the  species  accounts  fol- 
lows Stebbins  (1985).  Subspecific  nomenclature 
was  used  in  all  cases  where  subspecies  have  been 
described;  otherwise  the  currently  recognized 
species  names  were  used.  Common  names  used 
are  from  Collins  et  al.  (1978)  or  Stebbins  (1985). 

Derivation  of  Regional  Distribution  Patterns 

Two  different  methods  were  used  to  compare 
the  distributions  of  regional  species.  Following 
Savage  (1960)  and  Murphy  (1983a),  numerical 
techniques  of  analysis  were  used.  Peters  (1971) 
pointed  out  that  although  numerical  methods  in 
biogeography  are  capable  of  making  quantitative 
distinction  between  areas,  they  cannot  distin- 
guish whether  said  areas  have  a  real  basis  in  na- 
ture. This  problem  was  addressed  by  using  re- 
gional ecogeographic  formations  as  the  primary 
areas  for  numerical  comparison  instead  of  pre- 
established  faunal  or  vegetation  provinces  or  the 
equal  quadrants  technique  often  used  with  nu- 
merical faunal  analyses. 

Secondly,  a  two-phased  heuristic  method  was 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


FIGURE  3.  Location  of  collecting  stations  in  the  Sierra  San  Pedro  Martir  Region,  Baja  California  Norte,  Mexico.  Stations 
20,  21,  25,  30,  31,  36,  40-43,  51,  and  60  have  been  omitted;  these  stations  are  adjacent  to  the  next  highest  number,  e.g.,  station 
20  and  21  are  near  station  22.  Exact  locations  and  elevations  of  stations  are  in  Appendix  A.  Topographic  intervals  equal  610 
m  (2,000  ft). 


used  to  analyze  regional  distributions.  Phase  one 
consisted  of  a  comparison  of  distributions  across 
a  transect  of  regional  ecogeographic  formations, 
and  phase  two  was  a  comparison  of  overall  re- 
gional distributions  as  plotted  on  maps  based  on 
Figures  3,  4,  and  5. 

My  numerical-heuristic  strategy  of  analysis  was 
similar  to  the  approach  of  Morafka  ( 1 977)  except 
that  he  used  two  numerical  techniques,  faunal 
resemblance  coefficients  (two  formulas)  and  an 
equal  quadrants  analysis.  I  used  three  faunal  re- 
semblance coefficients  and  no  equal  quadrants 
analysis.  Both  the  numerical  and  heuristic  meth- 
ods are  discussed  in  more  detail  in  the  section 
on  zoogeographic  analysis. 


Distributions  Beyond  the  Martir  Region 

Distributional  congruency  beyond  the  Martir 
Region  for  those  taxa  linked  ecogeographically 
within  the  Martir  Region  was  considered  to  be 
indicative  of  possible  historical  links  and  a  shared 
evolutionary  history.  Evidence  for  such  extra- 
regional  congruency  was  sought  using  indepen- 
dent means.  The  method  of  establishing  and 
comparing  the  combined  distributions  of  related 
forms,  conspecific  or  congeneric,  or  both  (indi- 
vidual tracks),  in  order  to  establish  or  deny  broad 
geographical  congruency  between  unrelated 
taxa— and  thus  demonstrate  more  general  dis- 
tribution patterns  (generalized  tracks)— follows 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


the  principles  and  method  indicated  by  Croizat 
(1964). 

RESULTS 
THE  ENVIRONMENT 

The  Sierra  San  Pedro  Martir  Region  refers  to 
that  area  of  northern  Baja  California  between  the 
Pacific  Ocean,  the  Gulf  of  California  (Sea  of  Cor- 
tez),  and  latitudes  3 1°20'N  and  30°00'N  (Fig.  1), 
an  area  of  approximately  1 50  km2.  The  mountain 
range  central  to  the  Region,  the  Sierra  San  Pedro 
Martir,  hereafter  referred  to  as  the  Sierra,  is  over 
3,000  m  in  elevation,  and  transects  the  center  of 
the  Martir  Region  from  north  to  south.  The  Col- 
orado Desert  is  east  of  the  Sierra,  and  foothill 
ranges  and  the  San  Quintin  Plain  are  to  the  west. 
The  Martir  Region  is  at  the  southern  extreme  of 
the  north  temperate  zone  and  is  influenced  by 
two  diverse  climates,  the  Pacific  and  Gulf  re- 
gimes (Hastings  and  Turner  1965). 

Physiographic  Features 

The  topography  and  geography  of  the  Sierra 
San  Pedro  Martir  were  described  by  Nelson 
(1921)  and  Henderson  (1960).  Woodford  and 
Harris  (1938)  described  the  geology  and  geo- 
morphology  of  the  Region,  and  Allison  (1964) 
discussed  the  topographic  relationships  of  the 
Martir  Region  with  the  features  of  surrounding 
regions. 

Information  on  the  general  physiography  of 
the  Martir  Region  was  obtained  from  topograph- 
ic maps  (Estados  Unidos  Mexicanos,  Series  501), 
flight  navigation  charts  (U.S.  Dept.  Commerce 
CH-22,  1979),  and  Robinson  (1972).  Based  on 
these  sources  and  my  field  investigations,  I  sub- 
divided the  Region  into  nine  primary  topograph- 
ic elements,  each  with  a  unique  and  relatively 
homogeneous  physiography  (Fig.  4). 

The  Sierra  consists  of  an  elevated,  granitic  fault 
block  dated  from  the  mid-Cretaceous  (Silver  et 
al.  1 956).  The  fault  block  is  a  segment  of  the  Baja 
California  Cordillera,  and  is  considered  part  of 
the  Peninsular  Range  Physiographic  Province 
(Allison  1964),  which  extends  from  the  Trans- 
verse Ranges  of  southern  California  (at  the  lat- 
itude of  Los  Angeles)  south  along  the  length  of 
Baja  California. 

The  northern  extremity  of  the  Sierra  is  at  San 
Matias  Pass  (980  m),  which  is  probably  a  trans- 


verse fault  (Allen  et  al.  1956).  The  southern  ex- 
tremity of  the  Sierra  is  in  the  vicinity  of  Cerro 
Matomi  (30°24'N,  1 1 5°10'W),  south  of  which  the 
Sierra  gradually  decreases  in  elevation  and  is 
covered  by  lava-capped  mesas.  Here,  the  moun- 
tains grade  into  the  North  Central  Desert.  The 
area  from  Cerro  Matomi  south  to  30°N  latitude 
was  designated  the  North  Central  Desert  To- 
pographic Element  (Fig.  4). 

The  eastern  slope  of  the  Sierra,  designated  the 
East  Scarp  Topographic  Element  (Fig.  4),  is  a 
steep  and  fairly  straight  fault  scarp  rising  1,220- 
2,120  m  above  the  desert  floor.  A  series  of  steep 
canyons  transect  this  scarp,  and  terminate  in 
piedmont  alluvial  fans,  opening  onto  Valle  de 
San  Felipe  and  Valle  Chico,  two  continuous  des- 
ert valleys  paralleling  the  scarp  base  at  about  600 
m  elevation.  These  valleys  are  part  of,  and  con- 
tiguous with,  the  southern  Colorado  Desert.  The 
Colorado  Desert  extends  from  California  and 
Arizona,  south  along  the  Gulf  coast  of  Baja  Cali- 
fornia to  the  vicinity  of  Bahia  de  Los  Angeles 
(29°00'N,  1 13°15'W).  This  desert  from  31°20'N 
latitude  south  to  Bahia  de  Los  Angeles  was  des- 
ignated the  South  Colorado  Desert  Topographic 
Element  (Fig.  4).  All  topographic  elements  from 
the  ridge  of  the  East  Scarp,  east  to  the  Gulf  of 
California,  comprise  the  Gulf  Slope. 

The  west  slope  of  the  Sierra  consists  of  a  series 
of  lesser  fault  scarps  with  intervening  steps  be- 
tween that  drop  more  gradually  than  the  East 
Scarp.  They  range  from  915  m  to  1,525  m.  This 
area  is  designated  the  West  Scarp  Topographic 
Element.  Below  the  West  Scarp  Topographic  Ele- 
ment is  a  series  of  foothill  ranges  that  drop  grad- 
ually to  a  coastal  plain.  The  foothill  ranges  and 
the  coastal  plain  are  designated  the  Western 
Foothills,  and  the  San  Quintin  Plain  Topograph- 
ic Elements,  respectively  (Fig.  4). 

The  corrugated  Main  Scarp  of  the  Sierra  San 
Pedro  Martir  is  13-16  km  wide,  64  km  long,  and 
is  bordered  and  transected  by  irregular,  boulder- 
stacked  ridges.  The  Main  Scarp  contains  four 
major,  and  numerous  minor,  alluvial-filled, 
meadow  expanses;  several  of  these  meadows 
contain  year-round  surface  water.  The  Main  Scarp 
was  subdivided  into  the  Upper  Main  Scarp  To- 
pographic Element,  consisting  of  those  areas  north 
of  Tasajera  Ridge  and  above  2,300  m  (7,550  ft), 
and  the  Lower  Main  Scarp  Topographic  Ele- 
ment, consisting  of  those  areas  south  of  Tasajera 
Ridge  and  below  2,300  m  (Fig.  4).  Tasajera  Ridge 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


FIGURE  4.  Geographic  relationships  of  the  major  topographic  elements  of  the  Sierra  San  Pedro  Martir  Region,  Baja  California 
None,  Mexico.  (1)  San  Quintin  Plain;  (2)  Western  Foothills;  (3)  West  Scarp;  (4)  Upper  Main  Scarp;  (5)  Lower  Main  Scarp;  (6) 
East  Scarp;  (7)  South  Colorado  Desert;  (8)  North  Central  Desert;  (9)  Western  Arroyos. 


transects  the  Main  Scarp  at  about  midlength,  and 
runs  west  to  east  at  30°57'N  latitude. 

The  Pacific  Slope  is  west  of  the  crest  of  the 
Sierra  San  Pedro  Martir  and  north  of  the  North 
Central  Desert  Topographic  Element  (Fig.  4).  It 
is  transected  by  a  system  of  westerly  directed 
canyons,  carrying  partly  subterranean  runoff  from 
the  Sierra,  72  km  west  to  the  Pacific.  These  ar- 
royos  were  designated  the  Western  Arroyos  To- 
pographic Element  (Fig.  4).  In  subsequent  sec- 
tions, the  suffix  "Topographic  Element"  has  been 
omitted,  and  the  capitalized  geographic  entity  is 
used  to  designate  these  topographic  divisions  (e.g., 
West  Scarp  Topographic  Element  =  West  Scarp). 


Regional  Climates 

The  climate  of  the  Martir  Region  results  from 
the  interaction  of  regional  topography  with  two 
major  climatic  regimes  that  govern  weather  con- 
ditions over  most  of  Baja  California.  The  cli- 
matic regimes  consist  of  the  relatively  cool,  moist 
Pacific  Coastal  Regime,  produced  by  the  cold 
California  Current  with  its  prevailing  westerly 
winds,  and  the  considerably  warmer  drier  Gulf 
Regime  (Hastings  and  Turner  1965;  Meigs  1966). 

Hastings  and  Turner  (1965)  postulated  that 
seasonal  differences  in  variability  and  amount  of 
precipitation  produce  the  distinctive  vegetation- 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


al  associations  of  their  phytogeographic  prov- 
inces (Fig.  2b).  Lacking  quantitative  data,  they 
lumped  the  Sierra  San  Pedro  Martir  with  all  of 
northwestern  Baja  California  in  the  San  Pedro 
Martir  Phytogeographic  Province.  My  observa- 
tions within  this  province  indicated  more  com- 
plex vegetational  relationships.  I  found  this  mon- 
tane area  to  have  a  climate  and  vegetation 
sufficiently  unique  to  warrant  separate  consid- 
eration. 

The  Sierra  has  predictable  winter  and  spring 
precipitation,  often  falling  as  snow  in  the  winter 
(J.  Alonso,  Mexican  National  Observatory,  pers. 
comm.).  A  summer  rainfall  pattern  results  from 
a  thermal  low  pressure  cell  to  the  east  of  the 
Sierra,  causing  moist  air  to  flow  east  across  the 
mountains.  Afternoon  thunder  showers  occur  al- 
most daily  from  late  June  through  September  as 
a  result  of  orographic  uplift.  Data  from  San  Juan 
de  Dios  (elevation  1 ,400  m)  in  the  Sierra  Juarez, 
1 44  km  north  of  the  Sierra  San  Pedro  Martir, 
show  the  highest  annual  precipitation  (33.4  cm) 
yet  reported  for  northern  Baja  California  (Has- 
tings and  Humphrey  1969).  The  Main  Scarp  of 
the  Sierra  San  Pedro  Martir  is  considerably  higher 
in  elevation  than  the  Sierra  Juarez  and  thus  prob- 
ably receives  the  highest  annual  precipitation  in 
Baja  California. 

The  precipitation  percolates  down  through  the 
granitic  sand  and  the  soil  surface  dries  rapidly  in 
the  warm  summer  air.  Flash-flooding  from  un- 
usually heavy  rainfall  occurs  occasionally  on  all 
scarps.  Lightning  regularly  accompanies  these 
storms  and  is  responsible  for  fires  on  the  forested 
plateau.  Underbrush  is  scarce  throughout  much 
of  the  forest,  perhaps  due  to  fires. 

The  Main  Scarp,  and  the  upper  reaches  of  the 
West  and  East  scarps,  areas  influenced  by  the 
high  elevation  weather  system  of  the  Sierra,  have 
been  designated  the  Montane  Climatic  Area  (Fig. 
2c). 

Those  areas  on  the  west  side  of  the  Sierra  San 
Pedro  Martir,  encompassing  the  lower  elevations 
of  the  West  Scarp,  the  Western  Foothills,  and  the 
Western  Arroyos  were  designated  the  Pacific  Cli- 
matic Area  (Fig.  2c).  This  area  is  characterized 
by  frequent  fog,  cool  oceanic  temperatures,  and 
relatively  predictable  winter  and  spring  precip- 
itation (Table  1 ,  Western  Foothills). 

Those  areas  on  the  east  slope  of  the  Sierra  San 
Pedro  Martir,  and  corresponding  with  the  south- 
ern portion  of  the  Lower  Colorado  Valley  Phy- 


togeographic Province  (Table  1),  were  designated 
the  Gulf  Climatic  Area;  it  encompasses  the  lower 
East  Scarp  and  the  South  Colorado  Desert  To- 
pographic Elements  (Fig.  2c;  Table  1).  This  area 
is  one  of  the  hottest  places  in  the  world  (Meigs 
1953)  and  is  characterized  by  a  lack  of  predict- 
able rainfall.  Shreve  (1934)  remarked  that  this 
area  may  be  the  most  sparsely  vegetated  of  any 
in  North  America. 

The  area  designated  the  Central  Desert  Cli- 
matic Area  corresponds  with  the  northern  por- 
tion of  the  Vizcaino  Phytogeographic  Province 
(Table  1),  and  encompasses  the  North  Central 
Desert  Topographic  Element  (Fig.  2c;  Table  1). 
This  area  has  the  driest  summers  of  the  four 
climatic  areas.  The  lack  of  rainfall  in  the  Central 
Desert  is  mitigated  by  cool  Pacific  air  and  fre- 
quent fogs,  which  provide  considerable  ground 
moisture.  Bostic  (1971)  described  the  climate  of 
the  Central  Desert  in  greater  detail. 

The  narrow  San  Quintin  Plain  Topographic 
Element,  extending  northward  to  the  vicinity  of 
latitude  30°20'N,  was  included  in  the  Central 
Desert  Climatic  Area  (Fig.  2c).  Hastings  and 
Turner  (1965)  included  this  area  with  north- 
western Baja  California,  placing  it  in  their  San 
Pedro  Martir  Phytogeographic  Province.  Cli- 
matic data  from  this  coastal  plain  were  analyzed 
and  compared  with  similar  data  from  adjacent 
topographic  elements  to  the  northeast  and  south 
(Table  1).  The  climate  of  the  San  Quintin  Plain 
falls  between  that  of  the  two  adjacent  areas,  but 
seasonal  precipitation  patterns  are  most  similar 
to  the  Central  Desert  Climatic  Area.  Floral  and 
faunal  characteristics  of  the  San  Quintin  Plain 
indicate  that  this  area,  and  the  proximate  inland 
valleys,  are  within  the  influence  of  the  Central 
Desert  Climate  (Short  and  Crossin  1967;  Bostic 
1968;  this  study). 

References  to  the  four  climatic  areas  of  the 
Martir  Region  will  use  this  format:  Gulf  Climatic 
Area  =  Gulf  Climate. 

Vegetative  Elements 

Following  the  approach  of  a  similar  regional 
study  conducted  by  Schoenherr  ( 1 9  7  6)  in  the  San 
Gabriel  Mountains  of  southern  California,  the 
structure  and  distribution  of  Martir  Region  vege- 
tation communities  were  determined.  Nelson 
(1921)  reported  on  an  extensive  natural  history 
survey  of  Baja  California,  and  described  vege- 


10 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


TABLE  1 .    SUMMARY  OF  CLIMATIC  DATA  FROM  THE  SIERRA  SAN  PEDRO  MARTIR  REGION,  BAJA  CALIFORNIA  NORTE,  MEXICO; 
REPORTED  BY  PHYTOGEOGRAPHIC  PROVINCE  AND  TOPOGRAPHIC  ELEMENT  (Fie.  4). 


Geographic  area 


Winter 


Spring 


Summer 


Fall 


Annual 


Number  of 
weather 
stations' 


Phytogeographic  province2 
Lower  Colorado  Valley 
Vixcaino 
San  Pedro  Martir 

Topographic  clement3 
South  Colorado  Desert 
San  Quintin  Plain 
North  Central  Desert 
Western  Foothills 
Sierra  Scarps4 


Mean  precipitation  (cm) 


1.9 
5.6 
9.2 

2.3 
8.1 
5.5 
9.7 


0.8 

1.0 
4.1 


0.7 
2.8 
1.3 
3.5 


1.4 
1.1 
1.8 


1.3 
1.6 
0.4 
4.0 


2.0 

2.8 

2.7 


2.0 

2.1 
2.1 
4.3 


Mean  temperature  (°C) 


6.1 
10.5 
17.8 

6.5 
13.5 

9.5 

18.9 

(33.4) 


13  + 

27 


Topographic  element 

South  Colorado  Desert 

15.0 

21.8 

31.8 

24.8 

23.2 

2 

San  Quintin  Plain 

13.2 

15.2 

19.3 

18.0 

16.4 

4 

North  Central  Desert 

13.4 

16.3 

23.3 

19.5 

19.1 

3 

Western  Foothills 

11.9 

16.6 

22.7 

18.7 

17.2 

7 

Sierra  Scarps 

(no  data) 

0 

1  See  Hastings  and  Humphrey  (1969)  for  exact  locations. 

2  Data  from  Hastings  and  Turner  (1965). 

3  Mean  precipitation  and  temperature  data  were  summarized  from  weather  stations  within  each  topographic  element;  data 
are  from  Hastings  and  Humphrey  (1969). 

4  Data  from  Sierra  Juarez;  see  text. 


tational  characteristics  and  listed  dominant 
species  in  accordance  with  Merriam's  (1 898)  life 
zones.  Wiggins  (1944)  surveyed  the  Sierra  San 
Pedro  Martir  and  surrounding  lowlands,  Shreve 
(1936)  described  the  complex  vegetation  west  of 
the  Sierra,  and  Chambers  (1955)  described  the 
flora  of  a  canyon  on  the  East  Scarp. 

Seven  vegetational  associations  are  recognized 
within  the  Martir  Region  (Fig.  5):  Coniferous 
Forest,  Chaparral,  Pinyon-Juniper  Woodland, 
Creosote  Bush  Scrub,  Coastal  Sage  Scrub,  Cen- 
tral Desert  Scrub,  and  Riparian  Woodland.  These 
vegetational  associations  are  modeled  after  the 
communities  of  Munz  and  Keck  (1949,  1959) 
except  where  noted  below.  The  dominant  species 
of  each  association  are  listed  in  Table  2. 

Coniferous  Forest 

Coniferous  Forest  vegetation  occurs  above 
1,830  m  on  the  Sierra  Main  Scarp,  within  areas 
influenced  by  the  Montane  Climate  (Fig.  5,  6). 
The  forest  is  open  and  parklike,  with  sparse  un- 
dergrowth of  low  shrubs,  forbs,  and  grasses. 

At  higher  elevations  (over  2,400  m),  vegeta- 


tional elements  characteristic  of  Merriam's 
(1898)  Canadian  and  Hudsonian  zones  (lodge- 
pole  pine  [Pinus  murrayana],  white  fir  [Abies 
concolor],  and  quaking  aspen  [Populus  tremu- 
loides])  are  intermixed  with  the  dominant  ele- 
ments of  Coniferous  Forest.  This  impoverished 
boreal  vegetation  is  best  developed  in  deep  can- 
yons and  on  north-facing  slopes  of  the  Upper 
Main  Scarp  at  the  north  end  of  the  Sierra. 

Chaparral 

Chaparral  covers  extensive  areas  of  the  Pacific 
Slope  of  the  Sierra,  between  1,220  m  and  2,120 
m,  and  occurs  in  a  narrow  belt  on  the  East  Scarp 
between  1,670  m  and  2,120  m  (Fig.  5,  7).  The 
dense  and  often  impenetrable  vegetation  is  com- 
monly 2.5-3.0  m  in  height.  Distribution  of  Chap- 
arral on  both  versants  corresponds  with  the  steep 
scarp  faces  where  the  Montane  Climate  inter- 
grades  with  the  drier  and  warmer  Pacific  and 
Gulf  climates. 

Within  the  Chaparral,  and  particularly  in 
shaded  canyons  and  areas  with  surface  and  sub- 
surface water,  are  patches  of  oak  woodland.  Oak 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


11 


miles  7 
6    km  II 


Creosote  Bush  Scrub 
Pinyon-Juniper  Woodland 
Coastal  Sage  Scrub 
Chaparral 
Coniferous  Forest 
Central  Desert  Scrub 


FIGURE  5.  Vegetational  associations  of  the  Sierra  San  Pedro  Martir  Region,  Baja  California  Norte,  Mexico.  Map  is  based 
on  U.S.  Dept.  Commerce,  CH-22  (1979).  Distribution  of  vegetation  is  based  on  Shreve  and  Wiggins  (1964),  Short  and  Crossin 
(1967),  and  field  investigations.  Riparian  Woodland  Association  occurs  in  all  other  vegetation  types. 


woodland  in  the  Sierra  San  Pedro  Martir  occurs 
primarily  as  riparian  habitat,  and  is  considered 
with  the  Riparian  Woodland  association. 

Pinyon-Juniper  Woodland 

This  vegetational  type  occurs  between  920  m 
and  1,830  m  on  the  East  Scarp  of  the  Sierra  (Fig. 
5,  8);  this  distribution  coincides  with  areas  of 
higher  elevation  dominated  by  the  Gulf  Climate. 


Being  high  on  the  east  versant  of  the  Sierra,  Pin- 
yon-Juniper Woodland  is  probably  also  influ- 
enced to  some  degree  by  the  Montane  Climate. 
The  trees  in  the  Pinyon-Juniper  Woodland  vary 
from  approximately  3.0  to  10.0  m  in  height,  and 
are  widely  dispersed.  Open  shrub  growth  occurs 
throughout  the  Woodland.  This  association  ex- 
tends around  the  southern  and  northern  ends  of 
the  range  above  920  m  and  intergrades  with 
Chaparral. 


12 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


TABLE  2.    DOMINANT  SPECIES  OF  THE  VEGETATIONAL  ASSOCIATIONS  OF  THE  SIERRA  SAN  PEDRO  MARTIR  REGION,  BAJA  CA- 
LIFORNIA NORTE,  MEXICO. 


Vegetation  association 


Dominant  species 


Coniferous  Forest 


Coastal  Sage  Scrub 


Pinyon-Juniper  Woodland 


Central  Desert  Scrub 


Chaparral 


Creosote  Bush  Scrub 


Riparian  Woodland 
(Species  composition  var- 
ies greatly  with  elevation 
and  slope;  abbreviations 
refer  to  correspondence 
with  above  associations.) 


jeffery  pine  (Pi nus  Jeffrey i),  sugar  pine  (P.  lambertiana),  incense-cedar  (Libocedrus  decurrens), 

rose  sage  (Salvia  pachyphylla),  greenleaf  manzanita  (Arctostaphylos  patula),  pink-bracted 

manzanita  (A.  pringlei),  snow  bush  (Ceanothus  cordulatus),  scrub  oak  (Quercus  spp.) 
flattop  buckwheat  (Eriogonumfasciculatum),  California  sagebrush  (Artemisia  californica), 

white  sage  (Savia  apiana),  century  plant  (Agave  shawii),  prickly  pear  (Opuntia  littoralis), 

coastal  cholia  (Opuntia  prolifera),  Ephedra  californica 
pinyon  pine  (Finns  monophylla  and  P.  quadrifolia),  California  juniper  (Juniperus  californica), 

Mojave  yucca  ( Yucca  schidigera),  barrel  cactus  (Ferocactus  acanthodes),  Agave  aurea,  scrub 

oak  (Quercus  turbinella),  sugar  bush  (Rhus  ovata) 
datilillo  (Yucca  valida),  elephant  tree  (Pachycormus  discolor),  century  plant  (Agave  shawii}, 

cardon  (Pachycereus  pringlei),  cholia  (Opuntia  spp.),  Franseria  chenopodiafolia,  Ocotillo 

(Fourquiera  splendens),  Dudleya  spp. 
bigberry  manzanita  (Arctostaphylos  glauca),  Mexican  manzanita  (A.  pungens),  chamise  (Ade- 

nostoma  fasciculatum),  ceanothus  (Ceanothus  leucodermis  and  C.  greggi),  toyon  (Hetero- 

meles  arbutifolia),  red  shank  (Adenostoma  sparifolium),  scrub  oak  (Quercus  dumosa),  sugar 

bush  (Rhus  ovata) 
creosote  bush  (Larrea  tridentata),  white  bursage  (Franseria  dumosa),  ocotillo  (Fourquieria 

splendens),  mesquite  (Prosopis  juliflora),  agave  (Agave  spp.),  palo  verde  (Cericidium  micro- 

phyllum),  yucca  (Yucca  spp.),  ironwood  (Olneya  tesota),  cholia  (Opuntia  spp.),  catclaw 

(Acacia  greggi),  cardon  (Pachycereus  pringlei) 
willow  (Salix  spp.),  wild  rose  (Rosa  californica),  jeffery  pine  (Pinus  jefferyi),  incense-cedar 

(Libocedrus  decurrens)  (CF);  Oak  (Quercus  spp.),  coffeeberry  (Rhamnus  californica)  (Chp); 

Cottonwood  (Populus  fremonti)  (P-J,  CSS);  Sycamore  (Platamus  racemosa),  ceanothus  (C. 

leucodermis  and  C.  greggi  (CSS);  mesquite  (Prosopis  juliflora),  desert  willow  (Chilopsis  lin- 

earis)  (CBS  and  CDS);  fan  palm  (Washingtonia filifera)  (CDS) 


Creosote  Bush  Scrub 

Creosote  Bush  Scrub  occurs  below  1 ,070  m  on 
the  eastern  side  of  the  Sierra,  in  the  influence  of 
the  Gulf  Climate  (Fig.  5,  9).  The  physiognomy 
of  this  vegetation  is  predominantly  shrub,  0.7- 
4.8  m  tall,  and  widely  dispersed.  Creosote  Bush 
Scrub  vegetation  becomes  more  luxuriant  along 
the  base  of  the  East  Scarp,  and  to  a  lesser  degree 
along  the  immediate  Gulf  coast  of  the  South  Col- 
orado Desert.  Qualitative  observations  suggest 
more  plant  species  occur  in  these  two  marginal 
parts  of  the  South  Colorado  Desert  and  the  shrubs 
and  cacti  attain  greater  size  here  than  in  much 
of  the  central  area  of  the  desert.  This  phenom- 
enon may  be  a  result  of  increased  surface  or  sub- 
surface moisture,  as  Gulf  Slope  runoff  (from  the 
East  Scarp  and  South  Colorado  Desert),  which 
is  primarily  subterranean,  is  nearest  the  surface 
in  these  areas. 

Fieldwork  in  the  South  Colorado  Desert  was 
restricted  primarily  to  the  basal  slopes  of  the 
Sierra,  and  the  two  proximate  desert  valleys,  Valle 
de  San  Felipe  and  Valle  Chico.  The  Creosote 


Bush  Scrub  classification  is  expanded  to  include 
the  entire  South  Colorado  Desert,  south  to  lati- 
tude 29°N,  following  Shreve  and  Wiggins  (1 964). 
At  the  north  end  of  the  Sierra,  Creosote  Bush 
Scrub  extends  over  San  Matias  Pass  (980  m)  into 
Valle  de  Trinidad,  where  it  intergrades  with 
Coastal  Sage  Scrub.  South  of  the  Sierra  San  Pedro 
Martir  (beyond  30°25'N),  Creosote  Bush  Scrub 
intergrades  with  Central  Desert  Scrub  along  the 
Baja  California  Cordillera. 

Coastal  Sage  Scrub 

The  Pacific  Slope  below  1 ,220  m,  from  slightly 
beyond  the  north  limit  of  the  Martir  Region, 
south  to  the  vicinity  of  latitude  30°15'N,  is  vege- 
tationally  complex.  Shreve  (1936)  described  this 
area  from  Arroyo  Santo  Tomas  to  Arroyo  So- 
corro  as  being  within  a  10-mi  transition  zone 
(from  north  to  south)  between  Chaparral  and 
Sonoran  Desert  vegetation.  However,  the  clas- 
sification Coastal  Sage  Scrub  was  adopted  as  best 
reflecting  the  life-forms  and  species  composition 
that  dominate  most  of  this  area  (Fig.  5,  10).  The 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


13 


FIGURE  6.    The  Lower  Main  Scarp  Ecogeographic  Formation  at  Rancho  Viejo  (station  37)  in  Coniferous  Forest  adjacent 
montane  meadow. 


plants  are  mostly  "half-shrubs,  one  to  five  feet 
tall  or  somewhat  woodier  and  larger,  forming  a 
more  open  community  than  chaparral"  (Munz 
and  Keck  1949:97).  There  are  numerous  loca- 
tions along  the  lower  Pacific  Slope  of  the  Region 
where  elements  of  desert  vegetation  dominate. 
The  most  extensive  of  these  areas  is  the  San 
Quintin  Plain  (Fig.  1 1)  and  the  valleys  and  low 
slopes  immediately  to  the  east.  The  Chaparral 
occurring  on  the  lower  Pacific  Slope,  with  the 
exception  of  scattered  stands  at  high  elevations 
in  the  Western  Foothills,  is  mostly  dwarfed  and 
depauperate.  The  desert  scrub  and  Chaparral 
complexes  are  not  the  primary  associations  of 
the  lower  Pacific  Slope. 

Central  Desert  Scrub 

The  scrub  vegetation  changes  noticeably  in  the 
vicinity  of  latitude  30°15'N  on  the  Pacific  Slope. 
The  Creosote  Bush  Scrub  of  the  lower  East  Scarp 
and  Colorado  Desert  drainages,  the  Coastal  Sage 
Scrub  of  the  Western  Foothills,  and  the  Chap- 
arral of  the  southern  Sierran  slopes  interdigitate 
with  a  Sonoran  vegetation  characterized  by  sar- 
cophyllous  forms.  Wiggins  (1960)  described  this 
area  as  part  of  the  Central  Desert  Phytogeograph- 


ic  Area  of  the  Vizcaino  Province.  Following  the 
work  of  Aschman  (1959),  Wiggins  (1960),  and 
Bostic  (1971),  I  also  considered  this  area  to  have 
a  unique  biota.  The  plant  association  of  this  area 
is  referred  to  as  Central  Desert  Scrub  in  order  to 
locate  it  geographically  and  to  indicate  its  close 
structural  relationship  to  the  aforementioned 
scrub  vegetational  associations  (Fig.  5,  12). 

This  classification  is  applicable  to  the  vege- 
tation of  the  San  Quintin  Plain,  to  the  vicinity 
of  latitude  31°20'N.  Central  Desert  Scrub  vege- 
tation is  denser  and  more  lush  on  the  San  Quintin 
Plain  than  in  the  North  Central  Desert.  Sarco- 
phyllous  forms  dominate  in  many  areas  here, 
often  occurring  in  impenetrable  stands.  Bostic 
(1971)  noted  that  many  of  these  plants  are  adapt- 
ed structurally  to  collect  moisture  from  the  air 
(Yucca  valida,  Agave  spp.,  Dudleya  spp.).  The 
regular  fog  and  dew  along  this  Plain  probably 
provide  a  major  source  of  water  for  these  plants 
and  explain  the  rich  vegetation  here. 

Riparian  Woodland 

This  plant  association  (Fig.  13)  is  based  pri- 
marily on  similarities  of  growth  form  and  phys- 
ical setting  rather  than  simply  on  associations  of 


14 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


FIGURE  7.  The  West  Scarp  Ecogeographic  Formation  in  Arroyo  San  Rafael  (station  1 0).  The  vegetation  is  primarily  Chaparral, 
with  elements  of  Pinyon-Juniper  Woodland  and  Riparian  Woodland  evident  on  the  lower  slopes  and  in  the  canyon  bottom, 
respectively.  The  Upper  Main  Scarp  Ecogeographic  Formation  is  in  the  background. 


plant  species.  Each  of  the  previously  described 
vegetational  associations  contains  a  riparian  ele- 
ment, with  only  a  few  species  common  to  all 
riparian  elements.  The  species  vary  considerably 
relative  to  elevation,  however  the  presence  of 
closed  overstory,  dense  understory  thickets,  and 
lush,  moist  ground  vegetation  is  common  to  the 
Riparian  Woodland  at  all  elevations  of  the  Mar- 
tir  Region.  The  protective  cover  and  microcli- 
matic  conditions  provided  by  riparian  areas  are 
important  for  many  of  the  reptiles  and  amphib- 
ians in  the  Region. 

The  most  prevalent  plant  species  in  riparian 
associations  are  willows  (Salix  spp.).  They  com- 
pose most  of  the  vegetation  in  the  dense  under- 
story thickets  characteristic  of  the  riparian  cor- 
ridors, and  they  are  also  a  major  component  of 
the  overstory.  In  proximity  to  Creosote  Bush 
Scrub  and  Central  Desert  Scrub,  willows  are  rare 
and  are  generally  replaced  by  mesquite  (Prosopis 
juliflord)  and  desert  willow  (Chilopsis  linearis). 

Grasses,  sedges,  and  herbs  form  a  dense  mat 
of  streamside  meadow  throughout  riparian  as- 
sociations. This  meadow  is  most  highly  devel- 
oped within  the  Coniferous  Forest,  where  exten- 
sive areas  of  wet  meadow  occur  with  little  or  no 


overstory.  In  scrub  associations,  meadow  vege- 
tation is  reduced  to  narrow  strips  at  streamside, 
rarely  more  than  1  m  wide. 

Ecogeographic  Formations 

On  the  basis  of  the  physiographic,  climatic, 
and  vegetational  data  presented  above,  I  have 
divided  the  Martir  Region  into  nine  component 
ecosystems  or  ecogeographic  formations  (Fig.  14). 
Boundaries  of  the  nine  topographic  elements 
closely  approximate  the  ecotonal  boundaries  of 
the  ecogeographic  formations.  For  practical  pur- 
poses, these  boundaries  will  subsequently  be 
treated  as  synonymous  and  each  ecogeographic 
formation  will  be  identified  by  its  topographic 
component  (Fig.  4).  For  the  zoogeographic  anal- 
ysis, specimens  were  recorded  first  by  vegeta- 
tional association  and  secondarily  by  geographic 
locale.  Thus,  a  specimen  present  in  Chaparral  on 
the  Lower  Main  Scarp  was  scored  in  ecotonal 
vegetation  on  the  Lower  Main  Scarp  Ecogeo- 
graphic Formation. 

From  west  to  east  and  from  north  to  south, 
the  nine  ecogeographic  formations  of  the  Sierra 
San  Pedro  Martir  Region  are:  San  Quintin  Plain 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


15 


FIGURE  8.    The  East  Scarp  Ecogeographic  Formation  in  upper  Canon  El  Cajon  (near  station  5 1)  in  Pinyon-Juniper  Woodland. 


Ecogeographic  Formation  (northwest)  (Fig.  11), 
North  Central  Desert  Ecogeographic  Formation 
(southwest)  (Fig.  1 2),  Western  Foothills  Ecogeo- 
graphic Formation  (Fig.  10),  Western  Arroyos 
Ecogeographic  Formation  (Fig.  1 3),  West  Scarp 
Ecogeographic  Formation  (Fig.  7),  Upper  Main 
Scarp  Ecogeographic  Formation  (north)  (see  Fig. 
7),  Lower  Main  Scarp  Ecogeographic  Formation 
(south)  (Fig.  6),  East  Scarp  Ecogeographic  For- 
mation (Fig.  8),  and  South  Colorado  Desert  Eco- 
geographic Formation  (Fig.  9). 

SPECIES  ACCOUNTS 
(see  Appendix  A  for  a  list  of  collecting  stations) 


CAUDATA 


Plethodontidae 


Batrachoseps  pacificus  major  (Camp,  1915), 
Garden  Slender  Salamander 

Batrachoseps  pacificus  ssp. 

San  Pedro  Martir  Slender  Salamander 

Batrachoseps  pacificus  major  is  known  in 
northwestern  Baja  California  south  to  Arroyo  El 
Rosario  (K.  Yanev,  pers.  comm.).  It  is  known 
from  Colonett  (LACM  3433 1-35)  and  from  near 
Santo  Tomas  (MVZ  55102-12).  Other  localities 
are  more  common  northward. 


16 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


FIGURE  9.    The  South  Colorado  Desert  Ecogeographic  Formation  near  the  Junction  of  Mexico  Hwy  3  and  5  in  Creosote 
Bush  Scrub  vegetation. 


An  undescribed  subspecies  of  B.  pacificus  (see 
Yanev  1978)  is  known  locally  only  from  the  up- 
per scarps  of  the  Sierra  San  Pedro  Martir  down 
to  1,785  m  elevation. 

Brame  and  Murray  (1968)  speculated  that  Ba- 
trachoseps  pacificus  major  and  Batrachoseps  pa- 
cificus ssp.  are  sympatric  in  the  Sierra  San  Pedro 
Martir.  I  found  no  records  of  B.  p.  major  in  the 
Martir  Region,  east  of  Santo  Tomas,  20  km  from 
the  Pacific  coast.  Apparently  suitable  habitat  for 
Batrachoseps  occurs  inland  along  the  arroyos  of 
the  Western  Foothills,  but  surface  moisture  in 
these  arroyos  is  seldom  permanent,  or  depend- 
able. 

I  collected  a  gravid  female  of  Batrachoseps  pa- 
cificus ssp.  on  24  June  1973  at  station  25  (MVZ 
140701).  Two  individuals  were  observed  be- 
neath a  log  on  moist  sand  in  a  clump  of  willow 
on  21  June  1974  at  station  36.  Other  specimens 
are  known  from  Arroyo  Encantada  (station  4 1 ; 
CAS  57219-32),  Encantada  Meadow  (SDNHM 
4 1 32-36),  and  La  Grulla  (SU  1 2943).  All  of  these 
localities  correspond  with  riparian  woodland 
habitat  of  the  coniferous  forest.  B.  pacificus  ssp. 
also  occurs  in  riparian  habitat  in  the  chaparral 
on  the  West  Scarp.  I  collected  two  specimens 


inside  rotting  logs  in  oak  woodland  on  1 1  April 
1974  near  station  29  at  1,785  m  elevation  (MVZ 
140702-03);  a  series  from  La  Zanja  Creek  (SU 
12933-42)  is  from  similar  habitat. 


SALIENTIA 


Pelobatidae 


Scaphiopus  couchii  Baird,  1854, 
Couch's  Spadefoot 

Wasserman  (1970)  indicates  that  Scaphiopus 
couchii  is  restricted  to  the  Colorado  Desert,  and 
southward  along  the  Gulf  Slope  to  the  Cape  Re- 
gion of  southern  Baja  California.  He  lists  no  rec- 
ords on  the  Pacific  slope  of  Baja  California  north 
of  the  Magdalena  Plain.  A  specimen  collected  on 
Hwy  1,  5.3  km  south  junction  of  Hamilton  Ranch 
Road,  Valle  de  San  Quintin  (CAS  94808)  and 
reported  by  Welsh  (1 91 6a)  as  possibly  S.  couchii, 
is  actually  a  juvenile  S.  hammondii. 

Scaphiopus  hammondii  Baird,  1854, 
Western  Spadefoot 

This  toad  is  known  from  several  localities  on 
the  Pacific  Slope  of  Baja  California  north  of  the 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


17 


FIGURE  10.    The  Western  Foothills  Ecogeographic  Formation  12  km  west  of  Rancho  San  Jose  (station  27)  in  Coastal  Sage 
Scrub  vegetation. 


Martir  Region.  Within  the  Martir  Region,  Lins- 
dale  (1932)  reports  it  from  Rancho  San  Jose  (sta- 
tion 27)  in  the  Western  Foothills;  it  is  known 
also  from  the  northern  San  Quintin  Plain  at  San 
Antonio  del  Mar  (SDNHM  42095-96),  32  km 
south  of  San  Vicente  (LACM  90967-68),  along 
Mexican  Hwy  1,  5.3  km  south  junction  of  Ham- 
ilton Ranch  Road,  Valle  de  San  Quintin  (CAS 
94808),  and  10  km  south  of  San  Quintin 
(SDNHM  41404).  These  are  areas  of  coastal  sage 
and  central  desert  scrub. 


Bufonidae 

Bufo  boreas  halophilus  Baird  and  Girard,  1 852<2, 
California  Toad 

This  toad  is  abundant  in  the  riparian  meadows 
and  woodlands  of  the  coniferous  forest,  occur- 
ring up  to  2,800  m  in  or  near  riparian  areas  of 
the  entire  Pacific  Slope;  it  is  uncommon  in  the 
Western  Foothills  and  in  coastal  areas.  It  fre- 
quents areas  near  standing  or  slow-flowing  water. 


18 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


FIGURE  1 1.    The  San  Quintin  Plain  Ecogeographic  Formation  between  Mexican  Hwy  1  and  San  Antonio  Del  Mar  (station 
9).  The  vegetation  is  Central  Desert  Scrub  with  elements  of  Coastal  Sage  Scrub. 


Its  southernmost  locality  is  El  Rosario  (CAS 
136714-15). 

Adults  are  primarily  nocturnal;  diurnal  surface 
activity  in  the  summer  was  observed  on  the  Main 
Scarp,  but  mostly  during  afternoon  cloud  cover 
or  showers.  Newly  metamorphosed  juveniles, 
common  in  late  spring  in  the  wet  meadows  of 
the  Main  Scarp,  were  continuously  active,  often 
in  direct  sun.  Two  incidences  of  cannibalism  were 
observed  among  these  juveniles. 

Adult  specimens  were  collected  at  stations  3 
(MVZ  140712),  26  (MVZ  140704-05),  37  (MVZ 
140706-07),  and  43  (MVZ  140710);  adults  were 
observed  at  stations  27,  29,  44,  47,  57,  58,  and 


62.  Juveniles  were  collected  29  June  1973  at  sta- 
tion 43  (MVZ  140708-09)  and  23  July  1973  at 
station  15  (MVZ  14071 1).  They  were  observed 
between  these  dates  at  stations  12,  15,  27,  39, 
and  43.  Larvae  were  observed  or  collected  25 
June  1973  at  station  22  (MVZ  140727)  and  31 
July  1973  at  station  15. 

Bufo  microscaphus  calif  or  nicus  (Camp,  1915), 
Arroyo  Toad 

This  toad  is  uncommon  on  the  Lower  Main 
Scarp  in  riparian  habitats  of  the  coniferous  for- 
est; 2,300  m  was  the  highest  recorded  elevation. 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


FIGURE  12. 
vegetation. 


The  North  Central  Desert  Ecogeographic  Formation  at  San  Juan  de  Dios  (station  64)  in  Central  Desert  Scrub 


In  riparian  habitats  in  the  chaparral  of  the  West 
Scarp,  they  are  common  and  also  abundant  in 
riparian  habitats  in  coastal  sage  scrub  at  the  West 
Scarp  base.  Tevis  (1 944)  found  this  species  at  the 
Hamilton  Ranch  (Rio  Santo  Domingo)  on  the 
San  Quintin  Plain;  this  apparently  is  the  south- 
ernmost record  in  Baja  California. 

In  the  San  Bernardino  Mountains  of  California 
where  B.  microscaphus  is  sympatric  with  B.  bo- 
reas,  the  latter  species  occurs  in  cooler,  moister 
habitats  (Cunningham  1962).  My  observations 
of  the  microhabitat  affinities  of  B.  boreas  and  B. 
microscaphus  in  the  Martir  Region  concur  with 


those  reported  by  Cunningham.  Additionally,  I 
found  B.  microscaphus  only  in  or  near  fast-flow- 
ing water,  whereas  B.  boreas  occurred  only  in  or 
near  standing  or  slow-flowing  water.  It  is  appar- 
ently not  known  whether  these  two  species  of 
Bufo  deposit  their  eggs  in  different  habitats  or 
their  larvae  show  different  habitat  preferences. 
Larvae  of  both  species  were  found  together  in 
several  streams  of  moderate,  steady  flow  (sta- 
tions 15,  37,  43,  and  57).  A  pair  of  B.  micro- 
scaphus was  observed  in  amplexus  in  fast-flow- 
ing water  at  0905  on  3  May  1974  at  station  10. 
Adults  were  active  at  night;  toads  were  twice 


20 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


FIGURE  13.     Riparian  Woodland  vegetation  in  Arroyo  San  Rafael  at  station  14. 


observed  far  from  water,  in  coastal  sage  scrub 
and  chaparral.  On  three  occasions,  they  were  ob- 
served returning  to  riparian  habitat  at  dawn. 

Adults  were  collected  at  stations  14  (MVZ 
140719),  43  (MVZ  140713,  140738),  49  (MVZ 
140715-18,  1475 12),  and  56  (MVZ  1407 14).  Ju- 
veniles were  collected  or  observed  13  July  1973 
at  station  57  (MVZ  140739)  and  3  August  1973 
at  station  49.  Larvae  were  observed  1 3  July  1973 
at  station  43  and  31  July  1973  at  station  15. 

Bufo  punctatus  Baird  and  Girard,  18520, 
Red-spotted  Toad 

I  observed  this  toad  in  riparian  habitats  with 
flowing  water  at  three  localities  on  the  East  Scarp 
surrounded  by  rocky  creosote  bush  scrub  habitat. 
Three  specimens  are  known  from  the  Main  Scarp 
at  La  Grulla  (SU  12945-47).  The  species  also  is 
known  from  the  Sierra  Juarez  (CAS  1 2 1 20 1-09). 
Bostic  (1971)  reported  the  species  in  the  Central 
Desert. 

Adults  were  observed  or  collected  at  stations 
17, 46  (MVZ  140720,  140747),  and  60;  juveniles 
were  collected  on  24  April  1974  at  station  17 
(MVZ  140748).  Larvae  were  observed  or  col- 


lected 14  April  1974  at  station  46  (MVZ  140740) 
and  1 6  April  1 974  at  station  59.  Larvae  also  were 
observed  on  29  April  1974,  1  km  above  sta- 
tion 4. 

Hylidae 

Hyla  cadaver ina  Cope,  1866, 
California  Treefrog 

I  found  this  frog  in  riparian  woodland  through- 
out the  Sierra,  but  it  was  rare  on  the  Main  Scarp; 
the  highest  elevation  recorded  was  2,300  m.  The 
species  is  common  on  the  West  and  East  scarps 
and  in  the  arroyos  of  the  Western  Foothills.  Adult 
males  were  heard  chorusing  at  station  64,  a  spring 
in  the  North  Central  Desert.  Duellman  (1970) 
reported  Hyla  cadaverina  from  32  km  east  of  El 
Rosario.  South  of  the  Martir  Region,  it  is  known 
from  Bahia  de  Los  Angeles  (LACM  1 305,  1 3 1 63- 
69),  4.8  km  N  of  Rancho  Catavina  (LACM 
1 3 1 73-95),  and  from  Rancho  Santa  Ynez  (LACM 
107923-26,  1 13730-31);  these  localities  are  the 
southernmost  records  for  the  species  in  Baja  Ca- 
lifornia. 

These  treefrogs  were  usually  found  among  steep 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


21 


PACIFIC  CLIMATE- 


MONTANE 
CLIMATE 


•GULF  CLIMATE 


CENTRAL  DESERT  CLIMATE 


2800m  - 


1400m  - 


San 

QuintiYi 

Plain 


West  Foothills 
West  Arroyos 


Upper  & 

Lower 

Main 

rarp    I 


East 
Scarp 


Colorado  Desert 


Central 

Desert 

Scrub 


Coastal  Sage 
Scrub 

: — ecotones  — 


Chaparral 


Confer 
Forest 


Chaparral 
Pinyon-Juniper 

Creosote  Bush 
Scrub 


0 


75  km 


100 


125 


150 


FIGURE  14.  The  ecogeographic  formations  of  the  Sierra  San  Pedro  Martir  Region,  Baja  California  Norte,  Mexico  at  latitude 
3 1°N.  An  ecogeographic  formation  is  comprised  of  a  topographic  element  (Fig.  4),  a  climatic  element(s)  (Fig.  2c,  Table  1),  and 
a  vegetation  association  (Fig.  5.  Table  2).  Riparian  Woodland  Associations  are  present  within  all  vegetations  and  form  the 
primary  vegetation  in  the  Western  Arroyos.  A  ninth  ecogeographic  formation,  the  North  Central  Desert,  occurs  south  of  latitude 
31°N;  it  is  continuous  with  the  San  Quintin  Plain  Ecogeographic  Formation  and  contains  similar  vegetation. 


granitic  boulders  and  cliff  facings  above  fast- 
flowing  water.  Cunningham  (1964)  observed  the 
aquatic  garter  snake,  Thamnophis  hammondii, 
eating  this  frog.  The  substrate  preference  of  this 
frog  may  be  related  to  the  predatory  habits  of  T. 
hammondii,  which  was  twice  observed  attempt- 
ing to  climb  streamside  rock  faces  in  the  near 
vicinity  of  perched  treefrogs.  On  two  occasions 
I  observed  these  frogs  on  sand,  gravel,  and  aquat- 
ic vegetation  at  streamside.  These  localities  were 
isolated  riparian  sites  created  by  springs  on  the 
lower  East  Scarp;  T.  hammondii  was  not  ob- 
served at  these  localities  (above  stations  4  and 
46).  Boulders  and  rock  facings  were  present  in 
riparian  habitat  at  these  localities. 

Groups  of  5-30  individuals  were  commonly 
observed  congregating  in  direct  sunlight  in  the 
mist  above  cataracts  and  waterfalls  in  the  after- 


noon. Individuals  moved  back  and  forth  from 
shade  to  sunlight,  perhaps  for  thermoregulation. 
Adults  were  collected  at  station  35  (MVZ 
140746),  41  (MVZ  140721),  49  (MVZ  140729- 
31),  and  56  (MVZ  140722, 140724-25);  they  were 
observed  at  or  near  stations  4,  10,  16,  17,  20, 
21,  23,  27,  46,  60,  and  64.  Juveniles  were  col- 
lected 13  July  1973  at  station  56  (MVZ  140723); 
larvae  were  collected  on  14  April  1974  at  station 
46  (MVZ  140741),  and  were  observed  through 
29  April  1974  at  stations  4,  16,  20,  21,  23,  and 
59. 

Hyla  regilla  hypochondriaca  Hallowell,  1854, 
Pacific  Treefrog 

Hyla  regilla  hypochondriaca  ranges  to  2,750 
m  in  riparian  habitat  in  the  coniferous  forest, 


22 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


and  was  found  in  the  riparian  woodland  in  the 
Arroyos  of  the  West  Scarp  and  Western  Foot- 
hills, west  to  the  Pacific  coast.  Bostic  (1971)  re- 
ported H.  r.  desert icola  (=hypochondriaca)  from 
two  localities  in  the  Central  Desert. 

All  Hyla  regilla  were  observed  near  standing 
and  slow-flowing  water.  Most  animals  were  in 
aquatic  vegetation  at  streamside  or  floating  in 
the  water.  H.  regilla  was  rarely  observed  more 
than  1 5  cm  above  the  ground  or  more  than  1  m 
from  water.  This  contrasts  with  H.  cadaverina, 
which  was  found  in  areas  of  fast  water  flowing 
over  granite  boulders  where  the  frogs  inhabited 
rock  surfaces  often  2-5  m  above  the  water.  These 
two  treefrog  species  seldom  occurred  in  sym- 
patry  within  the  Region  as  a  result  of  the  unique 
riparian  microhabitats  occupied  by  each;  I  found 
them  together  at  a  single  locality  (station  27).  H. 
r.  hypochondriaca  was  active  diurnally  only  un- 
der conditions  of  deep  shade  or  cloud  cover;  most 
activity  was  observed  at  twilight  and  at  night. 

Adults  were  collected  at  stations  44  (MVZ 
1 1 142-56)  and  62  (MVZ  140746),  and  observed 
at  stations  1 1,  12,  14,  19,  27,  29,  35,  37,  38,  39, 
57,  and  62.  In  excess  of  several  hundred  juveniles 
were  observed  4  July  1973  at  station  39.  Larvae 
were  observed  27  July  1973  at  station  12  and  1 1 
April  1974  at  station  29. 

Ranidae 

Rana  aurora  draytoni  Baird  and  Girard,  1852a, 
California  Red-legged  Frog 

This  species  was  abundant  on  the  Lower  Main 
Scarp,  in  riparian  woodland,  and  wet  meadow 
habitats  of  the  coniferous  forest  up  to  an  eleva- 
tion of  2,200  m.  They  also  were  found  in  Arroyos 
of  the  West  Scarp  and  Western  Foothills  to  the 
Pacific  coast.  Linsdale  (1932)  reported  this  species 
from  San  Ramon  at  the  mouth  of  Rio  Santo 
Domingo,  which  is  apparently  the  southernmost 
record  for  the  species  in  Baja  California. 

I  observed  Rana  aurora  basking  and  feeding 
during  the  day  in  streamside  and  aquatic  vege- 
tation associated  with  fast,  slow,  and  standing 
water.  It  was  not  observed  in  rocky  areas  with 
cascading  water.  Most  of  the  R.  aurora  were  found 
near  deep  pools  which  were  used  for  escape  cov- 
er. 

Adult  specimens  were  collected  at  stations  38 
(MVZ  140732)  and  39  (MVZ  140733-35);  ani- 
mals were  observed  at  stations  14,  27,  35,  37, 


39,  42,  44,  and  49.  Larvae  with  well-developed 
legs  were  collected  on  10  August  1973  at  station 
10  (MVZ  140736). 

Rana  boylii  Baird,  1854, 
Foothill  Yellow-legged  Frog 

On  several  occasions  I  made  extensive  search- 
es of  the  Lower  Main  Scarp  near  the  western  edge 
at  La  Grulla  (station  39)  where  Loomis  (1965) 
reported  three  specimens  of  R.  boylii.  I  did  not 
find  this  species  here  or  along  any  other  wa- 
tersheds in  the  Sierra.  The  preferred  habitat  of 
R.  boylii— fast- flowing  water  over  gravel  or  cob- 
ble streambeds— is  rare  in  the  Martir  Region, 
and  I  suggest  that  R.  boylii  may  be  a  marginal 
species  in  the  Region  as  a  result  of  habitat  lim- 
itations. The  abundance  of  R.  aurora  in  montane 
riparian  habitat  may  also  contribute  to  the  scar- 
city ofR.  boylii.  Stream  alterations  due  to  mining 
activities  and  acid  rain  are  other  possible  expla- 
nations for  the  scarcity  or  loss  of  this  species  from 
the  Region. 


TESTUDINATA 


Emydidae 


Clemmys  marmorata  (Baird  and  Girard,  1852a), 
Pacific  Pond  Turtle 

Linsdale  (1932)  reported  Clemmys  marmo- 
rata from  Valle  de  Trinidad  (probably  in  Rio 
San  Antonio)  and  Arroyo  San  Telmo,  on  the 
Pacific  Slope.  A  record  from  3.2  km  south  of 
Valladares  (CAS  56884-85)  placed  the  species 
in  a  third  arroyo  to  the  south.  P.  Haneline  (pers. 
comm.)  collected  a  specimen  of  C.  marmorata 
from  30  mi  east  of  El  Rosario  (Rancho  El  Metate, 
station  65).  In  May  1974, 1  was  informed  by  the 
ranch  foreman  at  this  locality  that  this  animal 
was  brought  from  a  more  northern  population. 
I  found  C.  marmorata  in  the  deep  pools  of  a 
perennial  stream  at  Rancho  San  Antonio  on  Ar- 
royo Santo  Domingo,  at  the  base  of  the  West 
Scarp  (station  49).  This  site  is  the  southern  dis- 
tribution limit;  incongruously,  giant  cardon  (Pa- 
chycreus  pringlei)  and  barrel  cactus  (Ferocactus 
sp.)  occur  a  few  meters  from  streamside.  This 
turtle  is  uncommon  in  the  Martir  Region,  prob- 
ably as  a  result  of  alteration  of  riparian  habitat 
for  agricultural  purposes. 

Seeliger  (1 945)  reported  that  the  six  specimens 
she  examined  from  Baja  California  are  not  sim- 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


23 


ilar  to  either  described  subspecies  in  California. 
Bury  (1970)  suggested  that  an  analysis  of  geo- 
graphic variation  in  this  species  is  needed. 

Five  specimens  were  collected  at  station  49 
(MVZ  207759-63).  An  adult  turtle  was  seen  div- 
ing in  a  pool  on  the  West  Scarp,  at  1,600  m  in 
Arroyo  de  San  Rafael  (station  1 0). 


SAURIA 


Gekkonidae 


Coleonyx  switaki  (Murphy,  1974), 
Barefoot  Gecko 

Fritts  et  al.  (1982)  reported  Coleonyx  switaki 
from  eastern  San  Diego  County,  California  and 
from  Bahia  de  Los  Angeles  south  of  the  Martir 
Region.  These  and  previous  records  (Murphy 
1974)  and  the  ecological  preferences  noted  by 
Fritts  et  al.  (1982)  suggest  a  contiguous  range  for 
this  gecko  along  the  eastern  flank  of  the  Penin- 
sular Range  Mountains,  although  no  specimens 
have  been  reported  from  the  Martir  Region. 

Coleonyx  variegatus  variegatus  (Baird,  1859), 
Desert  Banded  Gecko 

Coleonyx  variegatus  abbotti  Klauber,  1945, 
San  Diego  Banded  Gecko 

The  Desert  Banded  Gecko,  Coleonyx  varie- 
gatus variegatus,  occurs  in  the  Colorado  Desert 
and  south  to  San  Felipe  (Klauber  1945).  A  single 
specimen  from  the  vicinity  of  Bahia  de  Los  An- 
geles (CAS  121182)  appears  to  be  C.  v.  penin- 
sularis  Klauber  (A.  Leviton,  pers.  comm.).  This 
specimen  is  apparently  an  unreported  northern 
range  extension  for  the  subspecies  on  the  Gulf 
Slope. 

Coleonyx  variegatus  abbotti  Klauber,  the  San 
Diego  Banded  Gecko,  occurs  on  the  Pacific  Slope 
of  the  Martir  Region.  This  lizard  was  collected 
at  Rancho  San  Jose  (SDNHM  24390)  and  east 
of  San  Telmo  (SDNHM  42542)  in  the  Western 
Foothills;  Bostic  (1971)  reported  two  specimens 
from  the  Central  Desert. 

I  collected  two  specimens  of  Coleonyx  varie- 
gatus variegatus  in  Valle  de  San  Felipe  near  the 
east  side  of  San  Matias  Pass  (MVZ  140742-43). 
These  specimens  resembled  C.  v.  variegatus 
(Klauber  1945)  except  that  they  had  a  clear  and 
distinct  nuchal  light  loop,  characteristic  of  C.  v. 
abbotti.  Klauber  ( 1 945)  postulated  that  these  races 
intergrade  across  some  low  passes  in  the  Pen- 


insular Range,  and  the  present  specimens  sup- 
port this  contention. 

Phyllodactylus  nocticolus  (Dixon,  1964), 
Baja  California  Leaf-toed  Gecko 

Dixon  (1969)  indicated  that  Phyllodactylus 
nocticolus  ranges  along  the  east  scarps  of  the  Pen- 
insular Range  from  California  southward  down 
the  length  of  Baja  California.  He  noted  that  P. 
n.  nocticolus  occurs  through  the  Martir  Region. 
Bostic  (1971)  described  a  new  subspecies,  P.  n. 
sloani,  from  the  Central  Desert,  thus  extending 
the  species  range  to  the  Pacific  Slope,  south  of 
the  Martir  Region.  The  northernmost  locality  he 
reported  was  39  km  SE  Rosario  (29°48'N, 
115°33'W). 

I  collected  two  specimens  of  Phyllodactylus 
nocticolus  at  station  65  (MVZ  140744-45)  about 
25  km  NW  of  the  above  locality.  This  site  is  a 
volcanic  slope  with  boulders,  in  Central  Desert 
scrub  vegetation.  The  geckos  were  both  under 
exfoliating  slabs  of  volcanic  rock,  1.3  m  above 
ground.  These  two  specimens  had  21  paraver- 
tebral  tubercules,  within  the  range  of  both  P.  n. 
nocticolus  and  P.  n.  sloani,  but  only  28  and  30 
longitudinal  rows,  within  the  range  of  P.  n.  noc- 
ticolus and  below  that  of  P.  n.  sloani  (Dixon 
1964;  Bostic  1971). 

Iguanidae 

Callisaurus  draconoides  Blainville,  1835, 
Zebra-tailed  Lizard 

Callisaurus  draconoides  occurs  on  both  Gulf 
and  Pacific  slopes  of  the  Martir  Region.  It  in- 
habits all  scrub  vegetations,  and  is  most  abun- 
dant in  the  creosote  bush  scrub  of  Valle  de  San 
Felipe  and  Valle  Chico.  It  ranges  through  San 
Matias  Pass  into  Valle  de  Trinidad  (SDNHM 
1 6869-8 1 ),  and  occurs  at  Rancho  San  Jose  (Lins- 
dale  1932),  and  16  km  west  of  station  10  in  the 
Western  Foothills.  C.  draconoides  is  uncommon, 
with  a  spotty  distribution  on  the  Pacific  Slope  of 
the  Region;  distribution  and  abundance  increase 
southward  into  the  Central  Desert  (stations  63, 
64;  Bostic  1971).  North  of  the  Region,  it  occurs 
at  Punta  Banda  (LACM  94409)  and  east  of  En- 
senada(LACM  107136). 

This  species  was  found  in  sandy  areas  of  open 
benches  and  slopes,  or  on  sandy  soil  around  scat- 
tered outcrops.  Specimens  were  collected  at  sta- 


24 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


tions  2  (MVZ  140753),  4  (140750-51),  and  63 
(MVZ  140752);  animals  were  observed  at  sta- 
tions 1,  5,  46,  54,  61,  64,  and  65. 

Crotaphytus  insularis  vestigium 

Smith  and  Tanner,  1972, 

Baja  California  Black-collared  Lizard 

Smith  and  Tanner  (1972)  described  Crotaphy- 
tus insularis  vestigium  and  indicated  a  Gulf  Slope 
range  for  the  species  through  south  central  Baja 
California.  It  is  known  from  the  lower  East  Scarp 
of  the  Martir  Region  (SU  17048;  Linsdale  1932). 

I  collected  two  adult  males  at  stations  49  (MVZ 
140754)  and  65  (MVZ  140755),  on  the  Pacific 
Slope  of  the  Region.  These  records  confirm  the 
validity  of  two  unreported,  and  apparently  ig- 
nored, specimens  collected  in  1 935  from  Rancho 
San  Jose  (station  27;  SDNHM  24391-92).  Van 
Denburgh  (1922)  reported  this  species  at  Trini- 
dad, west  of  San  Matias  Pass,  the  northernmost 
record  on  the  Pacific  Slope.  The  lizard  is  now 
known  from  the  Martir  Region,  from  the  Pacific 
Slope  of  the  Central  Desert  at  Mesa  de  San  Carlos 
(Bostic  1971),  and  near  Rancho  Catavina  (LACM 
16993,  63176).  In  Baja  California  Sur,  it  occurs 
at  Arroyo  La  Purisima,  12.8  km  southwest  of 
Canipole  (SU  1 1545);  4.0  km  northeast  San  Jose 
de  Comondu;  at  Comondu;  at  La  Purisima;  and 
9.6  km  and  52.8  km  north  of  Canipole  (Smith 
and  Holland  1971).  These  records  indicate  a  much 
greater  distribution  for  this  lizard  in  Baja  Cali- 
fornia than  the  Gulf  Slope  range  proposed  by 
Smith  and  Tanner  (1972,  1974). 

Both  specimens  I  collected  were  basking  on 
boulders  during  the  midafternoon.  The  site  at 
station  65  is  a  rocky  volcanic  slope  grown  to 
Central  Desert  scrub;  the  site  at  station  49  is  a 
granite  outcrop  in  coastal  sage  scrub. 

Dipsosaurus  dorsal  is  dorsal  is 
Baird  and  Girard,  1852a, 
Desert  Iguana 

This  lizard  is  known  from  numerous  localities 
in  the  Colorado  Desert,  south  to  Bahia  de  Los 
Angeles  on  the  Gulf  Slope  (Murray  1955).  Lins- 
dale (1932)  reported  it  from  the  Vizcaino  Desert 
near  San  Ignacio.  A  second  race  is  recognized 
south  of  the  Vizcaino  Desert  through  the  Cape 
Region  (Schmidt  1922).  The  species  is  absent 
from  the  Pacific  Slope  of  Baja  California  north 
of  latitude  29°N  in  the  Central  Desert  Region. 


I  observed  Dipsosaurus  dorsalis  dorsalis  along 
the  base  of  the  East  Scarp  in  Valle  de  San  Felipe 
and  Valle  Chico  (stations  5  and  54),  and  in  a 
canyon  bottom  west  of  the  scarp  base  (station 
61).  All  observations  of  D.  d.  dorsalis  occurred 
in  creosote  bush  scrub  on  open  benches  and  slopes 
with  a  sandy  substrate. 

Gambelia  wizlizenii  copei  (Yarrow,  1882), 
Cope's  Leopard  Lizard 

Gambelia  wizlizenii  wizlizenii 
(Baird  and  Girard,  1852*), 
Longnose  Leopard  Lizard 

Banta  and  Tanner  (1968)  indicated  that  Gam- 
belia wizlizenii  copei  occupies  most  of  the  pen- 
insula south  of  the  Martir  Region,  ranging  north 
to  the  vicinity  of  San  Felipe  on  the  Gulf  coast, 
and  that  G.  w.  wizlizenii  occurs  in  the  Colorado 
Desert  of  northeastern  Baja,  from  the  vicinity  of 
San  Felipe  northward.  On  the  Pacific  Slope,  G. 
w.  copei  ranges  on  the  coast  to  just  north  of  the 
Martir  Region  and  inland  north  just  into  the 
United  States  (Banta  and  Tanner  1968;  Mahrdt 
1973). 

I  collected  four  specimens  of  Gambelia  wiz- 
lizenii copei  on  the  Pacific  Slope  of  the  Martir 
Region.  I  found  two  active  males  during  mid- 
morning  at  station  27  (MVZ  140756-57).  They 
were  on  sandy  soil  under  shrubs  in  an  area  of 
scattered  granite  outcrops,  in  coastal  sage  scrub. 
A  third  adult  male  was  captured  basking  on  a 
rock  in  creosote  bush  scrub  at  station  2  (MVZ 
1 40759).  A  subadult  female  was  captured  emerg- 
ing from  a  burrow  at  0815  hr  in  coastal  sage 
scrub/oak  woodland  ecotone  at  the  head  of  Ar- 
royo San  Telmo  (1,480  m  elevation),  station  19 
(MVZ  140758).  I  observed  one  probable  G.  w. 
copei  on  the  lower  East  Scarp  in  creosote  bush 
scrub  at  station  46. 

Petrosaurus  mearnsi  (Stejneger,  1894), 
Banded  Rock  Lizard 

This  species  has  been  reported  from  numerous 
East  Scarp  localities  (Van  Denburgh  1922;  Lins- 
dale 1932).  It  ranges  south  on  the  Gulf  Slope  to 
Bahia  de  Los  Angeles  (CAS  85338),  and  occurs 
north  into  California  as  far  as  Riverside  County. 
Petrosaurus  is  absent  from  the  Pacific  Slope  north 
of  the  Central  Desert  region. 

I  observed  Petrosaurus  mearnsi  only  in  rocky 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


25 


areas  of  both  creosote  bush  scrub  and  pinyon- 
juniper  woodland  up  to  1,220  m.  Few  individ- 
uals were  seen  on  the  ground  as  P.  mearnsi  seems 
to  prefer  steep  rock  faces  from  1 .5  to  5.0  m  above 
ground.  The  angles  of  slope  of  perching  sites 
measured  for  P.  mearnsi  were  greater  than  those 
of  all  other  species  of  lizards  in  the  Region,  com- 
monly exceeding  90°.  These  lizards  were  active 
all  day,  with  a  midmorning  peak.  The  greatest 
densities  of  lizards  were  at  riparian  edges  in  can- 
yons of  the  East  Scarp. 

Specimens  were  collected  at  stations  1 7  (M  VZ 
140762)  and  59  (MVZ  140760-61).  I  observed 
P.  mearnsi  at  stations  4,  16,  20,  21,  23,  46,  48, 
59,  60,  and  61. 

Phrynosoma  coronatum  (Blainville,  1835), 
Coast  Horned  Lizard 

Reeve  (1952)  indicated  that  the  race  Phryno- 
soma coronatum  schmidti  occurs  on  the  Pacific 
Slope  of  the  Martir  Region,  south  into  the  Cen- 
tral Desert,  and  northward  to  near  the  United 
States  border.  Bostic  (1971)  reported  the  species 
from  the  Central  Desert,  but  he  did  not  indicate 
subspecific  affinities.  P.  coronatum  appears  to  be 
absent  on  the  Gulf  Slope  north  of  latitude  29°N. 

I  observed  lizards  in  chaparral  up  to  1 ,400  m 
at  station  18,  and  at  many  localities  in  coastal 
sage  scrub.  They  frequented  sand,  soil,  and  rock 
rubble  substrates.  P.  coronatum  was  active  in 
direct  sunlight  and  on  substrates  with  unusually 
high  surface  temperatures  (up  to  49°C),  when 
other  lizard  species  were  relatively  inactive  and 
seeking  shelter  in  shade. 

Specimens  were  collected  at  stations  10  (MVZ 
140824),  18  (MVZ  140826),  27  (MVZ  140823, 
140825),  28  (MVZ  140821),  and  49  (140822). 

Phrynosoma  mcallii  (Hallowell,  1852), 
Flat-tailed  Horned  Lizard 

Reeve  (1952)  reported  three  specimens  of 
Phrynosoma  mcallii  from  extreme  northeastern 
Baja  California  (Colorado  Desert).  One  record 
exists  from  further  south,  at  the  mouth  of  Gua- 
dalupe  Canyon,  at  the  foot  of  the  Sierra  Juarez 
(CAS  1 1 9077).  This  species  has  yet  to  be  reported 
from  the  desert  immediately  east  of  the  Sierra 
San  Pedro  Martir.  P.  mcallii  occurs  in  the  vicin- 
ity of  sand  dunes  and  sandy  flats  (Stebbins  1985); 
such  habitat  is  present  in  the  Martir  Region,  but 
is  not  continuous  with  similar  habitat  to  the  north. 


Phrynosoma  platyrhinos  calidiarum  (Cope,  1896), 
Southern  Desert  Horned  Lizard 

Reeve  (1952)  reported  seven  records  for  Phry- 
nosoma platyrhinos  calidiarum  in  northeastern 
Baja  California,  four  from  the  South  Colorado 
Desert  in  the  vicinity  of  San  Felipe.  Welsh  and 
Bury  (1984)  reported  it  from  three  localities  in- 
land from  the  Gulf.  Other  records  of  note  in- 
clude: Valle  de  San  Felipe  (station  5;  SU  18588), 
Monies  de  Media  (SDNHM  25252),  and  15  km 
north  of  Bahia  de  San  Luis  Gonzaga  (CAS  91631). 
P.  platyrhinos  may  occur  south  to  the  vicinity  of 
Bahia  de  Los  Angeles.  Apparently  it  is  absent  on 
the  Pacific  Slope  of  the  Martir  Region  and  in  the 
Central  Desert. 

Sauromalus  obesus  obesus  (Baird,  1859), 
Western  Chuckwalla 

Gates  (1968)  first  reported  Sauromalus  obesus 
in  northern  Baja  California  from  specimens  col- 
lected at  six  localities  on  the  Gulf  Slope.  The 
northernmost  locality  was  San  Matias  Pass  (sta- 
tion 1);  the  southernmost  record  was  36  km 
southwest  of  Bahia  de  San  Luis  Gonzaga.  I  col- 
lected an  adult  female  at  station  1 6  (MVZ  1 40749) 
and  observed  individuals  at  stations  1 7  and  46 
on  the  east  scarp.  Apparently  this  species  is  ab- 
sent on  the  Pacific  Slope  of  the  Region. 

Recent  records  from  about  5  km  north  of  Pun- 
ta  Prieta  (MVZ  1 17420  and  1 17467-69)  extend 
the  known  range  of  5".  obesus  south  into  the  Cen- 
tral Desert.  A  single  record  from  the  Sierra  Juarez 
at  Canon  Camillas  (SDNHM  43974)  fills  some 
of  the  gap  between  Baja  California  populations 
and  those  known  from  southern  California.  This 
saxicolous  species  probably  occurs  throughout 
the  east  scarps  of  the  Peninsular  Range  from  Cal- 
ifornia into  the  Central  Desert  of  Baja  California. 
Bostic  (1971)  reported  S.  australis  from  50  km 
south  of  Punta  Prieta  in  the  Central  Desert,  thus 
indicating  that  a  zone  of  sympatry  between  S. 
australis  and  S.  obesus  probably  exists  in  the 
Central  Desert  in  the  vicinity  of  Punta  Prieta. 
Seib  (1980)  implies  that  this  zone  of  sympatry 
does  not  exist  in  the  Central  Desert,  ignoring  S. 
australis  and  treating  all  populations  in  the  Cen- 
tral Desert  as  S.  obesus.  More  work  is  needed  to 
clarify  relationships  among  the  chuckwallas  of 
Baja  California. 

Sauromalus  obesus  was  unusually  wary  in  the 
Martir  Region.  Individuals  sought  cover  in  re- 


26 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


sponse  to  movement  at  distances  as  great  as  70 
m.  The  lizards  occurred  on  high  cliff  faces  and 
ledges,  and  seldom  ventured  far  from  crevices 
where  they  sought  cover.  These  large  dark  lizards 
contrast  greatly  with  the  pale  granite  surface  of 
the  Martir  Region.  The  wariness  of  S.  obesus 
may  allow  use  of  these  pale  surfaces  in  an  area 
with  high  densities  of  raptorial  birds  (Short  and 
Crossin  1967).  I  submit  that  the  dark  color  may 
be  a  therm oregulatory  adaptation,  which  has  a 
higher  selective  value  than  does  cryptic  color- 
ation for  predator  avoidance.  The  extreme  war- 
iness of  S.  obesus  may  explain  why  this  species 
remained  unreported  for  so  long  in  this  area. 

The  specimen  captured  1 2  June  1 974  at  station 
16  contained  six  well-developed  ova.  Scutella- 
tion  most  closely  matched  the  description  for  the 
race  Sauromalus  obesus  obesus  (Shaw  1945). 


Sceloporus  graciosus  vandenburghianus 

Cope,  1896, 

Southern  Sagebrush  Lizard 

This  lizard  was  the  most  frequently  observed 
reptile  on  the  Main  Scarp  of  the  Sierra.  Its  dis- 
tribution was  limited  to  coniferous  forest,  coni- 
fer-chaparral ecotone,  and  related  riparian  edge 
habitats  of  the  upper  scarps,  from  1 ,980  to  2,820 
m.  Lizards  were  active  from  0500  to  1800  hr  on 
boulders,  logs,  and  occasionally  on  the  forest  floor; 
juveniles  occurred  mostly  on  the  ground.  In  the 
Martir  Region,  Sceloporus  graciosus  vanden- 
burghianus exhibited  frequent  climbing  behav- 
ior. Elsewhere,  they  are  found  mostly  on  the 
ground  (Stebbins  1966). 

The  distribution  of  S.  graciosus  in  the  Martir 
Region  may  be  influenced  by  interspecific  com- 
petition. I  expected  it  to  exist  in  the  chaparral  of 
the  East  and  West  scarps  of  the  Region,  a  seem- 
ingly appropriate  habitat.  In  fact,  I  found  it  in 
stands  of  open  chaparral  only  above  2,100  m  on 
the  Main  Scarp.  The  closely  related  51.  occiden- 
talis  occurs  in  chaparral  on  the  East,  West,  and 
Main  scarps  below  2,100  m.  This  suggests  that 
51.  occidentalis  may  be  limited  to  lower  areas  with 
higher  temperatures,  whereas  S.  graciosus  is 
found  in  cooler  climes  above  2,100  m.  The 
mechanism  of  exclusion  could  be  simple  aggres- 
sive dominance  of  the  larger  species  (S.  occiden- 
talis) over  the  smaller  (S.  graciosus),  driving  the 
latter  away  from  critical  resources  at  lower  ele- 
vations (interference  competition).  Feeding 


strategies  may  also  play  a  role  in  segregating  these 
two  lizards.  In  the  ecotonal  vegetation  between 
1,970  and  2,100  m  where  these  two  lizards  co- 
exist, limited  observations  indicated  a  difference 
in  their  choice  of  perching  and  foraging  sites. 
Adult  S.  graciosus  were  most  often  on  elevated 
perches  and  descended  to  forage  on  the  ground. 
S.  occidentalis  was  always  observed  on  the  ground 
or  on  low  boulders. 

Specimens  were  collected  at  stations  8  (MVZ 
140773-74),  25  (MVZ  140763),  38  (MVZ 
140768-69,  147514),  39  (MVZ  140764-67, 
147513),  47  (MVZ  140770-71),  55  (MVZ 
140772),  and  58  (MVZ  140785-86);  observa- 
tions were  made  at  stations  12,  22,  24,  25,  30, 
32,  36,  37,  42,  44,  47,  and  55. 

Sceloporus  magister  rufidorsum  Yarrow,  1883, 
Red-backed  Spiny  Lizard 

Sceloporus  magister  uniformis 
Phelan  and  Brattstrom,  1955, 
Yellowback  Spiny  Lizard 

Phelan  and  Brattstrom  (1955)  reported  four 
subspecies  of  Sceloporus  magister  in  Baja  Cali- 
fornia. Hall  (1973)  indicated  that  these  taxa  can 
be  divided  into  two  distinct  karyological  groups 
indicating  at  least  two,  and  possibly  four,  distinct 
species.  Murphy  (1983a),  citing  Hall  and  un- 
published electrophoretic  data,  treated  the  forms 
as  separate  species.  I  reserve  judgement  until  more 
information  is  available,  treating  the  northern 
forms  as  subspecies.  Two  forms,  one  from  each 
karyological  group,  occur  in  the  Martir  Region: 
S.  m.  uniformis  (2n  =  26  chromosomes)  in  the 
Colorado  Desert  south  to  San  Felipe  on  the  Gulf 
Slope  and  S.  m.  rufidorsum  (2n  =  30  chromo- 
somes) on  the  Pacific  Slope,  from  just  north  of 
the  Martir  Region  south  through  the  Central 
Desert.  Subspecific  morphological  differences  are 
primarily  based  on  dorsal  pattern  and  coloration 
of  adult  males,  with  the  number  of  femoral  pores 
as  a  secondary  distinguishing  character  (Phelan 
and  Brattstrom  1955). 

I  collected  three  adult  males.  A  male  from  sta- 
tion 2  (MVZ  140799)  had  a  uniform  beige  dor- 
sum  and  a  femoral  pore  count  of  13-13,  which 
closely  matches  the  description  of  S.  m.  unifor- 
mis. An  adult  female  from  station  2  (MVZ 
1 40798)  also  fits  this  description.  The  other  males, 
collected  at  station  27  (MVZ  140793-94)  were 
not  similar  to  the  described  forms  (Phelan  and 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


27 


Brattstrom  1955).  These  two  males  had  a  faint 
and  diffuse  rust  color  dorsally,  and  lacked  any 
pattern  or  side  bars.  Femoral  pore  counts  were 
17-18  and  16-16,  which  corresponds  to  S.  m. 
rufidorsum  (15-20),  not  to  S.  m.  uniformis  (10- 
1 5).  Other  specimens  of  the  S.  magister  complex 
collected  were  an  adult  female  from  station  62 
(MVZ  140797),  and  three  juveniles,  two  males, 
and  one  female  from  station  64  (MVZ  140795- 
96, 1475 1 7).  All  of  these  matched  the  description 
for  S.  m.  rufidorsum. 

This  small  sample  suggests  the  possibility  of 
integration  between  two  forms  occurring  in  the 
Martir  Region.  Similarly,  Bostic  (1971)  exam- 
ined 1 3  males  from  the  Central  Desert  and  found 
only  one  specimen  with  a  typical  Sceloporus  ma- 
gister rufidorsum  pattern.  Further  biochemical 
work  is  needed  to  describe  variation  in  the  S. 
magister  complex  in  this  part  of  Baja  California. 

Sceloporus  magister  was  observed  in  all  scrub 
vegetation  on  both  the  Gulf  and  Pacific  slopes 
of  the  Martir  Region,  up  to  1,220  m,  and  south 
into  the  Central  Desert.  It  was  usually  seen  in 
association  with  plant  forms  that  afforded  pro- 
tective cover.  Lizard  burrows  were  located  at  the 
base  of  plant  species  such  as  Mojave  Yucca  (Yuc- 
ca schidigera),  Agave  sp.,  and  Cholla  (Opuntia 
sp.).  Lizards  often  were  observed  up  to  1.5  m 
above  ground  in  a  wide  variety  of  woody  plants. 
I  saw  S.  magister  in  all  regional  edaphic  types, 
including  riparian  edge,  and  on  all  substrates.  S. 
magister  was  observed  at  stations  5,  9,  10,  16, 
18,  63,  and  65. 


Sceloporus  occidentalis  biseriatus 

Hallowell,  1854, 

Great  Basin  Fence  Lizard 

This  lizard  reaches  its  southern  limit  in  the 
Martir  Region,  where  it  occurs  primarily  along 
the  Pacific  Slope.  It  was  abundant  in  the  chap- 
arral of  the  West  and  Upper  East  scarps,  and 
uncommon  in  conifer-chaparral  ecotones  on  the 
Lower  Main  Scarp  (stations  43  and  47).  This 
species  was  absent  above  2, 1 20  m.  On  the  lower 
Pacific  Slope,  in  areas  of  coastal  sage  scrub,  it 
was  found  only  in  riparian  habitat. 

In  riparian  woodland,  Sceloporus  occidentalis 
was  highly  arboreal,  commonly  found  1.5  m  or 
more  above  ground  on  trees  and  fenceposts.  Liz- 
ards foraged  on  the  ground  from  these  elevated 
perches;  when  startled,  they  returned  to  their 


perches.  In  the  chaparral  and  conifer-chaparral 
ecotone,  S.  occidentalis  was  primarily  ground- 
dwelling.  In  the  conifer-chaparral  ecotone,  S.  oc- 
cidentalis is  sympatric  with  S.  graciosus  (see 
above). 

Specimens  were  collected  at  stations  1 0  (MVZ 
140790),  19  (MVZ  140792,  147518),  27  (MVZ 
140791),  29  (MVZ  140788),  35  (MVZ  140789), 
43  (MVZ  140781),  47  (MVZ  140782-83),  53 
(MVZ  140784),  55  (MVZ  140787,  147515),  and 
58  (MVZ  140785-86);  observations  were  at  sta- 
tions 14,  19,  56,  and  57. 

Sceloporus  orcutti  Stejneger,  189 3 a, 
Granite  Spiny  Lizard 

Sceloporus  orcutti  occurs  on  the  rocky  scarps 
of  the  Martir  Region  in  scrub  vegetations,  pin- 
yon-juniper  woodland,  and  chaparral.  It  reaches 
the  conifer-chaparral  ecotone  on  the  Lower  Main 
Scarp  (station  39)  at  2,080  m.  It  is  most  abundant 
on  the  lower  East  Scarp  in  rocky  creosote  bush 
scrub.  This  species  is  entirely  saxicolous,  and  is 
associated  with  large  boulders,  outcrops,  or  rock 
faces. 

Sceloporus  orcutti  was  extremely  wary  and  dif- 
ficult to  approach,  reacting  to  my  presence  at  a 
distance  of  about  30-40  m.  I  observed  one  lizard 
taken  by  a  Harris'  Hawk  (Parabuteo  unicinctus) 
and  another  by  a  Red-tailed  Hawk  (Buteo  ja- 
maicensis).  The  abundance  of  large  raptorial 
birds  in  the  Martir  Region,  the  dark  body  color 
of  this  spiny  lizard,  and  the  pale  granite  surfaces 
frequented  by  this  species  may  together  explain 
the  extreme  wariness  displayed  by  S.  orcutti. 
Mayhew  (1963)  mentioned  this  wariness  and 
noted  a  southward  increase  in  its  intensity.  S1. 
orcutti  is  easily  captured  by  noose  at  the  north 
end  of  its  range  in  the  San  Jacinto  Mountains  of 
California,  but  it  becomes  increasingly  difficult 
to  approach  further  southward,  and  is  virtually 
impossible  to  noose  in  the  Martir  Region.  Buteo 
populations  at  the  northern  extreme  of  this  liz- 
ard's range  may  be  less  dense,  due  to  human 
interference,  or  there  may  be  a  greater  choice  of 
prey  animals  to  select  from.  Both  factors  would 
result  in  less  intense  selective  pressure  for  alert- 
ness in  S.  orcutti. 

Sceloporus  orcutti  was  collected  at  stations  1 7 
(MVZ  140776-78),  27  (MVZ  140779,  147516), 
46  (MVZ  140775),  and  65  (MVZ  140780);  ob- 
servations were  made  at  stations  2,  4,  10,  1 1,  14, 
29,  35,  37,  39,  48,  49,  51,  56,  59,  60,  and  62. 


28 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


Uma  notata  notata  Baird,  1859, 
Desert  Fringe-toed  Lizard 

Schmidt  (1922)  noted  two  records  for  Uma 
notata  notata  from  the  desert  east  of  the  Sierra 
Juarez,  approximately  160  km  north  of  the  Mar- 
tir  Region.  Norris  (1958)  suggested  this  are- 
nicolous  lizard  is  restricted  to  contiguous  sand 
dune  habitat,  which  may  explain  its  apparent 
exclusion  from  the  Martir  Region,  where  dune 
habitat  is  not  contiguous.  On  the  other  hand,  the 
lack  of  records  for  U.  notata  may  be  an  artifact 
of  limited  collecting. 

Urosaurus  graciosus  graciosus  Hallowell,  1854, 
Western  Brush  Lizard 

Urosaurus  graciosus  graciosus  is  known  from 
the  South  Colorado  Desert  at  San  Felipe  (Lins- 
dale  1932);  it  ranges  northward  through  the  des- 
erts of  California  and  Arizona.  Stebbins  (1966) 
indicated  that  U.  g.  graciosus  ranges  through  much 
of  northeastern  Baja  California.  I  collected  two 
adult  males  in  creosote  bush  scrub  at  the  base  of 
the  East  Scarp,  one  in  Valle  Chico  (station  61) 
(MVZ  140808),  and  the  other  in  Valle  de  San 
Felipe  near  station  1 7  (MVZ  1 40809).  Both  spec- 
imens were  1-2  m  above  ground  in  iron  wood 
trees  (Olneya  tesotd).  These  localities  extend  the 
known  range  of  this  species  slightly  south  and 
east. 

Urosaurus' graciosus  graciosus  is  absent  from 
the  East  Scarp  and  the  Pacific  Slope  of  the  Martir 
Region,  and  is  also  unknown  from  the  Central 
Desert.  U.  microscutatus  and  U.  lahtelai  occur 
in  these  respective  areas  (see  below  and  Rau  and 
Loomis  1977);  presence  of  these  congeners  may 
exclude  U.  g.  graciosus. 

Urosaurus  microscutatus  (Van  Denburgh,  1894), 
Small-scaled  Lizard 

Urosaurus  microscutatus  was  common  on  the 
rocky  scarps  of  the  Martir  Region  in  all  scrub 
vegetations,  chaparral,  and  pinyon-juniper 
woodland;  it  occurs  up  to  2, 1 20  m  on  the  East 
Scarp,  and  1 ,525  m  on  the  West  Scarp.  The  species 
is  not  present  on  open  sandy  desert  and  areas 
with  few  outcrops;  it  is  absent  in  the  South  Col- 
orado Desert  and  San  Quintin  Plain,  and  un- 
common in  the  Western  Foothills.  Bostic  (1971) 
reported  it  as  uncommon  in  the  Central  Desert. 

This  species  is  arboreal,  occurring  up  to  2.5  m 
above  ground.  It  commonly  descends  from  ele- 


vated perches  to  forage  on  the  ground.  I  found 
Urosaurus  microscutatus  on  boulders  and  in 
vegetation,  with  females  and  young  occasionally 
observed  on  the  ground.  Greatest  densities  were 
reached  in  the  rocky  creosote  bush  scrub  areas 
of  the  East  Scarp  along  the  riparian  edge. 

Specimens  were  collected  at  stations  10  (MVZ 
140814,  140819),  27  (MVZ  140820),  46  (MVZ 
140815-16,  147519),  49  (MVZ  140812-13),  51 
(MVZ  14081 1),  53  (MVZ  140810,  147520),  and 
59  (MVZ  140817-18);  observations  were  made 
at  stations  4,  11,  17,  19,  23,  48,  60,  61,  62,  and 
65. 

Uta  stansburiana  elegans  (Yarrow,  1883), 
California  Side-blotched  Lizard 

Ballinger  and  Tinkle  (1972)  indicated  that  a 
single  race  of  the  Side-blotched  Lizard  occurs 
throughout  peninsular  Baja  California.  This 
species  is  the  most  common  and  ubiquitous  liz- 
ard in  the  Martir  Region  at  all  elevations  and  in 
all  vegetations  except  coniferous  forest.  On  the 
Lower  Main  Scarp  it  ranges  up  to  2,120  m  in 
chaparral. 

Uta  stansburiana  elegans  showed  no  special 
edaphic  or  substrate  preference.  It  is  mostly  a 
ground  lizard  but  occasionally  occurs  up  to  1.5 
m  above  ground  on  rock  outcrops. 

Specimens  were  collected  at  stations  10  (MVZ 
140806),  27  (MVZ  140807),  35  (MVZ  140802), 
45  (MVZ  140803),  46  (MVZ  140805),  49  (MVZ 
140804,  147522),  54  (MVZ  140800,  147521), 
and  56  (MVZ  140801);  observations  were  made 
at  stations  1,  2,  4,  9,  14,  17,  19,  47,  48,  57,  59, 
60,  61,  62,  63,  64,  and  65. 

Xantusiidae 

Xantusia  henshawi  henshawi  Stejneger,  1893&, 
Granite  Night  Lizard 

This  saxicolous  species  occurs  on  all  the  rocky 
scarps  of  the  Sierra  San  Pedro  Martir,  up  into 
the  coniferous  forest  on  the  Lower  Main  Scarp 
at  Arroyo  Encantada  (2,100  m;  CAS  57294-96), 
and  the  conifer-chaparral  ecotone  at  La  Grulla 
(station  39;  Murray  1 955).  I  collected  a  specimen 
in  chaparral  at  station  19  (MVZ  140828),  and 
three  specimens  in  coastal  sage  scrub  at  station 
27  (MVZ  140827-28,  140830).  I  saw  one  on  the 
East  Scarp  in  pinyon-juniper  woodland  at  sta- 
tion 5 1 .  The  species  undoubtedly  occurs  to  the 
desert  floor  at  the  base  of  the  East  Scarp  (Stebbins 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


29 


1966;  Lee  1975)  but  its  secretive  habits  and  the 
ideal  protective  habitat  of  the  cliffs  and  canyons 
on  the  East  Scarp  make  collecting  difficult.  All 
animals  were  found  under  exfoliating  slabs  on 
large  granite  boulders.  The  species  is  as  yet  un- 
reported  in  the  Central  Desert,  possibly  because 
of  the  marked  change  in  geomorphology  that  oc- 
curs at  this  latitude.  The  primarily  granitic  for- 
mations that  dominate  in  the  Martir  Region  and 
provide  cover  for  this  species  give  way  to  vol- 
canic formations  in  the  Central  Desert. 

Xantusia  vigilis  vigilis  Baird,  1858, 
Desert  Night  Lizard 

Xantusia  vigilis  wigginsi  Savage,  1952, 
Baja  California  Night  Lizard 

This  lizard  occurred  most  commonly  in  areas 
of  open  benches  and  slopes  in  creosote  bush  and 
Central  Desert  scrub  on  both  the  Gulf  and  Pacific 
slopes.  Apparently  it  is  absent  in  the  coastal  sage 
scrub  of  the  Western  Foothills.  I  collected  one 
specimen  on  the  rocky  lower  West  Scarp  at  1 ,400 
m  in  pinyon-juniper  woodland  (station  1 0;  M  VZ 
140833). 

Savage  (1952)  indicated  that  the  Martir  Re- 
gion was  the  probable  area  of  intergradation  be- 
tween the  races  Xantusia  vigilis  wigginsi  of  cen- 
tral Baja  California,  and  X.  v.  vigilis  of  the 
southwestern  United  States.  Two  adult  males 
from  San  Matias  Pass  (station  1)  matched  Sav- 
age's description  of  X.  v.  wigginsi  on  the  basis  of 
tail  pattern,  but  showed  a  dorsal  pattern  found 
in  both  forms.  A  juvenile  from  station  1  had 
dorsal  characteristics  of  X.  v.  vigilis.  An  adult 
from  station  1 0  completely  lacked  a  pattern,  and 
did  not  fit  any  previous  descriptions.  A  juvenile 
from  the  South  Colorado  Desert  near  station  54 
had  the  dorsal  pattern  of  A',  v.  vigilis  and  the  tail 
pattern  of  X.  v.  wigginsi.  One  of  each  of  two 
juveniles  from  station  62  in  the  north  Central 
Desert  fits  closely  the  descriptions  of  X.  v.  wig- 
ginsi and  X.  v.  vigilis.  A  series  of  specimens  at- 
tributed to  X.  v.  wigginsi  (SDNHM  41340-52), 
was  collected  at  San  Telmo,  a  lowland  locality 
of  the  Pacific  Slope  dominated  by  Central  Desert 
scrub.  The  eight  specimens  I  collected  show  char- 
acteristics of  both  subspecies,  and  failed  to  match 
either  satisfactorily.  These  findings  support  Sav- 
age's speculation  of  intergradation  in  this  Re- 
gion. 

All  animals  I  collected  were  taken  from  rubble 


and  dead  fall  of  Mojave  yucca  (Agave  sp.),  barrel 
cactus,  or  Echinocactus  sp.  Specimens  were  col- 
lected at  stations  1  (MVZ  140831,  147523),  2 
(MVZ  140836-37),  10  (MVZ  140833),  54  (MVZ 
140832),  64  (MVZ  140834-35);  observations 
were  recorded  at  stations  14  and  65. 

Scincidae 

Eumeces  gilberti  rubricaudatus  Taylor,  1935, 
Western  Redtail  Skink 

Jones  (1985)  indicated  that  Eumeces  gilberti 
rubricaudatus  ranges  south  to  the  Sierra  San  Pe- 
dro Martir  in  Baja  California,  but  with  an  inland 
distribution.  Rogers  and  Fitch  (1947)  reported 
E.  g.  rubricaudatus  from  the  Pacific  coast  of  the 
Martir  Region  at  San  Antonio  del  Mar  (station 
9).  A  record  from  La  Grulla  (station  40;  SDNHM 
44291)  indicates  that  they  range  into  the  conif- 
erous forest-chaparral  ecotone.  I  collected  three 
specimens,  from  stations  14  (MVZ  140842),  27 
(MVZ  140843),  and  49  (MVZ  140841)  and  ob- 
served one  at  station  1 9.  All  were  in  deep  riparian 
woodland;  light  intensity  range  50-85  FC,  rela- 
tive humidity  33^1%  (Welsh  1976a).  Jones 
(1985)  reported  this  species  in  the  chaparral 
vegetation  of  the  Martir  Region  but  failed  to  cite 
specimens,  suggesting  this  may  be  an  assumption 
based  on  habitat  associations  observed  else- 
where. Evidence  to  date  suggests  that  this  species 
is  restricted  to  the  riparian  corridors  in  the  Re- 
gion, and  may  occur  along  the  moist  coastal  cor- 
ridor of  the  San  Quintin  Plain.  A  record  from 
Guadalupe  Canyon  in  the  Sierra  Juarez  (CAS 
119087)  suggests  it  may  also  occur  in  riparian 
vegetation  on  the  East  Scarp  of  the  Martir  Re- 
gion. 

Eumeces  skiltonianus  skiltonianus 
(Baird  and  Girard,  18526), 
Western  Skink 

This  lizard  is  common  in  the  coniferous  forest 
of  the  Sierra,  occurring  up  to  2,800  m  (station 
22).  It  also  occurs  on  the  West  Scarp  and  at  lower 
elevations  on  the  Pacific  Slope;  records  include 
Rancho  San  Jose  (Linsdale  1932)  and  6.6  km 
east  of  San  Telmo  (SDNHM  42741).  Coastal  rec- 
ords for  Eumeces  skiltonianus  in  the  Martir  Re- 
gion include:  San  Antonio  del  Mar  (SDNHM 
41990),  10  km  southeast  Punta  Colnett  (CAS 
57556),  San  Quintin  (CAS  55802),  and  Colonia 
Guerrero  (LACM  107935). 


30 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


On  the  lower  Pacific  Slope,  in  areas  dominated 
by  scrub  vegetation,  Eumeces  skiltonianus  ap- 
peared to  be  restricted  to  riparian  habitat  and 
the  moist  coastal  corridor.  Limited  observations 
suggest  this  lizard  reaches  greatest  densities  in 
areas  of  riparian  edge  habitat. 

Specimens  were  collected  at  stations  37  (MVZ 
140838),  39  (MVZ  140839,  147524),  and  50 
(MVZ  1 40840).  Sightings  occurred  at  stations  1 0, 
19,  22,  and  38. 

A  site  recorded  from  Catavina  in  the  north 
Central  Desert  has  recently  come  to  my  attention 
(D.  Schmoldt,  pers.  comm.  and  35-mm  slide). 
This  would  constitute  a  new  southern  record  for 
the  subspecies  if  confirmed  by  a  specimen. 

Teiidae 

Cnemidophorus  hyperythrus  schmidti 
(Van  Denburgh  and  Slevin,  1921), 
Orangethroat  Whiptail 

Cnemidophorus  hyperythrus  schmidti  occurs 
on  the  Pacific  Slope  from  San  Diego  County, 
California,  south  through  the  Central  Desert, 
where  it  ranges  across  to  the  Gulf  coast,  and 
through  the  Cape  Region  (Burt  1931).  In  the 
Martir  Region  these  whiptails  were  found  in  scrub 
vegetation  of  the  western  foothills  and  north 
Central  Desert.  Bostic  (1968)  found  C.  hypery- 
thrus on  the  San  Quintin  Plain,  but  he  considered 
habitat  there  to  be  suboptimal;  C.  labialis  was 
far  more  abundant  than  C.  hyperythrus  in  this 
area. 

I  found  Cnemidophorus  hyperythrus  to  be  ter- 
restrial and  diurnal,  showing  preference  for  areas 
of  fine-textured,  soft  soils  with  shrub  cover.  It 
was  most  abundant  in  riparian/scrub  ecotone.  I 
saw  American  Kestrels  (Falco  sparverius)  cap- 
ture individuals  of  C.  hyperythrus  on  two  occa- 
sions. Specimens  were  collected  at  stations  27 
(MVZ  147525),  49  (MVZ  140844-46),  62  (MVZ 
140849-50),  and  64  (MVZ  140847^18). 

Three  subspecies  of  Cnemidophorus  hyp- 
erythrus have  been  described  from  Baja  Califor- 
nia: C.  h.  hyperythrus,  C.  h.  schmidti,  and  C.  h. 
beldingi  (Linsdale  1932).  The  northern  race,  C. 
h.  beldingi,  is  generally  distinguished  by  the  pres- 
ence of  two  middorsal  stripes  that  are  lacking  in 
the  central  race,  C.  h.  schmidti  (Murray  1955). 
Murray  noted  a  difference  between  races  in  the 
number  of  supraoculars  that  were  separated  by 
granules  from  the  frontals,  which  he  considered 
a  more  diagnostic  character.  Bostic  (1971)  de- 


scribed 15  specimens  from  the  Central  Desert, 
the  presumed  range  of  C.  h.  schmidti,  that  match 
Murray's  characters  for  C.  h.  beldingi. 

The  eight  specimens  I  collected  appear  to  be 
of  the  race  C.  h.  schmidti.  Available  data  (Table 
3)  indicate  only  slight  differences  between  the 
subspecies.  I  suggest  that  there  is  insufficient  dif- 
ferentiation to  warrant  three  distinct  races.  With 
most  morphological  characters  showing  a  north 
to  south  clinal  variation,  I  consider  the  central 
peninsular  region  to  be  an  area  of  intergradation 
for  the  ecotypes  C.  h.  beldingi  and  C.  h.  hyp- 
erythrus. Burt  (1931)  recognized  no  subspecies. 

Cnemidophorus  labialis  Stejneger,  1890, 
Baja  California  Whiptail 

Cnemidophorus  labialis  is  endemic  to  north 
central  Baja  California,  occurring  primarily  along 
the  San  Quintin  Plain  and  the  Pacific  coast  of 
the  Central  Desert  Region.  The  northernmost 
record  for  C.  labialis  is  Punta  San  Jose,  about 
40  km  NW  San  Vicente  (Bostic  1968).  It  ranges 
southward  in  the  Central  Desert  to  the  vicinity 
of  latitude  28°13'N.  Bostic  (1971)  considered 
this  to  be  near  the  southern  limit  of  distribution. 
Vegetation  in  the  range  of  C.  labialis  on  the  San 
Quintin  Plain  most  closely  approximates  Central 
Desert  scrub,  with  plant  density  increased  and 
species  composition  altered  slightly  by  the  fog 
and  cool,  moist  air.  I  did  not  collect  any  speci- 
mens of  C.  labialis. 

Cnemidophorus  tigris  multiscutatus  Cope,  1892a, 
Coastal  Whiptail 

Cnemidophorus  tigris  tigris 
Baird  and  Girard,  1852#, 
Great  Basin  Whiptail 

Cnemidophorus  tigris  multiscutatus  occurs  up 
to  1,980  m  in  the  Martir  Region,  on  both  sides 
of  the  peninsular  divide  in  all  types  of  scrub  vege- 
tation, pinyon-juniper  woodland,  and  chaparral. 
It  ranges  to  the  coast  on  the  Pacific  Slope  and 
south  through  the  Central  Desert  (Bostic  1971). 
On  the  Gulf  Slope,  C.  t.  multiscutatus  occurs  in 
the  south  Colorado  Desert  in  Valle  de  San  Felipe 
and  Valle  Chico,  along  the  base  of  the  East  Scarp. 
Murray  (1955)  collected  the  Great  Basin  Whip- 
tail  (C.  t.  tigris)  from  three  Gulf  coastal  localities 
in  the  South  Colorado  Desert  near  San  Felipe. 

I  found  Cnemidophorus  tigris  multiscutatus 
only  at  ground  level  on  sand,  soil,  rock,  rubble, 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


31 


TABLE  3.    VARIATION  IN  MORPHOLOGICAL  CHARACTERS  AMONG  SUBSPECIES  OF  CNEMIDOPHORUS  HYPERYTHRUS.  (See  footnotes 
for  sources  of  published  data.) 


Character 

Subspecies 

Martir 
specimens 

beldingi 

schmidti 

hyperythrus 

Number  of  granules  around  midbody 

72.8  ±  0.8 

75.2  ±  0.6 

77.6  ±  0.6 

75.0 

(66-79)  17' 

(66-83)  542 

(69-90)  45' 

(70-80)  8 

Anterior  supraoculars  separated  from 

the  frontal  by  granules: 

only  part  of  third 

53 

10"             192 

285 

0 

all  of  third 

0 

48"             142 

445 

8 

only  part  of  second 

253 

3"             162 

45 

— 

all  of  second 

363 

152 

0 

- 

Number  of  middorsal  lines: 

three 

96 

0 

48" 

1 

two 

1046 

24              32 

15" 

2 

one  (forked  anteriorly) 

466 

594            562 

34 

5 

1  Data  from  Walker  and  Taylor  (1968). 

2  Data  from  Bostic  (1971). 

3  Data  from  Van  Denburgh  (1922). 

4  Data  from  Linsdale  (1932). 

5  Combined  data  from  Murray  (1955)  and  Linsdale  (1932). 

6  Data  from  Burt  ( 1931). 


or  leaf  litter.  I  observed  it  foraging  throughout 
the  day,  mostly  in  the  broken  shade  of  the  shrub 
understory.  When  startled  or  chased,  these  whip- 
tails  sought  shelter  in  burrows. 

Specimens  were  collected  at  stations  2  (MVZ 
140857),  14  (MVZ  140854),  27  (MVZ  140855), 
48  (MVZ  140853),  49  (MVZ  140852,  147526), 
56  (MVZ  140851),  and  62  (MVZ  140856, 
1 47527).  These  animals  resembled  Bostic's  (1971) 
collection  of  Cnemidophorus  tigris  multiscutatus 
from  the  Central  Desert.  C.  t.  multiscutatus  was 
observed  at  stations  4,  5,  10,  17,  18,  19,  35,  46, 
48,  49,  52,  57,  59,  60,  61,  62,  63,  64,  and  65. 

Anguidae 

Elgaria  multicarinata  webbi  (Baird,  1858), 
San  Diego  Alligator  Lizard 

Elgaria  multicarinata  webbi  is  known  from  co- 
niferous forest  areas  on  the  Lower  Main  Scarp 
at  La  Grulla  (Linsdale  1932),  Arroyo  Encantada 
(CAS  57038^2),  El  Alcatraz  (CAS  57330),  and 
at  station  37.  This  lizard  was  found  in  chaparral 
on  the  West  Scarp  at  stations  10  (MVZ  140860) 
and  35  (MVZ  140859)  and  on  the  East  Scarp  at 
station  53  (MVZ  1 40858).  Western  Foothills  rec- 
ords include:  Valladares  (Linsdale  1932);  Mes- 
quite  Spring,  Valle  de  Trinidad  (SDNHM  1 6864); 
and  Rancho  San  Jose  (Linsdale  1932).  These 
Western  Foothill  records  are  all  from  sites  con- 


taining riparian  habitat.  The  chaparral  habitat 
occupied  by  E.  m.  webbi  at  higher  elevations  is 
scarce  below  1,220  m  in  the  Western  Foothills, 
and  I  suspect  that  its  distribution  below  1,220 
m  is  localized  primarily  around  the  riparian  cor- 
ridors. Records  from  Canon  del  Diablo  (SU 
17293-95)  and  Canon  El  Cajon  (Linsdale  1932) 
indicate  a  possible  similar  riparian  distribution 
of  E.  m.  webbi  on  the  East  Scarp  below  the  chap- 
arral association.  Coastally,  on  the  San  Quintin 
Plain,  I  observed  this  lizard  in  Central  Desert 
scrub  at  1.7  km  east  of  San  Antonio  del  Mar 
(station  9),  and  it  is  also  known  from  Colonia 
Guerrero  (LACM  75501,  107933).  Bostic  (1971) 
indicated  that  this  race  occurs  along  the  moist 
coastal  belt  of  the  Central  Desert  Region. 

All  specimens  I  collected  were  in  areas  of  rocky 
outcrops,  on  granitic  sand,  soil,  or  rock  rubble, 
except  the  individual  near  station  9,  which  was 
on  open  sandy  ground  near  vegetation. 

Anniellidae 

Anniella  geronimensis  Shaw,  1940, 
Baja  California  Legless  Lizard 

Anniella  geronimensis  occurs  coastally  on  the 
San  Quintin  Plain  from  Colonia  Guerrero  south 
to  the  vicinity  of  El  Rosario  (Shaw  1953;  Bezy 
et  al.  1 977).  This  species  is  endemic  to  the  Martir 
Region  and  a  single  near-shore  island.  Shaw 


32 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


(1953)  reported  A  geronimensis  sympatric  with 
A.  pulchra  in  the  vicinity  of  Colonia  Guerrero. 

Anniella  pulchra  Gray,  1852, 
Silvery  Legless  Lizard 

Anniella  pulchra  occurs  on  the  Pacific  Slope 
within  the  Martir  Region  at  San  Jose  (SDNHM 
40394),  Socorro  (SDNHM  8842),  east  of  San 
Telmo  (SDNHM  42738),  San  Antonio  del  Mar 
(SDNHM  4900-10),  and  west  of  Colonia  Guer- 
rero (Bezy  et  al.  1977).  A  resident  told  me  that 
A.  pulchra  occurred  in  sandy  fields  at  the  bottom 
of  the  arroyo  at  station  19.  This  site  at  1,450  m 
may  be  the  highest  record  for  the  species  in  Baja 
California.  Stebbins  (1954:303)  reported  A.  pul- 
chra from  1,950  m  in  the  Giant  Forest,  Sequoia 
National  Park  in  California.  Its  preference  for 
moist,  fine-textured  soils,  as  reported  by  Miller 
( 1 944),  probably  limits  A.  pulchra  to  arroyo  bot- 
toms and  coastal  localities  within  the  Martir  Re- 
gion; its  upward  distribution  in  the  Sierra  is  prob- 
ably restricted  by  the  steep,  granitic  west  scarp. 

Bury  (1983)  reported  Anniella  pulchra  from 
an  arroyo  on  the  West  Scarp  of  the  Sierra  Juarez, 
60  km  east  southeast  of  Ensenada.  Klauber  ( 1 932) 
recorded  A.  pulchra  at  Canon  San  Salado,  at  the 
base  of  the  East  Scarp  of  the  Sierra  Juarez.  Bury 
(1983)  maintained  that  this  locality  is  probably 
Arroyo  Salado,  1 9  km  south  of  San  Vincente  near 
the  Pacific  coast;  he  indicated  that  a  second  rec- 
ord from  the  Gulf  Slope  near  San  Felipe  is  also 
probably  invalild. 

SERPENTES 
Leptotyphlopidae 

Leptotyphlops  humilis  cahuilae  Klauber,  19316, 
Desert  Worm  Snake 

Leptotyphlops  humilis  humilis 
(Baird  and  Girard,  1853), 
Western  Worm  Snake 

Klauber  (1940)  reported  two  subspecies  of 
Leptotyphlops  humilis  in  northern  Baja  Califor- 
nia: L.  h.  humilis  occurs  on  the  Pacific  Slope  from 
San  Diego  County,  California,  south  into  the 
Central  Desert  Region,  and  L.  h.  cahuilae  occurs 
along  the  Gulf  Slope  from  California  into  central 
Baja  California.  Klauber  also  noted  a  record  of 
L.  h.  cahuilae  from  the  Vizcaino  Desert. 

I  collected  two  specimens  of  Leptotyphlops  hu- 
milis on  the  lower  Pacific  Slope  of  the  Martir 


Region:  one  was  taken  at  1950  hr,  as  it  moved 
rapidly  across  a  dirt  road  at  1,100  m  in  open, 
sandy,  coastal  sage,  6.2  km  west  of  station  28 
(MVZ  140861);  the  other  was  found  at  station 
27  (MVZ  140862),  crushed  in  a  cow  track.  Both 
specimens  agreed  in  dorsal  pigmentation  and  in 
number  of  pigmented  scale  rows  with  the  de- 
scription of  L.  h.  humilis  (Klauber  1940).  How- 
ever, the  specimen  from  station  27  had  a  dorsal 
scale  count  (285)  exceeding  that  established  for 
L.  h.  humilis  (range  257-283),  and  in  the  range 
for  L.  h.  cahuilae  (280-305).  This  suggests  a  zone 
of  intergradation  on  the  Pacific  Slope  of  the  Mar- 
tir Region. 

Previously  unpublished  records  of  Lepto- 
typhlops humilis  humilis  on  the  Pacific  Slope  in- 
clude 25  km  east  of  San  Telmo  (CAS  123717) 
and  8.3  km  south  of  Socorro  Ranch  (SDNHM 
19231),  San  Telmo  (SDNHM  42746). 

Leptotyphlops  humilis  cahuilae  was  first  re- 
ported from  the  south  Colorado  Desert  at  Punta 
San  Felipe  (Murray  1 955);  additional  records  are 
known  from  58.5  km  north  of  San  Felipe 
(SDNHM  44262-63)  and  8.3  km  north  of  San 
Felipe  (CAS  1 36368-69).  There  are  three  records 
for  this  subspecies  from  Bahia  de  Los  Angeles, 
at  the  south  end  of  the  Colorado  Desert  (CAS 
103465,  LACM  2167,  and  SDNHM  19998). 
Klauber  (1940)  reported  this  subspecies  as  far 
south  as  the  Vizcaino  Desert. 

All  records  for  the  races  of  Leptotyphlops  hu- 
milis from  the  Martir  Region  are  from  areas  of 
scrub  vegetation. 

Boidae 

Lichanura  trivirgata  roseofusca  Cope,  1868, 
Coastal  Rosy  Boa 

Three  forms  of  the  Rosy  Boa  are  now  recog- 
nized on  peninsular  Baja  California.  Klauber 
(1933),  Lowe  (1964),  and  Stebbins  (1966)  indi- 
cated that  there  are  three  subspecies  (Lichanura 
trivirgata  trivirgata,  L.  t.  gracia,  and  L.  t.  roseo- 
fusca). Klauber  (193  la)  implied  that  L.  t.  gracia 
was  intermediate  between  L.  t.  trivirgata  and  L. 
t.  roseofusca  on  the  basis  of  scale  counts  and  color 
description.  Gorman  (1965)  found  no  evidence 
of  intergradation  between  L.  t.  trivirgata  and  L. 
t.  gracia,  and  recognized  two  species:  L.  trivirgata 
and  L.  roseofusca. 

Bostic  (1971)  reported  two  specimens  from  the 
Central  Desert  of  Baja  California  that  were  sim- 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


33 


ilar  to  Klauber's  (193 la)  description  of  Licha- 
nura  trivirgata  gracia,  but  with  dorsal  and  ven- 
tral scale  counts  matching  L.  t.  trivirgata.  One 
of  these  specimens  was  collected  167  km  north 
of  San  Ignacio  in  the  Vizcaino  Desert,  previously 
the  northernmost  locality  known  for  L.  t.  trivir- 
gata in  Baja  California  (Gorman  1 965).  This  form 
is  known  from  the  Cape  Region  of  Baja  Califor- 
nia and  the  west  coast  of  Mexico. 

In  the  Martir  Region,  the  Rosy  Boa  is  known 
from  the  lower  Pacific  Slope  at  numerous  local- 
ities in  the  Western  Foothills  and  on  the  San 
Quintin  Plain,  all  corresponding  to  areas  of  coastal 
sage  and  Central  Desert  scrub.  Specimens  from 
this  area  match  the  description  for  Lichanura 
trivirgata  roseofusca,  which  ranges  north  into 
California  along  the  Pacific  coast.  Gorman  (1 965) 
ascribed  a  single  specimen  from  the  South  Col- 
orado Desert,  25  km  south  of  San  Felipe,  to  L. 
roseofusca.  Welsh  and  Bury  (1984)  reported  a 
specimen  matching  Klauber's  ( 1 93 1  a}  description 
for  L.  t.  gracia  from  56  km  northwest  of  San 
Felipe  in  the  Sierra  San  Felipe  of  the  South  Col- 
orado Desert.  L.  t.  gracia  occurs  in  the  high  des- 
erts of  California  and  Arizona  between  the  coast- 
al California  range  of  L.  t.  roseofusca,  and  the 
extreme  southern  Arizona  and  Sonora  range  of 
L.  t.  trivirgata.  The  northern  Mojave  Desert  range 
of  L.  t.  gracia  is  about  335  km  away  from  the 
Central  Desert  and  South  Colorado  Desert  re- 
gions of  Baja  California  where  Bostic  (1971)  and 
Welsh  and  Bury  (1984)  collected  L.  t.  gracia. 

In  view  of  this  great  distance,  and  the  lack  of 
low  desert  records  for  the  species  north  of  Baja 
California,  it  seems  highly  unlikely  that  a  con- 
tiguous population  of  Lichanura  trivirgata  gracia 
exists  across  this  area.  The  disjunct  distribution 
of  L.  t.  gracia  between  populations  of  L.  t.  ro- 
seofusca and  L.  t.  trivirgata  prompted  my  sug- 
gestion (Welsh  \916a)  that  L.  t.  gracia  is  a  re- 
current intermediate  form,  a  thesis  more  recently 
entertained  by  Yingling  (1982). 


Colubridae 

Arizona  elegans  eburnata  Klauber,  \946b, 
Desert  Glossy  Snake 

Arizona  elegans  occidentalis  Blanchard,  1 924, 
Western  Glossy  Snake 

Arizona  elegans  pacata  Klauber,  1 9466, 
Peninsula  Glossy  Snake 


Klauber  (19466)  reported  Arizona  elegans  oc- 
cidentalis at  numerous  localities  along  the  Pacific 
Slope  from  the  Martir  Region  north  into  Cali- 
fornia. It  has  been  reported  as  far  south  as  El 
Rosario  on  the  Pacific  coast  (SDNHM  44155). 
Seifert  (1980)  reported  two  specimens  from  the 
Central  Desert  south  of  Laguna  Chapala. 

A  second  subspecies,  Arizona  elegans  pacata, 
is  known  from  the  Central  Desert  Region.  Rec- 
ords are  known  from  near  Santo  Domingo 
(Klauber  19466),  43.4  km  northwest  of  San  Ig- 
nacio (Murray  1955),  and  50  km  west  northwest 
of  El  Arco  (Banta  and  Leviton  1963);  more  re- 
cently collected  specimens  are  at  CAS  and  MVZ 
(T.  Papenfuss,  pers.  comm.). 

Klauber  (19466)  reported  Arizona  elegans 
eburnata  in  the  Colorado  Desert  of  California 
and  Arizona,  and  implied  its  presence  in  north- 
eastern Baja  California.  In  June  1974, 1  collected 
a  juvenile  male  at  2030  hr,  3.2  km  southeast  of 
Rancho  Rosarito  (station  62;  MVZ  140868).  The 
site  was  a  sandy  arroyo  with  creosote  bush  scrub 
vegetation.  The  pattern  and  scutellation  matched 
those  described  for  A.  e.  eburnata  (Klauber 
1 9466).  This  specimen,  the  first  for  the  subspe- 
cies in  Baja  California,  extended  the  range  to  80 
km  southwest  of  San  Felipe,  and  is  also  the  first 
record  of  the  desert  glossy  snake  from  the  Pacific 
Slope.  Welsh  and  Bury  (1984)  reported  six  more 
specimens  of  this  race  from  Martir  Region  lo- 
calities northwest  of  San  Felipe.  All  records  of 
A.  elegans  from  the  Martir  Region  occurred  in 
areas  of  scrub  vegetation. 


Chilomeniscus  cinctus  Cope,  1861, 
Banded  Sand  Snake 

Chilomeniscus  cinctus  is  known  from  six  lo- 
calities on  the  Pacific  Slope  of  the  Martir  Region: 
Arroyo  Seco  (SDNHM  48150-52),  San  Antonio 
River,  northeast  of  Arroyo  Seco  (SDNHM 
43378),  14  km  east  of  San  Telmo  (SDNHM 
41325),  20  km  east  of  San  Telmo  (SDNHM 
42737),  and  Valle  de  Trinidad  (SDNHM  3037 1). 
The  first  five  sites  are  in  low-lying  canyons  that 
are  immediately  east  and  contiguous  with  the 
San  Quintin  Plain.  These  areas  are  dominated 
by  Central  Desert  and  coastal  sage  scrubs.  Valle 
de  Trinidad  is  at  the  northwest  end  of  the  Sierra 
and  is  dominated  by  creosote  bush  scrub  that  is 
contiguous  through  San  Matias  Pass  into  the  Col- 
orado Desert.  The  Valle  de  Trinidad  record  in- 


34 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


dicates  that  C.  cinctus  is  probably  present  in  the 
sandy  habitats  of  the  South  Colorado  Desert. 

R.  Seib  (pers.  comm.)  reported  this  species  from 
two  localities  in  the  north  Central  Desert:  Mis- 
sion San  Fernando  3.3  km  west  of  El  Progresso 
(MVZ  1 17449)  and  5.5  km  north  of  Santa  Ines 
parador,  Mexican  Hwy  1  (MVZ  117303).  Two 
other  unpublished  records  exist  for  Chilome- 
niscus  cinctus  in  this  area:  16.7  km  south  of  Ca- 
tavina  (SDNHM  42054),  and  50  km  south  of  El 
Marmol  (SDNHM  38663).  Bostic  (1971)  re- 
ported a  specimen  from  0.8  km  north  of  San 
Javier.  Numerous  additional  records  exist  for  the 
Central  Desert  southward  throughout  the  pen- 
insula (R.  Seib,  pers.  comm.),  including  three 
unpublished  records  from  San  Ignacio  (SDNHM 
3828-30)  in  the  Vizcaino  Desert. 

Chionactis  occipitalis  annulata  (Baird,  1859), 
Colorado  Desert  Shovel-nosed  Snake 

The  subspecies  Chionactis  occipitalis  annulata 
is  known  from  sandy  alluvial  soil  in  a  limited 
coastal  area  north  of  San  Felipe,  within  3.2  km 
of  the  Gulf  of  California  along  Mexican  Hwy  5 
(Wake  19660;  Cross  1970).  Welsh  and  Bury 
(1984)  reported  four  more  specimens  from  this 
area  and  two  specimens  from  1 1.2  km  and  14.2 
km  west  of  the  Gulf  of  California.  This  are- 
nicolous  snake  is  not  known  south  of  the  Martir 
Region;  the  discontinuity  of  suitable  sandy  hab- 
itat may  account  for  its  absence. 

Coluber  flagellum  fuliginosus  (Cope,  1895), 
Baja  California  Coachwhip 

Coluber  flagellum  piceus  (Cope,  1875), 
Red  Coachwhip 

Wilson  (1973)  indicated  two  subspecies  of  Col- 
uber flagellum  in  the  Martir  Region:  C.  /  piceus, 
ranging  south  through  the  Colorado  Desert  to 
the  vicinity  of  Bahia  de  San  Felipe,  and  C.  f. 
fuliginosus,  occurring  at  localities  on  the  lower 
Pacific  Slope,  and  south  through  the  Central  Des- 
ert. These  areas  contain  scrub  vegetation. 

The  dark  phase  of  C.  f.  fuliginosus  was  ob- 
served twice  during  midmorning  in  open  sandy 
coastal  scrub  at  station  27  on  the  Pacific  Slope. 
Both  snakes  were  emerging  from  burrows. 
Another  dark  racer  was  observed  above  station 
4  on  the  lower  East  Scarp  but  could  not  be  iden- 
tified to  subspecies. 

Wilson  (1973)  indicated  a  single  record  for 


Coluber  flagellum  piceus  in  the  South  Colorado 
Desert,  near  San  Felipe.  Welsh  and  Bury  (1984) 
reported  seven  records  from  inland  localities 
north  and  west  of  San  Felipe.  These  specimens 
were  all  red  phase  snakes. 


Coluber  lateralis  lateralis  (Hallowell,  1853), 
California  Striped  Racer 

Two  records  exist  for  Coluber  lateralis  lateralis 
on  the  East  Scarp:  Canon  el  Cajon  (Linsdale 
1932),  and  Canon  del  Diablo  (SU  17298).  On 
the  Pacific  Slope,  this  race  is  known  from  five 
localities:  San  Antonio  Mine  (Murray  1955), 
Valladares  Creek  (CAS  56881),  Socorro 
(SDNHM  10524),  and  San  Jose  (SDNHM  5130 
and  8856).  I  collected  two  specimens  on  the  Pa- 
cific Slope:  one  dead  on  the  road  in  rocky  chap- 
arral at  1,320  m  near  station  29  (MVZ  140863), 
and  another  at  station  19  (MVZ  140864).  An 
individual  was  observed  at  1 140  hr  in  a  hollow 
oak  trunk  in  oak  woodland;  this  snake  was  feed- 
ing on  a  Peromyscus  sp.  Available  records  in- 
dicate a  preference  for  chaparral  and  riparian 
habitats  on  the  east  and  west  scarps  and  on  the 
lower  Pacific  Slope. 

Diadophis  punctatus  similis  Blanchard,  1923, 
San  Diego  Ringneck  Snake 

Schmidt  (1922)  recorded  Diadophis  punctatus 
similis  from  San  Matias  Pass.  This  specimen  was 
probably  collected  at  a  spring  south  of  the  Pass, 
and  some  150  m  higher,  in  pinyon-juniper 
woodland.  This  race  is  known  from  Rancho  San 
Jose  (SDNHM  36534),  where  it  occurs  in  ripar- 
ian habitat  (Aida  Meling,  pers.  comm.),  and  from 
an  irrigated  slope  of  Arroyo  San  Telmo,  south 
of  Colnett  (R.  Marlow,  pers.  comm.).  These  rec- 
ords, and  habitat  preferences  of  the  species  in 
other  areas  (Stebbins  1966),  indicate  it  is  prob- 
ably restricted  to  the  Pacific  Slope  of  the  Martir 
Region,  in  riparian  woodland  corridors  below 
1,500  m  elevation. 

Elaphe  rosaliae  (Mocquard,  1899), 
Baja  California  Rat  Snake 

Elaphe  rosaliae,  long  known  from  San  Bartolo 
and  Santa  Rosalia  in  Baja  California  Sur  (Schmidt 
1922;  see  Ottley  and  Jacobsen  1983  for  addi- 
tional localities),  has  been  reported  more  recently 
from  localities  to  the  north.  Hunsaker  (1965)  re- 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


35 


ported  two  specimens  at  Canon  Guadalupe  on 
the  east  scarp  of  the  Sierra  Juarez,  and  Stebbins 
(1985)  reported  a  specimen  from  Mt.  Spring,  Im- 
perial Co.,  California.  These  records  indicate  E. 
rosaliae  ranges  in  the  intervening  area  on  the 
eastern  slopes  of  the  Peninsular  Range  through 
the  Martir  Region  (Stebbins  1985). 

Hypsiglena  torquata  deserticola  Tanner,  1946, 
Desert  Night  Snake 

Hypsiglena  torquata  klauberi  Tanner,  1946, 
San  Diego  Night  Snake 

Tanner  ( 1 966a)  reviewed  available  material  of 
Hypsiglena  torquata  and  proposed  that  the  Pa- 
cific Slope  distribution  for  H.  t.  klauberi  in  Baja 
California  was  through  the  Martir  Region  to  the 
North  Central  Desert.  Bostic  (1971)  reported  a 
specimen  of  this  race  from  near  El  Rosario.  Tan- 
ner ( 1 966fl)  suggested  that  H.  t.  deserticola  occurs 
in  northeastern  Baja  along  the  Gulf  Slope.  This 
was  based  on  five  specimens  from  the  Central 
Desert  Region  that  showed  evidence  of intergra- 
dation  between  H.  t.  klauberi  and  H.  t.  desertico- 
la. Tanner  also  indicated  that  the  species  occurs 
near  San  Felipe.  The  above  records  and  known 
habitat  preferences  of  this  species  (Stebbins  1966) 
indicate  that  it  probably  inhabits  all  scrub  and 
xeric  woodland  areas  on  both  slopes  of  the  Re- 
gion. 

Lampropeltis  getulus  californiae 
(Blainville,  1835), 
California  Kingsnake 

This  snake  is  known  from  the  East  Scarp  at 
Canon  El  Cajon  (Linsdale  1932).  I  observed  a 
juvenile  on  the  East  Scarp  at  station  20.  It  was 
active  in  shade  at  0840  hr  on  a  rocky  slope  of 
pinyon-juniper  woodland.  This  race  is  also 
known  on  the  Gulf  Slope  at  Bahia  de  San  Luis 
Gonzaga  (CAS  90227),  an  area  of  creosote  bush 
scrub.  Lower  Pacific  Slope  records  are  relatively 
numerous,  indicating  that  the  snake  occurs  in 
coastal  sage  and  Central  Desert  scrub  south  to 
El  Rosario  (SDNHM  1070,  41264).  I  observed 
two  specimens  on  the  lower  Pacific  Slope,  at  sta- 
tion 27;  one  was  in  coastal  sage  scrub,  and  the 
other  under  a  cotton  wood  tree  (Populus  sp.)  in  a 
grassy  field  at  the  edge  of  riparian  habitat.  This 
snake  has  not  been  reported  from  the  central 
South  Colorado  Desert  or  from  the  upper  Sierran 
scarps. 


Schmidt  ( 1 922)  reported  Lampropeltis  getulus 
yumensis  from  Volcano  Lake  in  the  Colorado 
Desert  of  extreme  northeastern  Baja  California. 
Two  records  from  1 1.6  km  east  of  Cerro  Prieto 
(32°25'N)  are  also  in  this  area  (MVZ  10761-62). 

Lampropeltis  zonata  agalma 
Van  Denburgh  and  Slevin,  1923, 
Baja  California  Mountain  Kingsnake 

I  collected  a  single  Lampropeltis  zonata  agal- 
ma active  at  1 635  hr  in  an  area  of  rocky  chaparral 
at  1,600  m,  1.7  km  west  of  station  29  (MVZ 
140871).  This  locality  was  about  10  km  down 
slope  from  the  nearest  coniferous  forest  habitat. 
All  other  locality  records  for  this  snake  in  the 
Martir  Region  (Zweifel  1952,  1974;  SDNHM 
46017;  and  R.  Moran,  pers.  comm.)  are  from 
higher  elevations,  up  to  2,800  m,  in  areas  of 
coniferous  forest. 

Throughout  its  range,  Lampropeltis  zonata  oc- 
curs sympatrically  with  the  wide-ranging  L.  ge- 
tulus, whose  range  circumscribes  the  areas  oc- 
cupied by  L.  zonata  (Stebbins  1966).  Storer  and 
Usinger  (1963)  indicated  that  L.  getulus  is  rare 
or  absent  in  coniferous  forests  of  the  Sierra 
Nevada  of  California,  habitat  where  L.  zonata  is 
found  most  commonly.  L.  getulus  occurs  up  to 
2,100  m  (Stebbins  1966),  but  is  uncommon  in 
coniferous  forest  habitats,  even  where  L.  zonata 
is  absent.  Perhaps  L.  getulus  is  limited  by  an 
elevational  factor  such  as  temperature,  and  L. 
zonata  replaces  it  in  these  higher,  cooler  areas. 
Out  of  coniferous  forest  habitats  L.  zonata  is  un- 
common, but  it  does  occur  marginally  in  chap- 
arral throughout  its  range  (Zweifel  1952,  1974; 
this  study). 

Phyllorhynchus  decurtatus  perkinsi 
Klauber,  1935, 
Western  Leafnose  Snake 

This  snake  is  known  in  the  Martir  Region  only 
from  the  South  Colorado  Desert  of  the  Gulf 
Slope;  records  exist  from  21.8  km  north  of  San 
Felipe  (SDNHM  18917)  and  1.6  km  north  of 
San  Felipe  (SDNHM  38108).  Welsh  and  Bury 
(1984)  reported  four  additional  specimens  from 
localities  50  km  northwest  of  San  Felipe.  The 
subspecies  P.  d.  decurtatus  occurs  from  south  of 
the  Martir  Region  through  central  Baja  Califor- 
nia and  south  into  the  Cape  Region.  Specimens 
are  known  from  16  km  north  of  Rancho  Cata- 


36 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


vina  (LACM  20751),  16.2  km  south  of  Punta 
Prieta  (Bostic  1971),  16.7  km  southeast  of  Mes- 
quital  (MVZ  50171)  in  the  Central  Desert,  and 
from  Bahia  de  Los  Angeles  (SU  2 1 800)  on  the 
Gulf  coast.  R.  Seib  (pers.  comm.)  reported  14 
additional  records  from  the  Central  Desert 
southward. 


Pituophis  melanoleucus  annectens 
Baird  and  Girard,  1853, 
San  Diego  Gopher  Snake 

Pituophis  melanoleucus  bimaris  Klauber,  1 946a, 
Baja  California  Gopher  Snake 

Pituophis  melanoleucus  affinis  Hallowell,  1852, 
Sonoran  Gopher  Snake 

Klauber  (\946d)  reported  Pituophis  melano- 
leucus annectens  from  numerous  localities  on  the 
Pacific  Slope  of  the  Martir  Region  up  into  co- 
niferous forest,  and  north  into  California.  Mur- 
ray (1955)  reported  this  form  at  1,400  m  in  chap- 
arral on  the  West  Scarp.  It  is  known  along  the 
coast  at  San  Antonio  del  Mar  (SDNHM  23800 
and  32149),  and  Bostic  (1971)  reported  this  race 
in  the  Central  Desert,  31.8  km  southeast  of  El 
Rosario. 

I  collected  three  snakes  that  all  correspond  in 
pattern  and  color  with  Pituophis  melanoleucus 
annectens.  A  snake  was  found  crossing  a  road  at 
1730  hr,  15.2  km  east  of  station  7  (MVZ  140869); 
a  second  individual  was  found  on  the  road  at 
1400  hr,  4.8  km  west  of  station  29  (destroyed  in 
transit);  and  a  third  was  found  dead  on  the  road 
in  midafternoon  at  station  28  (MVZ  140870). 
The  first  locality  was  in  an  area  of  Central  Desert/ 
coastal  sage  scrub,  and  the  other  two  sites  were 
in  areas  of  coastal  sage  scrub/chaparral  ecotone. 

The  subspecies  Pituophis  melanoleucus  bi- 
maris occurs  throughout  the  Central  Desert 
(Klauber  1 946a),  whereas  P.  m.  affinis  is  known 
from  two  localities  in  extreme  northeastern  Baja 
(Klauber  \946a).  The  latter  form  has  not  yet 
been  reported  in  the  desert  east  of  the  Sierra  San 
Pedro  Martir,  but  Welsh  and  Bury  (1984)  re- 
ported a  specimen  from  9.1  km  east  of  San  Ma- 
tias  Pass  that  appears  to  be  an  intergrade  between 
P.  m.  affinis  and  P.  m.  annectens. 

Klauber  (1946«)  noted  overlap  in  the  ranges 
of  Pituophis  melanoleucus  annectens  and  P.  m. 
bimaris  on  the  Pacific  Slope  through  the  Martir 
Region,  from  El  Rosario  north  to  Alamo;  he  found 


no  evidence  of  intergradation  in  the  specimens 
he  examined  from  this  area. 

Rhinocheilus  lecontei  lecontei 
Baird  and  Girard,  1853, 
Western  Longnose  Snake 

Klauber  (1941)  reported  two  records  from  ex- 
treme northeastern  Baja  California.  Other  rec- 
ords from  this  area  include:  51.2  km  east  of  Mex- 
icali  (LACM  36576),  east  of  Mexicali  (SDNHM 
19884-85;  LACM  20821),  and  Tajo  Canyon 
(SDNHM  1 8875).  Few  records  exist  for  the  east- 
ern Martir  Region;  Medica  (1975)  reported  one 
specimen  from  San  Felipe,  and  Welsh  and  Bury 
(1984)  reported  a  specimen  from  50  km  north 
and  12  km  west  of  San  Felipe.  There  are  several 
records  from  the  Pacific  Slope,  both  north  into 
California,  and  within  the  Martir  Region  as  fol- 
lows: 3.2  km  and  11.2  km  south  of  Santo  Tomas 
(LACM  59125,  102711),  8.3  km  east  of  Cabo 
Colnett  (SDNHM  34003),  3.3  km  south  of  San 
Simon  (SDNHM  42094),  3.3  km  north  of  San 
Simon  (SDNHM  42632),  and  12.7  km  north  of 
Arroyo  Seco  (SDNHM  42543).  These  Pacific 
Slope  localities  are  in  areas  of  coastal  sage  and 
Central  Desert  scrub,  whereas  the  records  from 
the  Gulf  Slope  correspond  with  areas  of  creosote 
bush  scrub.  Records  from  1 .7  km  west  of  El  Mar- 
mol  (SDNHM  42439),  24.2  km  south  of  Rancho 
Santa  Ines  (CAS  143725),  and  San  Borja 
(SDNHM  42439)  confirm  this  species  in  the 
Central  Desert.  Lockington  (1880)  reported  this 
snake  from  the  vicinity  of  Magdelena  Bay  in  Baja 
California  Sur,  but  Klauber  (1941)  doubted  the 
validity  of  this  record.  The  two  recent  specimens 
from  the  Central  Desert  Region,  and  a  record 
from  Cerralvo  Island  off  the  Cape  Region  (Soule 
and  Sloan  1 966),  give  renewed  credence  to  Lock- 
ington's  record.  Medica  (1975)  stated  that  this 
species  will  probably  be  found  to  inhabit  all  of 
Baja  California. 

Salvadora  hexalepis  hexalepis  (Cope,  1867), 
Desert  Patchnose  Snake 

Salvadora  hexalepis  klauberi  Bogert,  1945, 
Baja  California  Patchnose  Snake 

Salvadora  hexalepis  virgultea  Bogert,  1939, 
Coast  Patchnose  Snake 

A  specimen  of  Salvadora  hexalepis  was  col- 
lected at  0707  hr  crossing  a  sandy  road  in  chap- 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


37 


arral-coastal  sage  ecotone  at  915  m,  5.8  km  east 
of  station  27  (MVZ  140865).  A  second  individ- 
ual was  found  at  0906  hr  in  chaparral  at  station 
11  (MVZ  140866).  Both  specimens  fit  the  de- 
scription of  S.  h.  virgultea  (Bogert  1945).  This 
subspecies  is  known  from  the  Pacific  Slope  of 
northern  Baja  California,  from  California  to  the 
vicinity  of  El  Rosario.  Records  in  the  Martir  Re- 
gion indicate  an  affinity  to  areas  of  chaparral, 
coastal  sage,  and  Central  Desert  scrub  vegeta- 
tions. 

Bogert  (1945)  described  Salvadora  hexalepis 
klauberi  from  the  Central  Desert,  and  indicated 
a  zone  of  intergradation  between  S.  h.  klauberi 
and  S.  h.  virgultea  in  the  vicinity  of  latitude  30°N 
(El  Rosario).  His  speculation  that  S.  h.  hexalepis 
might  range  into  northeastern  Baja  in  the  Col- 
orado Desert  was  confirmed  by  a  specimen  from 
the  Gulf  Slope  of  the  Martir  Region  reported  by 
Welsh  and  Bury  (1984). 

Sonora  semiannulata  Baird  and  Girard,  1853, 
Western  Ground  Snake 

Frost  (1983)  placed  the  two  banded  ground 
snakes  of  Baja  California,  Sonora  bancroftae  and 
S.  mosaueri,  in  synonymy  with  the  unbanded  S. 
semiannulata;  at  that  time  he  concluded  that  in- 
sufficient evidence  existed  to  describe  subspe- 
cies. Despite  Frost's  conclusion  I  think  it  im- 
portant to  bear  in  mind  that  the  banded  forms 
are  superficially  different  from  the  unbanded  S. 
semiannulata,  and  that  their  respective  distri- 
butions are  not  inconsistent  with  other  unique 
elements  of  the  Baja  California  herpetofauna. 
Because  of  the  possible  zoogeographic  implica- 
tions of  these  populations  I  have  chosen  to  treat 
the  banded  forms  here  as  subspecies. 

Stickel  (1943)  indicated  that  S.  semiannulata 
occurred  through  the  Colorado  and  Central  Des- 
ert Regions  of  Baja  California  to  Santa  Rosalia; 
however  Frost  (1983)  indicated  a  lack  of  records 
to  confirm  the  species  in  the  Colorado  Desert 
east  of  the  Sierra  San  Pedro  Martir. 

Klauber  ( 1 943)  described  the  banded  form  So- 
nora bancroftae  from  a  single  specimen  collected 
in  Arroyo  San  Telmo,  3.3  km  east  of  San  Jorge, 
a  site  on  the  lower  Pacific  Slope  of  the  Martir 
Region,  in  an  area  of  Central  Desert/coastal  sage 
scrub  ecotone.  A  second  individual  is  known  from 
Rancho  Buena  Vista  (SDNHM  36527),  27  km 
further  east  and  slightly  higher  in  the  same  arroyo 
in  coastal  sage  scrub. 


Stickel  (1938)  described  another  banded  form, 
Sonora  mosaueri,  from  three  specimens  collect- 
ed at  Comondu  in  Baja  California  Sur.  It  has 
since  been  reported  from  the  Vizcaino  Desert  in 
the  vicinity  of  Santa  Rosalia  (SDNHM  44468 
and  46808),  and  in  the  Central  Desert,  37  km 
south  of  El  Rosario  (SDNHM  46802).  Stickel 
(1 943)  noted  that  S.  mosaueri  was  closely  related 
to,  and  might  prove  to  be  a  subspecies  of  5". 
semiannulata.  The  distribution  of  the  three  forms 
of  Sonora  are  in  disparate,  yet  contiguous,  areas 
of  desert  scrub  vegetations. 

Tantilla  planiceps  (Blainville,  1835), 
California  Blackhead  Snake 

Cole  and  Hardy  (1981)  recognized  no  subspe- 
cies of  Tantilla  planiceps.  Tanner  (1966&)  re- 
ported five  specimens  from  the  Pacific  Slope  of 
Baja  California,  three  from  north  of  the  Martir 
Region,  one  from  6.4  km  south  of  San  Vicente 
in  the  Western  Foothills,  and  one  from  the  Cen- 
tral Desert.  Bostic  (1971)  reported  a  specimen 
from  the  Central  Desert,  39  km  southeast  of  El 
Rosario.  These  localities  indicate  that  T.  plani- 
ceps occurs  in  coastal  sage  and  Central  Desert 
scrub  vegetations  within  the  Martir  Region.  A 
single  record  exists  from  the  Gulf  Slope  at  Ar- 
royo El  Tajo  on  the  east  scarp  of  the  Sierra  Juarez 
(SDNHM  45001),  indicating  T.  planiceps  is 
probably  also  present  on  the  Gulf  Slope  of  the 
Martir  Region. 

Thamnophis  elegans  hueyi 

Van  Denburgh  and  Slevin,  1923, 

San  Pedro  Martir  Garter  Snake 

This  endemic  subspecies  of  Thamnophis  ele- 
gans is  found  only  on  the  Main  Scarp  of  the  Sierra 
San  Pedro  Martir  (Van  Denburgh  and  Slevin 
1923)  and  occurs  in  riparian  meadow  and  nearby 
moist  woodland  habitats  of  the  coniferous  forest. 

Elsewhere,  Thamnophis  elegans  occurs  at  all 
elevations  up  to  3,000  m,  and  is  associated  with 
a  variety  of  terrestrial  and  riparian  habitats 
(Stebbins  1966).  In  the  Martir  Region,  T.  elegans 
is  found  only  in  and  around  marshy  meadows 
and  slow  streams  above  1,820  m.  T.  hammondii 
is  also  present,  but  uncommon  in  these  areas;  it 
occurs  more  frequently  in  and  around  the  fast- 
flowing  streams  in  the  steep  riparian  areas  below 
the  Main  Scarp  where  T.  elegans  is  absent.  My 
observations  of  foraging  of  Thamnophis  spp.  in 


38 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


the  Martir  Region  suggest  that  both  species  may 
be  localized  to  riparian  habitats  because  of  the 
moisture  requirements  of  their  amphibian  prey 
species.  The  usually  eurytopic  T.  elegans  may  be 
further  limited  to  marshy  habitats  of  the  main 
scarp  by  the  presence  of  T.  hammondii.  Foraging 
strategies  may  be  an  important  factor  in  segre- 
gating these  two  species  in  areas  of  sympatry  such 
as  the  Lower  Main  Scarp.  T.  hammondii  is  no- 
ticeably more  aquatic,  and  was  seldom  observed 
out  of  water.  It  was  observed  in  pursuit  of  anuran 
larvae  and  Hyla  cadaverina.  T.  elegans  seemed 
to  prefer  moist  or  damp  habitats  but  was  not 
observed  in  the  water.  It  was  observed  feeding 
on  toads  on  marshy  ground. 

An  adult  female  collected  22  June  1974  (sta- 
tion 24)  (MVZ  140876)  contained  eight  small 
ova.  An  adult  female  found  under  a  log  at  1916 
hr  had  recently  ingested  an  adult  Bufo  boreas. 
Specimens  were  collected  at  stations  12  (MVZ 
140875),  25  (MVZ  140872),  37  (MVZ  140873), 
and  40  (MVZ  140874);  observations  were  made 
at  stations  24  and  39. 

Thamnophis  hammondii  (Kennicott,  1 860), 
Two-striped  Garter  Snake 

This  snake  occurs  on  the  East  Scarp  and  the 
Pacific  Slope  of  the  Region,  frequenting  only  ri- 
parian habitat.  It  occurs  to  2, 1 20  m  on  the  Lower 
Main  Scarp  in  coniferous  forest,  on  the  East  Scarp 
in  creosote  bush  scrub  and  pinyon-juniper 
woodland,  and  on  the  Pacific  Slope  in  arroyos 
dominated  by  chaparral  and  coastal  sage  scrub, 
ranging  west  to  the  San  Quintin  Plain.  A  juvenile 
was  observed  at  station  65  in  an  irrigation  ditch 
that  flowed  from  an  apparently  permanent  spring 
in  Arroyo  Edwardo,  an  area  of  Central  Desert 
scrub.  Records  from  El  Rosario  (SDNHM  41593; 
LACM  1 04069-7 1 )  are  the  southernmost  for  this 
species. 

Thamnophis  hammondii  was  observed  under- 
water, along  streamsides,  on  streamside  rock 
faces,  and  up  to  3.4  m  above  ground  in  riparian 
overstory  (usually  willow).  I  twice  observed 
snakes  in  apparent  pursuit  of  Hyla  cadaverina. 
On  three  occasions,  T.  hammondii  were  ob- 
served underwater  stalking  and  lunging  at  anuran 
larvae. 

Specimens  were  collected  at  stations  1 7  (MVZ 
140880),  37  (MVZ  140887,  147528),  and  49 
(MVZ  140878-79).  Observations  were  made  at 
stations  10,  16,  20,  21,  23,  27,  38,  and  65. 


Trimorphodon  biscutatus  vandenburghi 
Klauber,  1924, 
California  Lyre  Snake 

An  adult  male,  matching  Gehlbach's  (1971) 
description  for  Trimorphodon  biscutatus  van- 
denburghi, was  found  dead  on  a  dirt  road  on  the 
morning  of  22  June  1973,  at  1,520  m  in  rocky 
chaparral,  3.5  km  west  of  station  29  (MVZ 
140877).  This  subspecies  is  known  in  north- 
western Baja  California  from  six  other  speci- 
mens: 1 6  km  north  of  Arroyo  Seco  on  the  San 
Quintin  Plain  (Murray  1955),  2.9  km  north  of 
El  Rosario  de  Arriba  (LACM  121552),  4.8  km 
east  of  La  Rumerosa  (SDNHM  121 149),  2.2  km 
south  of  Catavina  (CAS  146577),  27  km  south 
of  Santa  Ines  (MVZ  161561),  and  32  km  south- 
east of  Punta  Prieta  (LACM  126261).  Scott  and 
McDiarmid  (1984)  showed  a  lack  of  records  from 
the  Gulf  Slope  of  northeastern  Baja  California; 
Welsh  and  Bury  (1984)  reported  the  first  speci- 
men from  34  km  east  of  San  Matias  Pass  in  the 
South  Colorado  Desert.  Habitat  preferences 
(Stebbins  1 966)  suggest  the  probable  occurrence 
of  this  snake  in  pinyon-juniper,  chaparral,  and 
all  scrub  vegetations  of  the  Martir  Region.  The 
absence  of  contiguous  rocky  habitat  may  limit 
its  distribution  in  parts  of  the  Colorado  Desert. 
Gehlbach  (1971)  indicated  that  T.  b.  lyrophanes 
of  southern  and  central  Baja  California  is  known 
as  far  north  as  San  Ignacio  in  the  Vizcaino  Des- 
ert. 

Vipiridae 

Crotalus  atrox  atrox  Baird  and  Girard,  1853, 
Western  Diamondback  Rattlesnake 

This  snake  ranges  into  the  Colorado  Desert  of 
extreme  northeastern  Baja  California  (Schmidt 
1 922;  Linsdale  1 932),  and  has  been  reported  from 
Central  Baja  California  (Alvarez  and  Huerta 
1 974).  It  has  yet  to  be  found  in  the  Martir  Region. 
Suitable  habitat  for  this  species  occurs  in  the 
South  Colorado  Desert;  its  apparent  absence  may 
be  due  to  limited  collecting  in  this  remote  region. 

Crotalus  cerastes  laterorepens  Klauber,  1 944, 
Colorado  Desert  Sidewinder 

Crotalus  cerastes  laterorepens  is  known  from 
numerous  localities  in  the  South  Colorado  Des- 
ert from  San  Matias  Pass  to  the  vicinity  of  San 
Felipe  (Linsdale  1932;  Murray  1955;  Welsh  and 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


39 


Bury  1984),  south  to  San  Francisquito  on  the 
Gulf  coast  (Klauber  1 944).  Residents  of  the  Valle 
de  Trinidad  area  informed  me  that  it  occurs  at 
that  Pacific  Slope  locality.  All  regional  localities 
for  this  species  are  areas  of  creosote  bush  scrub. 
It  has  not  been  reported  from  the  Pacific  Slope 
south  of  the  Martir  Region  and  the  Gulf  Slope 
south  of  San  Francisquito.  It  may  be  excluded 
from  these  deserts  by  the  presence  of  the  closely 
related  form  C.  enyo. 

Crotalus  enyofurvus  Lowe  and  Norris,  1954, 
Dusky  Baja  California  Rattlesnake 

Crotalus  enyo  furvus  was  described  by  Lowe 
and  Norris  (1 954)  from  four  specimens  collected 
along  the  San  Quintin  Plain,  north  of  El  Rosario. 
Their  proposed  range  for  C.  e.  furvus,  based  on 
vegetative  and  edaphic  affinities,  corresponds 
with  Sonoran  habitat  of  the  Pacific  coast  of  the 
Martir  Region  (Short  and  Crossin  1967),  herein 
considered  as  part  of  the  Central  Desert  scrub 
vegetational  association.  Lowe  and  Norris's  pro- 
posed range  also  corresponds  closely  to  the  San 
Quintin  Plain.  A  record  from  4.6  km  north  of 
Arroyo  Seco  (SDNHM  46196)  is  just  north  of 
this  proposed  range,  in  Central  Desert  scrub.  Short 
and  Crossin  (1967)  indicated  that  this  habitat 
extends  north  to  the  vicinity  of  Rio  San  Antonio 
along  the  base  of  the  foothills  and  as  far  as  Punta 
Banda  on  the  coast.  Future  work  in  the  area  may 
reveal  that  the  northern  limit  of  this  snake  cor- 
responds closely  with  this  habitat. 

Crotalus  enyo  enyo  is  known  in  more  arid  hab- 
itats as  far  north  as  17.2  km  southeast  of  El  Ro- 
sario (Bostic  1971),  and  from  numerous  southern 
localities  of  the  Central  Desert,  Vizcaino,  and 
Cape  Regions  (Lowe  and  Norris  1954). 

Crotalus  mitchellii  pyrrhus  (Cope,  1867), 
Southwestern  Speckled  Rattlesnake 

Crotalus  mitchellii  pyrrhus  occurs  on  the  Gulf 
and  Pacific  slopes  of  the  Martir  Region.  I  ob- 
served two  snakes  on  the  rocky  East  Scarp  at 
stations  17  and  21  (MVZ  140884),  in  creosote 
bush  scrub  and  pinyon-juniper,  respectively.  A 
third  individual  was  observed  on  the  lower  West 
Scarp  at  station  28  in  coastal  sage-chaparral  eco- 
tone.  Klauber  (1936)  indicated  the  occurrence  of 
other  records  from  the  Martir  scarps,  Western 
Foothills,  and  northward  on  both  slopes  into  the 
southern  U.S.  The  three  snakes  I  observed  were 


active  during  midmorning  among  boulders.  Lit- 
erature records  and  my  observations  indicate  that 
C.  m.  pyrrhus  is  primarily  saxicolous,  although 
some  records  exist  for  other  than  rocky  habitat 
(Klauber  1972). 

South  of  the  Martir  Region  on  the  Gulf  Slope, 
Crotalus  mitchellii  pyrrhus  is  known  from  Bah ia 
de  Los  Angeles  (SDNHM  1 97 1 3-14).  Specimens 
from  4.8  km  north  of  (LACM  20005)  and  28.8 
km  south  of  (LACM  20006)  Rancho  Catavina 
are  apparently  the  first  records  of  this  species 
from  the  Pacific  Slope  of  the  Central  Desert.  The 
subspecies  C.  m.  mitchellii  occurs  in  the  Vizcaino 
Desert  and  south  into  the  Cape  Region  of  the 
Peninsula. 

Crotalus  ruber  ruber  Cope,  18926, 
Red  Diamondback  Rattlesnake 

This  subspecies  occurs  from  San  Diego  Coun- 
ty, California,  south  through  the  Martir  Region 
and  the  Central  Desert  (Klauber  1972).  It  has 
been  reported  from  numerous  localities  on  the 
lower  Pacific  Slope  in  the  Western  Foothills  and 
on  the  San  Quintin  Plain  (Klauber  1 972).  I  found 
it  in  the  sandy  desert  at  the  base  of  the  East  Scarp. 
Welsh  and  Bury  (1984)  reported  it  from  a  Gulf 
coast  locality  north  of  San  Felipe. 

I  observed  two  specimens  at  the  base  of  the 
East  Scarp  in  creosote  bush  scrub.  One  was  found 
active  in  late  afternoon  shade  on  an  open  sandy 
slope  at  station  54,  the  other  at  station  17,  coiled 
in  the  midafternoon  shade  of  an  overhanging 
rock  among  talus.  I  collected  one  specimen  at 
night  on  the  Pacific  Slope,  beside  a  dirt  road,  5 
km  north  of  El  Burro  (station  3)  (MVZ  140885), 
in  an  area  of  chaparral-piny on-juniper  ecotone. 

Crotalus  viridis  helleri  Meek,  1905, 
Southern  Pacific  Rattlesnake 

I  found  nine  records  of  Crotalus  viridis  helleri 
on  the  interior  lower  Pacific  Slope  of  the  Martir 
Region,  all  corresponding  with  areas  containing 
riparian  woodland  habitat:  La  Toya,  Valladares 
Creek,  and  Rancho  San  Jose.  Other  records  from 
the  Martir  and  Central  Desert  regions  indicate 
that  this  species  has  an  affinity  for  coastal  hab- 
itats: San  Antonio  del  Mar  (CAS  57555);  West 
of  San  Quintin  Bay  (SU  12141-42);  Camalu 
Plains,  8.3  km  north  of  Santo  Domingo  (SU 
1 1574);  Punta  Camalu  (CAS  40140);  Playa  Ma- 
ria Bay  (Klauber  1949);  and  16.7  km  south  of 


40 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


Jesus  Maria  (Bostic  1971).  In  the  Sierra  San  Pe- 
dro Martir,  I  found  C.  v.  helleri  up  to  2, 1 20  m 
on  the  Lower  Main  Scarp  at  stations  37  (MVZ 
140822),  38  (MVZ  140883),  and  39  (MVZ 
147529),  in  coniferous  forest  and  in  conifer- 
chaparral  ecotone.  This  snake  was  most  often 
encountered  in  riparian  edge  habitat  in  these 
areas.  I  also  found  a  specimen  in  open  chaparral 
at  1,300  m.  It  was  coiled  in  sand  on  a  road  at 
2008  hr.  C.  v.  helleri  has  not  been  reported  from 
the  Gulf  Slope  of  the  Martir  Region. 

This  species  frequents  riparian,  coastal,  and 
upland  habitats  but  is  apparently  absent  from 
inland  scrub  areas  of  the  Martir  and  Central  Des- 
ert regions.  However,  two  records  are  known  from 
arid  situations  in  the  Central  Desert:  33.4  km 
northwest  of  Bahia  de  Los  Angeles  (Klauber 
1949),  and  33.4  km  southeast  of  Mezquital 
(Murray  1 955).  Specimens  from  the  Central  Des- 
ert are  described  by  Murray  (1955)  and  Bostic 
(1971)  as  light  in  color.  Specimens  I  found  in  the 
Martir  Region  were  dark,  almost  black. 

Jones  (1981)  and  Ottley  and  Hunt  (1981)  re- 
ported two  records  ofCrotalus  viridis  helleri  from 
Baja  California  del  Sur;  these  specimens  repre- 
sent the  southernmost  records  for  the  species. 

ZOOGEOGRAPHIC  ANALYSES 

I  used  both  a  numerical  analysis  and  a  heuristic 
comparison  of  distributions  (based  on  locality 
records  listed  in  Species  Accounts)  across  the 
matrix  of  ecogeographic  formations  (Fig.  1 4)  to 
determine  distributional  patterns  within  the  Re- 
gion. Regional  patterns  were  then  extrapolated 
beyond  the  Region  to  include  the  entire  distri- 
bution of  each  species  in  order  to  test  for  overall 
geographic  congruency  between  species  linked  by 
a  common  ecogeographic  pattern  within  the  Re- 
gion (analysis  of  tracks).  Evidence  of  such  overall 
distributional  congruency  was  considered  indic- 
ative of  a  common  evolutionary  history  among 
linked  species. 

Numerical  Analysis 

Cheetham  and  Hazel  (1969)  reviewed  and  dis- 
cussed the  use  of  numerical  methods  (Faunal  Re- 
semblance Factors  or  FRFs)  in  biogeography, 
noting  the  inherent  advantages,  disadvantages, 
and  assumptions  associated  with  binary  coeffi- 
cients. They  listed  some  22  possible  formulas  for 
expressing  the  relationship  of  the  number  of  taxa 


in  common  between  paired  geographic  units  to 
the  total  number  of  taxa  present.  In  order  to 
ameliorate  the  disadvantages  inherent  with  any 
particular  FRF,  I  employed  three  different  FRFs 
that  tend  to  complement  one  another  by  em- 
phasizing different  aspects  of  faunal  assemblages. 
An  FRF  that  emphasizes  similarities  can  indicate 
historical  affinities  and  evolutionary  relation- 
ships between  geographic  areas,  whereas  an  FRF 
that  emphasizes  differences  can  test  for  sampling 
error  and  the  effects  of  faunas  of  unequal  size, 
situations  minimized  by  similarity  coefficients. 
In  addition  to  using  a  coefficient  of  similarity 
and  a  coefficient  of  difference,  I  used  a  third  coef- 
ficient, one  that  emphasizes  ecological  relation- 
ships between  geographic  units— a  community 
coefficient  (Jaccard  1902).  Subspecific  differ- 
ences were  ignored  for  this  analysis;  distributions 
were  determined  by  species. 

All  available  records  (see  Species  Accounts) 
were  plotted  on  a  schematic,  southwest  to  north- 
east, regional  transect  in  the  vicinity  of  latitude 
3 1°N  (all  localities  used  are  between  30°30'N  and 
3 1°30'N  latitude).  The  ecotones  between  regional 
ecogeographic  formations  (Fig.  1 4)  were  used  to 
delineate  the  units  along  the  transect  used  for 
numerical  comparison.  A  species  was  considered 
to  occur  within  an  ecogeographic  formation  only 
if  it  ranged  across  the  ecotone;  marginal  species, 
those  found  only  to  occur  in  the  ecotone  and  not 
in  a  particular  ecogeographic  formation,  were  not 
included.  The  fauna  of  each  formation  was  com- 
pared with  those  of  all  other  formations  using 
the  following  three  FRFs:  Savage's  (1960)  Coef- 
ficient of  Difference  (CD  =  1  -  C/N2  x  100); 
Simpson's  (1960)  Similarity  Coefficient  (SC  = 
C/N,  x  100),  and  Jaccard's  (1902:  in  Morafka 
1977)  Community  Coefficient  (CC  =  [(C/N,  + 
N2  -  C)  x  100]),  where  C  =  the  number  of 
species  in  common,  N,  =  the  number  of  species 
in  the  area  with  the  fewest  species,  and  N2  =  the 
number  of  species  in  the  area  with  the  most 
species. 

Faunal  resemblance  factors  are  descriptive 
rather  than  statistical  indices  (Cheetham  and  Ha- 
zel 1969)  and  thus  do  not  have  levels  of  signif- 
icance. I  established  a  "level  of  significance"  for 
each  coefficient  on  the  basis  of  usage  by  previous 
workers,  and  the  relative  correlation  of  the  dif- 
ferent coefficient  values  derived  from  the  com- 
parison of  geographic  units  where  N,  approaches 
or  equals  N2.  My  primary  emphasis,  when  eval- 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


41 


uating  FRF  values,  was  to  weigh  all  three  values 
for  each  pair  of  geographic  units,  treating  all  units 
relative  to  one  another,  and  to  look  for  obvious 
clusters  occurring  in  ranked  coefficient  values. 
Savage  (1960)  arbitrarily  selected  CD  =  50%  or 
greater  as  indicative  of  a  unique  faunal  grouping 
and  I  follow  his  usage  as  a  basis.  Employing  the 
50%  level  of  significance  with  Jaccard's  CC  and 
Simpson's  SC  indicated  to  me  that  they  are  more 
sensitive  and  less  sensitive,  respectively,  than  the 
CD  for  distinguishing  unique  faunal  groupings. 
I  have  therefore  partially  weighted  these  coeffi- 
cients. Those  ecogeographic  formation  pairs  with 
at  least  two  of  three  coefficients  meeting  the  fol- 
lowing criteria:  CD  <  50%,  CC  >  40%,  and  SC 
>  60%,  were  considered  to  be  related  at  a  sig- 
nificant level. 

Thirteen  pairs  met  my  criteria  for  significant 
relatedness  (Table  4).  The  relationships  between 
these  1 3  pairs  are  depicted  phenographically  (Fig. 
15a-c).  Subsequent  pairwise  analyses  between 
ecogeographic  formations  using  both  average 
linkage  (UPGMA)  and  single  linkage  (nearest 
neighbor)  methods  (SAS  1985)  yielded  results 
nearly  identical  to  those  depicted  in  Figure 
15a-c. 

Results  of  the  numerical  comparison  indicate 
the  presence  of  four  distinct  herpetofaunal  areas 
within  the  Martir  Region,  each  showing  some 
degree  of  overlap  with  one  or  more  of  the  ad- 
jacent areas  (Fig.  1 5a-c).  These  areas  are: 

( 1 )  The  Upper  Main  Scarp  and  the  Lower  Main 
Scarp— these  two  formations  support  the  same 
montane  herpetofauna  (SC  =  100%)  with  nu- 
merous species  dropping  out  with  increasing 
elevation  (CD  =  60%).  The  unequal  fauna  sizes 
between  these  two  formations  (7  vs.  20)  account 
for  the  high  SC  and  low  CC  values  (CC  =  40%). 

(2)  The  Lower  Main  Scarp,  West  Scarp,  and 
the  Western  Arroyos— these  three  formations 
support  a  fauna  that  apparently  shares  some  eco- 
logical characteristics  and  possibly  a  convergent 
history  with  that  of  (1)  above,  but  differ  in  their 
adaptation  to  the  warmer  and  drier  habitats  of 
lower  elevations,  extending  into  some  relatively 
extreme  xeric  areas  where  these  species  occur 
along  the  riparian  corridors  of  the  Western  Ar- 
royos, and  similar  habitats  of  the  East  Scarp. 

(3)  East  Scarp  and  West  Scarp— the  high  SC 
and  CC  and  the  low  CD  values  between  these 
formations  indicate  strong  faunal  affinities  be- 
tween these  two  rocky  xeric  woodland  areas. 


However,  each  of  these  two  formations  dem- 
onstrates only  slightly  less  significant,  differing 
affinities,  with  proximal  formations;  a  situation 
that  initially  seems  somewhat  contrary  and  con- 
fusing. The  East  Scarp  shows  close  faunal  links 
with  the  Colorado  Desert  and  the  Western  Foot- 
hills, while  the  West  Scarp  shows  close  affinities 
with  the  Western  Arroyos  and  the  Western  Foot- 
hills (the  FRF  comparisons  of  the  West  Scarp 
and  the  Western  Foothills  are  not  depicted  phe- 
nographically in  Fig.  5a-c  because  they  fell  just 
below  my  levels  of  significance:  CD  =  49%,  SC 
=  55%,  and  CC  =  36%).  I  interpret  these  data  to 
mean  that  two  distinct  overlapping  faunas  are 
present  on  the  East  and  West  scarps,  each  con- 
tinuing with  slightly  differing  distributions  onto 
adjacent  formations. 

(4)  The  Colorado  Desert,  North  Central  Des- 
ert, San  Quintin  Plain,  and  the  Western  Foot- 
hills—all coefficient  values  between  these  four 
relatively  open,  scrub  desert  formations  indicate 
strong  faunal  ties. 

Further  analysis  of  these  numerical  compari- 
sons follows  in  conjunction  with  the  results  of 
the  heuristic  analysis  below. 

Heuristic  Analysis 

The  heuristic  analysis  of  distributions  of  re- 
gional species  consisted  of  two  parts:  ( 1 )  a  com- 
parison of  ecogeographic  distributions  along  a 
schematic,  southwest  to  northeast,  regional  tran- 
sect between  latitudes  30°30'N  and  31°30'N  (re- 
sults depicted  in  Table  5);  (2)  a  comparison  of 
overall  regional  distributions  established  by  plot- 
ting available  literature  and  field  records  (Species 
Accounts)  on  maps  based  on  Figures  3,  4,  and  5 
(results  illustrated  in  Fig.  1 6).  A  species  that  oc- 
curred at  two  or  more  localities  within  a  contig- 
uous vegetation  or  physiographic  area  was  as- 
sumed to  occur  throughout  the  contiguous  habitat 
where  suitable  microhabitat  existed,  but  not 
above  or  below  the  highest  or  lowest  elevation 
records.  I  included  locality  records  from  the  Sier- 
ra Juarez  Region  to  the  north  and  the  Central 
Desert  Region  to  the  south  to  discern  distribu- 
tional trends  over  a  wider  area. 

Three  distinct  patterns  were  apparent  from  the 
transect  analysis  (Table  5):  (1)  the  High  Sierran 
Fauna  (I),  with  distributions  corresponding 
closely  to  those  of  the  Upper  and  Lower  Main 
Scarp  formations,  areas  linked  by  numerical 


42 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


43 


TABLE  4.  COMPARISON  OF  REGIONAL  HERPETOFAUNAS  BY  ECOGEOGRAPHIC  FORMATION  (Fie.  14)  USING  THREE  FAUNAL  COEF- 
FICIENT FACTORS.'  Number  of  species  present  in  parentheses.  C  =  number  of  species  in  common.  Formation  pairs  with  two  of 
three  coefficients  as  follows  are  considered  to  be  related  at  a  significant  level:  CD  <  50%,  SC  >  60%,  CC  ^  40%. 


East  Scarp 
(26) 

Upper  Main  Scarp 
(8) 

Lower  Main  Scarp 
(20) 

West  Scarp 

(31) 

West  Foothills 
(29) 

C 
CD 
SC 
CC 
C 
CD 
SC 
CC 
C 
CD 
SC 
CC 
C 
CD 
SC 
CC 
C 
CD 
SC 

18 
49% 
69% 

41% 

0 
100% 
0% 
0% 
3 
92% 
15% 
7% 
12 
66% 
38% 
22% 
22 
38% 
75% 

0 
100% 
0% 
0% 
9 
66% 
45% 
24% 
20 
36% 
76% 
54% 
16 
45% 
61% 

8 
60% 
100% 
40% 
6 
81% 
75% 
18% 
1 
97% 
12% 

16 

49% 
80% 
45% 
5 
83% 
25% 

16 
49% 
55% 

CC 

52% 

46% 

3% 

11% 

36% 

West  Arroyos 

C 

2 

9 

4 

12 

16 

5 

(22) 

CD 

94% 

66% 

82% 

46% 

49% 

83% 

SC 

9% 

40% 

50% 

60% 

72% 

22% 

CC 

4% 

23% 

15% 

37% 

43% 

10% 

San  Quintin  Plain 

C 

19 

13 

2 

5 

15 

22 

8 

(30) 

CD 

46% 

57% 

94% 

84% 

52% 

27% 

74% 

SC 

63% 

50% 

25% 

25% 

50% 

75% 

36% 

CC 

41% 

30% 

6% 

11% 

32% 

59% 

18% 

Central  Desert 

C 

26 

15 

2 

6 

14 

23 

6 

23 

(32) 

CD 

27% 

54% 

94% 

82% 

57% 

29% 

82% 

29% 

SC 

81% 

57% 

25% 

30% 

45% 

79% 

27% 

76% 

CC 

63% 

34% 

5% 

13% 

28% 

60% 

12% 

58% 

Colo. 

East 

Upper 

Lower 

West 

West 

West 

San 

Desert 

Scarp 

Main 

Main 

Scarp 

Foot- 

Ar- 

Quintin 

Scarp 

Scarp 

hills 

royos 

Plain 

(35) 

(26) 

(8) 

(20) 

(31) 

(29) 

(22) 

(30) 

1  CD  =  Coefficient  of  Difference  (CD  =  1  -  C/N2  x  100)  (Savage  1960);  SC  =  Similarity  Coefficient  (SC  =  C/N,  x  100) 
(Simpson  1960);  CC  =  Community  Coefficient  [CC  =  (C/N,  +  N2  -  C)  x  100]  (Jaccard  1902,  in  Morafka  1977).  N,  =  number 
of  species  in  the  area  having  the  fewest  species,  N2  =  the  number  of  species  in  the  area  with  the  most  species. 


comparison  above,  and  indicating  a  montane 
fauna;  (2)  the  Californian  Fauna  (II)  with  distri- 
butions corresponding  closely  to  those  of  the  sec- 
ond faunal  area  indicated  numerically  above,  and 
encompassing  primarily  the  Lower  Main  and 
West  scarps  and  the  Western  Arroyos;  and  (3) 


the  Colorado  Desert  Fauna  (III).  The  endemic 
Colorado  Desert  Fauna  is  apparently  too  small, 
relative  to  the  total  number  of  xerophilic  species 
present  within  the  South  Colorado  Desert  for- 
mation, to  have  been  detected  by  the  FRF  anal- 
ysis. The  considerable  overlap  in  ecogeographic 


FIGURE  15.  Relationships  between  herpetofaunas  of  the  ecogeographic  formations  (Fig.  14)  based  on  (a)  Coefficients  of 
Difference  (CD)  (Savage  1 960),  (b)  Coefficients  of  Similarity  (SC)  (Simpson  1 960),  and  (c)  Community  Coefficients  (CC)  (Jaccard 
1902,  in  Morafka  1977).  Width  of  bars  is  directly  proportional  to  degree  of  relationship,  length  of  solid  bars  is  inversely 
proportional  to  degree  of  relationship.  Formation  pairs  with  CD  >  0.50  not  included  (except  the  Upper  Main  Scarp  for  which 
the  smallest  value  is  included).  Formation  pairs  with  SC  <  0.60  not  included.  Formation  pairs  with  CC  <  0.37  not  included. 
See  Table  4  for  all  pairwise  values. 


44 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


TABLE  5.  ECOGEOGRAPHIC  DISTRIBUTION  OF  THE  HERPETOFAUNA  ON  A  SCHEMATIC  TRANSECT  OF  THE  SIERRA  SAN  PEDRO 
MARTIR  REGION,  BAJA  CALIFORNIA  NORTE,  MEXICO,  IN  THE  VICINITY  OF  LATITUDE  3 1°N.  Faunal  groupings  are  based  on  coincident 
regional  (Fig.  16)  and  overall  areal  (Fig.  11-21)  distributions.  Solid  lines  indicate  occurrence  throughout  a  formation;  X  without 
lines  indicates  occurrence  restricted  within  a  formation.  (X)  =  occurrence  only  within  Western  Arroyos  where  they  transect  the 

Western  Foothills  (not  considered  a  part  of  the  foothills  fauna).  ? ?  =  expected  occurrence.  *  =  resident  of  Pacific  Northwest 

Track  (see  text  and  Fig.  17).  **  =  resident  of  Madrean  Track  (see  text  and  Fig.  18). 


SPECIES 


ECOGEOGRAPHIC  FORMATION 


N       S       W       W       W 
C      Q      A      F       S 
Regional  Herpetof  aunas           D      P 
(CDS)   (CDS)    (RW)    (CSS)   (Chp) 

L      U 
M      M 
S       S 
(Con)   (Con) 

E      C 
S       D 
(P-J. 
CBS)   (CBS) 

I.   High  Sierran  Scarps  Fauna** 
Rana  boylii 

X 
X 

X 
X 

X 

X 

X 
X 
X 
X 
X 

X   ) 

) 

X 

{  
[  X 

[ 

X 

X 

: 

C 
X  X 

X 

X 

7 
X 

hx 

Batrachoseps  pacificus  ssp. 

Sceloporus  graciosus 

Lampropeltis  zonata 

X 

Thamnophis  elegans 

) 

;  x 

X 

X 
X 

II.   Calif  ornian  Fauna** 
Bufo  boreas* 

Bufo  microscaphus 

X 

X 

X  X 

Hyla  cadaverina 

X 
X 
X 

(X) 
(X) 
(X) 
X 

X 
X 
X 

X 
X 
X 

X    J 
X   2 
X    } 

C  X 

C   X 
C  X 
C   X 

X 

Hyla  regilla 
Rana  aurora* 

Scaphiopus  hanunondii 

Batrachoseps  p.  major 

X 

1C 
X 

X 

X 
X 
X 

X 

X 
X 
X 

(X) 
(X) 
(X) 
(X) 

Clemmys  marmorata* 

1C    } 

Anniella  pulchra 

Eumeces  gilberti 

Eumeces  skiltonianus 

Elgaria  multicarinata 

X 

X  

Sceloporus  occidentalis 

X--- 
X 

X 

.-?__. 

X 
X 

X 

(X) 

X 
X 

X 

X   ) 

Xantusia  vigilis 

X 

) 

Crotalus  viridis 

X 
X 

X 
X 
X 
X 

X 

Diadophis  punctatus 
Coluber  lateralis 

Xj 

? 

X 

X 
X 
X 

(X) 
(X) 
(X) 

e< 

X 
X 
X 

0 

.  .•> 

•>_  _. 

X---' 

> 
i  X 

Thamnophis  hammondii 

X 

X    ) 

:ones-  • 

?--x 

III.   Colorado  Desert  Fauna 
Phrynosoma  platyrhinos 

Urosaurus  graciosus 

X 

Chionactis  occipitalis 

X 

Crotalus  cerastes 

X 

NCD  -  North  central  Desert;  SQP  -  San  Quintin  Plain;  WA  -  Western  Arroyos;  WF  =  Western 
Foothills;  WS  =  West  Scarp;  LMS  =  Lower  Main  Scarp;  UMS  =  Upper  Main  Scarp;  ES  =  East  Scarp; 
CD  »  Colorado  Desert;  see  figure  4. 
(  )  »  Dominant  vegetation;  see  table  2. 


distribution  among  the  xerophilic  species  of  the 
Region  made  it  difficult  to  distinguish  additional 
distinct  faunal  assemblages  using  the  schematic 
ecogeographic  transect  format  alone. 


However,  comparison  of  overall  regional  dis- 
tributions indicated  a  total  of  seven  patterns 
among  the  regional  herpetofauna  (Fig.  1 6).  Pat- 
terns I,  II,  and  III  (Fig.  16)  correspond  with  the 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


45 


Table  5.  Continued. 


SPECIES 


ECOGEOGRAPHIC  FORMATION 


N      S 
C      Q 
Regional  Herpetofaunas         D      P 
(COS)     (CDS) 

W      W      W      L      U      E      C 
A      F      S      H      M      S      D 
S       S       (P-J. 
(RW)    (CSS)   (Chp)   (Con)   (Con)    CBS)   (CBS) 

IV.   Peninsular  Range  Fauna 
A.   Pacific  Slope  Forms: 
Anniella  geronimensis** 

X 

K  X 

X 

X  ' 

C 

i 
f 

1C 
! 

C 

? 

> 

X 

; 
* 

?-- 

X 

X 

7 
X 

Cnemidophorus  hyperythrus 

Cnemidophorus  labialis** 

X 

X 

: 

: 

[ 

X 

Phrynosoma  coronatum 

X 

} 

Crotalus  enyo 

X 

X 

X 

7 
1 

3 

Sonora  s.  bancroftae 

X 
X 
X 

X 

?   > 

3 

x   3 

B  .   Rocky  Scarps  Forms  : 
Crotaphytus  insularis 

Petrosaurus  mearnsi 

Phyllodactylus  nocticolus 

X 

X 

Sauromalus  obesus 

X 

X 

Sceloporus  orcutti 

X 

X 

Urosaurus  microscutatus 

7 

X 

X 

X 

X 

X 

X  
X  

»  ---X 

X 
X 

X 

Xantusia  henshawi** 

E 

i 
C 

<L 
i 

C 
C 
t 
i 
C 

X 
X 

X 
X 

X 

X 

X 
X 
X 

? 

X 

X 

3 

Crotalus  mitchellii 

X 

X-  — 

C.   Flat  land  Desert  Forms: 
Callisaurus  draconoides 

X 
X 

X 

X 

?-- 

( 

X 
X 
X 
X 

X 

X 
X 

Dipsosaurus  dorsalis 

Chilomeniscus  cinctus 

•> 

Crotalus  ruber 

X 

3 

_x—  ? 

-X  X 

Lichanura  trivirgata 

?  X 

3 

X—  - 

: 

Phyllorhynchus  decurtatus 

X 
X 

3 

X 

V.   North  American  Ubiquitous 
Scrubland  Fauna 
Bufo  punctatus 

X 
-X--? 

Schaphiopus  couchii 

Cnemidophorus  tigris 

X 

Coleonyx  variegatus 

X    J 

C    31 

X 

X 

?—  X 

X 

X 

X 

Gambelia  wislizenii 

X 

; 

X 

Sceloporus  magister 

X 

; 

X 

Uta  stansburiana 

X 

3 

: 

'-X 

X 

Arizona  elegans 

X 

3 

---X-- 

7 
X 

>  

x, 

X 

Hypsiglena  torquata 

X 

3 

X 

Lampropeltis  getulus 

7  
X 

X 

X--- 

'--X 

X 

7 
X 
X 

Leptotyphlops  humilis 

--X  

3 

.  —  ?  —  . 

[ 

X 

—  X— 

7 
X 

'--X 

? 

Coluber  flagellum 

Pituophis  melanoluecus 

X 

>  1 

X 
7 

X 

7 

-X—  ? 
--X-X 
X 

Rhinocheilus  lecontei 

X 

3 

: 
l 

ec 

X—  -' 

> 

Salvadora  hexalepis 

X 

] 

X 

7 

0 

Sonora  s.  semiannulata 

7  

X- 

7 

-X 

•-X  " 

.  7 
__  -•> 

Trimorphodon  biscutatus 

> 

—  —  - 

—  X-? 

:ones- 

_ 

46 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


FIGURE  16.  Distribution  patterns  of  the  herpetofauna  of 
the  Sierra  San  Pedro  Martir  Region,  Baja  California  Norte, 
Mexico.  Numerals  refer  to  Table  5.  In  some  cases  member- 
species  occur  within,  but  not  across  the  entire  pattern  range 
depicted:  see  Table  5  for  exact  regional  distributions  of  indi- 
vidual species.  Broken  lines  indicate  fewer  species  of  grouping 
indicated  occur  northward. 

three  derived  above  and  numerically.  The  re- 
maining species  represent  two  relatively  distinct 
groupings  that  occur  throughout  the  xeric  scrub- 
land formations  on  both  the  Pacific  and  Gulf 
slopes  (Fig.  16,  pattern  V),  or  only  within  some 
of  the  scrubland  formations  on  one  or  the  other, 
or  both,  slopes  (Fig.  16,  patterns  IVa,  IVb,  IVc). 
Several  species  subsequently  assigned  to  pattern 
IVc  (Fig.  16)  (Callisaurus  draconoides,  Chilo- 
meniscus  cinctus,  Crotalus  ruber,  and  Lichanura 
trivirgatd)  and  pattern  IVb  (Fig.  16)  (Crotalus 
mitchellii)  had  regional  distributions  that  were 
indistinguishable  from  members  of  pattern  V  (Fig. 
1 6)  at  this  stage  of  analysis.  Final  assignment  of 
these  species,  and  those  assigned  to  pattern  V, 
was  not  possible  until  overall  distribution  pat- 
terns were  compared  (see  Analysis  of  Tracks  be- 
low). 

The  remaining  three  species  assemblages  (Fig. 
16,  patterns  IVa,  IVb,  and  IVc)  are  centered  pri- 
marily along  the  East  or  West  scarps  or  both,  and 


the  Western  Foothills,  corresponding  with  the 
third  faunal  area  derived  numerically  above  (Fig. 
15).  This  third  FRF-derived  faunal  area  is  the 
least  clearly  defined  probably  due  to  the  complex 
overlapping  ecogeographic  variations  apparent 
among  the  several  groups  of  species  within  this 
faunal  area  and  their  overlap  with  many  of  the 
species  of  pattern  V.  Pattern  V  (Table  5)  corre- 
sponds closely  with  the  fourth  faunal  area  de- 
rived by  the  FRF  analysis  and  covering  the  four 
scrubland  desert  formations  (Fig.  15).  Most 
species  of  pattern  IVa  (Fig.  1 6;  Table  5)  occur 
only  on  the  Pacific  Slope  of  northern  or  central 
Baja  California.  Sonora  semiannulata  bancroftae 
and  Anniella  geronimensis  are  endemic  to  the 
Martir  Region;  Crotalus  enyo  and  Cnemidoph- 
orus  labialis  have  their  northern  limits  slightly 
north  of  the  Martir  Region.  The  remaining  two 
members  of  pattern  IVa  (Phrynosoma  corona- 
turn  and  Cnemidophorus  hyperythrus)  range  north 
across  the  U.S.  border  on  the  Pacific  Slope.  All 
of  these  species  avoid  rocky  substrates,  preferring 
sandy  soils.  Members  of  pattern  IVc  also  show 
preference  for  sandy,  fine-textured  substrates,  but 
demonstrate  wider  regional  and  overall  distri- 
butions than  members  of  pattern  IVa,  occurring 
throughout  the  xeric  and  subxeric  scrub  habitats 
on  both  the  Pacific  and  Gulf  slopes  and  ranging 
extensively  beyond  the  Martir  Region  (Fig.  16; 
Table  5).  Pattern  IVb  species  (Fig.  16;  Table  5) 
are  saxicolous  species  associated  with  the  Pen- 
insular Range  scarps  and  bordering  rocky  scrub- 
land habitats.  Three  of  these  species  (Petrosaurus 
mearnsi,  Phyllodactylus  nocticolus,  and  Sauro- 
malus  obesus)  occur  only  along  the  Gulf  Slope 
north  of  30°N  latitude.  The  remaining  pattern 
IVb  members  occur  on  both  slopes  of  the  Pen- 
insular Range  regionally  and  northward,  except 
Crotaphytus  insularis,  which  is  absent  from  the 
Pacific  Slope  north  of  32°N  latitude.  Xantusia 
henshawi  is  absent  south  of  30°N  latitude. 

Analysis  of  Tracks 

Lastly,  I  conducted  an  analysis  of  the  entire 
distribution  of  each  taxon  in  Baja  California  us- 
ing the  Species  Accounts  and  the  following  lit- 
erature: Van  Denburgh  (1922),  Schmidt  (1922), 
Savage  (1960,  1967),  Soule  and  Sloan  (1966), 
Stebbins  (1966,  1985),  Bostic  (1971),  Loomis  et 
al.  (1974),  Conant  (1975),  Murphy  (1983a),  and 
Murphy  and  Ottley  (1983,  1984).  This  last  stage 
of  the  analysis  sought  to  discern  the  generalized 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


47 


tracks  of  the  Baja  California  herpetofauna  by 
comparing  the  geographic  distributions  of  con- 
sanguineous forms  (monophyletic  lineages), 
seeking  spatial  congruency  among  unrelated  taxa, 
the  panbiogeographic  method  of  Croizat  ( 1 964; 
see  also  Craw  [1983]  and  Craw  and  Weston 
[1984]).  Distributions  were  examined  and  com- 
pared for  each  lineage  at  the  subspecific,  specific, 
and  generic  levels.  In  each  case  of  a  proposed 
generalized  track,  the  number  of  individual  tracks 
indicative  of  the  described  generalized  track  ex- 
ceeded the  minimum  three  taxon/area  state- 
ments (three  individual  tracks)  considered  as  es- 
sential to  support  a  hypothesis  of  a  generalized 
track  or  pattern  (Craw  1983). 

I  found  evidence  of  at  least  five  generalized 
tracks  among  the  herpetofauna  of  the  Sierra  San 
Pedro  Martir  Region:  a  Northwest  Pacific  Track, 
a  Madrean  Track,  a  Colorado  Desert  Track,  a 
Peninsular  Range  Track,  and  a  North  American 
Desert  and  Plains  Track. 

The  Pacific  Northwest  Track  (Fig.  1 7) 

The  20-plus  species  on  this  track  are  found 
along  the  western  Pacific  rim  from  Baja  Califor- 
nia north  through  California,  Oregon,  Washing- 
ton, and  into  Canada,  with  a  few  forms  occurring 
north  to  Alaska.  Many  species  of  this  track  have 
allopatric  congeners  in  the  northeastern  United 
States.  This  track  is  characterized  by  endemic 
salamanders,  most  of  which  occur  north  of  35°N 
latitude  in  California;  only  the  genera  Ensatina 
andAneides  occur  in  Baja  California.  In  southern 
California  and  Baja  California,  members  of  the 
Pacific  Northwest  Track  (Table  6)  occur  as  dis- 
junct populations  and  are  often  isolated  on 
mountain  ranges.  Clemmys  marmorata  is  the 
only  reptile  of  the  Martir  Region  that  I  believe 
is  a  member  of  this  track.  The  anurans  Bufo 
boreas  and  Rana  aurora  are  included  on  this 
track,  but  their  taxonomic  and  distributional 
patterns  do  not  match  the  pattern  as  well  as  the 
other  amphibians  indicated.  This  may  be  due  to 
their  euryecious  nature,  and  they  may  in  fact 
have  a  common  zoogeographic  history  with  oth- 
er track  members.  They  may  also  represent 
another,  more  recent  overlapping  track,  as  they 
appear  to  be  relatively  young  forms  in  North 
America  that  may  have  relatively  recently  dif- 
fused (Pielou  1979)  south  across  cool  temperate 
regions  (Savage  1960).  The  Martir  faunal  ele- 
ments of  the  Pacific  Northwest  Track  are  mem- 


FIGURE  17.    Generalized  tracks  of  the  herpetofauna  of  the 
Baja  California  Region:  Pacific  Northwest  Track. 


bers  of  either  the  High  Sierran  or  Californian 
faunas  (Table  5;  Fig.  16, 1  and  II)  and  have  their 
southern  terminus  in  the  Martir  Region. 

The  Madrean  Track  (Fig.  1 8) 

The  Madrean  Track  lineages  exhibit  a  complex 
pattern  of  distribution  across  southwestern  North 
America  often  involving  two  or  more  disjunct 
forms,  for  each  lineage,  within  Baja  California 
and  additional  disjunct,  closely  related  forms,  to 
the  north  and  east  in  California  and  Arizona, 
with  many  lineages  ranging  into  southwestern 
New  Mexico  and  south  into  the  Sierra  Occidental 
of  Mexico  (not  all  member  lineages  presently  oc- 
cur across  the  full  range  of  the  track).  These  forms 
avoid  lowland  xeric  habitats  for  the  most  part, 
or  if  they  occur  in  such  areas,  they  have  adapted 
to  microhabitats,  and  temporal  activity  pat- 
terns that  ameliorate  against  the  extremes  of  low 
moisture  and  high  temperature  (e.g.,  Scaphiopus, 
Xantusia  spp.). 

Within  Baja  California  many  of  these  forms 
occur  in  disjunct  populations,  with  some  occur- 


48 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


TABLE  6.    RESIDENT  TAXA  OF  THE  GENERALIZED  TRACKS  OF  THE  HERPETOFAUNA  OF  THE  BAJA  CALIFORNIA  REGION.  Includes 
only  those  track  members  that  occur  within  the  Baja  California  Region.  ( )  refer  to  Figures  17-21. 


Aneides  lugubris 
Ensatina  eschscholtzii 
Bufo  boreas 

Batrachoseps  pacificus  major 

B.  pacificus  spp. 

Scaphiopus  hammondii 

Bufo  microscaphus 

Hyla  cadaver ina 

H.  regilla 

Rana  boylii 

Sceloporus  graciosus 

S.  occidentalis 

Xantusia  vigilis  (2  spp.) 

X.  henshawi 

Phrynosoma  mcallii 
P.  p/atyrhinos 
Uma  notata 


Chrysemys  scripta 

Bipes  biporus 

Coleonyx  switaki 

Phyllodactylus  angelensis 

P.  apricus 

P.  bugastrolepis 

P.  homolepidurus  nolascoensis 

P.  nocticolus  (9  ssp). 

P.  partidus 

P.  santacruzensis 

P.  tinklei 

P.  unctus  (3  ssp.) 

P.  xanti 

Ca/lisaurus  draconoides  (6  ssp.) 

Crotaphytus  insularis  (2  ssp.) 

Ctenosaura  hemilopha 

Petrosaurus  complex 

P.  mearnsi 

P.  repens 

P.  thalassinus 
Phrynosoma  coronatum  complex 

P.  cerroense 

P.  coronatum  (3  ssp.) 


Bufo  punctatus 
Scaphiopus  couchii 
Coleonyx  variegatus  (6  ssp.) 
Uta  complex 

U.  antigua 

U.  nolascensis 

U.  pa/meri 

U.  squamata 

U.  stellata 

U.  stansburiana  (2  ssp.) 


Pacific  Northwest  Track  (17) 
Rana  aurora 
Clemmvs  marmorata 


Madrean  Track  (18) 
Eumeces  gilberti 
E.  skiltonianus  complex 

E.  skiltonianus 

E.  lagunensis 
Cnemidophorus  labialis 
Elgaria  complex 

E.  cedrosensis 

E.  multicarinata 

E.  paucicarinata 
Anniella  pulchra 
A.  geronimensis 


Diadophis  punctatus 
Coluber  lateralis  complex 

C.  aurigulus 

C.  barbouri 

C.  lateralis 
Lampropeltis  zonata 
Thamnophis  hammondii  complex 

T.  hammondii 

T.  digueti 
T.  elegans 
Crotalus  viridis 


Colorado  Desert  Track  (19) 

Urosaurus  graciosus 
Chionactis  occipitalis 
Crotalus  cerastes 


Peninsular  Range  Track  (20) 
Sator  complex 

S.  an  gust  us 

S.  grandaevus 
Sauromalus  complex 

5.  ater  (2  ssp.) 

S.  australis 

S.  hispidus 

S.  klauberi 

S.  obesus 

S.  slevini 

S.  varius 
Sceloporus  magister  complex 

5.  m.  lineatulus 

S.  m.  monserratensis 

S.  m.  rufidorsum 

S.  m.  zosteromus 
Sceloporus  orcutti  complex 

S.  hunsakeri 

S.  licki 

S.  orcutti 
Urosaurus  complex 

U.  lahtelai 

U.  microscutatus 

U.  nigricaudus 


Cnemidophorus  deppei  complex 

C.  ceralbensis 

C.  hyperythrus  (3  ssp.) 
Lichanura  trivirgata  (3  spp.) 
Chilomeniscus  complex 

C.  cinctus 

C.  punctissimus 

C.  savagei 

C.  stramineus  (2  ssp.) 
Etaphe  rosaliae 
Eridiphas  slevini  (2  ssp.) 
Nerodia  valida 

Phyllorhynchus  decurtatus  (3  ssp.) 
Sonora  semiannulata  bancroftae 
S.  s.  mosaueri 
Tantilla  planiceps 
Crotalus  atrox  complex 

C.  catalinensis 

C.  exsul 

C.  ruber  (2  ssp.) 

C.  tortugensis 
C.  enyo  (3  ssp.) 
C.  mitchellii  (4  spp.) 


North  American  Desert  and  Plains  Track  (21): 
Group  A:  old  vicariated  lineages 

Leptotyphlops  humilis  (5  ssp.) 
Hypsiglena  torquata  (7  ssp.) 
Lampropeltis  getulus  complex 
L.  catalinensis 
L.  getulus  (3  ssp.) 
L.  nitida 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


49 


TABLE  6.    CONTINUED. 


Gainbelia  wizlizenii 
Sceloporus  magister  uniformis 
Arizona  elegans  (3  ssp.) 
Coluber  flagellum  (2  ssp.) 
Pituophis  melanoleucus  (3  ssp.) 


North  American  Desert  and  Plains  Track  (21): 
Group  B:  post-vicariant  lineages 

Rhinocheilus  leconti 
Salvadora  hexalepis  (3  ssp.) 
Sonora  s.  semiannulata 
Trimorphodon  biscutatus  (2  ssp.) 
Crotalus  atrox 


ring  as  allopatric  species,  occupying  patches  of 
relatively  mesic  habitats  scattered  sporadically 
down  the  length  of  the  peninsula,  with  several 
lineages  present  on  one  or  two  Gulf  or  Pacific 
islands  (e.g.,  Hyla  regilla,  Eumeces  skiltonianus- 
E.  lagunensis,  Elgaria  multicarinata-E.  pauci- 
carinata,  E.  cedrosensis,  Coluber  lateralis-C. 
barbouri  and  C.  aurigulus,  Thamnophis  ham- 
mondii-T.  digueti,  Xantusia  vigilis,  and  Crotalus 
viridis).  The  remaining  lineages  of  this  track  (Ta- 
ble 6)  do  not,  for  the  most  part,  occur  south  of 
30°N  latitude  in  Baja  California.  North  of  30°N 
latitude  into  California  and  northward,  and  east 
across  the  Mojave  Desert  through  Arizona,  New 
Mexico,  and  south  into  Mexico  in  the  Sierra 
Madre  Occidental,  these  taxa  demonstrate  a  sim- 
ilar pattern  of  allopatric  distributions  with  many 
of  these  lineages  represented  by  two  or  three 
species  across  the  scope  of  the  track  (e.g.,  Bufo 
microscaphus;  Hyla  cadaverina,  H.  regilla-H. 
exima;  H.  arenicolor,  Elgaria  multicarinata-E. 
panamintina-E.  kingi;  Xantusia  v.  vigilis-X.  v. 
arizonae,  X.  h.  bolsonae,  etc.;  Lampropeltis  zo- 
nata-L.  pyromeland). 

The  remaining  lineages  included  on  this  track 
(Table  6)  are  somewhat  less  clearcut  in  their  pat- 
terns of  disjunction  and  taxonomic/geographic 
correlation  with  the  above  listed  taxa,  but  suf- 
ficient aspects  of  the  pattern  are  apparent  in  their 
distributions  to  support  their  inclusion  as  ele- 
ments of  the  Madrean  Track. 

Within  the  Martir  Region,  elements  of  the 
Madrean  Track  demonstrate  ecogeographic  af- 
finities coincident  with  the  mesic  habitats  of  the 
Gulf  and  Pacific  slopes  (Table  5,  I,  II,  and  IV). 

Colorado  Desert  Track  (Fig.  1 9) 

These  species  have  distributions  coincident 
with  the  hottest  and  driest  portions  of  the  south- 
western deserts,  the  area  where  the  Mojave,  Col- 
orado, and  Sonoran  deserts  merge.  Martir  faunal 


elements  present  on  this  track  occur  regionally 
only  in  the  Colorado  Desert  formation  (Table  5, 
III;  Fig.  16,  III).  Species  of  the  Colorado  Desert 
Track  demonstrate  the  most  limited  distribu- 
tions and  the  lowest  levels  of  geographic  and 
taxonomic  diversity  within  lineages  of  all  the 
faunal  assemblages  described  here  for  Baja  Ca- 
lifornia. Six  species  are  included;  the  Colorado 
Desert  fauna  of  the  Martir  Region  (Table  5,  III), 
plus  Uma  notata  and  Phrynosoma  mcallii. 


Peninsular  Range  Track  (Fig.  20) 

Lineages  of  this  track  have  their  overall  dis- 
tributions coincident  with  all  or  much  of  pen- 
insular Baja  California  and  the  Gulf  of  California 
islands,  extending  north  along  the  Peninsular 
Range  to  San  Gorgonio  Pass  in  southern  Cali- 
fornia. A  few  taxa  of  this  track  have  radiated 
northeast  into  the  Mojave  Desert  (e.g.,  Crotalus 
mitchellii,  Lichanura  trivirgata,  and  Dipsosaurus 
dorsalis)  and  southeast  into  the  Sonoran  Desert 
(e.g.,  Lichanura  trivirgata,  Chilomeniscus  cinc- 
tus,  Phyllorhynchus  decurtatus,  Callisaurus  dra- 
conoides,  Dipsosaurus  dorsalis).  Several  interre- 
lated characteristics  of  the  lineages  assigned  to 
this  track  are:  (1)  pronounced  endemism;  (2) 
presence  of  most  primitive  extant  forms;  and  (3) 
greatest  levels  of  geographic  and  taxonomic  di- 
versity within  lineages,  all  within  insular  and 
peninsular  Baja  California.  Twenty-six  species  or 
species-complexes  meet  these  criteria  (Table  6). 
Twenty  species-groups  of  the  Peninsular  Range 
Track  occur  in  the  Martir  Region  (Table  5,  IV), 
and  exhibit  three  variations  in  regional  distri- 
bution (Fig.  16,  IVa,  IVb,  and  IVc).  Six  taxa 
occur  further  south  in  peninsular  and  insular  Baja 
California  in  the  Central  and  Vizcaino  deserts 
and  the  Cape  Region  (Chrysemys  scripta,  Bipes 
biporus,  Ctenosaura  hemilopha,  Sator  spp., 
Eridiphas  slevini,  and  Nerodia  valida). 


50 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


FIGURE  18.    Generalized  tracks  of  the  herpetofauna  of  the 
Baja  California  Region:  Madrean  Track. 


FIGURE  19.    Generalized  tracks  of  the  herpetofauna  of  the 
Baja  California  Region:  Colorado  Desert  Track. 


North    American    Desert    and    Plains    Track 
(Fig.  21) 

Species  of  this  track  possess  broad  distribu- 
tions throughout  the  North  American  Desert  Re- 
gion, and  into  the  Great  Plains  Region.  These 
species  are  widespread  in  Baja  California,  oc- 
curring throughout  most  insular  and  peninsular 
xeric  habitats.  The  1 8  taxa  of  the  North  Amer- 
ican Desert  and  Plains  Track  present  in  the  Mar- 
tir  Region  demonstrate  broad  ecogeographic  dis- 
tributions throughout  the  scrub  and  woodland 
habitats  of  the  Region  (Table  5;  Fig.  16,  V). 

The  lineages  of  this  track  can  be  divided  into 
two  subsets  on  the  bases  of  distributional  vari- 
ations on  Gulf  of  California  islands,  and  the 
level  of  divergence  they  show  within  the  Baja 
California  Region  as  a  whole:  (1)  those  lineages 
that  show  extensive  divergence  within  the  Baja 
California  Region,  including  endemic  taxa  on 
some  or  all  of  the  "old  islands"  (Soule  and  Sloan 
1966;  see  also  Murphy  [1983&]  and  Murphy  and 
Ottley  [1984])  of  the  Gulf  of  California  (Table  6, 


Desert  and  Plains  Track— group  A);  and  (2)  those 
lineages  that  show  little  divergence  throughout 
peninsular  Baja  California  and  are  absent  from 
the  "old  islands"  (Soule  and  Sloan  1966)  (Table 
6,  Desert  and  Plains  Track— group  B).  The  taxa 
of  group  A  appear  to  have  a  shared  evolutionary 
history  within  the  Baja  California  Region  with 
the  lineages  of  the  Peninsular  Range  Track  (Ta- 
ble 6).  The  taxa  of  group  B  appear  to  be  relatively 
recent  (post-Pliocene)  emigrants  in  the  Baja  Ca- 
lifornia Region  despite  sharing  contemporary 
overall  continental  distributions  with  members 
of  group  A. 

Geographic  diversity  and  endemism  within  the 
lineages  of  this  track,  when  considered  within 
peninsular  and  insular  Baja  California,  are  not 
noticeably  more  extensive  than  in  other  parts  of 
the  North  American  Region  (in  contrast  with 
member  lineages  of  the  Peninsular  Range  Track). 
On  the  contrary,  with  many  lineages  of  this  track, 
geographic  diversity  and  endemism  are  greater 
beyond,  than  within,  insular  and  peninsular  Baja 
California. 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


51 


FIGURE  20.  Generalized  tracks  of  the  herpetofauna  of  the 
Baja  California  Region:  Peninsular  Range  Track.  The  extent 
of  this  track  in  southwest  Mexico  is  greater  than  indicated, 
particularly  southward. 


FIGURE  2 1 .  Generalized  tracks  of  the  herpetofauna  of  the 
Baja  California  Region:  North  American  Desert  and  Plains 
Track. 


DISCUSSION 

Theories  on  the  historical  biogeography  of  the 
herpetofauna  of  the  Baja  California  Region  have 
advanced  on  three  parallel  fronts:  (1)  with  in- 
creased knowledge  of  contemporary  herpeto- 
faunal  distributions  and  their  ecogeographic  af- 
finities; (2)  with  new  information  about  past 
climates  and  vegetation  assemblages— the  geo- 
floral  literature;  and  (3)  with  the  advent  of  a  new 
paradigm  in  geology— plate  tectonics— depicting 
a  dynamic  rather  than  static  landscape.  Nelson 
(1921)  and  Schmidt  (1922)  compiled  the  first 
syntheses  based  on  early  distributional  data  of 
the  peninsular  biotas.  Savage  ( 1 960)  constructed 
paleoecological  scenarios  using  data  in  the  geo- 
floral  literature,  and  in  conjunction  with  analyses 
of  contemporary  herpetofaunal  distributions, 
presented  a  classic  synthesis.  Soule  and  Sloan 
( 1 966)  and  Savage  ( 1 967)  discussed  evolutionary 
relationships  of  the  island  herpetofaunas  of  the 
Gulf  of  California  and  proximate  Pacific  Ocean 


islands,  respectively.  However,  Savage  and  ear- 
lier investigators  based  their  analyses  on  the  as- 
sumption of  a  static  or  fixed  peninsula,  requiring 
Recent  distributions  to  have  been  derived  pri- 
marily from  north  to  south  dispersals  in  the  case 
of  species  present  on  the  peninsula,  or  tempo- 
rary land  bridges  or  chance  dispersal  across  water 
barriers  in  the  case  of  species  present  on  islands. 

More  recently,  in  conjunction  with  developing 
tectonic  theory,  Leviton  and  Tanner  (1960), 
Robinson  (1973),  and  Smith  and  Tanner  (1974) 
examined  biogeographic  relationships  within 
single  taxa,  and  indicated  that  particular  species 
of  the  southern  Baja  California  herpetofauna  ex- 
hibit distributions  more  consistent  with  a  theory 
of  vicariant  origins  resulting  from  past  tectonic 
events  than  with  a  theory  of  north  to  south  dis- 
persals down  a  fixed  peninsula.  Murphy  (1975), 
following  this  lead,  proposed  a  "trans-gulfian" 
vicariance  model,  based  on  the  dislodging  of  pre- 
peninsular  land  masses  from  further  south  along 


52 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


coastal  Mexico  during  the  Miocene,  to  explain 
contemporary  distributions  of  many  of  the  sub- 
tropical associated  elements  of  the  Cape  Region 
and  associated  deep-water  islands  (Bipes  biporus, 
Nerodia  valida,  Ctenosaura  hemilopha,  Pseud- 
emys  scripta,  Eridiphas  sleveni,  Sator  angustus, 
Crotalus  catalinensis,  and  three  subspecies  of 
Leptotyphlops  humilis).  Welsh  (197 '6a,  b,  1981) 
indicated  that  distributional  data  for  most  of  the 
Sonoran  Desert  herpetofauna  of  Baja  California 
showed  patterns  of  distribution  and  differentia- 
tion on  peninsular  and  insular  land  masses  in- 
dicative of  long-term  isolation  and  evolution 
consistent  with  a  tectonic  vicariant  origin.  Ele- 
ments of  the  herpetofauna  whose  distributions 
are  consistent  with  this  scenario  include  some  or 
all  members  of  the  following  genera  within  the 
Baja  California  Region:  Callisaurus,  Crotaphy- 
tus,  Dipsosaurus,  Petrosaurus,  Phrynosoma, 
Sauromalus,  Sceloporus,  Ufa,  Urosaurus,  Cole- 
onyx,  Phyllodactylus,  Cnemidophorus,  Lichan- 
ura,  Leptotyphlops,  Chilomeniscus,  Hypsiglena, 
Lampropeltis,  Phyllorhynchus,  Sonora,  Tantilla, 
Trimorphodon,  and  Crotalus.  Murphy  (1983a) 
developed  a  comprehensive  evolutionary  scenar- 
io based  on  geomorphological  and  paleobotan- 
ical  evidence  and  tectonic  theory  to  further  elu- 
cidate the  possible  vicariant  events  responsible 
for  many  contemporary  herpetofaunal  distribu- 
tion patterns.  Kim  et  al.  (1976),  Wyles  and  Gor- 
man (1978),  Murphy  and  Papenfuss  (1979),  and 
Murphy  (1983a)  presented  electrophoretic  data 
that,  assuming  the  validity  of  the  "biochemical 
evolutionary  clock"  (Sarich  1977),  support  the 
thesis  of  a  vicariant  origin  from  Miocene  tectonic 
events  to  account  for  distributional  patterns 
among  major  components  of  the  contemporary 
herpetofauna  of  Baja  California. 

Paleogeographic  and  Paleobotanic  History  of 
the  Baja  California  Region 

The  following  scenario  is  a  shortened  version 
that  follows  closely  the  synthesis  developed  by 
Murphy  (19830)  but  with  some  modifications  as 
indicated  below.  Berggren  and  Van  Couvering 
(1974)  and  Van  Couvering  (1978)  present  a  geo- 
logic time  scale  that  differs  from  the  generally 
accepted  time  scale  by  placing  the  Miocene-Plio- 
cene boundary  at  5.2  MYBP  (million  years  be- 
fore present),  rather  than  at  1 1-13  MYBP,  and 
the  Pliocene-Pleistocene  boundary  at  1.6-2.0 
MYBP  rather  than  at  1  MYBP.  The  Berggren 


and  Van  Couvering  time  scale  will  be  followed 
in  subsequent  discussion. 

Miocene  and  Earlier 
Paleogeography 

The  "two-fault"  hypothesis  of  Atwater  (1970), 
which  indicates  that  two  major  migrating  land 
masses  have  been  active  along  Pacific  coastal 
Mexico  since  at  least  the  Miocene,  is  central  to 
the  hypothesis  of  a  tectonic  origin  for  many  pres- 
ent-day herpetofaunal  distributions  of  Baja 
California.  Gastil  et  al.  ( 1 972)  supported  the  "two- 
fault"  theory  and  indicated  further  that  "Pen- 
insular California"  was  probably  separated  from 
coastal  Mexico  as  several  large  fragments.  These 
two  land  masses  are:  (1)  "Alta  California,"  de- 
fined here  as  consisting  of  present-day  California 
from  about  Santa  Barbara  to  San  Francisco  and 
west  of  the  San  Andreas  fault  (modified  from 
Gastil  et  al.  1972),  and  (2)  "Peninsular  Califor- 
nia," those  areas  south  of  the  Transverse  Ranges 
of  southern  California  and  west  of  the  Gulf  of 
California  (Gastil  et  al.  1972).  "Alta  California" 
probably  began  its  traverse  20-25  MYBP  when 
it  was  torn  from  "Peninsular  California"  (then 
south  of  its  present  location,  and  part  of  coastal 
Mexico)  from  about  Bahia  San  Sebastian  Viz- 
caino northward.  The  Pacific  islands  of  Baja  Ca- 
lifornia north  of  Vizcaino  Bay  and  the  Channel 
Islands  of  California  are  probably  fragments  that 
resulted  from  the  traverse  of  "Alta  California." 

The  dating  for  the  separation  of  "Peninsular 
California"  from  coastal  Mexico  varies  consid- 
erably from  early  Tertiary  (60  MYBP;  Gastil  et 
al.  1972)  to  early  Pliocene  (5  MYBP;  Moore 
1 973).  Geological  evidence  (Atwater  and  Molnar 
1973;  Gastil  and  Jensky  1973)  suggests  that  the 
southern  end  of  Baja  California  from  about  La 
Paz  south— the  Cape  fragment— originated  from 
coastal  Mexico  near  the  Jalisco-Nayarit  border 
in  the  mid-Miocene  (12-14  MYBP).  This  and 
other  separate  island  and  island  fault  blocks  have 
been  simultaneously  moving  northwestward  since 
mid-Miocene  through  Pliocene  times.  By  the  close 
of  the  Miocene,  "Peninsular  California"  had  mi- 
grated 260  km  northwest  to  the  region  of  the 
Islas  Las  Tres  Marias  (Gastil  and  Jensky  1973). 
Thus  it  appears  that  from  the  mid-Miocene,  be- 
ginning about  14  MYBP,  much  of  "Peninsular 
California"  existed  as  an  island  archipelago.  This 
archipelago  included  much  of  the  present-day 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


53 


Cape  Region  and  associated  deep-water  islands 
(Santa  Catalina,  Santa  Cruz,  San  Diego,  and  Isla 
Cerralvo— hereafter  referred  to  collectively  with 
the  proto-Cape  Region  as  the  Cape  Islands),  as 
well  as  upland  areas  of  the  present-day  Vizcaino 
peninsula  (possibly  including  Isla  Cedros).  Si- 
multaneously, the  northern  Peninsular  Ranges 
were  beginning  to  uplift  and  separate  as  a  result 
of  subduction  (Murphy  1983a).  Murphy  (19830, 
fig.  6)  indicated  that  the  northern  Peninsular 
Ranges  were  not  totally  separated  from  western 
Mexico  at  this  time.  A  proto-Gulf  of  California 
was  beginning  to  form  about  1 4  M  YBP  but  may 
not  have  become  a  permanent  geographic  feature 
until  around  the  Miocene-Pliocene  boundary,  6- 
5  MYBP  (Gastil  and  Jensky  1973.) 

Subduction  along  the  fault  system  continued 
to  elevate  the  northern  Peninsular  Ranges,  and 
the  central  Peninsular  Ranges  began  to  emerge 
as  a  result  of  volcanism  (Durham  and  Allison 
1 960;  Karig  and  Jensky  1 972).  The  Sierra  Madre 
Occidental,  the  Sierra  Madre  Oriental,  and  the 
Mexican  Plateau  were  also  being  elevated  as  a 
result  of  Miocene  tectonic  activities  (King  1959). 
"Thus  the  only  geographically  isolated  portions 
of  the  extant  peninsula  in  the  middle  Miocene 
were  the  islands  near  (and  including)  the  current 
Cape  area,  the  Sierra  Vizcaino,  and  several 
southern  peninsula-associated  island  localities" 
(Murphy  19830).  Murphy  (19830,  fig.  8,  and  text) 
indicated  that,  by  the  late  Miocene,  the  Cape 
Islands  were  connected  temporarily  to  more 
northern  portions  of  the  peninsula,  at  least  until 
about  the  Miocene-Pliocene  boundary,  when  this 
connection  was  apparently  inundated. 


Paleobotany 

Prior  to  the  Eocene,  neotropical  floral  elements 
dominated  western  Mexico;  from  the  Eocene  on- 
ward, however,  declining  temperatures  and  in- 
creasing aridity  progressively  restricted  the  neo- 
tropical-Tertiary geofloral  elements  southward 
(Axelrod  1 979).  Oak-pinyon  woodlands  and  arid 
tropic-scrub  habitats  (Madro-Tertiary  vegeta- 
tion associations)  were  developing  but  were  re- 
stricted to  higher  elevations  of  the  young  devel- 
oping mountain  regions  in  the  interior.  By  the 
close  of  the  Oligocene  the  Madro-Tertiary  geo- 
flora  predominated  along  the  mountain  chains 
and  uplands  of  central  Mexico  (Axelrod  1975). 
The  declining  temperatures  and  increasing  arid- 


ity continued  to  facilitate  the  spread  of  Madro- 
Tertiary  geoflora  in  upland  areas.  Lowland  areas 
of  Mexico  west  of  the  Sierra  Madre  Occidental 
were  still  dominated  by  a  neotropical  flora  (Ax- 
elrod 1975,  1979). 


Pliocene 
Paleogeography 

During  the  Pliocene,  beginning  about  5  MYBP, 
"Peninsular  California"  began  a  more  rapid 
northwestward  movement,  from  the  present  re- 
gion of  the  Islas  Las  Tres  Marias  to  its  present 
location  (Larsen  1972;  Atwater  and  Molner  1973; 
Moore  1973).  The  resurgence  of  plate  interac- 
tions and  the  union  of  "Peninsular  California" 
with  the  Pacific  Plate  (Atwater  and  Molner  1 973) 
resulted  in  further  uplift  of  the  Peninsular  Ranges. 
Volcanism  continued  in  the  southern  Peninsular 
Ranges,  including  the  Sierra  de  la  Giganta  (Mina 
1957).  The  proto-Gulf  of  California  had  com- 
pletely formed  by  5  MYBP  (Karig  and  Jensky 
1972;  Gastil  et  al.  1975)  and  extended  from  the 
region  of  the  San  Gorgonio  Pass  in  southern  Cal- 
ifornia (Allen  1 957)  to  its  opening  into  the  Pacific 
Ocean  between  the  Islas  Las  Tres  Marias  and 
mainland  Mexico  (Moore  1973).  In  the  central 
and  southern  peninsula,  the  Sierra  Vizcaino  and 
possibly  the  Cape  Islands  remained  isolated  from 
the  northern  Peninsular  Ranges.  During  the  Plio- 
cene the  Los  Angeles  Basin  was  submerged  be- 
neath the  Pacific  Ocean.  The  combination  of  the 
flooding  of  the  Los  Angeles  Basin  and  the  for- 
mation of  the  proto-Gulf  of  California  greatly 
reduced  terrestrial  access  to  the  northern  Pen- 
insula (Durham  and  Allison  1960);  it  may  have 
in  fact  been  totally  isolated  (Murphy  19830). 
Murphy  (19830,  fig.  9)  refers  to  this  probable 
restricted  or  eliminated  terrestrial  access  to  the 
northern  Peninsular  Ranges  as  the  San  Gorgonio 
Barrier. 

Pliocene  orogenic  events  brought  the  Sierra 
Madre  Occidental,  the  Sierra  Madre  Oriental, 
and  the  Mexican  Plateau  to  near  their  present 
elevations  (Eardley  1951;  King  1959).  This  re- 
sulted in  at  least  a  partial  barrier  to  gene  exchange 
between  populations  of  xerophilic  herpetofaunas 
in  the  proto-Chihuahuan  and  proto-Sonoran  xe- 
ric  habitats  on  either  side  at  the  northern  end  of 
the  Sierra  Madre  Occidental;  an  area  referred  to 
as  the  Cochise  Filter  Barrier  (Morafka  1977). 


54 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


Paleobotany 

The  newly  formed,  moderately  high  montane 
regions  initiated  the  establishment  of  easterly 
rainshadows.  Axelrod  (1975)  indicated  that  by 
the  late  Pliocene,  climates  in  the  middle  latitudes 
(25-40°N)  were  essentially  equivalent  to  those  of 
the  present.  The  deserts  as  we  know  them  today, 
however,  did  not  develop  until  the  last  intergla- 
cial  period  (Axelrod  1979).  The  only  paleobo- 
tanical  data  from  "Peninsular  California"  are 
from  the  Mount  Eden  Beds  at  the  north  end  of 
the  Peninsular  Range  in  southern  California. 
These  are  characterized  by  a  mixture  of  Arcto- 
Tertiary  and  Madro-Tertiary  geofloral  assem- 
blages. Eight  habitats  were  represented:  (1)  des- 
ert, (2)  arid  subtropical  scrub,  (3)  coastal  sage, 
(4)  grassland,  (5)  chaparral,  (6)  live  oak  and  wal- 
nut woodland,  (7)  digger  pine  woodland,  and  (8) 
big-cone  pine  conifer  forest  (Axelrod  1937,  1950). 
Murphy  ( 1 9830)  assumes  that  the  desert  and  arid 
subtropical  scrub  habitats  were  marginal  and  that 
the  area  was  predominantly  mesic  in  character. 
This  interpretation  is  supported  by  vertebrate 
fossils  from  the  same  area  (Frick  1933).  Terres- 
trial and  aquatic  vertebrate  fossil  remains  from 
the  Pliocene  in  southern  Baja  California  (Miller 
1977,  1980)  indicate  a  warmer  climate  than  the 
present  one  with  year-round  fresh  water,  possibly 
a  subtropical  savanna  habitat. 

Pleistocene 
Paleogeography 

During  this  epoch,  the  most  significant  process 
of  change— from  the  point  of  view  of  contem- 
porary herpetofaunal  distributions— appears  to 
have  been  the  rise  and  fall  in  sea  level  resulting 
from  glacial  fluctuations  that  alternately  isolated 
and  then  connected  many  islands  in  the  Gulf  of 
California  to  larger  peninsular  land  masses  (those 
islands  separated  from  the  peninsula  by  an  ocean 
depth  of  less  than  1 30  m)  (Auffenburg  and  Mil- 
stead  1965).  Tectonic  events  continued  as  well, 
separating  the  Angel  de  la  Guarda  island  block 
(Angel  de  la  Guarda,  Partida  Norte,  San  Lorenzo 
Norte  and  Sur)  from  the  peninsula,  prior  to  1 
MYBP  (Moore  1 973).  Tectonic  activity  also  con- 
tinued the  elevation  of  the  Peninsular  Ranges 
(Gastil  et  al.  1975),  completing  the  Coast  Range 
Corridor  in  California  (Peabody  and  Savage 
1958),  and  forcing  the  waters  of  the  Gulf  of  Cal- 
ifornia to  recede  to  their  present  position.  The 


Sierra  Vizcaino  (Minch  et  al.  1 976)  and  the  Cape 
islands  (Mina  1957)  were  unified  with  the  pen- 
insula. 

Paleobotany 

Geofloral  data  from  Baja  California  and  prox- 
imate areas  of  the  southwest  from  the  Pleistocene 
indicate  a  progressive  drying  and  warming  trend 
with  each  of  the  four  interglacial  periods  (Axel- 
rod  1966,  1975,  1979;  VanDevender  and 
Spaulding  1979).  Axelrod  (1966)  characterized 
the  region  at  the  north  end  of  the  Peninsular 
Ranges  during  the  interglacials  as  progressing 
from  subhumid  to  semiarid,  to  subdesert,  and 
finally  to  desert  environments.  Increased  eleva- 
tion of  the  northern  Peninsular  Ranges  and  other 
mountain  regions  of  the  southwest,  created  east- 
erly rainshadows  that  promoted  increased  aridity 
and  the  subsequent  expansion  of  desert  habitats 
during  the  Pleistocene  (Axelrod  1979).  Evidence 
suggests  that  climates  in  southern  Baja  California 
continued  to  be  relatively  mesic  during  this  ep- 
och compared  with  those  of  northern  peninsular 
areas.  A  fossil  Pleistocene  terrestrial  vertebrate 
fauna  from  southern  Baja  California  (Miller  1 977) 
indicates  both  wooded  and  grassland  habitats  oc- 
curred in  the  area,  indicative  of  denser  vegeta- 
tions than  those  presently  in  existence.  In  gen- 
eral, excepting  the  evolving  desert  plant 
communities,  evidence  indicates  that  mostly  mi- 
nor distributional  perturbations  in  vegetation  as- 
sociations, rather  than  significant  changes  in 
vegetational  components,  occurred  during  the 
Pleistocene  in  southwestern  North  America  (Ax- 
elrod 1966,  1975,  1979).  Such  perturbations  were 
characterized  by  shifts  in  elevation  of  the  vege- 
tation communities  of  Baja  California  concom- 
itant with  glacial  advances  and  retreats  in  more 
northern  areas. 

The  Paleobiogeography  of  the  Herpetofauna: 
Previous  Scenarios 

The  broad  paleobiogeographic  scenarios  of 
Savage  (1960)  and  Murphy  (1983«)  provide  the 
most  current  and  comprehensive  framework  from 
which  to  discuss  the  evolutionary  trends  of  the 
Baja  California  herpetofauna.  Savage  (1960)  and 
Murphy  (1983#)  each  developed  a  set  of  perti- 
nent data  essential  to  an  overall  understanding 
of  the  evolutionary  mechanisms  shaping  her- 
petofaunal relationships  in  Baja  California.  Mur- 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


55 


phy's  paleobiogeographic  scenario,  based  on  cur- 
rent understanding  of  plate  tectonics,  clearly 
outdates  Savage's  hypothesis  as  to  the  primary 
mechanisms  shaping  the  peninsular  herpeto- 
faunas.  However,  Savage  developed,  in  broad 
outline,  a  powerful  model  to  describe  evolving 
North  American  herpetofaunal  assemblages,  the 
progenitors  of  the  contemporary  peninsular  and 
insular  herpetofaunas.  He  relied  on  the  geolog- 
ical history  (King  1958;  Durham  and  Allison 
1 960),  and  more  particularly  the  paleoecological 
history,  in  his  reconstruction  of  Tertiary  and 
Quaternary  environments.  This  included  the 
geofloral  history  as  described  in  the  work  of  Cha- 
ney,  Axelrod,  and  MacGinitie  (see  Murphy 
[1983a]  or  Savage  [1960]  for  complete  citations) 
and  the  concept  of  interdigitating  and  evolving 
geofloras.  Savage's  inherent  assumption  was  that 
herpetofaunas  evolve  in  conjunction  with  broad 
changes  in  phytogeographic  relationships.  He  de- 
nned four  contemporary  herpetofaunal  areas  in 
Baja  California  (Savage  1960)  (Fig.  2a  plus  the 
Cape  Region);  these  areas  closely  approximate 
the  four  contemporary  phytogeographic  prov- 
inces of  the  peninsula  as  denned  by  Shreve  and 
Wiggins  (1964;  see  also  this  paper  Fig.  2b).  Sav- 
age considered  each  of  his  herpetofaunal  areas 
to  be  dominated  by  assemblages  of  species  with 
a  common  evolutionary  history.  He  noted  that 
a  given  Recent  herpetofauna  is  composed  of  ".  .  . 
a  compliment  of  species  originating  from  several 
sources.  A  considerable  portion  of  the  species  in 
each  unit,  however,  originated  in  or  near  the  area 
now  occupied  by  the  faunal  assemblages.  For 
these  reasons,  present-day  herpetofaunas  are 
mixtures  of  species  of  diverse  historical  origins, 
although  a  particular  fauna  is  dominated  by  de- 
rivatives of  one  of  the  historical  elements  de- 
nned" (Savage  1960:189). 

Murphy  (1983a:14)  challenged  the  assump- 
tion that  particular  herpetofaunas  evolved  in 
conjunction  with  and  parallel  to  particular  geo- 
floras. His  argument  is  supported  by  two  ex- 
amples that  illustrate  the  absence  of  complete 
distributional  correlation  between  groups  of  her- 
petofaunal species  and  particular  contemporary 
vegetational  associations  of  Madro-Tertiary 
vegetation  in  Baja  California.  These  arguments 
are  weak  and  based  on  a  few  rigid  and  very  ex- 
acting applications  of  what  was  intended  only  to 
be  a  very  general  correlation.  An  exact  correla- 
tion between  a  particular  vegetational  associa- 


tion (e.g.,  the  Sonoran  Desert  flora)  and  a  par- 
ticular herpetofauna  (e.g.,  the  Colorado  Desert 
herpetofauna)  was  never  assumed,  intended,  or 
implied  in  my  understanding  of  Savage's  model. 
Savage  did  not  rule  out  the  possible  ameliorating 
effects  of  physiography  or  of  interspecific  com- 
petition, for  example,  in  determining  distribu- 
tion. He  used  the  evolving  geofloras  as  evidence 
of  existing  paleoclimatic  conditions  and  in  so 
doing  made  the  assumption  only  that  contem- 
porary and  phylogenetic  antecedent  herpeto- 
faunal assemblages  occupy  similar  climatic  re- 
gimes. This  seems  a  valid  assumption  considering 
the  taxa  concerned  are  terrestrial  ectotherms  and 
have  physiological  limitations  that  translate  di- 
rectly into  ecogeographic  restrictions  (Bartho- 
lomew 1958).  This  assumption  also  seems  well 
supported  by  the  fossil  record.  Murphy's  argu- 
ments are  further  weakened  by  the  fact  that 
throughout  his  own  paleobiogeographic  scena- 
rio, he  employs  the  same  paleobotanical  litera- 
ture and  makes  the  same  general  assumption  that 
geofloras  are  evidence  of  paleoclimatic  condi- 
tions that  are  indicative  of  certain  historical  her- 
petofaunal assemblages  in  a  given  region  during 
a  given  time. 

Generalized  Tracks  and  Ecogeographic 
Patterns 

The  strength  of  Murphy's  (1983#)  presenta- 
tion lies  in  the  excellent  paleobiogeographic  syn- 
thesis he  constructs  to  explain  contemporary  dis- 
tribution patterns,  and  in  the  genetic  evidence 
he  presents  to  support  many  of  his  conclusions. 
However,  I  find  his  herpetofaunal  assemblages 
("tracks")  to  have  serious  shortcomings.  Appar- 
ently this  is  a  result  of  his  misapplication  of  the 
concept  of  track  (Croizat  1 964),  and  the  concom- 
itant lack  of  distinction  between  contemporary 
distribution  patterns  and  the  diverse  historical 
assemblages  from  which  they  derived. 

It  is  important  to  clarify  some  terms  and  con- 
cepts critical  to  subsequent  discussion.  Murphy 
( 1 983#)  describes  six  herpetofaunal  areas  in  pen- 
insular Baja  California:  Californian,  Colorado 
Desert,  Vizcaino  Desert  North,  Coastal  Vizcaino 
Desert,  Peninsular  Area  (with  three  subdivi- 
sions), and  the  Cape  Area.  These  herpetofaunal 
areas  are  basically  broad  phytogeographic  units 
defined  on  the  basis  of  major  physiographic  or 
geomorphological  transformations  and  concom- 
itant differences  in  the  general  botanical  com- 


56 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


position  (based  on  Nelson  1921;  Savage  1960; 
Shreve  and  Wiggins  1964;  Bostic  1971;  Loomis 
et  al.  1 974;  and  Murphy's  own  fieldwork).  They 
are  comparable  in  concept  to  ecogeographic  for- 
mations (Fig.  14),  but  lack  as  fine  a  resolution 
because  they  each  encompass  a  larger  geographic 
area  containing  complex  geomorphological  for- 
mations with  numerous  vegetational  commu- 
nities. They  are  probably  best  characterized  as 
ecogeographic  provinces. 

Murphy  (1983a)  compiles  the  herpetofaunal 
composition  of  these  ecogeographic  provinces  and 
compares  them  quantitatively.  From  this  anal- 
ysis, he  derives  what  he  terms  "tracks."  He  then 
states  "phylogenetic  relationships  of  the  herpe- 
tofaunal elements  are  then  superimposed  on  these 
tracks  of  distribution  to  suggest  hypotheses  ex- 
plaining the  evolutionary  relationships  of  the 
herpetofauna  of  Baja  California"  (Murphy 
1983#:3).  The  use  of  the  term  "track"  in  this 
sense  is  misleading  and  confusing.  The  original 
concept  of  "track"  as  developed  and  employed 
by  Croizat  (1964;  see  also  Croizat  et  al.  1974; 
Craw  1983;  Craw  and  Weston  1984)  has  two 
components.  The  first  concept  is  that  of  the  in- 
dividual track;  this  concept  depicts  the  spatial 
(geographical)  expression  of  relatedness  among 
disparate  forms.  The  geographic  distributions  of 
all  related  forms  of  a  given  taxon  (e.g.,  a  species 
complex,  members  of  a  genus  or  family)  from 
disparate  geographic  areas,  when  considered  to- 
gether and  depicted  spatially,  comprise  an  in- 
dividual track.  The  second  concept,  that  of  gen- 
eralized or  standard  track  (op.  cit.)  describes  the 
phenomenon  of  congruency  in  space  of  more  than 
one  individual  track.  Both  individual  and  gen- 
eralized tracks  are  spatial  expressions  of  evolu- 
tionary time.  A  generalized  track,  because  of  the 
spatiotemporal  congruency  it  expresses,  indi- 
cates historical  homology  among  the  individual 
tracks,  and  thus  among  the  life  forms  represent- 
ed. In  other  words,  a  generalized  track  is  a  spatial 
and  temporal  representation  of  a  group  of  phy- 
logenetically  unrelated  organisms  that  shared 
proximate  space  through  time,  or  a  common 
evolutionary  history. 

Murphy  states  (198 3a:l  1):  "I  re-emphasize  that 
among  the  various  species  constituting  each  of 
these  five  major  generalized  tracks  of  distribu- 
tion, some  have  phylogenetic  affinities  with  trop- 
ical Mexico  and  others  with  territories  north  and 
east  of  the  peninsula."  In  my  understanding  of 


Croizat  (1964),  for  unrelated  life  forms  to  be 
members  of  the  same  generalized  track,  they  need 
to  demonstrate  distributional  similarities 
throughout  much  or  all  of  their  respective  ranges. 
In  other  words,  one  would  find  consistent  spatial 
overlap  among  the  phylogenetic  lineages  of  each 
of  the  various  member  taxa  of  the  generalized 
track.  From  the  lack  of  this  spatial  overlap  among 
represented  lineages,  one  would  have  to  conclude 
the  presence  of  more  than  a  single  generalized 
track.  Murphy's  (1983a)  use  of  the  term  "track" 
to  describe  herpetofaunal  taxa  that  have  present- 
day  geographic  overlap  within  Baja  California, 
but  lack  such  spatial  congruency  among  their 
respective  phylogenetic  lineages  elsewhere,  is  in- 
consistent with  the  concept  of  track. 

The  common  denominator  supporting  Mur- 
phy's faunal  groupings  is  consistent  spatial  over- 
lap among  species,  geographic  or  ecologic,  or  both, 
only  within  parts  of  Baja  California.  Lacking  the 
requisite  geographic  congruency  among  repre- 
sented lineages  beyond  the  Baja  California  Re- 
gion, one  would  have  to  conclude  that  these 
groupings  represent  contemporary  ecogeograph- 
ic patterns,  not  historical  patterns  or  tracks.  They 
appear  to  be  analogous  with  the  heretofore  de- 
scribed ecogeographic  patterns  of  the  herpeto- 
fauna of  the  Sierra  San  Pedro  Martir  Region  of 
northern  Baja  California  (Fig.  16;  Table  5;  and 
text). 

Subsequent  use  of  the  term  track  in  this  dis- 
cussion should  be  understood  in  the  sense  used 
by  Croizat  (1964).  When  indicating  Murphy's 
"tracks"  quotation  marks  will  be  used  to  distin- 
guish his  from  the  generalized  tracks  that  this 
investigation  indicates  exist  among  the  herpe- 
tofauna of  the  Baja  California  Region.  Savage's 
( 1 960)  historical  assemblages  appear  to  be  con- 
sistent with  the  temporal  aspect  of  generalized 
tracks;  these  parallels  will  be  indicated  below.  I 
consider  his  spatial  designation  of  province  con- 
sistent with  the  spatial  aspects  of  generalized 
tracks,  however,  like  Morafka  (1977)  I  consider 
Savage's  (1960)  provinces  to  equal  superprov- 
inces.  I  follow  Morafka  (1977)  in  my  use  of  prov- 
ince to  indicate  a  geographical  unit  with  uniform 
dominant  physiography,  vegetation,  and  fauna 
(see  Dice  1943).  My  use  of  the  term  track,  in  the 
spatial  sense,  is  synonymous  with  province  or 
superprovince  as  indicated. 

The  terms  track  and  generalized  track,  as  used 
in  the  temporal  sense  and  assemblage  or  histor- 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


57 


ical  assemblage,  are  used  interchangeably  and 
indicate  historical  biogeographic  groupings;  they 
represent  events  in  evolutionary  time.  The  term 
fauna  is  used  to  indicate  a  contemporary  group- 
ing of  forms  with  similar  ecogeographic  restric- 
tions; a  province  or  track  in  the  spatial  sense  may 
include  one  or  more  faunas  but  only  a  single 
fauna  is  considered  the  participant  fauna  (Mo- 
rafka  1977).  The  terms  fauna  and  track  as  used 
in  the  spatial  sense  represent  the  present,  in  eco- 
logical time;  here  considered  to  be  subsets  of  the 
historical  groupings. 

Extant  Faunas  and  Generalized  Tracks 

Savage  (1960)  did  not  list  all  of  those  species 
or  species-groups  he  considered  to  have  evolved 
together  as  part  of  a  particular  historical  assem- 
blage. But,  using  his  examples  of  species  with 
distributions  representative  of  his  historical  as- 
semblages, the  list  of  extant  vegetational  asso- 
ciations occupied  by  Recent  derivatives  of  his 
historical  complexes  and  components,  and  the 
contemporary  species  composition  of  Savage's 
Herpetofaunal  Areas,  I  assigned  each  contem- 
porary species  to  a  particular  historical  assem- 
blage. The  assignment  of  species  and  species- 
groups,  though  based  on  Savage's  (1960)  work, 
is  my  own  interpretation  and  is  thus  subject  to 
errors  for  which  I  alone  am  responsible.  Murphy 
likewise  did  not  list  the  member  species  for  each 
of  his  nine  possible  "tracks,"  leaving  the  reader 
to  decipher  from  his  table  2  (Murphy  1983a) 
those  species  that  fit  the  description  for  each  of 
his  "minor"  and  "major  generalized  tracks."  He 
was  unclear  about  proper  assignment  of  some 
forms  (e.g.,  on  page  6,  he  assigns  the  Sceloporus 
magister  complex,  minus  the  nominal  species, 
to  his  "Transpeninsular  Mesophilic  Track,"  but 
on  page  32  he  discusses  this  complex  as  part  of 
his  "Transpeninsular  Xerophilic  Track").  I  have 
attempted  to  interpret  Murphy's  work  and  assign 
species  in  accordance  with  his  descriptions;  once 
again  I  alone  am  responsible  for  any  errors. 

The  following  hypothesis  of  interacting  and 
overlapping  generalized  tracks  comprising  the 
contemporary  herpetofauna  of  the  Baja  Califor- 
nia Region  is  a  working  one  that  best  integrates 
our  current  knowledge  from  pertinent  fields. 
However,  it  is  a  hypothesis  to  be  tested  and 
changed  wherever  future  investigations  find  it 
lacking. 


The  Generalized  Tracks  of  the 
Baja  California  Herpetofauna 

Savage's  (1960)  and  Murphy's  (1983a)  Cali- 
fornian  Herpetofaunal  Area  (Fig.  2a),  consisting 
of  the  Pacific  Slope  of  Baja  California  north  of 
latitude  30°N,  is  the  most  biologically  complex 
area  of  the  Martir  Region,  and  probably  of  the 
entire  peninsula.  It  is  also  one  of  the  most  poorly 
investigated  by  previous  workers.  Savage  indi- 
cated that  this  area  was  dominated  by  an  assem- 
blage of  species  with  more  extensive  ranges 
northward  in  coastal  and  montane  California  that 
shared  an  evolutionary  history  associated  with 
derivatives  of  the  Madro-Tertiary  geoflora  (the 
Californian  Component  of  his  Madrean  Com- 
plex). In  addition,  he  indicated  that  some  ele- 
ments of  his  Sonoran  Desert  Component  of  the 
Desert  and  Plains  Complex  occur  within  the  Cal- 
ifornian Area.  This  study  indicates  that  elements 
of  four  contemporary  faunas  (High  Sierran  Scarps, 
Californian,  Peninsular  Range,  and  North  Amer- 
ican Ubiquitous  Scrubland;  Fig.  15,  16;  Table 
5),  representing  four  different  tracks  (Pacific 
Northwest,  Madrean,  Peninsular  Range,  and 
North  American  Desert  and  Plains:  Fig.  17,  18, 
20,  21),  overlap  on  the  Pacific  Slope  of  Baja  Cal- 
ifornia north  of  latitude  30°N. 

The  Californian  Fauna  together  with  the  High 
Sierran  Scarps  Fauna  (Table  5,  I  and  II)  match 
closely  the  species  Savage  indicated  as  endemic 
to  and  dominating  the  Californian  Herpetofau- 
nal Area.  Several  species  within  this  area,  as  not- 
ed by  Savage,  are  associated  primarily  with  de- 
rivatives of  the  Arcto-Tertiary  geoflora.  These 
species,  the  High  Sierran  Scarps  Fauna,  are  re- 
stricted to  Arcto-Tertiary  floral  derivatives  in 
northern  Baja  California  and  southern  Califor- 
nia. However,  outside  of  Baja  California,  mem- 
bers of  this  fauna  demonstrate  broader  ecological 
niches  and  occur  in  conjunction  with  more  me- 
sophilic  derivatives  of  the  Madro-Tertiary  geo- 
flora (see  Species  Accounts;  Stebbins  1 966,  1 985). 
The  variability  in  niche  breadth  in  different  parts 
of  their  range,  and  the  presence  of  sympatric  con- 
geners within  the  Baja  California  Region,  suggest 
that  the  geographic  limits  of  the  High  Sierran 
Scarps  Fauna  result  from  competition.  Overall 
distributions  of  these  species  closely  match  those 
of  the  Californian  Fauna,  suggesting  that  these 
two  faunas  have  a  very  similar  and  perhaps  par- 
allel evolutionary  history  in  western  North 


58 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


America.  The  relatively  low  Coefficient  of  Dif- 
ference and  high  Similarity  Coefficient  and  Com- 
munity Coefficient  values  (Table  4;  Fig.  1 5)  be- 
tween areas  occupied  by  these  two  faunas  are 
also  indicative  of  a  close  relationship.  With  the 
exception  of  three  species  of  the  Californian  Fau- 
na (Rana  aurora,  Bufo  boreas,  and  Clemmys 
marmorata)  that  are  also  associated  primarily 
with  Arcto-Tertiary  geofloral  derivatives,  the 
species  of  the  High  Sierran  Scarps  Fauna  and  the 
Californian  Fauna  are  considered  to  be  contem- 
porary members  of  the  Madrean  Track  (Fig.  18; 
Table  5,  6;  and  below). 

The  Pacific  Northwest  Track 

Rana  aurora,  Bufo  boreas,  and  Clemmys  mar- 
morata are  considered  to  be  members  of  the  Pa- 
cific Northwest  Track  (Fig.  17;  Table  5,  6).  This 
track  is  synonymous  with  Savage's  Western 
American  Complex  of  the  Old  Northern  Ele- 
ment. Pacific  Northwest  Track  members  have 
the  most  tenuous  toehold  on  the  predominantly 
xeric  Baja  California  peninsula.  Several  member 
species  occur  in  scattered  populations  in  a  few 
canyons  of  extreme  northwestern  Baja  (i.e., 
Aneides  lugubris  and  Ensatina  eschscholtzii). 
Members  of  this  track  probably  dispersed  into 
the  Region  following  completion  of  the  Coast 
Range  Corridor  (Peabody  and  Savage  1958)  dur- 
ing the  cooler  and  moister  Pleistocene  epoch. 
Two  of  these  species,  Rana  aurora  and  Bufo  bo- 
reas, appear  to  have  derived  from  Asian  forms, 
and  as  noted  by  Savage  (1960),  apparently  dis- 
persed across  the  Bering  Land  Bridge  and  down 
the  west  coast  of  North  America  during  the 
Pleistocene  (Savage's  Holarctic  Element).  It  is 
reasonable  to  propose  that  they  constitute  a  sec- 
ond, more  recent  generalized  track,  one  that 
overlaps  the  older  Pacific  Northwest  Track. 

The  Madrean  Track 

The  Madrean  Track  (Fig.  18;  Table  6)  is  con- 
sidered synonymous,  in  terms  of  process  and  pat- 
tern, with  Savage's  (1960)  Madrean  Complex  of 
the  Young  Northern  Element,  but  with  a  con- 
siderably different  species  composition.  Savage 
distinguished  a  Californian  Component  and  a 
San  Lucan  Component  in  his  Madrean  Complex. 
His  San  Lucan  Component  includes  most  of  the 
species  here  assigned  to  the  Peninsular  Range 
Track  (Table  6),  as  discussed  below.  Included»as 


contemporary  elements  of  a  San  Lucan  compo- 
nent of  my  Madrean  Track  (Table  6)  are  only 
those  forms  present  as  sibling  species  in  the  Cal- 
ifornian and  Cape  regions  (and  on  some  Pacific 
coastal  islands  of  Baja  California)  and  absent 
from  most  or  all  of  the  Central  and  Vizcaino 
deserts  (e.g.,  Eumeces,  Hyla,  Elgaria,  and  Col- 
uber "lateralis  complex"). 

Savage  (1960)  considered  his  Madrean  Com- 
plex the  oldest  assemblage  in  the  Baja  California 
Region.  Such  dating  probably  refers  to  those  taxa 
that  are  here  assigned  to  the  Peninsular  Range 
Track  fauna  (Table  6).  Distribution  of  popula- 
tions on  "old"  or  "deep-water"  islands  of  the 
Gulf  of  California  (Soule  and  Sloan  1966;  Mur- 
phy 19836),  and  the  concomitant  relatively 
greater  levels  of  divergence  within  lineages  (see 
Species  Accounts  for  taxonomic  references),  in- 
dicate that  the  members  of  the  Peninsular  Range 
Track  are  much  older  elements  on  the  land  masses 
of  Baja  California  than  are  Madrean  Track  forms. 
Biochemical  and  electrophoretic  analyses  (Kim 
et  al.  1976;  Wyles  and  Gorman  1978;  Murphy 
and  Papenfuss  1979;  Murphy  19830)  indicate  a 
probable  mid-Miocene  divergence  for  several  of 
the  lineages  of  the  Peninsular  Range  Track. 

The  Madrean  Track  fauna,  both  within  and 
beyond  the  Baja  California  Region,  appears  to 
be  a  previously  more  widespread,  vicariated  (see 
Figure  18)  assemblage  of  relatively  mesophilic 
species-groups.  The  level  of  differentiation  among 
disjunct  populations  of  these  forms  within  the 
Baja  California  Region  appears  to  match  Mayr's 
(1978)  criteria  for  sibling  species  or  semispecies. 
Divergence  and  endemism  within  lineages  of 
Madrean  Track  elements  beyond  Baja  California 
(in  California,  Arizona,  New  Mexico,  and  Mex- 
ico) are  considerably  greater  than  the  level  of 
sibling  species,  with  most  lineages  having  pop- 
ulations along  the  track  that  appear  to  meet  the 
criteria  of  superspecies  (Mayr  1978).  The  rela- 
tively greater  differentiation  within  lineages  of 
the  Madrean  Track  beyond,  compared  to  within 
Baja  California,  suggests  a  long-term  historical 
relationship  among  these  species-groups  prior  to 
their  inhabiting  Baja  California,  and  an  evolu- 
tionary history  distinct  from,  yet  probably  par- 
alleling, the  lineages  of  the  Peninsular  Range 
Track. 

Much  of  existing  Baja  California  appears  to 
have  been  inaccessible  to  southward  emigration 
of  terrestrial  vertebrates  (see  Paleobiogeographic 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


History  above),  possibly  from  the  mid-Miocene 
and  probably  from  about  the  Miocene-Pliocene 
boundary  until  the  Pleistocene.  If  so,  after  the 
mid-Miocene  or  the  Miocene-Pliocene  bound- 
ary and  prior  to  the  Pleistocene,  Madrean  Track 
elements  could  have  reached  Baja  California  only 
by  being  carried  along  on  the  migrating  land 
masses,  i.e.,  by  "transgulfian  vicariance"  (Mur- 
phy 1975,  19830).  The  complete  absence  of 
Madrean  Track  forms  on  the  "old"  or  "deep- 
water"  Gulf  islands  (Soule  and  Sloan  1966;  Mur- 
phy 1983fl,  b\  Murphy  and  Ottley  1983,  1984) 
discounts  this  possibility,  although  vicariance  and 
subsequent  extinction  cannot  be  entirely  ruled 
out.  An  alternative  possibility  is  that  some  of 
these  lineages  were  present  on  the  proto-northern 
Peninsular  Range  land  masses  during  the  Mio- 
cene and  vicariated  at  a  later  time  (e.g.,  Miocene- 
Pliocene  boundary)  than  did  lineages  on  the  Cape 
Islands.  However,  the  relatively  low  levels  of  di- 
vergence between  the  Baja  California  popula- 
tions of  these  lineages,  their  absence  from  the 
"old"  or  "deep-water"  Gulf  islands,  and  the  in- 
crease in  numbers  of  Madrean  Track  species 
present  along  a  south  to  north  transect  in  pen- 
insular Baja  California,  strongly  suggest  a  rela- 
tively recent  (i.e.,  Pleistocene)  diffusion— gradual 
movement  across  hospitable  areas  over  long  pe- 
riods (Pielou  1979)— southward.  The  disjunct 
distributions  (vicariance)  of  some  Madrean  Track 
elements  along  the  peninsula  suggests  even  more 
recent  (mid-Pleistocene  to  Recent)  fragmenta- 
tion of  ranges  in  conjunction  with  climatic  shifts 
toward  aridity  (Axelrod  1967,  1975,  1979)  or 
from  possible  periodic  inundation  of  low-lying 
areas  by  changes  in  sea  level,  or  both  (Flint  1971). 
At  least  two  possible  scenarios  are  suggested 
by  the  distributions  of  the  Madrean  Track  lin- 
eages within  Baja  California.  Firstly,  the  spatial 
relationships  between  the  congeners  of  the  Cal- 
ifornian  and  the  High  Sierran  Scarps  Faunas  (Ta- 
ble 5),  which  I  propose  to  have  resulted  from 
competition,  are  consistent  with  a  hypothesis  of 
Pleistocene  to  Recent  periodic  diffusions  by  these 
lineages  southward  into  Baja  California.  Late 
Pliocene  or  Pleistocene  formation  of  a  terrestrial 
connection  between  Baja  California  and  south- 
ern California  land  masses  and  the  concomitant 
erogenic  uplift  would  have  provided  southward 
emigration  routes  as  well  as  the  upland  retreats 
consistent  with  the  contemporary  distributions 
of  these  faunas.  These  contemporary  distribu- 


tions of  mesophilic  species,  and  those  of  more 
xerophilic  species  associated  with  other  tracks 
discussed  below,  indicate  that  for  at  least  part  of 
this  time,  barriers  to  gene  flow  existed.  Elements 
of  the  Madrean  Track  probably  dispersed  in  suc- 
cessive diffusions  across  a  partial  or  fluctuating 
barrier  (San  Gorgonio  Barrier)  during  this  time. 
Secondly,  prior  to  this  time,  from  mid-Miocene 
to  Pleistocene,  it  is  unclear  what  the  nature  of 
the  terrestrial  connection  at  the  northern  end  of 
the  extant  peninsula  was,  or  even  if  it  existed. 
Some  elements  I  have  assigned  to  the  Madrean 
Track  have  distributions  that  suggest  they  may 
have  been  present  on  the  northern  peninsular 
land  masses  at  mid-Miocene,  when  increased 
tectonic  activity  vicariated  their  ranges,  isolating 
them  from  ancestral  stocks  to  the  east  (e.g.,  An- 
niella  geronimensis,  Cnemidophorus  labialis,  and 
Xantusia  henshawi). 

A  hypothesis  that  incorporates  both  possible 
scenarios  is  preferred.  Species  like  Anniella  ge- 
ronimensis, Cnemidophorus  labialis,  and  Xan- 
tusia henshawi  are  not  good  candidates  to  sup- 
port a  Pleistocene  diffusion  hypothesis;  their 
endemism,  narrow  niches,  and  low  vagility  are 
all  indicative  of  species  that  are  evolutionarily 
relatively  conservative,  with  their  present  ranges 
probably  contracting  rather  than  expanding.  They 
appear  to  be  the  remains  of  an  ancestral  fauna, 
a  proto-Madrean  Track  fauna  that  may  have  vi- 
cariated in  the  Miocene.  If  this  were  indeed  the 
case,  these  species  could  arguably  be  evidence  of 
another,  earlier,  generalized  track,  possibly  the 
result  of  the  same  geomorphological  events  that 
shaped  the  Peninsular  Range  track  further  south 
as  discussed  below.  On  the  other  hand,  their  sym- 
patric  congeners  and  the  other  species  of  the 
Madrean  Track,  probably  made  their  appearance 
in  the  northern  peninsula  (and  beyond),  subse- 
quent to  Pleistocene  climatic  changes,  sea  level 
shifts,  and  related  phenomena  that  promoted  an 
ingress  across  the  San  Gorgonio  Barrier.  Under 
the  cool  moist  conditions  of  Pleistocene  glacial 
maxima,  the  more  hydrophilic  High  Sierran 
Scarps  Fauna  could  have  undergone  diffusion 
southward.  The  presence  of  Lampropeltis  zonata 
and  Batrachoseps  pacificus  on  Todos  Santos  Is- 
land off  the  Pacific  coast  of  Baja  California  sup- 
ports the  contention  that  cool,  moist  conditions 
were  more  widespread  in  northern  Baja  Califor- 
nia in  the  past.  With  glacial  retreat  and  the  long- 
term  drying  and  warming  trend  culminating  in 


60 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


the  late  Pleistocene  to  Recent  (Axelrod  1967, 
1975,  1979),  conditions  became  increasingly  fa- 
vorable for  the  westward  and  southward  dis- 
persal of  the  warm-mesophilic  Californian  Fau- 
na. Concomitant  with  these  southward  advances 
of  the  Californian  Fauna,  the  High  Sierran  Scarps 
Fauna  was  increasingly  restricted  and  relegated 
to  habitats  at  higher,  cooler  elevations.  With  the 
advancing  aridity,  even  the  Californian  Fauna 
became  restricted;  ranges  became  fragmented  and 
populations  in  arid  areas  relegated  to  protect  mi- 
crohabitat,  such  as  their  present-day  riparian 
corridor  distributions  at  lower  elevations  in  the 
Martir  Region  (Table  5;  Fig.  16).  This  scenario 
of  diffusing  and  fragmenting  faunas  is  consistent 
with  the  current  evidence  of  geologic  and  cli- 
matic conditions  during  Pleistocene  and  Recent 
times.  It  is  also  consistent  with  and  supportive 
of  the  thesis  that  sympatry  within  monophyletic 
lineages  is  evidence  of  dispersal  (Croizat  et  al. 
1974;  Platnick  and  Nelson  1978). 

Four  interrelated  criteria  were  used  to  distin- 
guish species  of  the  Madrean  Track  from  those 
of  the  Peninsular  Range  Track  (discussed  below). 
(1)  Madrean  Track  species  and  species  groups, 
for  the  most  part,  exhibit  extensive  ranges  be- 
yond the  Baja  California  Region.  Members  of 
the  Peninsular  Range  Track  have  their  greatest, 
and  in  many  cases  their  total,  distribution  within 
the  Baja  California  Region.  (2)  In  the  Baja  Ca- 
lifornia Region,  Madrean  Track  forms  have 
comparatively  less  differentiation  within  mono- 
phyletic lineages  compared  to  lineages  of  the 
Peninsular  Range  Track,  on  both  peninsular  and 
insular  land  masses.  (3)  Madrean  Track  forms 
are  present  only  on  some  of  the  land-bridged 
islands  (Soule  and  Sloan  1966;  Murphy  and  Ott- 
ley  1983,  1984)  of  the  Gulf  of  California  and  on 
some  of  the  Pacific  coastal  islands  of  Baja  Cali- 
fornia and  California.  They  are  absent  from  all 
of  the  "old"  or  "deep-water"  islands  (Soule  and 
Sloan  1966;  Murphy  and  Ottley  1983,  1984)  of 
the  Gulf,  which  are  inhabited  by  numerous  en- 
demic forms  of  the  Peninsular  Range  Track  lin- 
eages and  some  elements  of  the  North  American 
Desert  and  Plains  Track  (discussed  below).  (4) 
As  a  general  trend,  Madrean  Track  forms  appear 
to  have  their  closest  relationships  within  mono- 
phyletic lineages  with  forms  occurring  in  Cali- 
fornia, Arizona,  New  Mexico,  and  the  northwest 
central  uplands  of  Mexico  (e.g.,  Rana  boy  Hi 
[Zweifel  1955],  Lampropeltis  zonata  [Tanner 


1953],  Batrachoseps  [Wake  19666],  Xantusia 
[Bezy  1972;  Bezy  and  Sites  1987]).  Peninsular 
Range  Track  elements  are  most  closely  related 
to  forms  occurring  in  lowland  areas  from  central 
Mexico  southward  (e.g.,  Ctenosaura  hemilopha 
[Bailey  1928;  Smith  1972],  Eridiphas  [Duellman 
1958;  Leviton  and  Tanner  1960;  Cadle  1984; 
Dowling  and  Jenner  1987],  Phyllodactylus  [Dix- 
on  1964,  1969;  Murphy  and  Papenfuss  1979], 
Cnemidophorus  hyperythrus  and  C.  ceralbensis 
[Lowe  et  al.  1970;  Robinson  1973],  Phrynosoma 
coronatum  [Presh  1969],  Sator  [Wyles  and  Gor- 
man 1978],  Nerodia  [Conant  1969],  and  Bipes 
[Kim  et  al.  1976;  Papenfuss  1982]). 

Peninsular  Range  Track 

The  third  contemporary  fauna  with  elements 
in  Savage's  Californian  Herpetofaunal  Area  of 
the  Martir  Region  is  the  Peninsular  Range  Fauna 
(Table  5,  IV).  Some  of  these  taxa  occur  also  in 
the  Colorado  Desert  Herpetofaunal  Area  (Fig. 
2a)  and  southward,  throughout  the  length  of  the 
peninsula.  The  Peninsular  Range  Fauna  is  part 
of  the  larger  Peninsular  Range  Track  (Table  6; 
Fig.  20)  comprised  primarily  of  forms  endemic 
to  the  Baja  California  Region  and  associated  with 
warm-temperate  and  subtropical  floral  ele- 
ments. Many  forms  of  the  Peninsular  Range 
Track  occur  only  south  of  the  Martir  Region  in 
the  Central  and  Vizcaino  deserts,  in  the  Cape 
Region,  or  on  islands  in  the  Gulf  of  California 
(see  Murphy  1983a;  Murphy  and  Ottley  1983, 
1984). 

These  peninsular  and  insular  endemics  are  the 
oldest  lineages  present  in  the  Baja  California  Re- 
gion, dating  from  mid-Miocene  (Murphy  1983a). 
Savage  (1960)  indicated  the  members  of  this  fau- 
na were  part  of  two  of  his  southward  dispersing 
waves:  a  Sonoran  Component  of  his  Desert  and 
Plains  Complex,  and  a  Madrean  Complex.  How- 
ever, the  present  distributions  of  these  species 
and  species-groups  are  more  consistent  with  a 
theory  of  origin  in  Baja  California  based  on  vi- 
cariance,  each  distribution  appearing— at  least  in 
part— to  result  from  the  migration  and  fragmen- 
tation of  peninsular  and  insular  land  masses  from 
coastal  Mexico.  Murphy  (1975)  proposed  such  a 
theory  of  vicariated  origin  to  account  for  several 
of  these  lineages  found  in  the  Cape  Region  and 
on  some  associated  islands;  his  "trans-gulfian 
migration"  hypothesis.  Welsh  (1976<2,  b,  1981) 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


61 


proposed  a  similar  hypothesis  to  account  for  the 
distributions  of  most  of  the  Sonoran  Desert-as- 
sociated lineages  of  Baja  California  (see  list  at 
beginning  of  Discussion),  that  are  considered  to 
have  a  shared  evolutionary  history  with  the  more 
clearly  vicariated  subtropical  Cape  Island  en- 
demics of  the  Peninsular  Range  Track.  Murphy 
(1983#)  included  these  Sonoran  Desert-associ- 
ated lineages  and  added  several  more  mesophilic 
lineages  in  his  revised  "trans-gulfian  vicariance" 
model. 

Murphy  (1983a)  included  the  mesophilic  lin- 
eages Batrachoseps,  Hyla  (=H.  regilla  progeni- 
tor), Eumeces,  and  Elgaria  in  his  "trans-gulfian 
vicariance"  model.  Exception  is  taken  to  the  ad- 
dition of  these  lineages,  all  of  which  are  herein 
considered  to  be  part  of  the  Madrean  Track.  In 
adding  these  mesophilic  genera,  Murphy  ( 1 9  8  3  a) 
assumed  that  the  Cape  Region  or  Cape  Islands 
supported  both  Neotropical  and  Madro-Tertiary 
geofloral  elements  during  the  mid-Miocene  when 
the  area  became  separated  from  coastal  Mexico: 
"In  the  absence  of  fossil  data,  I  assume  that  the 
Cape  Islands  supported  a  mixture  of  Neotropi- 
cal-Tertiary and  Madro-Tertiary  geofloras  in  the 
respective  tropical  and  temperate  climatic  re- 
gimes; the  temperate  regime  of  the  Cape  resulting 
from  altitudinal  effects"  (Murphy  1983«:16). 
Murphy  was  making  the  same  kind  of  assump- 
tion (inferring  presence  of  historically  associated 
herpetofauna  from  presence  of  geoflora)  that  he 
criticized  Savage  for  making.  Despite  his  appar- 
ent contradiction,  and  "willingness"  to  join  Sav- 
age and  myself  in  making  this  kind  of  assump- 
tion, I  think  in  this  particular  case,  the  assumption 
is  invalid  as  I  shall  discuss  below. 

Axelrod  (1979:28)  indicated  the  presence  of 
both  Neotropical  geofloral  elements  (dry  tropic 
forest  and  arid  tropic  scrub)  and  Madro-Tertiary 
geofloral  elements  (oak-pinyon  woodland)  along 
coastal  western  Mexico  during  the  Miocene. 
However,  he  stated  that  ". .  .  woodland  presum- 
ably reached  southward  down  a  line  of  volcanos, 
linking  the  flora  of  the  Cape  Region  with  that  of 
Sonora-Sinaloa-Nayarit-Jalisco,  which  probably 
lived  above  scattered  dry-tropic  forest  and  thorn 
forest"  (1 979:29).  Whereas  it  seems  possible  that 
volcanism  could  have  provided  the  requisite  up- 
land habitat  in  some  parts  of  coastal  western 
Mexico  as  Axelrod  indicated,  the  Cape  Islands 
land  masses  are  areas  of  granitic  batholith,  not 
volcanic  rock  (Mina  1957;  Gastil  et  al.  1975). 


Furthermore,  there  is  no  evidence  that  this  area 
had  been  uplifted  to  sufficient  height  by  mid- 
Miocene  to  create  the  cooler  climatic  regime  req- 
uisite for  the  Madro-Tertiary  geoflora. 

Even  in  the  event  that  such  vegetational  ele- 
ments existed  at  the  right  time  and  in  the  right 
place  on  the  geomorphic  progenitors  of  the  Cape 
Islands,  these  vegetation  communities  would 
have  been  present  on  upland  areas  as  islands  of 
habitat  surrounded  by  great  expanses  of  Neo- 
tropical floral  elements  (Axelrod  1979:28).  As 
such,  (a)  these  "islands"  would  have  been  far 
removed  from  the  northwest  central  highland  and 
central  plateau  areas  of  Mexico  considered  to  be 
the  centers  of  differentiation  and  evolution  of  the 
Madro-Tertiary  geoflora  (Axelrod  1975,  1979), 
and  (b)  they  would  also  most  probably  be  highly 
depauperate.  Therefore,  it  seems  highly  unlikely 
that  these  "islands"  could  contain  and  support 
the  mesophilic  elements  of  the  ancestral  herpe- 
tofauna that  are  thought  to  have  evolved  in  as- 
sociation with  the  then-developing  Madro-Ter- 
tiary vegetation. 

The  possibility  that  Batrachoseps,  Hyla,  Eu- 
meces, and  Elgaria  vicariated  on  the  Cape  Is- 
lands in  upland  habitats  during  mid-Miocene 
cannot  be  entirely  discounted.  However,  it  seems 
highly  improbable  and  less  likely  than  the  more 
parsimonious  explanation  that  these  genera  are 
elements  of  the  Madrean  Track  (Fig.  18),  whose 
origins  in  the  Cape  Region  are  probably,  as  Sav- 
age (1960)  suggested,  the  result  of  late  Pliocene 
or  Pleistocene  diffusions  down  an  existing  pen- 
insula. 

The  fossil  herpetofaunal  evidence  that  exists 
from  the  Cape  Region  (Miller  1977)  dates  from 
the  late  Pliocene  and  includes  Boa,  Geochelone, 
and  Crocodylus,  etc.,  indicating  decidedly  trop- 
ical conditions  at  least  in  the  vicinity  of  this  low- 
land site. 

The  noninsular  forms  of  the  Peninsular  Range 
Track  show  three  general  patterns  of  distribution 
in  the  Baja  California  Region:  (1)  restricted  en- 
demic lineages— those  lineages  found  only  in  one 
or  more  of  the  four  southern  ecogeographic  prov- 
inces (Murphy  1983a)  of  the  peninsula  or  on 
associated  "deep-water"  islands  (Soule  and  Sloan 
1966;  Murphy  19836),  or  both;  generally  south 
of  latitude  28°N  through  the  Cape  Region  (e.g., 
Bipes  biporus,  Chrysemys  scripta,  Ctenosaura 
hemilopha,  Sator  complex,  Nerodia  vallida,  and 
Eridiphas  slevini);  (2)  endemic  lineages  (note: 


62 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


some  of  these  lineages  have  member-species  that 
when  considered  separately,  fit  the  description 
in  (1)  above)— those  lineages  that  range  through- 
out most  or  all  of  the  Peninsular  Ranges  but  not 
north  of  the  San  Gorgonio  Barrier  (e.g.,  Cne- 
midophorus  hyperythrus,  Crotaphytus  insularis, 
Petrosaurus  complex,  Phyllodactylus  complex, 
Sceloporus  magister  complex— those  members 
with  2n  =  30  chromosomes,  Sceloporus  orcutti 
complex,  Elaphe  rosaliae,  Crotalus  enyo,  and  C. 
rubef);  (3)  radiated  lineages— those  lineages  that 
have  expanded  beyond  the  Peninsular  Ranges 
into  habitats  of  the  Mojave  and  Sonoran  deserts 
and  north  and  east  (Callisaurus  draconoides,  Co- 
leonyx  switaki,  Dipsosaurus  dorsalis,  Phrynoso- 
ma  coronation,  Sauromalus  complex, 
Chilomeniscus  complex,  Crotalus  mitchellii,  Li- 
chanura  trivirgata,  and  Phyllorhynchus  decur- 
tatus). 

Murphy  (1983a)  outlined  a  plausible  paleo- 
biogeographic  scenario  to  account  for  the  distri- 
butional permutations  and  divergence  within 
lineages  among  the  Peninsular  Range  Track  fau- 
na. Lineages  with  member  elements  distributed 
in  accordance  with  each  of  the  above  described 
geographic  patterns  consist  of  sibling  species  along 
the  peninsula,  with  one  or  two  forms  in  the  Cape 
Region  and  another  form  or  forms  to  the  north 
across  the  Isthmus  of  La  Paz  (e.g.,  Petrosaurus, 
Urosaurus,  Sceloporus  orcutti  complex,  S.  ma- 
gister complex,  Phyllodactylus  complex,  and 
Chilomeniscus  complex).  Murphy  attributed  this 
species-pairs  phenomenon  to  a  temporary  or 
fluctuating  terrestrial  corridor  between  the  Cape 
Islands  (Region)  and  areas  to  the  north,  probably 
during  the  late  Miocene  or  early  Pliocene  (see 
his  "Transpeninsular  Xerophilic  Track").  It 
would  appear  that  more  than  a  single  vicariating 
event,  or  intermittent  gene  flow  between  more 
than  two  areas  is  involved  here,  based  on  the 
presence  of  more  than  two  distinct  forms  among 
several  of  these  lineages  on  the  southern  penin- 
sula. Murphy  (1983a,  b)  gives  detailed  account- 
ings of  the  numerous  unique  patterns  among  in- 
sular distributions  of  the  lineages  of  the  Peninsular 
Range  Track,  and  also  for  those  insular-occur- 
ring forms  I  have  assigned  to  the  North  Amer- 
ican Desert  and  Plains  Track  (see  Table  6  and 
below). 

Assuming  the  validity  of  the  hypothesis  that 
species  of  the  Peninsular  Range  Track  predate— 
in  Baja  California— those  species  of  the  Madrean 
Track,  then  future  electrophoretic  or  immuno- 


logical  distance  data  should  indicate  a  trend  of 
greater  genetic  distance  between  Peninsular  Range 
Track  members  and  their  southwest  Mexican 
siblings,  than  that  which  would  be  found  between 
the  Madrean  Track  forms  and  their  siblings  be- 
yond the  Baja  California  Region  (with  the  ex- 
ception of  some  of  the  possible  proto-Madrean 
Track  forms  like  Cnemidophorus  labialis  and 
Xantusia  henshawi).  Good  (1988)  presented  al- 
lozyme  data  for  the  genus  Elgaria  that  indicate 
a  closer  relationship  between  E.  paucicarinata  of 
the  Cape  Region  and  E.  panamintina  of  south- 
eastern California,  than  between  E.  paucicari- 
nata and  E.  kingi  of  western  Mexico.  Following 
Savage's  (1960)  scheme  of  historical  herpeto- 
faunas  of  North  America,  the  lineages  of  the  Pen- 
insular Range  Track  would  be  called  the  Pen- 
insular Range  Complex  of  the  Young  Northern 
Herpetofauna. 


The  Colorado  Desert  Track 

Peninsular  members  of  the  Colorado  Desert 
Track  (Fig.  19;  Table  6)  are  found  only  in  the 
Colorado  Desert  Herpetofaunal  Area  (Fig.  2a). 
The  Colorado  Desert  was  formed  during  the 
Pleistocene;  it  resulted  from  the  continued  ele- 
vation of  the  northern  Peninsular  Ranges,  which 
forced  the  Gulf  of  California  to  recede  (Atwater 
and  Molner  1973),  and  the  accumulation  of  al- 
luvial deposits  from  the  Colorado  River  filling 
in  the  head  of  the  Gulf  (Norris  1958).  Species  of 
the  Colorado  Desert  Track  have  the  most  re- 
stricted distributions  of  the  xerophilic  herpeto- 
fauna  of  the  Baja  California  Region;  monophy- 
letic  lineages  demonstrate  the  lowest  levels  of 
interpopulational  morphological  differentiation. 

The  limited  distributions  and  minimal  intra- 
lineal  differentiation  (see  Adest  1977),  together 
with  their  endemism,  suggest  a  Pleistocene  to 
Recent  isolation  and  in  situ  evolution  for  these 
forms.  The  presence  of  congeners  of  several  of 
these  species  among  the  Peninsular  Range  Track 
fauna  (Table  6),  suggests  that  competition  may 
limit  the  spread  of  these  species  further  south 
into  other  desert  habitats  of  Baja  California. 

This  is  an  incipient  track  and  hence  does  not 
demonstrate  the  characteristics  of  spatial  and 
temporal  fragmentation  among  forms  associated 
with  older  tracks  (except  see  the  distribution  of 
the  genus  Uma;  Stebbins  1985).  The  Colorado 
Desert  Track  becomes  apparent  when  one  ex- 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


63 


amines  the  pattern  of  distributions  of  sibling 
forms  in  the  Chihuahuan,  Sonoran,  and  Penin- 
sular deserts.  In  fact,  these  four  desert  areas,  each 
considered  here  as  a  generalized  track  with  its 
own  participant  fauna,  can  also  be  viewed  as 
subsets  of  the  more  extensive  and  older  North 
American  Desert  and  Plains  Track  (see  below). 

Morafka  (1977:187)  hypothesized  a  faunal 
assemblage  associated  with  the  Colorado  Desert 
he  called  the  California  Gulf  Arch  Assemblage. 
He  considered  this  assemblage  of  10  reptiles 
(Dipsosaurus  dorsalis,  Sauromalus  obesus,  Cal- 
lisaurus  draconoides,  Lichanura  trivirgata,  Chi- 
lomeniscus  cinctus,  Chionactis  occipitalis,  Phyl- 
lorhynchus  decurtatus,  Crotalus  cerastes,  Crotalus 
mitchellii,  and  Gopherus  agassizii)  to  be  a  sec- 
ondary, more  northerly  centered  ecological  unit 
of  the  Sonoran  Herpetofauna,  a  fauna  centered 
to  the  south  in  Sonora  and  Sinaloa,  Mexico.  I 
agree  that  the  fauna  of  the  Colorado  Desert  show 
strong  affinities  with  the  Sonoran  Herpetofauna, 
but  I  do  not  consider  it  a  subset  thereof.  Despite 
the  distributional  similarities  apparent  among 
these  forms,  track  analysis  indicates  his  is  a  syn- 
thetic assemblage.  It  appears  to  me  that  his  Cal- 
ifornia Gulf  Arch  Assemblage  is  comprised  of 
Peninsular  Range  and  Sonoran  herpetofaunal 
elements  that  have  radiated  into  amenable  hab- 
itats not  otherwise  occupied  in  each  of  the  other 
areas,  plus  Colorado  Desert  Track  forms  that 
have  evolved  in  situ;  in  both  cases  as  a  result  of 
post-Pleistocene  geomorphological  and  climatic 
changes  favoring  expansion  of  extreme  desert 
habitats. 

Adest  (1987)  analyzed  allozyme  data  for  Cal- 
lisaurus  draconoides  from  populations  in  the  Cape 
Region  of  Baja  California  throughout  much  of 
the  range  including  Nevada,  and  down  the  west 
coast  of  Mexico  to  the  vicinity  of  Mazatlan.  His 
results  indicated  low  genetic  diversity  through- 
out the  range,  with  some  minor  differences  in 
populations  near  Mazatlan  and  on  Isla  Ceralvo 
in  the  Gulf  of  California.  His  data  were  incon- 
clusive to  resolve  the  question  of  a  Pleistocene 
isolation  in  either  a  Sonoran  or  a  Cape  Region 
refugium  (Savage  1960)  or  a  California  Gulf  Arch 
refugium  (Morafka  1977).  However,  he  con- 
cluded that  his  data  did  not  support  a  model  of 
a  Pleistocene  Cape  Region  refuge  for  the  species. 
Unfortunately  he  did  not  include  samples  from 
the  population  on  Isla  Angel  de  La  Guarda,  a 
deep-water  island  isolated  from  central  penin- 
sular Baja  California  1  MYBP  (Murphy  1 983/7). 


Significant  differences  between  this  population 
and  those  tested  by  Adest  (1987)  would  add  cre- 
dence to  a  hypothesis  of  a  mid-latitude  penin- 
sular or  insular  refuge  for  Callisaurus  draco- 
noides. 

Following  Savage's  historical  groupings,  the 
species  of  the  Colorado  Desert  Track  might  be 
called  the  Colorado  Desert  Component  of  the 
Desert  and  Plains  Complex. 

The  riparian  herpetofauna  of  the  Colorado 
River  delta  region  (Bufo  alvarius,  Bufo  cognatus, 
Bufo  woodhousei,  Urosaurus  ornatus,  Thamno- 
phis  marcianus,  and  Trionyx  spiniferus)  were  not 
included  in  my  ecogeographic  analysis  because 
none  of  them  range  south  into  the  Martir  Region. 
With  the  exceptions  of  Bufo  alvarius  and  Uro- 
saurus ornatus,  which  have  distributions  char- 
acteristic of  Sonoran  Track  forms,  these  species 
appear  to  be  post-Pleistocene  to  Recent  radia- 
tions from  the  east.  Trionyx  spiniferus  and  Bufo 
woodhousei  appear  to  be  elements  of  a  western 
radiation  of  the  Austriparian  Herpetofauna  (Sav- 
age 1960)  centered  in  the  southeastern  U.S. 
Thamnophis  marcianus  appears  to  be  a  member 
of  the  Chihuahuan  Track  that  has  crossed  the 
Cochise  Filter  Barrier  (Morafka  1977),  and  Bufo 
cognatus  appears  to  be  a  member  of  the  North 
American  Desert  and  Plains  Track. 


The  North  American  Desert  and  Plains  Track 

The  North  American  Desert  and  Plains  Track 
(Fig.  21)  overlaps  four  less  extensive  herpeto- 
faunal tracks:  the  Peninsular  Range  Track,  the 
Colorado  Desert  Track,  the  Sonoran  Desert  Track 
(encompassing  parts  of  northern  Sinaloa,  Sono- 
ra, and  southern  Arizona),  and  the  Chihuahuan 
Desert  Track  (including  much  of  Mexico's  cen- 
tral plateau— see  Morafka  1977).  Each  of  these 
tracks  has  an  endemic  herpetofauna,  however 
these  tracks  also  share  many  species  and  species- 
groups.  As  here  employed,  the  term  North  Amer- 
ican Desert  and  Plains  Track  encompasses  the 
combined  areas  and  fauna  common  to  the  Pen- 
insular Range  Deserts,  Colorado  Desert,  Sono- 
ran Desert,  and  Chihuahuan  Desert  generalized 
tracks,  and  extends  into  the  high  deserts  and  plains 
of  the  southwest,  west,  and  central  North  Amer- 
ica. This  concept  of  an  older  (relative  to  the  true 
desert  tracks  listed  above),  more  comprehensive 
arid  regions  track  is  consistent  with  Morafka's 
(1977)  concept  of  a  Miocene-Pliocene  "Moha- 


64 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


via"  Biota,  and  the  Desert  and  Plains  super- 
province  of  Savage  (1960). 

The  fauna  of  the  North  American  Desert  and 
Plains  Track  occurs  within  all  of  both  Savage's 
and  Murphy's  herpetofaunal  areas  of  Baja  Ca- 
lifornia. Elements  of  this  ubiquitous  fauna  ap- 
pear to  have  populated  the  Baja  California  Re- 
gion by  two  distinct  means: 

(1)  The  majority  of  these  lineages  (Table  6, 
North  American  Desert  and  Plains  Track,  group 
A)  apparently  accompanied  the  progenitors  of 
the  lineages  of  the  Peninsular  Range  Track,  hav- 
ing vicariated  on  prepeninsular  and  insular  land 
masses  as  they  were  separated  from  coastal  Mex- 
ico in  the  mid-Miocene.   These  species  and 
species-groups  are  widely  distributed  on  the  "old" 
or  "deep-water"  islands  of  the  Baja  California 
Region  and  they  all  demonstrate  relatively  high 
degrees  of  morphological  differentiation  within 
monophyletic  lineages  on  the  peninsula  and  as- 
sociated islands  (Soule  and  Sloan  1966;  Savage 
1967;  Ballinger  and  Tinkle  1972;  Case  1975; 
Murphy  \983a,  b;  Murphy  and  Ottley  1983, 
1984). 

(2)  Other  elements  (Table  6,  North  American 
Desert  and  Plains  Track,  group  B)  dispersed 
southward  along  the  peninsula  after  Pleistocene 
to  Recent  climatic  changes,  which  promoted  cor- 
ridors of  xeric  habitats  at  the  northern  end  of  the 
Peninsular  Ranges,  allowing  for  diffusion  south- 
ward. Evidence  for  this  diffusion  hypothesis  are 
the  relatively  low  levels  of  morphological  differ- 
entiation among  all  peninsular  and  insular  pop- 
ulations of  each  lineage,  and  their  exclusive  oc- 
currence on  land-bridged  islands  of  the  Gulf  of 
California  and  on  some  apparently  recently  iso- 
lated land-bridged  islands  of  the  Pacific  side  of 
the  peninsula  (Soule  and  Sloan  1966;  Savage 
1967;  Ballinger  and  Tinkle  1972;  Case  1975; 
Murphy  19836;  Murphy  and  Ottley  1983,  1984). 
These  recently  invading  ubiquitous  xeric  forms 
could  be  considered  as  a  distinct,  overlapping 
generalized  track  analogous  to  the  situation  with 
the  earlier  Old  Northern  Element  and  the  more 
recent,  overlapping  Holarctic  Element  on  the  Pa- 
cific Northwest  Track.  These  more  recent  lin- 
eages might  be  considered  a  Holxeric  Element 
of  the  Desert  and  Plains  Track. 

The  hypothesis  for  the  evolution  of  the  North 
American  Desert  and  Plains  Fauna,  proposed  by 
Savage  ( 1 960),  incorporates  three  distinct  centers 
of  evolution  of  xerophilic  herpetofauna  of  North 


America.  Dispersal  and  genetic  introgression  be- 
tween these  areas  since  the  late  Pleistocene  and 
possibly  during  glacial  minima  has  resulted  in 
the  present,  relatively  continuous  distribution  of 
much  of  this  fauna.  The  endemic  xerophilic  her- 
petofaunas  of  the  Peninsular  Range  Track,  the 
Sonoran  Desert  Track,  and  the  Chihuahuan  Des- 
ert Track  appear  to  be  comprised  of  those  forms 
with  a  competitive  disadvantage  in  xeric  habi- 
tats, such  as  restricted  niches,  or  lower  vagility, 
or  both. 

SUMMARY 

Theories  of  the  historical  biogeography  of  the 
the  herpetofauna  of  the  Baja  California  Region 
have  developed  in  conjunction  with  the  growing 
geofloral  literature  and  the  development  of  a  new 
paradigm  in  geology,  that  of  plate  tectonics.  Past 
syntheses  by  Schmidt  ( 1 922),  Savage  ( 1 960),  and 
Murphy  (1983a)  reflect  this  evolution  of  ideas. 
Five  primary  historical  patterns  or  generalized 
tracks  are  proposed  to  describe  the  interaction 
of  historical  faunal  assemblages  that  produced 
contemporary  distributional  patterns:  the  Pacific 
Northwest  Track,  the  Madrean  Track,  the  Pen- 
insular Range  Track,  the  Colorado  Desert  Track, 
and  the  North  American  Desert  and  Plains  Track. 
Vicariance  of  peninsular  land  masses  from  west- 
ern Mexico  as  a  result  of  Miocene-Pliocene  tec- 
tonic activity  is  the  most  significant  process 
shaping  extant  patterns.  In  situ  evolution  during 
the  Pliocene,  Pleistocene,  and  Recent  epochs,  as 
well  as  radiations  northward  from  these  southern 
vicariated  areas,  and  biotic  dispersal— both  dif- 
fusion and  secular  migration  (Pielou  1 979)— from 
proximate  regions  to  the  north  and  east  played 
major  secondary  roles  influencing  present  dis- 
tributional patterns. 

RESUMEN 

Conduje  un  analisis  ecogeografico  de  la  her- 
petofauna de  la  region  de  San  Pedro  Martir  en 
Baja  California  Norte,  Mexico.  Se  analizaron  mas 
de  3,000  archives  municipales  conseguidos  de  la 
literatura,  los  museos,  y  investigaciones  en  el 
campo,  por  medio  de  una  matriz  de  formaciones 
ecogeograficas  basadas  en  los  climas  regionales, 
la  fisiografia,  y  la  vegetation  existente.  Metodos 
numericos  y  intuitivos  de  analisis  biografico  in- 
dican  que  las  65  especies  que  se  encuentran  ocu- 
rren  en  siete  distintos  modelos  de  distribution. 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


65 


Se  analizan  estos  modelos  contemporaneos  en 
un  contexto  geografico  mas  amplio  para  encon- 
trar  indicaciones  de  sus  origenes  historicos. 

Se  han  desarrollado  las  teorias  de  la  biogeo- 
grafia  historica  de  la  herpetofauna  de  la  region 
de  Baja  California  en  junto  con  la  creciente  li- 
teratura  geofloral  y  el  desarrollo  de  un  nuevo 
paradigmo  de  la  geologia— el  de  "plate  tecto- 
nics." Sintesis  anteriores  por  Schmidt  (1922), 
Savage  (1960),  y  Murphy  (1983a)  reflejan  esta 
evolucion  de  ideas.  Se  propone  cinco  patrones 
historicos,  o  "generalized  tracks"  (Croizat  1964) 
para  describer  la  interaction  de  "faunal  assem- 
blages" historicos  que  producen  los  modelos  de 
distribucion  comtemporaneos:  el  "Pacific 
Northwest  Track,"  el  "Madrean  Track,"  el  "Pe- 
ninsular Range  Track,"  el  "Colorado  Desert 
Track,"  y  el  "North  American  Desert  and  Plains 
Track."  La  "vicariance"  (Croizat  et  al.  1974)  de 
tierras  peninsulares  del  oeste  de  Mexico  como 
resultado  de  actividad  tectonica  es  al  proceso 
mas  significante  en  el  porceso  de  la  formation 
de  los  modelos  que  existen  hoy  dia.  Evolucion 
en  situ  durante  las  epocas  Pliocena,  Pleistocena, 
y  Reciente,  ademas  de  radiaciones  hacia  el  norte 
de  estas  areas  "vicariated"  del  sur,  y  "biotic  dis- 
persal" (diffusion  and  secular  migration— Pielou 
1979)  de  las  regiones  vecinas  del  norte  y  el  este 
tomaron  importantes  papeles  secundarios  que 
influyeron  los  modelos  de  distribucion  que  esis- 
ten  ahora. 


LITERATURE  CITED 

ADEST,  G.  A.  1977.  Genetic  relationships  in  the  genus  Uma 
(Iguanidae).  Copeia  1977(l):47-52. 

.  1987.  Genetic  differentiation  among  populations  of 

the  zebratail  lizard,  Callisaurus  draconoides  (Sauria:  Igua- 
nidae). Copeia  1987(4):854-859. 

ALLEN,  C.  R.  1957.  San  Andreas  Fault  zone  in  San  Gorgonio 
Pass,  southern  California.  Geol.  Soc.  Am.  Bull.  68:315-350. 

ALLEN,  C.  R.,  L.  T.  SILVER,  AND  F.  G.  STEHLI.  1956.  Agua 
Blanca  Fault— a  major  transverse  structure  of  northern  Baja 
California,  Mexico.  Geol.  Soc.  Am.  Bull.  67(Dec):1664. 

ALLISON,  E.  C.  1964.  Geology  of  areas  bordering  the  Gulf  of 
California.  Pp.  3-29  in  Marine  geology  of  the  Gulf  of  Cal- 
ifornia. T.  H.  Van  Andel  and  G.  G.  Shor,  Jr.,  eds.  American 
Association  of  Petroleum  Geologists,  Memoirs  3. 

ALVAREZ,  T.  AND  P.  HUERTA.  1974.  Nuevo  registro  de  Cro- 
talus  atrox  para  la  peninsula  de  Baja  California.  Rev.  Soc. 
Mexicana  Hist.  Nat.  35:1 13-1 15. 

ASCHMAN,  H.  1959.  The  Central  Desert  of  Baja  California: 
demography  and  ecology.  Ibero-Amer.  42.  University  of 
California  Press,  Berkeley.  282  pp. 

ATWATER,  T.    1970.    Implications  of  plate  tectonics  for  the 


Cenozoic  tectonic  evolution  of  western  North  America.  Geol. 
Soc.  Am.  Bull.  81:3513-3535. 

ATWATER,  T.  AND  P.  MOLNAR.  1973.  Relative  motion  of  the 
Pacific  and  North  American  plates  deduced  from  sea-floor 
spreading  in  the  Atlantic,  Indian  and  South  Pacific  oceans. 
Stanford  Univ.  Publ.  Geol.  Sci.  11:136-148. 

AUFFENBURG,  W.  AND  M.  W.  MiLSTEAD.  1965.  Reptiles  in 
the  Quaternary  of  North  America.  Pp.  557-568  in  The  Qua- 
ternary of  the  United  States.  H.  E.  Wright,  Jr.  and  D.  G. 
Frey.  eds.  Princeton  University  Press,  Princeton,  New  Jer- 
sey. 

AXELROD,  D.  I.  1937.  A  Pliocene  flora  from  the  Mount  Eden 
Beds,  southern  California.  Carnegie  Inst.  Washington  Publ. 
476:125-153. 

.  1950.  Further  studies  of  the  Mount  Eden  Flora, 

southern  California.  Carnegie  Inst.  Washington  Publ.  590: 
73-117. 

.    1966.  The  Pleistocene  Soboba  flora  of  southern  Cal- 


ifornia. Univ.  Calif.  Publ.  Geol.  Sci.  60:1-109. 

.    1967.  Geologic  history  of  the  California  insular  flora. 

Pp.  267-3 1 5  in  Proceedings  of  the  symposium  on  the  biology 
of  the  California  Islands.  R.  N.  Philbrick,  ed.  Santa  Barbara 
Botanic  Gardens,  Santa  Barbara,  California. 

.    1975.   Evolution  and  biogeography  of  the  Madrean- 

Tethyan  sclerophyll  vegetation.  Ann.  Mo.  Bot.  Card.  62(2): 
280-334. 

.  1979.  Age  and  origin  of  the  Sonoran  Desert  vege- 
tation. Occas.  Pap.  Calif.  Acad.  Sci.  132:1-74. 

BAILEY,  J.  W.  1928.  A  revision  of  the  lizards  of  the  genus 
Ctenosaura.  Proc.  U.S.  Natl.  Mus.  73(2733):  1-55. 

BAIRD,  S.  F.  1854.  Description  of  new  genera  and  species  of 
North  American  frogs.  Proc.  Acad.  Nat.  Sci.  Phila.  7(2):59- 
62. 

.    1858.  Description  of  new  genera  and  species  of  North 

American  lizards  in  the  museum  of  the  Smithsonian  Insti- 
tution. Proc.  Acad.  Nat.  Sci.  Phila.  7:253-256. 

— .  1859.  Reptiles  of  the  boundary.  In  Vol.  2,  Pt.  2, 
United  States  and  Mexican  Boundary  Survey,  U.S.  34th 
Congress,  1st  Session,  Exec.  Doc.  108:1-35. 

BAIRD,  S.  F.  AND  C.  GIRARD.  1852a.  Descriptions  of  new 
species  of  reptiles  collected  by  the  U.S.  Exploring  Expedition 
under  the  command  of  Capt.  Charles  Wilkes,  U.S.N.  Proc. 
Acad.  Nat.  Sci.  Phila.  6:174-177. 

.  18526.  Reptiles  in  Stanbury,  II.  Exploration  and  sur- 
vey of  the  Valley  of  the  Great  Salt  Lake  of  Utah.  App.  C.: 
336-353. 

.  1853.  Catalog  of  North  American  reptiles  and  am- 
phibians in  the  museum  of  the  Smithsonian  Institution.  Part 
I.  Serpentes.  Smithson.  Misc.  Collect.  2(5):  1-1 72. 

BALLINGER,  R.  E.  AND  D.  W.  TINKLE.  1972.  Systematics  and 
evolution  of  the  genus  Uta  (Sauria:  Iguanidae).  Misc.  Publ. 
Mus.  Zool.  Univ.  Mich.  (145):277-279. 

BANTA,  B.  AND  A.  LEVITON.  1963.  Remarks  upon  Arizona 
elcgans  pacata.  Herpetologica  18(4):277-279. 

BANTA,  B.  AND  W.  W.  TANNER.  1968.  The  systematics  of 
Crotaphytus  wizlizeni,  the  leopard  lizards  (Sauria,  Iguani- 
dae). Part  II.  A  review  of  the  status  of  the  Baja  California 
peninsular  populations  and  a  description  of  a  new  subspecies 
from  Cedros  Island.  Great  Basin  Nat.  28(4):  182-1 94. 

BARTHOLOMEW,  G.  A.  1958.  The  role  of  physiology  in  the 
distribution  of  terrestrial  vertebrates.  Pp.  81-95  in  Zooge- 
ography. C.  L.  Hubbs,  ed.  Am.  Assoc.  Advance.  Sci.,  Wash- 
ington, D.C. 


66 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


BERGGREN,  W.  A.  AND  J.  A.  VAN  COUVERING.  1 974.  The  late 
Neogene.  Palaeogeogr.  Palaeoclimatol.  Palaeoecol.  16:1-216. 

BEZY,  R.  L.  1972.  Karyotypic  variation  and  evolution  of  the 
lizard  family  Xantusiidae.  Los  Angeles  Co.  Mus.  Contr.  Sci. 
227:1-29. 

BEZY,  R.  L.,  G.  C.  GORMAN.  Y.  L.  KIM,  AND  J.  W.  WRIGHT. 

1977.  Chromosomal  and  genetic  divergence  in  the  fossorial 
lizards  of  the  family  Anniellidae.  Syst.  Zool.  26(1):57-71. 

BEZY,  R.  L.  AND  J.  W.  SITES,  JR.  1987.  A  preliminary  study 
of  allozyme  evolution  in  the  lizard  family  Xantusiidae.  Her- 
petologica  43(3):280-292. 

BLAINVILLE,  M.  H.  D.  DE  1835.  Description  de  quelques 
especes  de  reptiles  de  la  California  precedee  de  1'analyse  d'un 
systeme  general  d'erpetologie  et  d'amphibiologie.  Nouv.  Ann. 
Mus.  d'Hist.  Nat.  4:232-296. 

BLANCHARD,  F.  N.  1923.  Comments  on  ring-necked  snakes 
(genus  Diadophis),  with  diagnosis  of  new  forms.  Occ.  Pap. 
Mus.  Zool.  Univ.  Mich.  142:1-9. 

.  1924.  A  new  snake  of  the  genus  Arizona.  Occ.  Pap. 

Mus.  Zool.  Univ.  Mich.  150:1-5. 

BOGERT,  C.  M.  1939.  A  study  of  the  genus  Salvadora,  the 
patchnose  snakes.  Publ.  Univ.  Calif.  Los  Angeles  Biol.  Sci. 
1:177-236. 

.  1945.  Two  additional  races  of  the  patch-nosed  snake, 

Salvadora  hexalepis.  Am.  Mus.  Nat.  Hist.  Novit.  1285:1- 
14. 

BOSTIC,  D.  L.  1968.  Thermal  relations,  distributions,  and 
habitat  of  Cnemidophorus  labialis  (Sauria:  Teiidae).  Trans. 
San  Diego  Soc.  Nat.  Hist.  15(3):21-30. 

.  1971.  Herpetofauna  of  the  Pacific  coast  of  north 

central  Baja  California,  Mexico,  with  a  description  of  a  new 
subspecies  of  Phyllodactylus  xanti.  Trans.  San  Diego  Soc. 
Nat.  Hist.  16(10):237-264. 

BRAME,  A.  H.,  JR.  AND  K.  F.  MURRAY.  1968.  Three  new 
slender  salamanders  (Batrachoseps}  with  a  discussion  of  re- 
lationships and  speciation  within  the  genus.  Bull.  Los  An- 
geles Co.  Mus.  Nat.  Hist.  4:1-33. 

BURT,  C.  E.  1931.  A  study  of  the  teiid  lizards  of  the  genus 
Cnemidophorus  with  special  reference  to  their  phylogenetic 
relationships.  Bull.  U.S.  Natl.  Mus.  154:1-286. 

BURY,  R.  B.  1970.  Clemmys  marmorata.  Cat.  Am.  Amphib. 
Rept.  100.1-100.3. 

.  1983.  Geographic  distribution.  Annie/la  nigra  ar- 

gentea.  Herpetol.  Rev.  14(3):83-84. 

CADLE,  J.  E.  1984.  Molecular  systematics  of  neotropical  xe- 
nodontine  snakes:  II.  Central  American  xenodontines.  Her- 
petologica40(l):21-30. 

CAMP,  C.  L.  1915.  Batrachoseps  major  and  Bufo  cognatus 
californicus,  new  Amphibia  from  southern  California.  Univ. 
Calif.  Publ.  Zool.  12(12):327-334. 

CASE,  T.  J.  1975.  Species  numbers,  density  compensation, 
and  colonizing  ability  of  lizards  on  islands  in  the  Gulf  of 
California.  Ecology  56(1):3-18. 

CHAMBERS,  K.  L.  1955.  A  collection  of  plants  from  the  eastern 
flank  of  the  Sierra  San  Pedro  Martir,  Baja  California.  Dudley 
Herb.  Stanford  Univ.,  Contr.  4(8):323-330. 

CHEETHAM,  A.  H.  AND  J.  E.  HAZEL.  1969.  Binary  (presence- 
absence)  similarity  coefficients.  J.  Paleontol.  43: 1 1 30-1 1 36. 

COLE,  C.  J.  AND  L.  M.  HARDY.  1981.  Systematics  of  North 
American  colubrid  snakes  related  to  Tantilla  planiceps 
(Blainville).  Bull.  Am.  Mus.  Nat.  Hist.  17(3):  199-284. 

COLLINS,  J.  T.,  J.  E.  HUHEEY,  J.  L.  KNIGHT,  AND  H.  M.  SMITH. 

1978.  Standard  common  and  scientific  names  for  North 


American  amphibians  and  reptiles.  S.S.A.R.  Misc.  Publ. 

Herpetol.  Circ.  No.  7,  36  pp. 
CONANT,  R.    1969.   A  review  of  the  water  snakes  of  the  genus 

Natrix  in  Mexico.  Bull.  Am.  Mus.  Nat.  Hist.  142(1):  1-140. 
.    1975.    A  field  guide  to  the  reptiles  and  amphibians 

of  eastern  and  central  North  America.  Houghton-Mifflin, 

Cambridge,  Massachusetts.  429  pp. 
COPE,  E.  D.    1861.   Contributions  to  the  ophiology  of  Lower 

California,  Mexico,  and  Central  America.  Proc.  Acad.  Nat. 

Sci.  Phila.  13:292-306. 

.    1 866.  Fourth  contribution  to  the  herpetology  of  trop- 
ical America.  Proc.  Acad.  Nat.  Sci.  Phila.  18:123-132. 
.     1867.    On  Reptilia  and  Batrachia  of  the  Sonoran 

Province  of  the  Nearctic  Region.  Proc.  Acad.  Nat.  Sci.  Phila. 

18:300-314. 
.    1868.    Observations  on  some  specimens  of  Verte- 


brata  presented  by  Wm.  M.  Gabb,  of  San  Francisco,  which 
were  procured  by  him  in  western  Nevada  and  the  northern 
part  of  Lower  California.  Proc.  Acad.  Nat.  Sci.  Phila.  20:2. 
.  1875.  Checklist  of  North  American  Batrachia  and 


Reptilia.  Bull.  U.S.  Natl.  Mus.  1:1-104. 
.    1892a.    A  synopsis  of  the  species  of  the  teiid  genus 


Cnemidophorus.  Trans.  Am.  Philos.  Soc.  17(l):27-52. 
.    1892fc.    A  critical  review  of  the  characters  and  vari- 


ations of  the  snakes  of  North  America.  Proc.  U.S.  Natl. 
Mus.  14(882):589-694. 
.    1895.    On  some  new  North  American  snakes.  Am. 


Nat.  29:676-680. 

.  1 896.  On  two  species  of  lizards  from  southern  Cal- 
ifornia. Am.  Nat.  30:676-680. 

CRAW,  R.  C.  1983.  Panbiogeography  and  vicariance  cladis- 
tics:  are  they  truly  different?  Syst.  Zool.  32(4):43 1-438. 

CRAW,  R.  C.  AND  P.  WESTON.  1984.  Panbiogeography:  a 
progressive  research  program?  Syst.  Zool.  33(1):  1-1 3. 

CROIZAT,  L.  1964.  Space,  time,  form:  the  biological  synthesis. 
Published  privately  by  author.  Caracas,  Venezuela.  88 1  pp. 

CROIZAT,  L.,  G.  NELSON,  AND  D.  E.  ROSEN.  1974.  Centers  of 
origin  and  related  concepts.  Syst.  Zool.  23(3):265. 

CROSS,  J.  K.  1970.  The  shovel-nosed  snake  (Chionactis  oc- 
cipitalis)  in  Baja  California.  Herpetologica  26(1):  134- 140. 

CUNNINGHAM,  J.  D.  1962.  Observations  on  the  natural  his- 
tory of  the  California  Toad,  Bufo  californicus  Camp.  Her- 
petologica 17(4):255-260. 

.  1964.  Observations  on  the  ecology  of  the  canyon 

treefrog,  Hyla  californiae.  Herpetologica  20(1):55-61. 

DICE,  L.  R.  1943.  The  biotic  provinces  of  North  America. 
University  of  Michigan,  Ann  Arbor.  79  pp. 

DIXON,  J.  R.  1964.  The  systematics  and  distribution  of  the 
lizard  genus  Phyllodactylus  in  North  and  Central  America. 
New  Mexico  State  Univ.  Sci.  Bull.  64(1):  1-39. 

.  1969.  Phyllodactylus  xanti.  Cat.  Am.  Amphib.  Rept. 

79.1-79.2. 

DOWLING,  H.  G.  AND  J.  V.  JENNER.  1987.  Taxonomy  of 
American  xenodontine  snakes.  II.  The  status  and  relation- 
ships of  Pseudoleptodeira.  Herpetologica  43(2):  190-200. 

DUELLMAN,  W.  E.  1 958.  A  monographic  study  of  the  colubrid 
snake  genus  Leptodeira.  Bull.  Am.  Mus.  Nat.  Hist.  1 14(1): 
1-52. 

.  1970.  The  hylid  frogs  of  Middle  America.  Monogr. 

Mus.  Nat.  Hist.,  Univ.  Kansas  1:1-753. 

DURHAM,  J.  W.  AND  E.  C.  ALLISON.  1960.  The  geologic  his- 
tory of  Baja  California  and  its  marine  fauna.  Syst.  Zool.  9: 
47-9 1 . 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


67 


EARDLEY,  A.  J.  1951.  Structural  geology  of  North  America. 
Harper,  New  York.  624  pp. 

FLINT,  R.  F.  1971.  Glacial  and  Quaternary  geology.  John 
Wiley  and  Sons,  Inc.,  New  York.  892  pp. 

FRICK,  C.  1933.  New  remains  of  trilophodont-tetrabelodont 
mastodons.  Bull.  Am.  Mus.  Nat.  Hist.  59:505-652. 

FRITTS,  T.  H.,  H.  L.  SMELL,  AND  R.  L.  MARTIN.  1982.  An- 
arbylus  switaki  Murphy:  an  addition  to  the  herpetofauna  of 
the  United  States  with  comments  on  relationships  with  Co- 
leonyx.  J.  Herpetol.  16(l):39-52. 

FROST,  D.  R.  1 983.  Sonora  semiannulata.  Cat.  Am.  Amphib. 
Rept.  333.1-333.4. 

GASTIL,  R.  AND  W.  JENSK.Y.  1973.  Evidence  for  strike-slip 
displacement  beneath  the  trans-Mexican  volcanic  belt.  Stan- 
ford Univ.  Publ.  Geol.  Sci.  1 1:171-180. 

GASTIL,  R.,  R.  P.  PHILLIPS,  AND  E.  C.  ALLISON.  1975.  Re- 
connaissance geology  of  the  state  of  Baja  California.  Geol. 
Soc.  Am.  Mem.  14.  170  pp. 

GASTIL,  R.,  R.  P.  PHILLIPS,  AND  R.  RODRIGUEZ-TORRES.  1972. 
The  reconstruction  of  Mesozoic  California.  Twenty-fourth 
International  Geological  Congress  1972(3):2 17-229. 

GATES,  G.  O.  1968.  Geographical  distribution  and  character 
analysis  of  the  iguanid  lizard  Sauromalus  obesus  in  Baja 
California,  Mexico.  Herpetologica  24(4):285-288. 

GEHLBACH,  F.  R.  1971.  Lyre  snakes  of  the  Trimorphodon 
biscutatus complex:  a  taxonomic  resume.  Herpetologica  27(2): 
200-211. 

GOOD,  D.  A.  1988.  Allozyme  variation  and  phylogenetic 
relationships  among  the  species  of  Elgaria  (Squamata,  An- 
guidae).  Herpetologica  44(2):  154- 162. 

GORMAN,  G.  C.  1965.  The  distribution  of  Lichanura  trivir- 
gata  and  the  status  of  the  species.  Herpetologica  21(4):283- 
287. 

GRAY,  J.  E.  1852.  Descriptions  of  several  new  genera  of 
reptiles,  principally  from  the  collection  of  H.  M.  S.  Herald. 
Ann.  Mag.  Nat.  Hist.  2(10):437^40. 

HALL,  W.  P.  1973.  Comparative  population  cytogenetics, 
speciation,  and  evolution  in  the  iguanid  lizard  genus  Sce- 
loporus.  Ph.D.  Dissertation.  Harvard  University,  Cam- 
bridge, Massachusetts. 

HALLOWELL,  E.  1852.  Descriptions  of  new  species  of  reptiles 
inhabiting  North  America.  Proc.  Acad.  Nat.  Sci.  Phila.  6: 
177-182. 

.  1853.  On  some  new  reptiles  from  California.  Proc. 

Acad.  Nat.  Sci.  Phila.  6(7):236-238. 

.    1854.    Descriptions  of  new  reptiles  from  California. 


Proc.  Acad.  Nat.  Sci.  Phila.  7:91-97. 

HASTINGS,  J.  R.  AND  R.  R.  HUMPHREY,  EDS.  1969.  Clima- 
tological  data  and  statistics  for  Baja  California.  Technical 
reports  on  the  meteorology  and  climatology  of  arid  regions, 
no.  18.  Univ.  Arizona  Atmos.  Phys.,  Tucson. 

HASTINGS,  J.  R.  AND  R.  M.  TURNER.  1965.  Seasonal  precip- 
itation regimes  in  Baja  California,  Mexico.  Geografiska  An- 
naler  47  Ser.  A:204-223. 

HENDERSON,  D.  A.  1960.  Geography  of  the  Sierra  Juarez  and 
San  Pedro  Martir,  Baja,  California,  Mexico.  Calif.  Geogr.  1: 
21-28. 

HEYER,  W.  R.  1967.  A  herpetofaunal  study  of  an  ecological 
transect  through  the  Cordillera  de  Tilaran,  Costa  Rica.  Co- 
peia  1967(2):259-271. 

HUNSAKER,  D.  1965.  The  ratsnake  Elaphe  rosaliae  in  north- 
ern Baja  California.  Herpetologica  21:71-72. 


JACCARD,  P.  1902.  Lois  de  distribution  florale  dans  la  zone 
alpine.  Bull.  Soc.  Vaudoise  Sci.  Nat.  38:69-130. 

JONES,  K.  B.  1 985.  Eumecesgilberti.  Cat.  Am.  Amphib.  Rept. 
372.1-372.3. 

JONES,  L.  C.  1981.  Geographic  distribution:  Crotalus  viridis 
helleri.  Herpetol.  Rev.  12(2):65. 

KARIG,  D.  E.  AND  W.  JENSKY.  1972.  The  proto  Gulf  of  Cal- 
ifornia. Earth  Planet.  Sci.  Lett.  17:169-174. 

KENNICOTT,  R.  1860.  Descriptions  of  new  species  of  North 
American  serpents  in  the  museum  of  the  Smithsonian  In- 
stitution, Washington.  Proc.  Acad.  Nat.  Sci.  Phila.  12:328- 
338. 

KIM.  Y.  L,  G.  C.  GORMAN,  T.  J.  PAPENFUSS,  AND  A.  K. 
ROYCHOUDHURY.  1976.  Genetic  relationships  and  genetic 
variation  in  the  amphisbaenian  genus  Bipes.  Copeia  1976(1): 
120-124. 

KING,  P.  B.  1958.  Evolution  of  modern  surface  features  in 
western  North  America.  Am.  Assoc.  Advance.  Sci.  Publ.  5 1 : 
3-60. 

.  1959.  The  evolution  of  North  America.  Princeton 

University  Press,  Princeton,  New  Jersey.  189  pp. 

KLAUBER,  L.  M.  1924.  Notes  on  the  distribution  of  snakes 
in  San  Diego  County,  California.  Bull.  Zool.  Soc.  San  Diego 
1:1-23. 

.  1931a.  A  new  subspecies  of  the  California  boa  with 

notes  on  the  genus  Lichanura.  Trans.  San  Diego  Soc.  Nat. 
Hist.  6(20):305-318. 

.    1931/7.    Notes  on  the  worm  snakes  of  the  southwest 


with  descriptions  of  two  new  subspecies.  Trans.  San  Diego 
Soc.  Nat.  Hist.  6(23):333-352. 
.    1932.    Notes  on  the  silvery  footless  lizard,  A nniella 


pulchra.  Copeia  1932(l):4-6. 

.   1933.  Notes  on  Lichanura.  Copeia  1933(4):2 14-2 15. 

.     1935.    Phvllorhvnchus,  the  leaf-nosed  snake.  Bull. 


Zool.  Soc.  San  Diego  12:1-31. 
.    1936.   Crotalus  mitchellii,  the  speckled  rattlesnake. 


Trans.  San  Diego  Soc.  Nat.  Hist.  8(1 9):  149-1 84. 
.    1940.   The  worm  snakes  of  the  genus  Leptotyphlops 


in  the  United  States  and  northern  Mexico.  Trans.  San  Diego 
Soc.  Nat.  Hist.  9(18):87-162. 
.    1941.    The  long-nosed  snakes  of  the  genus  Rhino- 


cheilus.  Trans.  San  Diego  Soc.  Nat.  Hist.  9(29):289-332. 
-.    1 943.   A  new  snake  of  the  genus  Sonora  from  Lower 


California,  Mexico.  Trans.  San  Diego  Soc.  Nat.  Hist.  10(4): 
69-70. 
.    1944.    The  sidewinder,  Crotalus  cerastes,  with  the 


description  of  a  new  subspecies.  Trans.  San  Diego  Soc.  Nat. 
Hist.  10(8):91-126. 

1945.    The  geckos  of  the  genus  Co/eonyx  with  de- 


scriptions of  new  subspecies.  Trans.  San  Diego  Soc.  Nat. 
Hist.  10(11):  133-2 16. 

-.    1946a.    The  gopher  snakes  of  Baja  California,  with 


descriptions  of  new  subspecies  ofPituophis  catenifer.  Trans. 
San  Diego  Soc.  Nat.  Hist.  1  1(1):  1^10. 

-.    19466.   The  glossy  snakes,  Arizona,  with  descriptions 


of  new  subspecies.  Trans.  San  Diego  Soc.  Nat.  Hist.  10(17): 
311-398. 
.    1949.    Some  new  and  revised  subspecies  of  rattle- 


snakes. Trans.  San  Diego  Soc.  Nat.  Hist.  1 1(6):61-1 16. 
.    1972.    Rattlesnakes,  their  habits,  life  histories,  and 


influences  on  mankind.   University  of  California  Press, 
Berkeley.  1,533  pp. 
LARSON,  R.  L.    1972.    Bathymetry,  magnetic  anomalies,  and 


68 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


plate  tectonic  history  of  the  mouth  of  the  Gulf  of  California. 

Geol.  Soc.  Am.  Bull.  83:3345-3360. 
LEE,  J.C.   1975.  Theautecologyof-Ya/7Ms/a/!m$/zaw/(Sauria: 

Xantusiidae).  Trans.  San  Diego  Soc.  Nat.  Hist.  17(19):259- 

277. 

LEVITON,  A.  E.  AND  W.  W.  TANNER.    1960.    The  generic  al- 
location of  Hypsiglena  slevini  Tanner  (Serpentes:  Colubri- 

dae).  Occ.  Pap.  Calif.  Acad.  Sci.  27:1-7. 
LINSDALE,  J.  M.    1932.   Amphibians  and  reptiles  from  Lower 

California.  Univ.  Calif.  Publ.  Zool.  38(6):345-386. 
LOCKINGTON,  W.  N.    1880.    List  of  California  reptiles  and 

batrachia  collected  by  Mr.  Dunn  and  Mr.  W.  J.  Fisher  in 

1876.  Am.  Nat.  14:290-296. 
LOOMIS,  R.  B.    1965.    The  yellow-legged  frog,  Rana  boylei, 

from  the  Sierra  San  Pedro  Martir,  Baja  California  Norte, 

Mexico.  Herpetologica  21(1):78-80. 
LOOMIS,  R.  B.,  S.  G.  BENNETT,  S.  R.  SANBORN,  C.  H.  BARBOUR, 

AND  H.  WEINER.    1974.   A  handlist  of  the  herpetofauna  of 

Baja  California,  Mexico  and  adjacent  islands.  California  State 

University,  Long  Beach.  1 1  pp. 
LOWE,  C.  H.    1 964.    The  vertebrates  of  Arizona.  Annotated 

check  lists.  University  of  Arizona  Press,  Tucson. 
LOWE,  C.  H.  AND  K.  S.  NORRIS.    1954.    Analysis  of  the  her- 
petofauna of  Baja  California,  Mexico.  Trans.  San  Diego  Soc. 

Nat.  Hist.  12(4):47-64. 
LOWE,  C.  H.,  J.  W.  WRIGHT,  C.  J.  COLE,  AND  R.  L.  BEZY. 

1970.   Chromosomes  and  evolution  of  the  species  group  of 

Cnemidophorus  (Reptilia:  Teiidae).  Syst.  Zool.  19(2):  128- 

141. 
MACARTHUR,  R.  H.    1972.  Geographical  ecology:  patterns  in 

the  distribution  of  species.  Harper  and  Row,  New  York.  269 

pp. 
MAHRDT,  C.  R.    1973.   Geographic  distribution:  Crotaphytus 

wizlizenii  copei.  HISS  News- J.  1:98. 
MAYHEW,  W.  W.    1963.    Biology  of  the  granite  spiny  lizard, 

Sceloporus  orcutti.  Am.  Midi.  Nat.  69(2):3 10-327. 
MAYR,  E.    1978.  Origin  and  history  of  some  terms  in  system- 
atic and  evolutionary  biology.  Syst.  Zool.  27(1):88-104. 
MEDICA,  P.  A.   1975.  Rhinocheilus.  Cat.  Am.  Amphib.  Rept. 

175.1-175.3. 
MEEK,  S.  E.    1905.  An  annotated  list  of  a  collection  of  reptiles 

from  southern  California  and  northern  Lower  California. 

Field  Columb.  Mus.,  Zool.  Ser.  7(1):  1-1 9. 
MEIGS,  P.    1953.    World  distribution  of  arid  and  semi-arid 

homoclimates.  Pp.  202-210  in  Reviews  of  research  on  arid 

zone  hydrology.  Arid  Zone  Prog.  1 .  UNESCO,  Paris,  France. 
.    1 966.   Geography  of  coastal  deserts.  Arid  Zone  Re- 
search 28.  UNESCO,  Paris,  France.  140  pp. 
MERRIAM,  C.  H.     1898.    Life  zones  and  crop  zones  of  the 

United  States.  USDA  Div.  Biol.  Surv.,  Bull.  10. 
MILLER,  A.  H.    1951.    An  analysis  of  the  distribution  of  the 

birds  of  California.  Univ.  Calif.  Publ.  Zool.  50(6):53 1-643. 
MILLER,  C.  M.    1944.    Ecologic  relations  and  adaptations  of 

the  limbless  lizards  of  the  genus  Anniella.  Ecol.  Monogr. 

14(3):27 1-289. 
MILLER,  W.  E.    1977.   Pleistocene  terrestrial  vertebrates  from 

southern  Baja  California.  Geol.  Soc.  Am.  Abstract  9:468. 
— .    1 980.  The  late  Pliocene  Las  Tumas  Local  fauna  from 

southernmost  Baja  California.  J.  Paleontol.  54:762-805. 
MINA,  V.  F.    1957.    Bosquejo  geologico  del  Territorio  de  la 

Baja  California.  Boletin  de  la  Associacion  Mexicana  de  Geo- 

logos  Petroleros  9:141-269. 
MINCH,  J.  C.,  R.  G.  GASTIL,  W.  FINK,  J.  ROBINSON,  AND  A. 

H.  JAMES.    1976.    Geology  of  the  Vizcaino  Peninsula.  Pp. 


136-145  in  Aspects  of  the  geologic  history  of  the  California 
Continental  Borderland.  D.  G.  Howell,  ed.  Pacific  Section 
of  the  American  Association  of  Petroleum  Geologists,  Misc. 
Pubs.  24. 

MOCQUARD,  F.  1 899.  Contribution  a  la  faune  herpetologique 
de  la  Basse  California.  Nouv.  Arch.  Mus.  Nat.  Hist.  4(1): 
297-344. 

MOORE,  D.  G.  1973.  Plate-edge  deformation  and  crustal 
growth.  Gulf  of  California  structural  province.  Geol.  Soc. 
Am.  Bull.  84:1883-1906. 

MORAFKA,  D.  J.  1977.  A  biogeographical  analysis  of  the  Chi- 
huahuan  Desert  through  its  herpetofauna.  W.  Junk  B.  V., 
The  Hague,  Netherlands.  3 1 2  pp. 

MUNZ,  P.  A.  AND  D.  KECK.  1949.  California  plant  com- 
munities. El  Aliso  2(1):87-105. 

.  1959.  A  California  flora.  University  of  California 

Press,  Berkeley,  California.  1,681  pp. 

MURPHY,  R.  W.  1974.  A  new  genus  and  species  of  eu- 
blepharine  gecko  (Sauria:  Gekkonidae)  from  Baja  California. 
Mexico.  Proc.  Calif.  Acad.  Sci.  40:87-92. 

.  1975.  Two  new  blind  snakes  (Serpentes:  Leptotyph- 

lopidae)  from  Baja  California,  Mexico,  with  a  contribution 
to  the  biogeography  of  peninsular  and  insular  herpetofauna. 
Proc.  Calif.  Acad.  Sci.  40(5):93-107. 

.  1976.  The  evolution  of  a  peninsular  and  insular  her- 
petofauna: a  drift  based  alternative  hypothesis.  M.S.  Thesis, 
California  State  University,  San  Francisco.  7 1  pp. 

.  1983a.  Paleobiogeography  and  genetic  differentia- 
tion of  the  Baja  California  herpetofauna.  Occ.  Pap.  Calif. 
Acad.  Sci.  137:1-48. 

.    1983£.    The  reptiles:  origin  and  evolution.  Pp.  130- 


158  in  Island  biogeography  in  the  Sea  of  Cortez.  T.  J.  Case 
and  M.  L.  Cody,  eds.  University  of  California  Press,  Berke- 
ley. 508  pp. 

MURPHY,  R.  W.  AND  J.  R.  OTTLEY.  1983.  A  distributional 
checklist  of  the  reptiles  and  amphibians  on  the  islands  in 
the  Sea  of  Cortez.  Appendix  6.1  and  6.2,  pp.  429-437  in 
Island  biogeography  in  the  Sea  of  Cortez.  T.  J.  Case  and  M. 
L.  Cody,  eds.  University  of  California  Press,  Berkeley.  508 
pp. 

.  1984.  Distribution  of  amphibians  and  reptiles  on 

islands  in  the  Gulf  of  California.  Ann.  Carnegie  Mus.  53: 
207-230. 

MURPHY,  R.  W.  AND  T.  J.  PAPENFUSS.  1979.  Biochemical 
relationships,  identification,  and  variation  of  Phyllodactylus 
unctiisand  Phyllodactylus  paucituberculatus.  Biochem.  Syst. 
Ecol.  8:97-100. 

MURRAY,  K.  F.  1955.  Herpetological  collections  from  Baja 
California.  Herpetologica  1 1(l):33^t8. 

NELSON,  E.  W.  1921.  Lower  California  and  its  natural  re- 
sources. Nat.  Acad.  Sci.  16.  First  memoir.  194  pp. 

NORRIS,  K.  W.  1958.  The  evolution  and  systematics  of  the 
iguanid  genus  Uma  and  its  relation  to  the  evolution  of  other 
North  American  desert  reptiles.  Bull.  Am.  Mus.  Nat.  Hist. 
114:247-326. 

OTTLEY,  J.  R.  AND  L.  E.  HUNT.  1981.  Geographic  distribu- 
tion: Crotalus  viridis  helleri.  Herpetol.  Rev.  12(2):65. 

OTTLEY,  J.  R.  AND  E.  E.  JACOBSEN.  1983.  Pattern  and  col- 
oration of  juvenile  Elaphe  rosaliae,  with  notes  on  natural 
history.  J.  Herpetol.  17(2):189-191. 

PAPENFUSS,  T.  J.  1982.  The  ecology  and  systematics  of  the 
amphisbaenian  genus  Bipes.  Occ.  Pap.  Calif.  Acad.  Sci.  1 36. 
42pp. 

PEABODY,  F.  E.  AND  J.  M.  SAVAGE.    1958.    Evolution  of  the 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA 


69 


Coast  Range  corridor  in  California  and  its  effect  on  the  origin 
and  dispersal  of  living  amphibians  and  reptiles.  Am.  Assoc. 
Advance.  Sci.  Publ.  51:159-186. 

PETERS,  J.  A.  1971.  A  new  approach  in  the  analysis  of  bio- 
geographical  data.  Smithson.  Contr.  Zool.  107:1-28. 

PHELAN,  R.  L.  AND  B.  H.  BRATTSTROM.  1955.  Geographical 
variation  in  Sceloporus  magister.  Herpetologica  1 1(1):1-14. 

PIELOU,  E.  C.  1979.  Biogeography.  J.  Wiley  and  Sons,  New 
York.  351  pp. 

PLATNICK,  N.  I.  AND  G.  NELSON.  1 978.  A  method  of  analysis 
for  historical  biogeography.  Syst.  Zool.  27(1):1-16. 

PRESH,  W.  1969.  Evolutionary  osteology  and  relationships 
of  the  horned  lizard  genus  Phrynosoma  (family  Iguanidae). 
Copeia  1969(2):250-275. 

RAU,  C.  S.  AND  R.  B.  LOOMIS.  1977.  A  new  species  of  Uro- 
saurus  (Reptilia,  Lacertilia,  Iguanidae)  from  Baja  California, 
Mexico.  J.  Herpetol.  ll(l):25-29. 

REEVE,  W.  L.  1952.  Taxonomy  and  distribution  of  the  horned 
lizards  genus  Phrynosoma.  Bull.  Univ.  Kan.  Sci.  34(2):817- 
960. 

ROBINSON,  J.  W.  1972.  Camping  and  climbing  in  Baja.  La 
Siesta  Press,  Glendale,  California.  96  pp. 

ROBINSON,  M.  D.  1973.  Chromosomes  and  systematics  of 
the  Baja  California  whiptail  lizards  Cnemidophorus  hyp- 
erythrus  and  C.  ceralbensis  (Reptilia:  Teiidae).  Syst.  Zool. 
22(l):30-35. 

ROGERS,  T.  L.  AND  H.  S.  FITCH.  1 947.  Variation  in  the  skinks 
of  the  Skiltonianus  group.  Univ.  Calif.  Publ.  Zool.  48:169- 
219. 

ROTRAMEL,  G.  L.  1973.  The  development  and  application 
of  the  area  concept  in  biogeography.  Syst.  Zool.  22:227-232. 

SARICH,  V.  M.  1977.  Rates,  sample  sizes  and  neutrality  hy- 
pothesis for  electrophoresis  in  evolutionary  studies.  Nature 
265:24-28. 

SAS  INSTITUTE,  INC.  1985.  SAS  user's  guide:  statistics,  ver- 
sion 5  edition.  SAS  Institute,  Gary,  North  Carolina.  956  pp. 

SAVAGE,  J.  M.  1952.  Studies  on  the  lizard  family  Xantusiidae. 
I.  The  systematic  status  of  the  Baja  California  night  lizards 
allied  to  Xantusia  vigilis,  with  the  description  of  a  new  sub- 
species. Am.  Midi.  Nat.  48(2):467-479. 

.  1960.  Evolution  of  a  peninsular  herpetofauna.  Syst. 

Zool.  9(3-4):  184-2 12. 

.    1967.    Evolution  of  the  insular  herpetofaunas.  Pp. 


.    1945.    The  chuckwalla,  genus  Sauromalus.  Trans. 

San  Diego  Soc.  Nat.  Hist.  10(15):269-306. 

1953.    Anniella  pulchra  and  A.  geronimensis;  sym- 


219-227  in  Proceedings  of  the  symposium  on  the  biology 

of  the  California  islands.  Santa  Barbara  Botanic  Gardens, 

Santa  Barbara,  California. 
SCHMIDT,  K.  P.    1922.  The  amphibians  and  reptiles  of  Lower 

California  and  the  neighboring  islands.  Bull.  Am.  Mus.  Nat. 

Hist.  46:607-707. 

SCHOENHERR,  A.  A.    1976.   The  herpetofauna  of  the  San  Ga- 
briel Mountains,  Los  Angeles  County,  California.  Special 

Publication  of  the  Southwest  Herpetological  Society,  Feb.  1, 

1976.95pp. 
Scorr,  N.  J.  AND  R.  W.  McDiARMio.    1984.    Trimorphodon 

biscittatus.  Cat.  Am.  Amphib.  Rept.  353.1-353.4. 
SEELIGER,  L.  M.    1945.    Variation  in  the  Pacific  mud  turtle. 

Copeia  1945(3):  150-1 59. 
SEIB,  R.  L.    1 980.   Baja  California:  a  peninsula  for  rodents  but 

not  for  reptiles.  Am.  Nat.  1 15(4):6 13-620. 
SEIFERT,  W.    1980.    Geographical  distribution:  Arizona  ele- 

gans  occipitalis.  Herpetol.  Rev.  1 1(2):39. 
SHAW,  C.  E.    1940.   A  new  species  of  legless  lizard  from  San 

Geronimo  Island,  Lower  California.  Trans.  San  Diego  Soc. 

Nat.  Hist.  9(24):225-227. 


patric  species.  Herpetologica  8(4):  167-1  70. 
SHORT,  L.  L.  AND  R.  CROSSIN.    1967.   Notes  on  the  avifauna 

of  northwestern  Baja  California.  Trans.  San  Diego  Soc.  Nat. 

Hist.  14(20):28  1-300. 
SHREVE,  F.    1934.    Vegetation  of  the  northwestern  coast  of 

Mexico.  Bull.  Torrey  Bot.  Club  61(7):373-380. 

-  .    1936.  The  transition  from  desert  to  chaparral  in  Baja 
California.  Madrono  3:257-320. 

SHREVE,  F.  AND  L.  WIGGINS.    1964.    Vegetation  and  flora  of 

the  Sonoran  Desert.  Stanford  University  Press,  Palo  Alto, 

California.  1,740pp. 
SILVER,  L.  T.,  E.  G.  STEHLI,  AND  C.  R.  ALLEN.    1956.   Lower 

Cretaceous  prebatholith  rocks  of  northern  Baja  California. 

Calif.  Inst.  Tech.,  Div.  Geol.  Sci.,  Contr.  No.  799:1-1  1. 
SIMPSON,  G.  G.    1960.   Notes  on  the  measurement  of  faunal 

resemblance.  Am.  Sci.  258:300-31  1. 
SMITH,  H.  M.    1972.    The  Sonoran  subspecies  of  the  lizard 

Ctenosaura  hemilopha.  Great  Basin  Nat.  32(2):  104-1  1  1. 
SMITH,  H.  M.  AND  R.  L.  HOLLAND.    1971.  Noteworthy  snakes 

and  lizards  from  Baja  California.  J.  Herpetol.  5(l-2):56-69. 
SMITH,  H.  M.  AND  W.  W.  TANNER.   1  972.  Two  new  subspecies 

of  Crotaphytus  (Sauna,  Iguanidae).  Great  Basin  Nat.  32:25- 

34. 

-  .    1974.    A  taxonomic  study  of  the  western  collared 
lizards,  Crotaphytus  collaris  and  Crotaphytus  insularis.  Brig- 
ham  Young  Univ.  Sci.  Bull.,  Biol.  Series  19(4):  1-29. 

SOULE,  M.  AND  A.  J.  SLOAN.  1966.  Biogeography  and  dis- 
tribution of  the  reptiles  and  amphibians  on  islands  in  the 
Gulf  of  California,  Mexico.  Trans.  San  Diego  Soc.  Nat.  Hist. 


STEBBINS,  R.  C.    1954.    Amphibians  and  reptiles  of  western 
North  America.  McGraw-Hill  Co.,  New  York.  536  pp. 

-  .    1966.   A  field  guide  to  western  reptiles  and  amphib- 
ians. Houghton  Mifflin  Co.,  Boston,  Massachusetts.  279  pp. 

-  .    1985.   A  field  guide  to  western  reptiles  and  amphib- 


ians, 2nd  ed.,  revised.  Houghton  Mifflin  Co.,  Boston,  Mas- 
sachusetts. 336  pp. 

STEJNEGER,  L.  1890.  Description  of  a  new  lizard  from  Lower 
California.  Proc.  U.S.  Natl.  Mus.  12:643-644. 

.  1 893a.  Annotated  list  of  the  reptiles  and  batrachians 

collected  by  the  Death  Valley  Expedition  in  1891,  with  de- 
scription of  new  species.  N.  Am.  Fauna  7:159-228. 

.    1893ft.    Diagnosis  of  a  new  California  lizard.  Proc. 


U.S.  Natl.  Mus.  16(944):467. 

-.    1 894.   Description  of  Uta  mearnsi,  a  new  lizard  from 


California.  Proc.  U.S.  Natl.  Mus.  17:589-591. 
STICKEL,  W.  H.    1938.  The  snakes  of  the  genus  Sonora  in  the 

United  States  and  Lower  California.  Copeia  1 938(4):  182- 

190. 
.    1943.   The  Mexican  snakes  of  the  genus  Sonora  and 

Chionactis  with  notes  on  the  status  of  other  colubrid  genera. 

Proc.  Biol.  Soc.  Washington  56:109-128. 
STORER,  T.  I.  AND  R.  L.  USINGER.    1963.   Sierra  Nevada  nat- 
ural history.  University  of  California  Press,  Berkeley.  374 

PP. 
TANNER,  W.  W.    1946.   A  taxonomic  study  of  the  gen  us  Hyp- 

siglena.  Great  Basin  Nat.  5(3-4):25-92. 

.    1953.   A  study  of  taxonomy  and  phylogeny  of  Lam- 

propeltis  pyromelana  Cope.  Great  Basin  Nat.  13(l-2):47- 

66. 


70 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


.    1966a.    The  night  snakes  of  Baja  California.  Trans. 

San  Diego  Soc.  Nat.  Hist.  14(1 5):  189-1 96. 
.    1 966ft.    A  re-evaluation  of  the  genus  Tantilla  in  the 


southwestern  United  States  and  northwestern  Mexico.  Her- 
petologica  22(2):  134-1 52. 

TAYLOR,  E.  H.  1935.  A  taxonomic  study  of  the  cosmopolitan 
scincoid  lizards  of  the  genus  Eumeces  with  an  account  of 
the  distribution  and  relationships  of  its  species.  Kans.  Univ. 
Sci.  Bull.  23:1-643. 

TEVIS,  L.  1944.  Herpetological  notes  from  Lower  California. 
Copeia  1944(1):6-18. 

UDVARDY,  M.  D.  F.  1 969.  Dynamic  zoogeography.  Van  Nos- 
trand  Reinhold  Co.,  New  York.  445  pp. 

U.S.  DEPT.  COMMERCE.  1979.  Operational  navigation  chart 
(ONQCH-22.  Aeronautical  chart  and  information  center, 
St.  Louis,  Missouri. 

VAN  COUVERING,  J.  A.  1 978.  Status  of  late  Cenozoic  bound- 
aries. Geology  6:169. 

VAN  DENBURGH,  J.  1894.  Description  of  three  new  lizards 
from  California  and  Lower  California  with  a  note  on  Phry- 
nosoma  blainvillii.  Proc.  Cal.  Acad.  Sci.  4(2):296-301. 

.  1922.  The  reptiles  of  western  North  America.  Occ. 

Pap.  Calif.  Acad.  Sci.  10:1-1028. 

VAN  DENBURGH,  J.  AND  J.  R.  SLEVIN.  1921.  Preliminary 
diagnosis  of  more  new  species  of  reptiles  from  islands  in  the 
Gulf  of  California.  Proc.  Calif.  Acad.  Sci.  1 1(4):395-398. 

.  1923.  Preliminary  diagnosis  of  four  new  snakes  from 

Lower  California,  Mexico.  Proc.  Calif.  Acad.  Sci.  1(13): 
1-2. 

VAN  DEVENDER,  T.  R.  AND  W.  G.  SPAULDING.  1979.  Devel- 
opment of  vegetation  and  climate  in  the  southwestern  United 
States.  Science  204:701-710. 

WAKE,  D.  B.  1 966a.  The  colubrid  snake  genus  Chionactis  in 
Baja  California,  Mexico.  Copeia  1966(2):364. 

.  1966ft.  Comparative  osteology  and  evolution  of  the 

lungless  salamanders,  family  Plethodontidae.  Mem.  So.  Cal- 
if. Acad.  Sci.  4:1-111. 

WALKER,  J.  M.  AND  H.  L.  TAYLOR.  1968.  Geographical  vari- 
ation in  the  teiid  lizard  Cnemidophorus  hyperythrus.  I.  The 
caeruleus-like  subspecies.  Am.  Midi.  Nat.  80(1):  1-27. 

WASSERMAN,  A.  O.  1970.  Scaphiopus  couchii.  Cat.  Am.  Am- 
phib.  Rept.  85.1-85.4. 

WELSH,  H.  H.,  JR.  1976a.  Ecogeographic  distribution  of  the 
herpetofauna  of  the  Sierra  San  Pedro  Martir  Region,  Baja 


California  Norte,  Mexico.  M.S.  Thesis,  Humboldt  State  Uni- 
versity, Arcata,  California.  169  pp. 

.  1976ft.  Evolutionary  trends  and  zoogeographic  his- 
tory of  the  Baja  California  herpetofauna.  45  pp.  [Unpub- 
lished ms,  mimeo] 

.  1981.  Ecogeographic  distribution  of  the  northern  Baja 

California  herpetofauna  and  its  relationships  with  the  west- 
ern Pacific  herpetofauna.  Abstract,  paper  given  at  American 
Society  of  Ichthylogists  and  Herpetologists  Meeting.  Oregon 
State  University.  June  1981. 

WELSH,  H.  H.,  JR.  AND  R.  B.  BURY.  1984.  Additions  to  the 
herpetofauna  of  the  south  Colorado  Desert,  Baja  California, 
with  comments  on  the  relationships  ofLichanura  trivirgata. 
Herpetol.  Rev  15(2):53-56. 

WIGGINS,  I.  L.  1 944.  Notes  on  the  plants  of  northern  Baja 
California.  Dudley  Herb.  Stanford  Univ.  Contr.  3(9):289- 
312. 

.  1960.  The  origins  and  relationships  of  land  flora.  In 

Symposium:  the  biogeography  of  Baja  California  and  ad- 
jacent seas.  Syst.  Zool.  9(1-3):  148-1 65. 

WILSON,  L.  E.  1973.  Masticophis  flagellum.  Cat.  Am.  Am- 
phib.  Rept.  145.1-145.4. 

WOODFORD,  A.  O.  AND  T.  F.  HARRIS.  1938.  Geological  re- 
connaissance across  Sierra  San  Pedro  Martir,  Baja  Califor- 
nia. Bull.  Geol.  Soc.  Am.  9:1297-1336. 

WYLES,  J.  S.  AND  G.  C.  GORMAN.  1978.  Close  relationship 
between  the  lizard  genus  Sator  and  Sceloporus  utiformis 
(Reptilia,  Lacertilia,  Iguanidae):  electrophoretic  and  im- 
munological  evidence.  J.  Herpetol.  12(3):343-350. 

YANEV,  K.  P.  1978.  Evolutionary  studies  of  the  plethodontid 
salamander  genus  Batrachoseps.  Ph.D.  Dissertation,  Uni- 
versity of  California,  Berkeley. 

YARROW,  H.  C.  1883  (1882).  Descriptions  of  new  species  of 
reptiles  and  amphibians  in  the  United  States  National  Mu- 
seum. Proc.  U.S.  Natl.  Mus.  5(299):438-443. 

YINGLING,  P.  R.  1982.  Lichanura,  L.  trivirgata.  Cat.  Am. 
Amphib.  Rept.  294.1-294.2. 

ZWEIFEL,  R.  G.  1952.  Pattern  variation  and  evolution  of  the 
mountain  king  snake,  Lampropeltis  zonata.  Copeia  1952(3): 
152-168. 

.  1955.  The  ecology  and  systematics  of  the  Ranaboylei 

species  group.  Univ.  Calif.  Publ.  Zool.  54:207-292. 

.    1974.  Lampropeltis  zonata.  Cat.  Am.  Amphib.  Rept. 


174.1-174.4. 


WELSH:  BAJA  CALIFORNIA  HERPETOFAUNA  7 1 

APPENDIX  A.    COLLECTING  STATIONS  IN  THE  SIERRA  SAN  PEDRO  MARTIR  REGION,  BAJA  CALIFORNIA  NORTE,  MEXICO. 


Station 
number 

Elevation 

Primary 
vegetation* 

Name  and  coordinates 

ft 

m 

1 

San  Matias  Pass;  31°19'N,  115°30'W 

3,200 

980 

CBS 

2 

Vallede  Picacho;  31°15'N,  115°35'W 

3,400 

1,004 

CBS 

3 

Rancho  El  Burro;  31°14'N,  1  15°36'W 

4,450 

1,360 

P-J 

4 

Canon  Esperanza  (mouth);  31°12'N,  1  15°27'W 

1,640 

500 

CBS 

5 

Valle  de  San  Felipe  (center);  31°12'N,  1  15°20'W 

1,575 

480 

CBS 

6 

Canon  Copal  (mouth);  3l°09'N,  115°25'W 

1,970 

600 

CBS 

7 

San  Telmo;  31°08'N,  1  16°05'W 

400 

120 

CDS 

8 

Arroyo  Leon  (head);  31°08'N,  1  15031'W 

7,400 

2,260 

Conifer 

9 

San  Antonio  del  Mar;  31°07'N,  116°17'W 

75 

23 

CDS 

10 

Arroyo  de  San  Rafael,  6  mi  E  MS  Rancho;  31°07'N,  1  15°33'W 

5,000 

1,525 

P-J,  Chp 

11 

Mikes  Sky  Rancho;  31°06'N,  1  15°38'W 

4,150 

1,270 

Chp 

12 

Cerro  Venado  Blanco;  31°06'N,  1  15°29'W 

9,325 

2,820 

Conifer 

13 

Canon  Copal  (head);  31°06'N,  1  15°29'W 

7,375 

2,225 

Conifer 

14 

Arroyo  de  San  Rafael,  2  mi  E  MS  Rancho;  31°05'N,  1  15°37'W 

4,350 

1,330 

Chp 

15 

Arroyo  de  San  Rafael  (head);  31°05'N,  1  15°31'W 

7,275 

2,210 

Conifer 

16 

Canon  del  Diablo,  2  mi  W  mouth;  31°05'N,  1  15°24'W 

2,500 

760 

CBS 

17 

Canon  del  Diablo  (mouth);  31°05'N,  1  15°22'W 

2,175 

660 

CBS 

18 

Mikes  Sky  Rancho,  airstrip;  31°04'N,  1  15°38'W 

4,600 

1,400 

Chp 

19 

Rancho  Conception;  3I°04'N,  115°34'W 

4,775 

1,450 

Chp 

20 

Canon  del  Diablo,  4  mi  W  mouth;  31°04'N,  1  15°24'W 

2,675 

810 

P-J/CBS 

21 

Canon  del  Diablo,  6  mi  W  mouth;  31°03'N,  1  15°24'W 

3,150 

960 

P-J/CBS 

22 

Observatorio  Nacional  de  Mexico;  31°02'N,  1  15°27'W 

9,200 

2,800 

Conifer 

23 

Canon  del  Diablo,  8  mi  W  mouth;  31°02'N,  1  15°24'W 

4,000 

1,220 

P-J 

24 

Compamento  Forestal;  31°01'N,  115°28'W 

8,100 

2,460 

Conifer 

25 

Yerba  Buena  Spring;  31°00'N,  1  15°28'W 

8,000 

2,420 

Conifer 

26 

Vallecitos,  SE  end;  30°59'N,  115°28'W 

8,000 

2,420 

Conifer 

27 

Rancho  San  Jose;  30°58'N,  115°25'W 

2,300 

640 

CSS 

28 

Socorro;  30°58'N,  115°40'W 

4,200 

1,280 

CSS/Chp 

29 

Oak  Pasture;  30°58'N,  115°36'W 

5,900 

1,830 

Chp 

30 

Los  Llanitos;  30°58'N,  115°26'W 

7,900 

2,410 

Conifer 

31 

Picacho  Spring;  30°58'N,  1  15°25'W 

8,200 

2,500 

Conifer 

32 

Blue  Bottle  Mt.;  30°58'N,  115°24'W 

9,675 

2,950 

Conifer 

33 

Rancho  Viejo  trail  head,  La  Tasajera  ridge;  30°57'N,  1  15°30'W 

7,900 

2,410 

Conifer 

34 

Llanitos  Camp;  30°57'N,  115°26'W 

7,700 

2,350 

Conifer 

35 

La  Jolla;  30°56'N,  115°35'W 

5,250 

1,600 

Chp 

36 

La  Tasajera  Ridge,  2  mi  NE  R.  Viejo;  30°56'N,  1  15°28'W 

7,000 

2,120 

Conifer 

37 

Rancho  Viejo;  30°55'N,  1  15°29'W 

6,900 

2,200 

Conifer 

38 

San  Antonio  Creek,  1  mi  W  La  Grulla;  30°54'N,  1  15°30'W 

6,750 

2,060 

Conifer 

39 

La  Grulla,  W  end;  30°54'N,  1  15°29'W 

6,800 

2,080 

Conifer 

40 

La  Grulla,  center;  30°54'N,  1  15°27'W 

6,900 

2,100 

Conifer 

41 

La  Encantada  Creek;  30°54'N,  115°26'W 

7,050 

2,150 

Conifer 

42 

Arroyo  San  Antonio,  head;  30°53'N,  1  15°30'W 

6,575 

2,000 

Conifer 

43 

La  Grulla,  SE  end;  30°53'N,  1  15°26'W 

6,950 

2,110 

Conifer 

44 

La  Encantada;  30°53'N,  1  15°25'W 

7,100 

2,160 

Conifer 

45 

Arroyo  Valladares,  Santa  Cruz  Road;  30°52'N,  1  15°43'W 

2,025 

620 

CSS 

46 

Canon  El  Cajon  (mouth);  30°52'N,  1  15°14'W 

2,125 

650 

CBS 

47 

El  Alcatraz;  30°51'N,  1  15°26'W 

6,800 

2,080 

Conifer 

48 

Canon  El  Cajon,  6  mi  W  mouth;  30°51'N,  115°16'W 

4,000 

1,220 

P-J/CBS 

49 

Rancho  San  Antonio;  30°50'N,  1  15°38'W 

1,970 

600 

CSS 

50 

El  Remudero;  30°50'N,  115°22'W 

6,900 

2,200 

Conifer 

51 

Canon  Barrosa/Canon  El  Cajon  divide;  30°50'N,  1  15°17'W 

6,575 

2,000 

P-J 

52 

Cowpod  Camp,  Canon  El  Cajon;  30°50'N,  1  15°16'W 

5,200 

1,580 

P-J 

53 

Canon  Barrosa  (head);  30°49'N,  115°19'W 

7,550 

2,300 

Chp 

54 

Valle  de  Chico  (center);  30°49'N,  1  15°08'W 

1,575 

480 

CBS 

55 

Santa  Rosa;  30°48'N,  115°2r\V 

6,800 

2,080 

Conifer 

56 

Mission  San  Pedro  Martir;  30°47'N,  1  15°24'W 

5,500 

1,675 

Chp 

57 

Santo  Tomas;  30°46'N,  1  15°24'W 

5,900 

1,800 

Chp 

72 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 


APPENDIX  A.    CONTINUED. 


Elevation 

Station 

Primary 

number                                           Name  and  coordinates 

ft 

m 

vegetation* 

58            Arroyo  Alonso  (head);  30°46'N,  1  1  5°20'  W 

6,575 

2,000 

Conifer 

59            Canon  Agua  Caliente,  9  mi  W  mouth;  30°41'N,  1  15°16'W 

3,125 

950 

CBS 

60            Canon  Agua  Caliente,  4  mi  W  mouth;  30°40'N,  1  15°12'W 

2,225 

680 

CBS 

6  1            Canon  Auga  Caliente  (mouth);  30°40'N,  1  1  5°09'  W 

2,200 

670 

CBS 

62            Rancho  Rosarito;  30°28'N,  1  1  5°1  8'  W 

2,750 

840 

CSS/CBS 

63            Rio  de  Rosarito;  30°2  1'N,  1  1  5°22'  W 

2,200 

670 

CDS 

64            San  Juan  de  Dios;  30°1  1'N,  11  5°09'W 

2,175 

660 

CDS 

65            Rancho  El  Metate;  30°1  1'N,  1  15°08'W 

2,200 

670 

CDS 

*  Key  to  vegetations:  Conifer  =  Coniferous  Forest,  CSS  =  Coastal  Sage  Scrub,  CBS 

=  Creosote  Bush  Scrub, 

CDS  =  Central 

Desert  Scrub,  Chp  =  Chaparral,  and  P-J  =  Pinyon-Juniper  Woodland. 

CALIFORNIA  ACADEMY  OF  SCIENCES 

Golden  Gate  Park 
San  Francisco,  California  941 18 


PROCEEDINGS 

OF  THE 

CALIFORNIA  ACADEMY  OF  SCIENCES 


Vol.  46,  No.  2,  pp.  73-81,  5  figs.  December  7,  1988 


THREE  NEW  SPECIES  OF  HOLOGRAPHIS  (ACANTHACEAE) 

FROM  MEXICO 

By 

Thomas  F.  Daniel 

Department  of  Botany,  California  Academy  of  Sciences,  Golden  Gate  Park, 
San  Francisco,  California  94118 


ABSTRACT:  Three  new  species  of  Holographis,  H.  caput-medusae,  If.  leticiana,  and  //.  tolantongensis,  are 
described  from  Mexico.  Diagnostic  features  of  each  species  are  illustrated  and  the  known  distribution  of  the 
three  species  is  mapped.  A  revised  key  to  the  entire  genus,  incorporating  data  and  species  that  have  become 
known  since  the  last  monograph  of  Holographis,  is  presented.  Fifteen  species  are  now  known  for  this  entirely 
Mexican  genus. 

Received  February  16,  1988.  Accepted  May  11,  1988. 

additional  collections  of  previously  recognized 

Holographis  Nees  is  a  small  genus  of  perennial  species  of  Holographis  has  become  available, 

herbs  and  shrubs  endemic  to  Mexico.  Since  my  Some  species  (e.g.,  H.  argyrea,  H.  anisophylla, 

monograph  of  the  genus  (Daniel  1983),  in  which  and  H.  parayand)  are  now  known  from  twice  as 

10  species  were  recognized,  one  species  (H.  pe-  many  or  more  collections  than  they  were  in  1983. 

lorid)  has  been  transferred  into  Holographis  from  Fruiting  collections  have  been  located  for  four 

Stenandrium  Nees  (Daniel  1984)  and  an  addi-  species  (H.  argyrea,  H.  anisophylla,  H.  peloria, 

tional  species  (H.  websteri)  has  been  newly  de-  and  H.  pueblensis)  whose  capsules  and  seeds  were 

scribed  (Daniel  1986).  Continued  botanical  ex-  previously  undescribed.  The  known  geographic 

ploration  in  the  dry  forests  of  Mexico  has  revealed  distributions  of  several  species  (e.g.,  H.  aniso- 

the  existence  of  three  undescribed  and  distinctive  phylla  and  H.  peloria)  have  been  increased.  A 

species,  bringing  the  total  number  of  species  now  key  to  all  species  of  the  genus  is  presented  below 

known  for  the  genus  to  15.  These  three  new  that  includes  the  five  taxa  recognized  in  Holo- 

species  are  described  below.  graphis  since  the  monograph  of  1983  and  that 

Because  of  the  increased  botanical  activity  in  incorporates  information  from   other  species 

Mexico  in  recent  years,  a  significant  number  of  based  on  additional  collections  examined. 

KEY  TO  THE  SPECIES  OF  HOLOGRAPHIS 

1 .  Leaves  opposite  (rarely  subopposite). 

2.  Corolla  yellow  with  maroon  markings,  vertical  during  an  thesis,  the  upper  lip  1-2.5  mm 
long;  stamens  1.5  mm  long  with  thecae  1-1.3  mm  long;  style  4  mm  long;  Balsas  basin  in 

Guerrero. H.  argyrea 

[73] 


74  PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  2 

2.  Corolla  white  or  pinkish,  horizontal  during  anthesis,  the  upper  lip  2.5-5  mm  long;  stamens 
2.8-5.5  mm  long  with  thecae  1.5-2.5  mm  long;  style  5-7.5  mm  long;  plants  occurring  to 
the  north  of  Guerrero. 
3.  Abaxial  surface  of  bracts,  bractlets,  and  calyx  glabrous;  calyx  4-5  mm  long;  corolla 

white;  Tamaulipas H.  tamaulipica 

3.  Abaxial  surface  of  bracts,  bractlets,  and  calyx  pubescent;  calyx  5-10  mm  long;  corolla 
pinkish;  plants  occurring  to  the  south  and  west  of  Tamaulipas. 
4.  Corolla  glabrous  on  external  surface;  calyx  lobes  aristate  at  apex;  staminode  2  mm 

long;  Hidalgo H.  tolantongensis 

4.  Corolla  pubescent  on  external  surface;  calyx  lobes  acute  at  apex;  staminode  0.5-0.8 

mm  long;  Sonoran  Desert  region. H.  virgata 

1 .  Leaves  whorled,  four  per  node  (lower  leaves  of  H.  pueblensis  sometimes  opposite  to  subop- 
posite;  rarely  opposite  on  some,  but  not  all,  shoots  of  individuals  of  H.  ehrenbergiand). 
5.  Leaves  orbicular,  0.8-1.3  times  longer  than  wide,  the  margin  spinose  toothed;  bracts  al- 
ternate; Chihuahuan  Desert  region. H.  ilicifolia 

5.  Leaves  lanceolate  to  ovate  to  elliptic  to  obovate  to  oblanceolate,  1.5-6  times  longer  than 
wide,  margin  entire  to  undulate;  bracts  opposite;  plants  occurring  outside  of  the  Chihuahuan 
Desert  region. 
6.  Corolla  yellow,  the  upper  lip  6-9  mm  long;  stamens  9-1 1  mm  long;  seeds  glabrous. 

7.  Inflorescence  to  4.5  cm  long,  many-flowered;  bracts  5-1 1  mm  long,  1.5-2  mm  wide; 
calyx  7-10.5  mm  long H.  pueblensis 

1.  Inflorescence  to  2  cm  long,  usually  reduced  to  2  flowers;  bracts  1.5-5  mm  long,  0.5- 

1  mm  wide;  calyx  3.5-6.5  mm  long H.  ehrenbergiana 

6.  Corolla  white,  pinkish,  or  reddish  (color  unknown  in  H.  hintonii),  the  upper  lip  1.5- 
5.5  mm  long;  stamens  1.2-6.5  mm  long;  seeds  pubescent  or  with  hairlike  papillae 
(glabrous  in  H.  pelorid). 

8.  Leaves  up  to  13  mm  long;  corolla  6.5-7.5  mm  long,  the  lower  lip  2.5-3.5  mm  long; 
stamens  1.2  mm  long  with  thecae  1  mm  long;  style  1.8-2.5  mm  long;  Yucatan 
Peninsula. H.  websteri 

8.  Leaves  up  to  105  mm  long;  corolla  8-17  mm  long,  the  lower  lip  4-9  mm  long; 
stamens  2.5-6.5  mm  long  with  thecae  1.2-2.6  mm  long;  style  3.5-9.5  mm  long; 
plants  occurring  to  the  west  of  the  Yucatan  Peninsula. 

9.  Bracts  6-1 3  mm  long;  bractlets  6-10.5  mm  long;  corolla  pubescent  with  eglandular 
trichomes  on  external  surface;  seeds  glabrous,  covered  with  stout  papillae  or 
tubercules. 

1 0.  Cauline  trichomes  to  0. 1  mm  long;  bracts  lanceolate,  0.8-1 .2  mm  wide,  long- 
attenuate  and  recurved  at  apex;  bractlets  and  calyx  lobes  recurved;  capsule 
puberulent;  Guerrero H.  caput-  medusae 

10.  Cauline  trichomes  0.1-1  mm  long;  bracts  ovate-elliptic  to  obovate,  2.8-5 
mm  wide,  rounded  to  acute  and  erect  at  apex;  bractlets  and  calyx  lobes  erect; 
capsule  glabrous;  Durango  and  Zacatecas H.  peloria 

9.  Bracts  1.5-6  mm  long;  bractlets  1.5-4  mm  long;  corolla  pubescent  with  glandular 
(sometimes  absent  in  H.  parayana  and  H.  pallidd)  and  eglandular  trichomes  on 
external  surface;  seeds  pubescent  with  dendroid  trichomes. 

1 1 .  Inflorescence  axis  pubescent  with  crooked,  interwoven,  eglandular  trichomes 

obscuring  the  axis;  bracts  broadly  ovate  to  orbicular,  1.5-2  mm  long 

H.  hintonii 

1 1 .  Inflorescence  axis  pubescent  with  straight  to  ascendent  to  recurved  to  flex- 
uose,  eglandular  (and  often  glandular)  trichomes,  these  neither  crooked,  in- 
terwoven, nor  obscuring  the  axis;  bracts  triangular  to  subulate  to  lanceolate 
to  ovate,  2-6  mm  long. 
1 2.  Bracts  attenuate  to  a  point  but  lacking  a  distinct  mucro  at  apex;  upper 


DANIEL:  NEW  SPECIES  OF  HOLOGRAPHS 


75 


lip  of  corolla  3-5.5  mm  long;  staminode  0.6-1.2  mm  long;  capsule  1 1- 
17  mm  long. 

13.  Cauline  trichomes  ascendent-appressed,  0.1-0.5  mm  long;  corolla 
horizontal  during  an  thesis;  stamens  3-3.2  mm  long  with  thecae  1.2- 

1.5  mm  long;  style  5-6  mm  long;  seeds  2.5-3  mm  long;  Chiapas. 
H.  parayana 

13.  Cauline  trichomes  retrorse,  0.05-0.2  mm  long;  corolla  more  or  less 
vertical  during  anthesis;  stamens  5.5-6.5  mm  long  with  thecae  2- 

2.6  mm  long;  style  7-9.5  mm  long;  seeds  4.5-5.5  mm  long;  Oaxaca. 
H.  leticiana 

1 2.  Bracts  mucronate  at  apex;  upper  lip  of  corolla  1.5-3  mm  long;  staminode 
0.1-0.3  mm  long;  capsule  7-9  mm  long. 

14.  Bractlets  1-1.5  mm  wide;  corolla  8-10  mm  long,  the  upper  lip  1.5- 

2  mm  long;  stamens  2.5-3  mm  long;  style  3.5-4.5  mm  long;  capsule 
glabrous;  Jalisco  and  Colima H.  anisophylla 

14.  Bractlets  0.8-1  mm  wide;  corolla  10-12  mm  long,  the  upper  lip  2- 

3  mm  long;  stamens  3.5-4  mm  long;  style  5.3-6  mm  long;  capsule 
pubescent;  Sonora  and  Sinaloa H.  pallida 


Holographis  tolantongensis  T.  F.  Daniel,  sp.  nov. 

(Figures  1,  2) 

TYPE.  — MEXICO.  Hidalgo:  Mpio.  Cardonal,  Barranca  de 
Tolantongo,  lado  oeste,  45  km  NE  of  Ixmiquilpan,  10  February 
1982,  R.  Herndndez  M.  y  P.  Tenorio  L.  6948  (Holotype: 
MEXU!;  isotype:  CAS!). 

Suffrutex  usque  ad  6  dm  altus.  Caules  juniores 
pubescentes  trichomatibus  eglandulosis  0.05-0.7 
mm  longis.  Folia  opposita  petiolata,  lamina  lan- 
ceolata  vel  ovata  vel  elliptica  13-32  mm  longa 
3.5-14  mm  lata  2.3-4.7-plo  longior  quam  latior 
margine  revoluta.  Inflorescentia  spicata,  brac- 
teae  lanceolatae  6.5-9  mm  longae  1.5-2.5  mm 
latae  apice  acuminatae  erectae,  bracteolae  lanci- 
subulatae  5.5-7  mm  longae  1-1.5  mm  latae  apice 
subaristatae  erectae.  Calyx  7.5-10  mm  longus, 
corolla  rosea-purpurea  12-13  mm  longa  extus 
glabra,  stamina  4.5  mm  longa,  stylus  7-7.5  mm 
longus.  Capsula  ignota. 

Subshrub  to  6  dm  tall.  Young  stems  evenly 
pubescent  with  understory  of  straight  to  antrorse, 
eglandular  trichomes  0.05-0.2  mm  long  and 
sparse  overstory  of  straight  to  flexuose,  eglan- 
dular trichomes  0.3-0.7  mm  long.  Leaves  op- 
posite, petiolate,  petioles  2-6  mm  long,  blades 
coriaceous,  lanceolate  to  ovate  to  elliptic,  13-32 
mm  long,  3.5-14  mm  wide,  2.3-4.7  times  longer 
than  wide,  acute  at  apex,  acute  at  base,  adaxial 
surface  dark  green,  evenly  pubescent  with  flex- 
uose to  antrorse-flexuose  trichomes  to  1  mm  long 
(strigose),  abaxial  surface  light  green,  pubescent 
along  veins  and  glabrous  or  nearly  so  in  inter- 


costal regions,  margin  revolute.  Inflorescence  of 
terminal  spikes  to  3.5  cm  long,  flowers  opposite 
at  nodes,  rachis  strigose.  Bracts  lanceolate,  6.5- 
9  mm  long,  1.5-2.5  mm  wide,  acuminate  and 
erect  at  apex,  abaxial  surface  strigose.  Bractlets 
lance-subulate,  5.5-7  mm  long,  1-1.5  mm  wide, 
shorter  than  bracts,  subaristate  and  erect  at  apex, 
margin  often  hyaline.  Calyx  7.5-10  mm  long, 
lobes  narrowly  lanceolate  to  lance-subulate,  un- 
equal in  length  and  overlapping  one  another, 
sparsely  strigose  distally,  aristate  and  erect  at  apex, 
margins  often  hyaline.  Corolla  pinkish-purple, 
somewhat  intermediate  between  vertical  and 
horizontal  during  anthesis,  12-13  mm  long,  gla- 
brous on  external  surface,  tube  7-7.5  mm  long, 
upper  lip  4-4.5  mm  long  with  lobes  2-2.5  mm 
long,  lower  lip  6-7  mm  long  with  obovate  lobes 
5-6  mm  long.  Stamens  4.5  mm  long,  filaments 
pubescent,  thecae  1.4-1.5  mm  long,  pollen  (ob- 
served in  Diaphane)  prolate,  32-39  nm  long  (po- 
lar diameter),  25-27  nm  wide  (equatorial  di- 
ameter), 1.3-1.6  times  longer  than  wide,  colpi 
indistinct;  staminode  2  mm  long,  pubescent  at 
apex.  Style  7-7.5  mm  long,  glabrous;  stigma 
asymmetric,  0.2  mm  long;  ovary  glabrous.  Cap- 
sule unknown. 

DISTRIBUTION  AND  PHENOLOGY.— Known  only 
from  the  Barranca  de  Tolantongo  in  central  Hi- 
dalgo, Mexico  (Fig.  3).  The  plants  occur  abun- 
dantly in  rocky  soil  with  Pinus pinceana  Gordon, 
Quercus  opaca  Trel.,  Gochnatia  hypoleuca  (DC.) 
A.  Gray,  and  Juniperus  L.  at  an  elevation  of 
about  2,000  m.  Flowering:  February. 


76 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  2 


FIGURE  1 .    Holographis  tolantongensis  (Herndndez  M.  y  Tenorio  L.  6948).  a,  Habit,  x  0.5;  b,  Leaf  and  section  of  stem,  x  3.3; 
c,  Portion  of  inflorescence,  x3.5;  d,  Corolla  cut  open  to  show  androecium,  x3.5;  e,  Style,  x4.6. 


Holographis  tolantongensis  exhibits  the  fol- 
lowing unique  combination  of  characters:  op- 
posite leaves,  bracts  with  the  abaxial  surface  pu- 
bescent, pinkish-purple  corollas  with  the  external 
surface  glabrous,  and  calyces  7.5-10  mm  long. 
The  only  other  species  of  Holographis  known 
from  Hidalgo,  H.  ehrenbergiana,  has  mostly 
whorled  leaves  and  yellow  corollas. 


Holographis  caput-medusae  T.  F.  Daniel,  sp.  nov. 

(Figures  2,  4) 

TYPE.  — MEXICO.  Guerrero:  along  road  from  Petatlan  to 
Camalotito,  9.8  km  SW  of  Camalotito,  8  March  1987,  T.  F. 
Daniel  and  B.  Bartholomew  4918  (Holotype:  CAS!;  isotypes: 
K!,  MEXU!,  MICH!,  NY!,  US!). 

Herba  perennis  usque  ad  4  dm  alta.  Caules 
juniores  pubescentes  trichomatibus  inconspicuis 


DANIEL:  NEW  SPECIES  OF  HOLOGRAPHS 


77 


FIGURE  2.  Scanning  electron  micrographs  of  pollen  of  Holographis.  a,  H.  tolantongensis  (Hernandez  M.  and  Tenorio  L. 
6948),  scale  equals  6  nm;  b,  H.  caput- medusae  (Daniel  and  Bartholomew  4918),  scale  equals  6  /*m;  c,  Holographis  leticiana 
(Torres  C.  et  al.  637),  scale  equals  7.5  /mi. 


retrorsi  eglandulosis  usque  ad  0.1  mm  longis. 
Folia  verticillata  in  quoque  nodo  quatuor  sessilia 
vel  subsessilia,  lamina  anguste  elliptica  vel  ellip- 
tica  vel  ovato-elliptica  vel  obovato-elliptica  30- 
97  mm  longa  10—40  mm  lata  2.1-6-plo  longior 
quam  latior  margine  plana.  Inflorescentia  spi- 
cata,  bracteae  lanceolatae  6-8  mm  longae  0.8- 
1.2  mm  latae  apice  longiattenuatae  recurvatae, 
bracteolae  lanceolatae  6-8  mm  longae  0.5-1  mm 
latae  apice  longiattenuatae  recurvatae.  Calyx  8.5- 
10  mm  longus,  corolla  albida  et  rosea  11-14  mm 


longa  extus  pubescens,  stamina  3.5  mm  longa, 
stylus  6.5-7.5  mm  longus.  Capsula  8-9  mm  lon- 
ga puberula.  Semina  2.2-3  mm  longa,  1.8-2  mm 
lata. 

Perennial  herb  to  4  dm  tall.  Young  stems  in- 
conspicuously pubescent  with  retrorse,  eglan- 
dular  trichomes  up  to  0. 1  mm  long  concentrated 
in  two  vertical  lines  to  glabrous.  Leaves  whorled 
(four/node),  sessile  to  subsessile,  petioles  (if  pres- 
ent) up  to  1  mm  long,  blades  chartaceous,  nar- 
rowly elliptic  to  elliptic  to  ovate-elliptic  to  ob- 


CAPUT-MEDUSAE 

LETICIANA 

TOLANTONGENSIS 


FIGURE  3.     Map  showing  distribution  of  Holographis  caput- medusae,  H.  leticiana,  and  H.  tolantongensis. 


78 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  2 


FIGURE  4.    Holographys  caput-medusae  (Daniel  and  Bartholomew  4918).  a,  Habit,  xQ.5;  b,  Inflorescence,  x  3.4;  c,  Corolla, 
x3.6;  d,  Corolla  cut  open  to  show  androecium,  x4.3;  e,  Capsule,  x4;  f,  Seed,  x  1 1.5. 


ovate-elliptic,  30-97  mm  long,  10-40  mm  wide, 
2. 1-6  times  longer  than  wide,  acuminate  at  apex, 
attenuate  nearly  to  or  to  the  node  at  base,  adaxial 
surface  sparsely  pubescent  with  coarse,  antrorse, 
eglandular  trichomes  0.2-0.5  mm  long,  abaxial 
surface  glabrous  or  nearly  so,  margin  entire  to 
somewhat  undulate,  flat.  Inflorescence  of  axillary 


spikes  to  3  cm  long,  flowers  opposite  at  nodes, 
rachis  evenly  and  densely  pubescent  with  straight, 
eglandular  trichomes  to  0.05  mm  long.  Bracts 
lanceolate,  6-8  mm  long,  0.8-1.2  mm  wide, 
abaxial  surface  pubescent  like  rachis,  apex  long- 
attenuate  and  recurved.  Bractlets  lanceolate,  6- 
8  mm  long,  0.5-1  mm  wide,  approximately  equal 


DANIEL:  NEW  SPECIES  OF  HOLOGRAPHS 


79 


in  length  to  bracts,  pubescent  like  bracts,  apex 
long-attenuate  and  recurved.  Calyx  8.5-10  mm 
long,  lobes  equal  in  length  and  slightly  overlap- 
ping one  another,  pubescent  like  bracts,  long- 
attenuate  and  recurved  at  apex,  margin  hyaline 
proximally.  Corolla  whitish  with  pink  lines  on 
lower  lip,  horizontal  during  anthesis,  11-14  mm 
long,  pubescent  on  external  surface  with  eglan- 
dular  trichomes  0.05-0. 1  mm  long,  tube  5-7  mm 
long,  upper  lip  4.5-5.5  mm  long  with  linear- 
elliptic  lobes  4.5-5.5  mm  long,  lower  lip  6-7  mm 
long  with  elliptic  to  obovate  lobes  3-4.7  mm 
long.  Stamens  3.5  mm  long,  filaments  pubescent, 
thecae  1.3-1.6  mm  long,  pollen  (observed  in  Dia- 
phane)  prolate,  34-42  /um  long  (polar  diameter), 
25-27  jum  wide  (equatorial  diameter),  1.4-1.6 
times  longer  than  wide,  colpi  distinct;  staminode 
0.6  mm  long,  pubescent  at  apex.  Style  6.5-7.5 
mm  long,  sparsely  pubescent;  stigma  unequally 
bilobed,  0.2-0.3  mm  long.  Capsule  8-9  mm  long, 
puberulent  with  trichomes  less  than  0.05  mm 
long.  Seeds  subelliptical  to  subrectangular  in  out- 
line, 2.2-3  mm  long,  1.8-2  mm  wide,  surfaces 
covered  with  hairlike  papillae  to  0. 1  mm  long. 

DISTRIBUTION  AND  PHENOLOGY.— Known  only 
from  the  seaward  slopes  of  the  Sierra  Madre  Sur 
in  western  Guerrero  (Fig.  3)  in  a  region  of  tropical 
dry  forest  at  an  elevation  of  approximately  1 50 
m.  Flowering  and  fruiting:  March. 

Holographis  caput- medusae  is  unique  in  the 
genus  by  virtue  of  its  relatively  long,  attenuate- 
recurved  bracts,  bractlets,  and  calyx  lobes.  These 
give  the  dense  spikes  a  medusoid  appearance. 
Holographis  caput- medusae  is  the  third  species 
of  the  genus  to  be  found  in  Guerrero.  Unlike  the 
other  two,  H.  argyrea  and  H.  hintonii,  it  occurs 
in  the  Pacific  lowlands.  The  holotype  was  col- 
lected in  a  weedy  thicket  rich  in  Acanthaceae, 
including:  Aphelandra  scabra  (Vahl)  Smith,  Car- 
lowrightia  arizonica  A.  Gray,  Didiptera  resupi- 
nata  (Vahl)  Juss.,  Elytraria  imbricata  (Vahl)  Pers., 
Henrya  insularis  Nees,  Odontonema  auricula- 
turn  (Rose)  T.  Daniel,  Ruellia  albiflora  Fernald, 
Tetramerium  nervosum  Nees,  and  Tetramerium 
tenuissimum  Rose. 

Holographis  leticiana  T.  F.  Daniel,  sp.  nov. 

(Figures  2,  5) 

TYPE. -MEXICO.  Oaxaca:  Distr.  Tehuantepec,  subida  al 
Cerro  Guiengola  por  la  ladera  S  donde  esta  la  fabrica  de  cal, 
25  October  1986,  M.  Leticia  Torres  C.,  R.  Torres  C.,  y  C. 
Martinez  637  (Holotype:  MEXU!;  isotypes:  CAS!,  and  others 
to  be  distributed). 


Frutex  usque  ad  1  m  altus.  Caules  juniores 
puberuli  trichomatibus  eglandulosis  0.05-0.2  mm 
longis.  Folia  verticillata  in  quoque  nodo  quatuor 
petiolata,  lamina  ovata  vel  ovato-elliptica  17-82 
mm  longa  9-27  mm  lata  1.5-3-plo  longior 
quam  latior  margine  plana.  Inflorescentia  spi- 
cata,  bracteae  lanceolatae  vel  lanci-subulatae  2.5- 

4  mm  longae  0.9-1.1  mm  latae  apice  attenuatae 
erectae,  bracteolae  subulatae  2.3-3.5  mm  longae 
0.5-0.7  mm  latae  apice  attenuatae  erectae.  Calyx 
4-5  mm  longus,  corolla  rosea  9.5-14.5  mm  longa 
extus  pubescens,  staminia  5.5-6.5  mm  longa, 
stylus  7-9.5  mm  longus.  Capsula  1 1-17  mm  lon- 
ga glabra.  Semina  5-5.5  mm  longa,  3.8-4.5  mm 
lata. 

Shrub  to  2  m  tall.  Young  stems  evenly  puber- 
ulent with  retrorse,  eglandular  trichomes  0.05- 
0.2  mm  long.  Leaves  whorled  (four/node),  pet- 
iolate,  petioles  2-17  mm  long,  blades  charta- 
ceous,  ovate  to  ovate-elliptic,  6-90  mm  long,  4- 
42  mm  wide,  1 .5-3  times  longer  than  wide,  acute 
to  acuminate  at  apex,  acute  to  attenuate  at  base, 
surfaces  pubescent  with  straight  to  subflexuose 
to  antrorse,  eglandular  trichomes  to  0.2  mm  long, 
margin  entire  to  subundulate,  flat.  Inflorescence 
of  axillary  spikes  to  2.5  cm  long,  flowers  opposite 
at  nodes,  rachis  evenly  and  densely  pubescent 
with  erect  to  downward  pointing,  eglandular  tri- 
chomes 0.1-0.2  mm  long.  Bracts  lanceolate  to 
lance-subulate,  2-4  mm  long,  0.9-1.1  mm  wide, 
abaxial  surface  pubescent  with  erect  to  subflex- 
uose or  downward  pointing,  eglandular  tri- 
chomes to  0.2  mm  long,  apex  attenuate,  erect. 
Bractlets  subulate,  2-3.5  mm  long,  0.5-0.7  mm 
wide,  shorter  than  or  equal  to  bracts,  pubescent 
like  bracts,  apex  attenuate,  erect.  Calyx  3-5  mm 
long,  lobes  equal  in  length  and  slightly  overlap- 
ping one  another,  lanceolate,  pubescent  like 
bracts,  attenuate  and  erect  at  apex.  Corolla  green- 
ish-white with  pinkish  or  maroon  stripes  on  low- 
er lip,  more  or  less  vertical  during  anthesis,  9.5- 
14.5  mm  long,  pubescent  on  external  surface  with 
glandular  and  eglandular  trichomes  0.05-0.2  mm 
long,  tube  2.5-5  mm  long,  abruptly  expanded 
into  a  throat  2.5-3.5  mm  long,  upper  lip  3-5.5 
mm  long  with  triangular  lobes  1-1.5  mm  long, 
lower  lip  4.5-7  mm  long  with  obovate  lobes  2.5- 

5  mm  long.  Stamens  5.5-6.5  mm  long,  filaments 
pubescent,  thecae  2-2.6  mm  long,  pollen  (ob- 
served in  Diaphane)  prolate,  32-39  ^m  long  (po- 
lar diameter),  22-25  /im  wide  (equatorial  di- 
ameter), 1.3-1.8  times  longer  than  wide,  colpi 
distinct;  staminode  0.7-1.2  mm  long,  sparsely 


80 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  2 


FIGURE  5.  Holographs  leticiana.  a,  Habit  (Daniel  5386),  xQ.5;  b,  Inflorescence  (Daniel  5386),  x4;  c,  Corolla  cut  open  to 
show  androecium  (Torres  C.  637),  x4;  d,  Distal  portion  of  style  and  stigma  (Daniel  5386),  x  16;  e,  Capsule  (Torres  C.  773), 
x3.4;  f,  Seed  (Torres  C.  773),  xg;  g,  Seminal  trichome  (Torres  C.  773),  x42. 


pubescent  at  apex.  Style  7-9.5  mm  long,  very 
sparsely  pubescent  to  glabrous;  stigma  indistinct- 
ly bilobed,  0.1  mm  long.  Capsule  11-18  mm 
long,  glabrous.  Seeds  subrectangular  to  subellip- 
tical  in  outline,  4.5-5.5  mm  long,  3.5-4.5  mm 


wide,  surfaces  covered  with  golden  to  brown, 
dendritic  trichomes  to  0.4  mm  long. 

DISTRIBUTION  AND  PHENOLOGY. —Known  only 
from  the  vicinity  of  the  Cerro  Guiengola  in 
southeastern  Oaxaca  (Fig.  3),  where  the  plants 


DANIEL:  NEW  SPECIES  OF  HOLOGRAPHS 


81 


occur  in  tropical  deciduous  forest  dominated  by 
legumes  and  cereoid  cacti  at  elevations  from  60 
to  600  m.  Flowering  and  fruiting:  October-March. 
Holographs  leticiana  can  be  distinguished  from 
all  other  species  in  the  genus  by  the  following 
combination  of  characters:  whorled  leaves,  pink- 
ish corollas,  and  stamens  5.5-6.5  mm  long.  It  is 
most  similar  to  H.  parayana,  a  species  occurring 
approximately  200  km  to  the  east  in  Chiapas. 
The  contrasting  features  of  these  two  species  are 
summarized  in  the  key  above.  Plants  of  this 
species  collected  in  November  1987  (Daniel  5386) 
were  locally  common  understory  shrubs.  Flowers 
were  visited  by  both  bombyliid  flies  and  mega- 
chilid  bees.  However,  pollen  similar  to  that  of 
Holographis  was  not  located  on  individuals  of 
either.  Other  Acanthaceae  growing  with  H.  le- 
ticiana include  Tetramerium  oaxacanum  T. 
Daniel  and  species  of  Ruellia  L.  and  Siphono- 
glossa  Oersted. 

PARATYPES.— MEXICO.  Oaxaca:  along  road  to  Ruinas 
Guiengola,  1.3-1.6  km  N  of  Mex.  190  between  Jalapa  de  Mar- 
ques and  Tehuantepec,  T.  Daniel  5386  (CAS,  DUKE,  F,  GH, 
K,  MICH,  MEXU,  MO,  NY,  TEX,  US);  61.8  km  NW  of 
Tehuantepec,  P.  Fryxell  and  E.  Lott  3403  (CAS);  Distr.  Te- 
huantepec, Ruinas  del  Cerro  Guiengola,  M.  L.  Torres  C.  et  al. 
511  (CAS,  MEXU);  Distr.  Tehuantepec,  ladera  S  del  Cerro 
Guiengola  por  la  fabrica  de  cal,  M.  L.  Torres  C.  et  al.  773 
(CAS,  MEXU);  Distr.  Tehuantepec,  hacia  el  N  a  300  m  del 


"Mango,"  por  la  subida  al  Ocotal  Cerro  Guiengola,  M.  L. 
Torres  C.  et  al.  834  (CAS,  MEXU). 


ACKNOWLEDGMENTS 

Funds  for  field  studies  were  provided  by  NSF 
grants  BSR-8304790  and  BSR-8609852.  I  am 
most  appreciative  to  the  following  persons  for 
their  valuable  assistance:  B.  Bartholomew,  M. 
Baker,  and  V.  Lee  (field  assistance);  M.  L.  Torres 
C.,  E.  Lott,  and  A.  Sanders  (making  collections 
available);  C.  Sudekum  (illustrations);  M.  A. 
Tenorio  (SEM  operation  and  photographic  as- 
sistance); S.  Middleton  (photography);  and  F.  Al- 
meda  (selecting  loan).  I  am  also  grateful  to  the 
curators  of  the  following  herbaria  for  loans  or 
allowing  me  to  study  specimens  in  their  respec- 
tive institutions:  ASU,  CAS,  ENCB,  F,  G,  GH, 
K,  LL,  MEXU,  MO,  P,  POM,  TEX,  and  US. 


LITERATURE  CITED 

DANIEL,  T.  F.  1983.  Systematics  of  Holographis  (Acantha- 
ceae). J.  Arnold  Arbor.  64:129-160. 

.  1 984.  New  and  reconsidered  Mexican  Acanthaceae. 

Madrono  31:86-92. 

.    1986.   New  and  reconsidered  Mexican  Acanthaceae. 


II.  Southw.  Naturalist  31:169-175. 


CALIFORNIA  ACADEMY  OF  SCIENCES 

Golden  Gate  Park 
San  Francisco,  California  941 18 


PROCEEDINGS 

OF  THE 

CALIFORNIA  ACADEMY  OF  SCIENCES 


Vol.  46,  No.  3,  pp.  83-94,  6  figs.,  2  tables.  December  7,  1988 


THE  EELPOUT  GENUS  PACHYCARA  (TELEOSTEI:  ZOARCIDAE) 

IN  THE  NORTHEASTERN  PACIFIC  OCEAN,  WITH 

DESCRIPTIONS  OF  TWO  NEW  SPECIES 

By 
M.  Eric  Anderson 

Department  of  Ichthyology,  California  Academy  of  Sciences, 
Golden  Gate  Park,  San  Francisco,  California  94118 

and 
Alex  E.  Peden 

Vertebrate  Unit,  Royal  British  Columbia  Museum, 
Victoria,  British  Columbia  V8V 1X4,  Canada 


ABSTRACT:  The  eelpout  genus  Pachycara  Zugmayer,  1911,  is  recognized  from  the  North  Pacific  Ocean  for 
the  first  time.  Four  species  are  included,  two  of  which— P.  gymninium  and  P.  lepinium—aie  new.  Pachycara 
gymninium  differs  from  P.  lepinium  mainly  in  having  a  scaleless  nape,  the  mediolateral  lateral  line  originating 
in  the  pectoral  axil  versus  posterior  to  the  pectoral  margin,  more  numerous  precaudal  vertebrae,  and  shorter 
pectoral  and  pelvic  fins.  The  type  species  of  Pachycara— P.  obesum—from  the  abyssal  Atlantic,  is  shown  to 
be  synonymous  with  the  earlier  described  Maynea  bulbiceps  Carman,  1899,  from  the  eastern  tropical  Pacific. 
Another  species  described  by  Carman  (1899)—Phucocoetes  suspectus—is  placed  in  Pachycara  and  redescribed 
from  three  specimens. 

Received  April  27,  1988.  Accepted  July  11, 1988. 


Anderson  (1988)  included  an  Antarctic  species, 

The  eelpout  genus  Pachycara  was  erected  for  first  described  as  Lycodes  brachycephalus  by 
a  single  specimen  taken  in  the  Bay  of  Biscay  at  Pappenheim  (1912),  in  Pachycara  because  it 
a  depth  of  4,780  m  by  the  expeditions  of  Prince  lacked  the  three  apomorphic  characters  previ- 
Albert  I  of  Monaco  (Zugmayer  191  la,  b).  A  sec-  ously  established  for  Lycodes  (Anderson  1984): 
ond  (juvenile)  specimen  of  the  type  species,  P.  1)  possession  of  cartilaginous  laminae  along  the 
obesum,  was  described  from  off  Virginia  by  Mar-  ventral  surface  of  the  dentary  bones  ("mental 
kle  and  Sedberry  (1978).  Earlier,  Garman  (1899)  crests"  of  Andriashev  1 954);  2)  loss  of  the  pal- 
described  Maynea  bulbiceps  from  the  eastern  atine  membrane  (oral  valve);  and  3)  loss  of  head 
tropical  Pacific  (Gulf  of  Panama).  This  species  pores  in  adults  (except  for  some  paedomorphic 
was  placed  in  Pachycara  as  a  putative  Pacific  species).  Anderson  (1984)  included  11  species  in 
sister  species  of  P.  obesum  by  Anderson  (1984).  Pachycara,  only  three  of  which  had  been  de- 

[83] 


84 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  3 


FOB. 


FIGURE  1.    Pachycara  bulbiceps  (Garman),  BCPM  980-99,  525  mm  SL,  off  the  Queen  Charlotte  Islands,  British  Columbia. 


scribed  previously,  and  a  revision  of  the  genus 
is  in  progress. 

The  purpose  of  this  paper  is  to  describe  in 
advance  of  the  senior  author's  revision  two  new 
species  of  Pachycara  from  the  northeastern  Pa- 
cific Ocean  that  have  been  known  in  collections 
for  many  years.  In  addition,  we  redescribe  the 
type  species,  P.  obesum,  herein  synonymized  with 
P.  bulbiceps,  now  known  from  six  North  Atlantic 
and  32  northeastern  Pacific  specimens,  as  well 
as  the  rare  P.  suspectum,  known  from  three  spec- 
imens. 

METHODS 

Measurements  were  made  with  an  ocular  mi- 
crometer or  dial  calipers  to  the  nearest  0.1  mm. 
Definitions  of  characters  and  methodology  fol- 
lows Gosztonyi  (1977),  or,  in  cases  of  contradic- 
tion, Peden  and  Anderson  (1978)  and  Anderson 
(1982,  1984).  All  specimens  were  measured  and 
radiographed  to  create  Table  1,  except  16  of  38 
P.  gymninium  in  SIO  67-121.  Character  state 
modifiers  are  based  on  those  numerically  scored 
characters  of  Anderson  (1984).  Museum  abbre- 
viations follow  Leviton  et  al.  (1985).  Measure- 
ments are  in  standard  length  (SL).  Certain  pro- 
portions expressed  in  percent  head  length  (HL) 
orSL. 

KEY  TO  NORTHEASTERN  PACIFIC 
PACHYCARA  SPECIES 

1A.  Pelvic  fins  present __ 2 

IB.   Pelvic  fins  absent _ 

Pachycara  bulbiceps  (Garman,  1899) 

2 A.  Number  of  precaudal  vertebrae  21-23; 
lateral  line  with  complete,  ventral  branch 
only;  head  length  1 5.6-1 7.7%  SL;  anal  fin 
origin  associated  with  vertebrae  20-22 
Pachycara  suspectum  (Garman,  1899) 

2B.  Number  of  precaudal  vertebrae  23-31 
(usually  25  or  more);  lateral  line  with 
complete  mediolateral  and  ventral 
branches;  head  length  12.0-15.9%  SL; 


anal  fin  origin  associated  with  vertebrae 
23-31  3 

3A.  Scales  present  on  nape;  mediolateral 
branch  of  lateral  line  originating  posterior 
to  pectoral  fin  margin;  pectoral  fin  length 
72.5-89.9%  HL;  pelvic  fin  length  11.5- 
17.3%  HL;  number  of  precaudal  verte- 
brae 23-26 Pachycara  lepinium  n.  sp. 

3B.  Scales  absent  on  nape  (rarely  present  in 
small  patches  immediately  anterior  to 
dorsal  fin  origin);  mediolateral  branch  of 
lateral  line  originating  in  pectoral  axil; 
pectoral  fin  length  63.0-70.8%  HL;  pelvic 
fin  length  4.8-1 1.3%  HL;  number  of  pre- 
caudal vertebrae  28-3 1 

_ _ Pachycara  gymninium  n.  sp. 

Pachycara  bulbiceps  (Garman,  1899) 

(Figures  1-3) 

Maynea  bulbiceps  Garman,    1899:140,    141,  pi.   E,   fig.    1. 

McAllister  and  Rees  1964:106,  107.  Pearcy  et  al.  1982:387, 

399,  400. 
Pachycara  obesa  Zugmayer,  1 9 1 1  a:  1 2;  1 9 1 1  b:  1 34-1 36,  pi.  VI, 

fig.  6.  Markle  and  Sedberry  1978:22-25,  fig.  1.  Merrett  and 

Marshall  1981:240.  Andriashev  1973:547. 
Pachychara  obesa  Zugmayer  (lapsus  calami).  Fowler  1936: 

1056,  fig.  47.  Andriashev  1986:1 149,  text  fig. 

MATERIAL  EXAMINED.— Northeastern  Atlantic:  MOM,  un- 
cat.  (holotype  of  Pachycara  obesa);  Bay  of  Biscay;  45°27'N, 
06°05'W;  4,780  m;  PRINCESSE  ALICE  sta.  1554;  7-9  Sept. 
1903.  BMNH  1981.6.16:15  (1);  off  Cap  Blanc,  West  Africa; 
20°17.2'N,  21°42.3'W;  4,002-4,007  m;  DISCOVERY  sta.  9131; 
17Nov.  1976.IOS,uncat.(l);offSenegal;  18°08.6'N,20°11.7'W; 
3,120  m;  DISCOVERY  sta.  10884;  12  Aug.  1983.  IOS,  uncat. 
(1);  Porcupine  Sea  Bight;  49°37.1'N,  13°48.5'W;  3,990-3,920 
m;  DISCOVERY  sta.  51803;  26  Sept.  1983.  Northwestern 
Atlantic:  USNM  21561 1  (1);  off  Virginia;  36°42.0'N,  74°04.5'W; 
2,400-2,450  m;  EASTWARD  cr.  7506,  sta.  20;  24-25  July 
1975.  MCZ  57612  (1);  Middle  Atlantic  Bight;  38°44.0'N, 
70°10.0'W;  3,028  m;  OCEANUScr.  93,  sta.  91 1;  3 1  Mar.  1981. 
Gulf  of  Panama:  MCZ  28681  (holotype  of  Maynea  bulbiceps); 
06°10'N,  83°06'W;  2,690  m;  ALBATROSS  sta.  3361;  25  Feb. 
1890.  UMML  22860  (2);  06°53-49'N,  79°27-28'W;  3,193- 
3,201  m;  JOHN  E.  PILLSBURY  sta.  P-526;  5  May  1967. 
UMML  33484  (1);06°47-48'N,  79°13'-80°00'W;  3,04 5-3,218 
m;  JAMES  M.  GILLISS  sta.  GS-3;  14  Jan.  1972.  UMML 
33485  (2);  06°52'N,  79°28'W;  3,200-3,229  m;  JAMES  M.  GIL- 
LISS sta.  GS-1;  13-14  Jan.  1972.  SIO  81-148  (4);  05°09.8'N, 


ANDERSON  AND  PEDEN:  NORTH  PACIFIC  PACHYCARA 


85 


FIGURE  2.    Pachycara  bulbiceps  (Garman),  diagrammatic  of  CAS  55588,  338  mm  SL,  from  the  Gulf  of  Panama,  showing 
extent  of  lateral  lines. 


81°41.2'W;  3,900-4,000  m;  MELVILLE;  3  Oct.  1981.  SIO  85- 
33  (4);  05°20.6'N,  81°55.5'W;  3,866  m;  DSRV  ALVIN  dives 
1361,  1363;  9-12  Apr.  1984.  CAS  55588  (1);  recat.  from  SIO 
81-148.  Middle  American  Trench:  SIO  73-286  (3);  off  Costa 
Rica;  09°43.8'N,  86°02.0'W;  2,601  m;  ALEXANDER  AGAS- 
SIZ;  19-20  Apr.  1973.  SIO  73-268(1);  off  Guatemala;  13°27.0'N, 
91°59.2'W;  3,332  m;  ALEXANDER  AGASSIZ;  13-14  Apr. 
1973.  Mexico:  SIO  66-35  (1);  offGuadalupe  Isl.;  29°01.0'N, 
118°01.2'W;  3,310  m;  ALEXANDER  AGASSIZ;  5-6  Apr. 
1966.  Cascadia  Abyssal  Plain  (off  Oregon):  OSUO  2559  (2); 
45°27.8-27.0'N,  126°1 7.4-2 1.0'W;  2,606  m;  YAQUINA  sta. 
CP-2C,  BMT  157;  17  Jan.  1970.  OSUO  2560  (2);  46°02.2'N, 
1 26°33.5-28.5'W;  2,706  m;  YAQUINA  sta.  CP-2A,  BMT  257; 
15  Feb.  1971.  OSUO  2561  (2);  45°1 8.0-1 7.0'N,  126°26.4- 
30.5'W;  2,710  m;  YAQUINA  sta.  CP-2C,  BMT  266;  18  Feb. 


1971.  OSUO  2562  (1);  44°43.  l^t6.0'N,  1 27°29.3-28.6' W;  2,8 1 8 
m;  YAQUINA  sta.  CP-3E,  BMT  3 1 7;  3  Feb.  1 973.  CAS  6 1 220 
(2);  44°39.7-39.0'N,  1 26°39.6-44.8' W;  2,8 1 6  m;  CAYUSE  sta. 
CP-2E,  BMT  253;  30  Sept.  1970.  LACM  44316-1  (1);  45°20.6- 
20.9'N,  126°35.7-39.7'W;  2,750  m;  YAQUINA  sta.  CP-2C, 
BMT  264;  17  Feb.  1971.  LACM  44317-1  (1);  46°01.4-05.0'N, 
1 27°3 1 .0-32.5' W;  2,835  m;  YAQUINA  sta.  CP-3 A,  BMT  322; 
12  Mar.  1973.  British  Columbia:  BCPM  980-99  (1);  off  Queen 
Charlotte  Isl.,  W  of  Tasu  Sound;  2,780  m;  EASTWARD  HO 
set  17;  26  Aug.  1979. 

DIAGNOSIS.— A  species  of  Pachycara  as  de- 
fined by  Anderson  (1984)  distinguished  by  the 
following  combination  of  characters:  pelvic  fins 


FIGURE  3.    Group  of  Pachycara  sp.  (presumably  P.  bulbiceps)  near  a  baitfall  in  the  Gulf  of  Panama  (05°20'N,  8 1°55'W;  depth, 
3,850  m).  Photographed  by  Craig  R.  Smith,  University  of  Washington,  DSRV  ALVIN  dive  1 140,  28  Sept.  1981. 


86 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  3 


TABLE  1 .    COUNTS  AND  MEASUREMENTS  OF  NORTHEASTERN  PACIFIC  PACHYCARA  SPECIES,  INCLUDING  ATLANTIC  MATERIAL  OF 
P.  BULBICEPS. 


P.  bulbiceps 
(Atlantic) 

P.  bulbiceps 
(Pacific) 

P.  gymninium 

P.  lepinium 

P.  suspectum 

Measurements,  in  percent  SL 

Head  length 

12.3-15.0 

11.8-14.8 

12.0-15.6 

13.8-15.9 

15.6-17.7 

Head  width 

5.7-8.5 

5.5-10.1 

5.7-8.2 

6.5-11.8 

8.0-8.9 

Pectoral  length 

9.9-12.2 

8.7-12.9 

8.3-10.3 

10.2-13.5 

9.1-10.1 

Predorsal  length 

17.8-18.7 

16.0-19.8 

15.9-21.5 

15.8-23.2 

15.9-18.2 

Preanal  length 

34.9-10.1 

35.1-40.0 

38.5-44.1 

34.3-41.9 

35.8-36.0 

Body  height 

6.4-12.7 

5.7-11.6 

7.4-13.6 

8.3-16.9 

8.4-9.9 

Gill  slit  length 

4.9-5.9 

4.1-6.9 

3.3-5.2 

4.9-9.4 

7.0  (in  1) 

Caudal  fin  length 

2.4-4.2 

1.5-3.6 

1.9-5.2 

1.5-2.7 

3.2  (in  1) 

Measurements,  in  percent  HL 

Head  width 

38.8-73.1 

43.5-80.5 

37.3-64.1 

44.5-75.3 

49.7-52.7 

Upper  jaw  length 

33.9-13.2 

37.1-52.7 

33.3-52.8 

35.9-52.0 

38.4-41.4 

Pectoral  length 

67.8-99.5 

64.1-99.9 

63.0-70.8 

72.5-89.9 

57.4-58.6 

Snout  length 

19.4-22.7 

16.8-26.2 

15.8-25.5 

17.6-25.9 

18.5-22.3 

Eye  diameter 

16.2-26.4 

11.5-22.4 

13.6-23.3 

12.4-18.7 

16.8-19.5 

Gill  slit  length 

32.9-50.9 

31.4-49.8 

21.0-43.1 

33.9-59.8 

39.7  (in  1) 

Interorbital  width 

10.6-17.4 

7.0-11.5 

9.1-11.3 

7.1-13.8 

6.4-7.3 

Interpupillary  width 

26.4-38.0 

18.2-32.8 

26.9-35.7 

22.3^t4.4 

24.2  (in  1) 

Pelvic  fin  length 

absent 

absent 

4.8-11.3 

11.5-17.3 

9.7-11.7 

Counts 

Vertebrae 

25-31  + 

27-31  + 

28-31  + 

23-26  + 

21-23  + 

82-90  = 

83-91  = 

73-80  = 

80-94  = 

84-87  = 

113-118 

112-119 

102-109 

105-120 

105-110 

Dorsal  fin  rays 

104-109 

105-114 

96-103 

99-113 

100-106 

Anal  fin  rays 

86-94 

88-97 

77-84 

85-98 

85-89 

Caudal  fin  rays 

8-12 

10-12 

10-12 

9-12 

9 

Pectoral  fin  rays 

16-19 

16-18 

14-18 

15-18 

16 

Pelvic  fin  rays 

absent 

absent 

3 

3 

3 

Gill  rakers 

2-3  + 

0-4  + 

3-5  + 

2-5  + 

1-2  + 

11-13  = 

11-15  = 

10-14  = 

11-15  = 

14-15  = 

13-16 

11-19 

13-19 

12-19 

16 

Vomerine  teeth 

3-6 

2-9 

3-8 

5-18 

3-6 

Palatine  teeth 

2-7 

4-10 

3-14 

7^10 

9-12 

absent;  mediolateral  branch  of  lateral  line  orig- 
inating posterior  to  pectoral  fin  margin;  scales 
absent  on  nape;  vertebrae  25-3 1  +  82-9 1  =  112- 
1 19;  head  length  1 1.8-15.0%  SL;  anal  fin  origin 
associated  with  vertebrae  23-30. 

DESCRIPTION.  — Counts  and  measurements 
presented  in  Table  1  were  compiled  from  all 
known  specimens,  82-525  mm  SL  (no  gravid 
females).  The  following  description  is  based  on 
10  adult  males,  eight  adult  females,  10  juveniles 
of  both  sexes,  and  seven  eviscerated  specimens. 
Head  large,  ovoid,  wider  in  adults  than  juveniles. 
Smallest  specimen  (82  mm  SL)  with  most  dor- 
soventrally  depressed  head,  resulting  in  more 
acute  snout  than  adults.  Body  short,  deep,  broad- 
er in  cross  section  in  adults  than  juveniles.  Body 
more  slender  than  similarly  sized  P.  gymninium 


and  P.  lepinium;  body  height  at  anal  fin  origin 
of  smallest  specimen  (5.7%  SL;  Table  1)  mea- 
suring like  slender  genus  Lycenchelys,  however 
specimens  >160  mm  with  body  height  >7.8% 
SL.  Tail  laterally  compressed,  more  so  poste- 
riorly, tapering  gradually  to  tip.  Skin  firm,  not 
gelatinous,  but  pliable  around  head,  especially 
near  mouth,  cheeks,  and  nape  due  to  subdermal 
lipid  layer.  Scales  minute,  cycloid,  imbedded, 
covering  body,  tail,  and  vertical  fins  to  one-half 
to  three-fourths  their  height;  no  scales  on  head 
or  nape;  scales  present  on  pectoral  axil  and  base, 
and  on  breast  to  isthmus  in  largest  specimens. 
No  scales  developed  in  smallest  specimen.  Eye 
circular,  larger  in  proportion  to  head  in  smallest 
specimens  than  largest  adults.  Gill  slit  short,  usu- 
ally extending  ventrally  to  lower  margin  of  pec- 


ANDERSON  AND  PEDEN:  NORTH  PACIFIC  PACHYCARA 


87 


toral  base.  Very  slight,  squared-off  lobe  formed 
at  dorsal  margin  of  operculum,  this  best  devel- 
oped in  small  juveniles;  gill  slit  of  large  adults 
often  without  distinct  lobe.  Single  pair  of  nostrils 
at  snout  tip,  nasal  tube  not  reaching  upper  lip. 
Pectoral  fin  large,  its  origin  at  or  just  below  body 
midline;  posterior  margin  somewhat  wedge- 
shaped  to  almost  evenly  rounded;  lowermost  6- 
9  rays  shorter  and  thicker  than  others,  their  tips 
not  at  all  or  slightly  exserted;  pectoral  rays  usu- 
ally 17-1 8  (Table  2). 

Mouth  terminal,  upper  jaw  slightly  longer  in 
relation  to  head  in  adult  males  than  comparably 
sized  females;  upper  jaw  45.0-52.7%  HL  (X  = 
48.8%,  n  =  10)  in  males  over  330  mm  SL,  41.8- 
44.6%  HL  (x  =  43.1%,  n  =  5)  in  females  over 
330  mm  SL.  Minute  epidermal  prickles  on  lips 
and  occasionally  in  patches  on  snout.  Palatine 
membrane  (oral  valve)  moderately  developed  in 
adults,  not  reaching  vomer,  larger  in  juveniles, 
reaching  anterior  edge  of  vomer.  Teeth  in  jaws 
and  palate  small,  conical,  dentition  not  sexually 
dimorphic.  Jaw  teeth  in  single  row  in  smallest 
specimen;  premaxillae  of  adults  with  2-3  rows 
anteriorly,  blending  into  single,  posterior  row; 
dentary  with  3-5  rows  anteriorly,  blending  into 
single  or  double  posterior  row.  Vomerine  teeth 
in  irregular  patch;  palatine  teeth  in  single  row. 

Cephalic  lateralis  system  reduced,  numerical 
variation  in  pores  of  postorbital  and  suborbital 
canals  only.  Postorbital  pores  one,  three  and  four 
present  on  one  side,  pore  one  only  on  the  other 
in  one  specimen;  pores  one  and  four  present  in 
21  specimens.  Two  pairs  of  anterior  supraorbital 
(nasal)  pores,  one  set  mesial  to  nasal  tube,  the 
other  posteromesially.  Usually  six  suborbital 
pores,  all  emanating  from  ventral  ramus  of  sub- 
orbital  bone  chain,  but  a  seventh  pore  emanating 
from  ascending  ramus  behind  eye,  just  below  first 
postorbital  pore  on  one  side  in  one  specimen. 
Eight  (not  seven  as  reported  by  Markle  and  Sed- 
berry  1978)  preoperculomandibular  pores,  four 
emanating  from  dentary,  one  from  anguloartic- 
ular,  and  three  from  preopercle.  Interorbital  and 
occipital  (supratemporal)  pores  absent.  Body  lat- 
eral line  of  two  branches:  mediolateral,  origi- 
nating at  or  just  posterior  to  pectoral  fin  margin, 
complete  to  tail  tip  (not  evident  in  some  mate- 
rial), and  ventral,  originating  just  posterior  to 
fourth  postorbital  pore,  coursing  gradually  across 
body  to  just  above  anus,  then  running  just  above 
anal  fin  to  tail  tip  (Fig.  2). 


TABLE  2.    PECTORAL  FIN  RAY  COUNTS  OF  NORTHEASTERN 
PACIFIC  PACHYCARA  SPECIES. 


Fin  ray  number 

Species 

14 

15 

16 

17 

18 

19 

P.  gymninium,  n  =  30 

3 

12 

11 

3 

1 

0 

P.  lepinium,  n  = 

30 

0 

1 

3 

20 

6 

0 

P.  bulbiceps,  n  = 

34 

0 

0 

4 

20 

9 

1* 

P.  suspectus,  n  = 

3 

0 

0 

3 

0 

0 

0 

*  USNM  215611,  recounted  after  Markle  and  Sedberry  (1978) 
who  reported  it  as  17. 


Vertebrae  symmetrical,  no  geographical  vari- 
ation exhibited  numerically.  Caudal  fin  rays  vari- 
able, with  two  epural,  4-6  upper  hypural,  and  3- 
5  lower  hypural  rays.  Last  anal  ray  associated 
with  second  preural  vertebra,  last  dorsal  ray  as- 
sociated with  third  or  fourth  preural.  Dorsal  fin 
origin  associated  with  vertebrae  4-7,  with  no  free 
pterygiophores.  All  dorsal  elements  soft  rays,  first 
one  not  segmented,  but  bilaterally  divided  at  least 
near  base.  Anal  fin  origin  associated  with  ante- 
penultimate to  ultimate  precaudal  vertebrae,  with 
3-9  anal  fin  pterygiophores  inserted  anterior  to 
haemal  spine  of  first  caudal  vertebra. 

Gill  rakers  short,  roughly  triangular  in  adults, 
more  slender  and  longer  in  juveniles;  adults  with 
3-4  epibranchial  rakers  and  usually  12-14  on 
ceratobranchial;  raker  denticles  absent.  Bran- 
chiostegal  rays  six:  four  articulating  with  cera- 
tohyal  and  two  with  epihyal.  Two  nublike  pyloric 
caeca,  not  longer  in  smallest  specimens  than  larg- 
est, as  in  some  zoarcids.  Pseudobranch  filaments 
relatively  long,  about  one  eye  diameter  in  length 
in  adults  (less  in  smallest  specimens),  numbering 
3-7. 

Fresh  specimens  uniformly  dark  brown,  head 
and  most  of  pectoral  fins  darker,  but  not  black. 
Pectoral  base  pale  in  some  preserved  specimens, 
with  irregular  gray  mottling  in  IOS  DISCOV- 
ERY sta.  51803.  Dorsal  fin  of  this  juvenile  and 
a  large  adult  (IOS  DISCOVERY  sta.  10884)  with 
thin  white  bands  over  black  background  mostly 
at  mid-body  and  near  tail  tip;  anal  fin  with  fewer 
white  bands.  Orobranchial  chamber  pale  to  dark 
brown.  Abdomen  and  eyes  dark  blue  in  freshly 
preserved  specimens. 

DISTRIBUTION.— Presently  known  from  both 
sides  of  the  North  Atlantic  at  depths  of  2,400- 
4,780  m,  and  off  the  Queen  Charlotte  Islands, 
British  Columbia,  to  the  Gulf  of  Panama,  North 
Pacific,  at  depths  of  2,601-4,000  m  over  brown 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  3 


and  green  mud  bottoms,  where  the  species  may 
swarm  near  food  falls  or  baited  traps  (Fig.  3). 

Pachycara  suspectum  (Garman,  1899) 

(Figure  4) 
Phucocoetes  suspectus  Garman,  1899:137,  pi.  XXX,  fig.  3,  3a. 

MATERIAL  EXAMINED.  — MCZ  28683  (holotype);  offAcapul- 
co,  Mexico;  16°33.0'N,99°52.5'W;  1 ,207  m;  ALBATROSS  sta. 
3418;  11  Apr.  1890.  SIO  68-106  (1);  Gulf  of  California; 
25°56.6'N,  110°37.5'W;  1,280  m;  22  Jan.  1968.  ZMUC 
P-761085  (1);  Gulf  of  Panama;  07°22'N,  79°32'W;  915-975 
m;  GALATHEA  sta.  739;  15  May  1952. 

DIAGNOSIS.— A  species  of  Pachycara  as  de- 
fined by  Anderson  (1984)  distinguished  by  the 
following  combination  of  characters:  pelvic  fins 
present;  lateral  line  with  ventral  branch  only; 
scales  absent  on  nape;  vertebrae  21-23  +  84-87 
=  105-1 10;  head  length  15.6-17.7%  SL;  anal  fin 
origin  associated  with  vertebrae  20-22. 

DESCRIPTION.  — Counts  and  measurements 
presented  in  Table  1  were  compiled  from  all  three 
specimens  (females),  218-280  mm  SL;  holotype 
badly  damaged.  Head  large,  ovoid,  deepest  at 
occiput,  dorsoventrally  depressed.  Snout  acute, 
steeply  sloping  anteriorly.  Body  short,  deep,  ovoid 
in  cross  section.  Skin  firm,  not  gelatinous,  but 
pliable  around  mouth  and  cheeks  due  to 
subdermal  lipid  layer.  Scales  minute,  cycloid, 
imbedded,  sparse  anteriorly,  covering  body  and 
tail,  absent  on  unpaired  fins,  head,  nape,  and 
pectoral  base.  Eye  ovoid,  entering  dorsal  profile 
of  head  when  viewed  laterally.  Gill  slit  long,  ex- 
tending ventrally  to  near  pelvic  base.  Slight, 
squared-off  lobe  formed  at  dorsal  margin  of  oper- 
culum.  Single  pair  of  nostrils  at  snout  tip,  nasal 
tube  not  reaching  upper  lip.  Pectoral  fin  large, 
its  origin  well  below  body  midline;  posterior 
margin  wedge-shaped;  lowermost  6-7  rays  short- 
er and  thicker  than  others,  their  tips  very  slightly 
exserted. 

Mouth  terminal,  upper  jaw  extending  poste- 
riorly to  vertical  through  middle  of  eye.  No  epi- 
dermal prickles  on  lips  or  snout.  Palatine  mem- 
brane (oral  valve)  moderately  developed,  reaching 
anterior  margin  of  vomer.  Teeth  in  jaws  and  pal- 
ate small,  conical.  Jaw  teeth  in  two  rows  ante- 
riorly, blending  into  single  posterior  row  in  all 
three  specimens.  Vomerine  teeth  in  irregular 
patch;  palatine  teeth  in  single  or  double  (ZMUC 
P-761085)  row. 

Cephalic  lateralis  system  reduced,  numerical 
variation  not  detected  (two  specimens  damaged). 


Postorbital  pores  one,  three,  and  four  present  in 
SIO  68-106.  Two  pairs  of  anterior  supraorbital 
(nasal)  pores,  one  set  mesial  to  nasal  tube,  the 
other  dorsally.  Seven  suborbital  bones  present 
from  which  emanate  six  pores  from  ventral  ra- 
mus  of  chain  and  one  from  ascending  ramus  be- 
hind eye,  just  below  first  postorbital  pore.  Eight 
preoperculomandibular  pores,  four  emanating 
from  dentary,  one  from  anguloarticular,  and  three 
from  preopercle.  Interorbital  and  occipital  pores 
absent.  Body  lateral  line  of  single,  complete,  ven- 
tral branch  only;  lateral  line  originating  just  pos- 
terior to  fourth  postorbital  pore,  coursing  steeply 
downward  on  anterior  portion  of  abdomen,  then 
running  just  above  anal  fin  to  tail  tip  (Garman 
1899,  pi.  XXX). 

Vertebrae  symmetrical.  Caudal  fin  with  one 
epural  and  four  upper  and  lower  hypural  rays  (in 
two).  Last  anal  ray  associated  with  second  pre- 
ural  vertebra,  last  dorsal  ray  associated  with 
fourth  preural  vertebrae.  Dorsal  fin  origin  asso- 
ciated with  vertebrae  2-3,  with  no  free  pteryg- 
iophores.  All  dorsal  elements  soft  rays,  first  one 
not  segmented,  but  bilaterally  divided.  Anal  fin 
origin  associated  with  penultimate  precaudal 
vertebra,  with  2-4  anal  fin  pterygiophores  in- 
serted anterior  to  haemal  spine  of  first  caudal 
vertebra. 

Gill  rakers  short,  roughly  triangular,  ventral- 
most  blunt,  but  dorsalmost  5-6  with  sharp  tips; 
raker  denticles  absent.  Branchiostegal  rays  six: 
four  articulating  with  ceratohyal  and  two  with 
epihyal.  Two  nublike  pyloric  caeca.  Pseudo- 
branch  filaments  short,  numbering  3-4. 

Coloration  of  fresh  specimens  not  confirmed, 
although  Garman  (1899)  reported  holotype  as 
"blackish."  Recent  material  greatly  faded.  One 
gravid  female,  218  mm  SL  (SIO  68-106),  with 
ova  averaging  4.4  mm  in  diameter. 

DISTRIBUTION.— Currently  known  from  the 
Gulf  of  California  south  to  the  Gulf  of  Panama 
at  depths  of  9 1 5-1 ,280  m  over  green  and  brown 
mud  bottoms. 

Pachycara  gymninium  n.  sp. 

(Figure  5) 

Lycodes  sp.  Hubbs  et  al.,  1979:14. 
Lycenchelys  "E."  Pearcy  et  al.,  1982:387. 

HOLOTYPE. -USNM  280121  (female,  422  mm  SL);  British 
Columbia,  W  of  Tasu  Sound,  Queen  Charlotte  Islands; 
52°38.0'N,  132°05.8'W;  trap,  2,744  m;  TALAPUS  set  18;  4 
Feb.  1980. 


ANDERSON  AND  PEDEN:  NORTH  PACIFIC  PACHYCARA 


89 


FIGURE  4.    Pachycara  suspectum  (Garman),  SIO  68-106, 
218  mm  SL,  from  the  Gulf  of  California. 


PARATYPES.— British  Columbia:  BCPM  980-121  (1);  same 
as  holotype.  BCPM  979-11365  (2);  W  of  Tasu  Sound;  trap, 
2,744  m;  EASTWARD  HO  set  5;  16  Aug.  1979.  BCPM  979- 
1 1 369  (6);  W  of  Tasu  Sound;  trap,  2,889  m;  EASTWARD  HO 
set  12;  22  Aug.  1979.  BCPM  980-100  (1);  W  of  Tasu  Sound; 
trap,  2,926  m;  EASTWARD  HO  set  18;  Aug.  1979.  USNM 
221 1 13  (1);  recat.  from  BCPM  979-1 1369.  NMC  86-0444  (1); 
W  of  Tasu  Sound;  trap,  2,743-2,926  m;  EASTWARD  HO  set 
11;  21  Aug.  1979.  Cascadia  Abyssal  Plain:  CAS  60524  (1); 
45C18. 7-18. 5'N,  126°34.4-28.6'W;  trawl,  2,750  m;  YA- 
QUINA  sta.  CP-2C,  BMT  265;  18  Feb.  1971.  CAS  60525  (1); 
45-21. 5-21. 7'N,  127°33.0-38.5'W;  trawl,  2,800  m;  YA- 
QUINA  sta.  CP-3C,  BMT  280;  1 8  May  1971.  California:  OSUO 
2558  (1);  Mendocino  Ridge;  40°45.8'N,  127°31.6'W;  trap,  3,225 
m;  USGS  S.  P.  LEE  sta.  L2-86-5T;  20-21  July  1986.  SIO  67- 
121  (36);  SW  of  Cordell  Bank;  37°57.4'N,  123°40.5'W;  trap, 
2,707-3,2 19m;  THOMAS  WASHINGTON;  16-1 7  June  1967. 
Mexico:  SIO  65-69  (3);  off  Guadalupe  Isl.,  4  mi  E  of  southern 
tip;  trap,  1,829  m;  4-5  Mar.  1965.  SIO  68-120  (6);  Gulf  of 
California;  25°35.4'N,  109°46.0'W;  trap,  over  1,728  m; 
THOMAS  WASHINGTON;  24-25  Jan.  1968.  SIO  70-12  (3); 
offBaja  California  Norte;  31°10.0-06.9'N,  1 18°52.6-53.4'W; 
trap,  2,524  m;  MELVILLE;  16-17  Dec.  1969. 

ADDITIONAL  MATERIAL.  — SIO  66-19  (2);  off  Guadalupe  Isl., 
Mexico;  29°07.8'N,  118'12.5'W;  trap,  2,378  m;  ALEX- 
ANDER AGASSIZ;  1-2  Apr.  1966. 

DIAGNOSIS.— A  species  of  Pachycara  as  de- 
fined by  Anderson  (1984)  distinguished  by  the 
following  combination  of  characters:  pelvic  fins 
present,  their  lengths  4.8-1 1.3%  HL;  mediolat- 
eral  branch  of  lateral  line  originating  in  pectoral 
axil  just  posterior  to  vertical  through  pectoral 


base;  scales  absent  on  nape,  or,  if  present,  not 
extending  anterior  to  line  connecting  anterodor- 
sal  edges  of  gill  slits;  pectoral  fin  length  63.0- 
70.8%  HL;  vertebrae  28-31  +  73-80=  102-109; 
head  length  12.0-15.6%  SL;  anal  fin  origin  as- 
sociated with  vertebrae  27-3 1 . 

DESCRIPTION.  — Counts  and  measurements 
presented  in  Table  1  were  compiled  from  all 
known  specimens,  76-422  mm  SL  (including  one 
gravid  female).  The  following  description  is  based 
on  10  adult  males,  nine  adult  females,  42  juve- 
niles of  both  sexes,  and  two  eviscerated  adults. 
Head  large,  ovoid,  wider  in  adults  than  juveniles. 
Body  short,  deep,  broader  in  cross  section  in 
adults  than  juveniles.  Tail  laterally  compressed, 
more  so  posteriorly,  tapering  gradually  to  tip. 
Skin  firm,  not  gelatinous,  but  pliable  around  head, 
especially  near  mouth,  cheeks  and  nape,  owing 
to  subdermal  lipid  layer.  Scales  minute,  cycloid, 
imbedded,  covering  body,  tail  and  vertical  fins 
to  about  half  their  height;  usually  no  scales  on 
nape,  but  when  present,  restricted  to  small  patches 
immediately  anterior  to  dorsal  fin  origin,  or, 
rarely,  extending  anteriorly  to  line  connecting 
anterodorsal  edges  of  gill  slits.  Scales  usually  ab- 
sent on  pectoral  base,  when  present,  usually  in 
irregular,  ovoid  patch  not  extending  from  dorsal 
to  ventral  margin  of  fin  and  not  extending  onto 
fin  more  than  1 5%  of  its  length.  Scales  on  right 
cheek  in  small  patch  between  preoperculoman- 
dibular  pores  seven  and  eight  in  one  specimen 
(SIO  67-121,  235  mm  SL  female),  scales  other- 
wise absent  on  head.  No  scales  developed  in 
smallest  specimen,  76  mm  SL.  Eye  circular,  larg- 
er in  proportion  to  head  in  small  specimens  than 
largest  adults.  Gill  slit  somewhat  shortened,  usu- 
ally extending  ventrally  to  just  above  ventral 
margin  of  pectoral  base,  but  restricted  to  mid- 
height  of  pectoral  base  in  a  few  specimens,  and 
extending  to  below  ventral  margin  of  pectoral 
base  in  a  few  others.  Slight,  rounded  or  squared- 


TL.B. 


FIGURE  5.    Pachycara  gymninium  n.  sp.,  USNM  280121  (holotype,  422  mm  SL),  off  the  Queen  Charlotte  Islands,  British 
Columbia. 


90 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  3 


FIGURE  6.    Pachycara  lepinium  n.  sp.,  USNM  280120  (holotype,  465  mm  SL),  off  the  Queen  Charlotte  Islands,  British 
Columbia. 


off,  posteriorly-directed  lobe  formed  at  dorsal 
margin  of  operculum,  its  dorsalmost  margin  usu- 
ally straight,  but  deeply  concave  in  one  speci- 
men. Single  pair  of  nostrils  at  snout  tip,  nasal 
tube  very  short,  usually  not  reaching  upper  lip 
(reaching  upper  lip  only  in  few  juveniles  less  than 
220  mm  SL).  Pectoral  fin  large,  its  origin  just 
below  body  midline;  posterior  margin  nearly 
evenly  rounded  (except  in  few  injured  speci- 
mens); ventralmost  3-5  rays  shorter  and  thicker 
than  others,  their  tips  not  at  all,  or  just  slightly 
exserted;  pectoral  rays  usually  15-16  (Table  2). 

Mouth  terminal,  upper  jaw  slightly  longer  in 
relation  to  head  in  adult  males  than  comparably 
sized  adult  females;  upper  jaw  40.5-52.8%  HL 
(je  =  43.9%,  n  =  12)  in  males  over  280  mm  SL, 
37.7-41.8%  HL  (Jc  =  39.5%,  n  =  1 1)  in  females 
over  295  mm  SL.  Minute,  epidermal  prickles  on 
anterior  portions  of  lips  and  chin,  these  often 
nearly  rubbed  off  in  large  specimens.  Palatine 
membrane  (oral  valve)  moderately  developed, 
almost  reaching  anterior  margin  of  vomer  in  most 
specimens.  Teeth  in  jaws  and  palate  small,  con- 
ical, dentition  not  sexually  dimorphic.  Jaw  teeth 
in  single  row  anteriorly  in  smallest  specimens, 
in  4-5  anterior  rows  in  adults,  blending  into  sin- 
gle posterior  row  in  all.  Vomerine  teeth  in  irreg- 
ular patch  or  arched  row;  palatine  teeth  in  single 
row. 

Cephalic  lateralis  pore  system  reduced,  nu- 
merical variation  in  pores  of  postorbital  and  sub- 
orbital  canals  only.  Postorbital  pores  one,  three, 
and  four  present  on  both  sides  in  one  specimen; 
pores  one  and  four  only  present  in  1 8;  pores  one 
and  four  on  one  side,  pores  one,  three,  and  four 
on  the  other  in  nine  specimens.  Two  pairs  of 
nasal  pores,  one  set  mesial  to  nasal  tube,  the 
other  posteromesially.  Usually  six  suborbital 
pores,  all  emanating  from  ventral  ramus  of  sub- 
orbital  bone  chain,  but  a  seventh  pore  emanating 
from  ascending  ramus  behind  eye  just  below  first 
postorbital  pore  in  seven  specimens,  and,  in  one, 


from  just  above  sixth  pore  on  cheek.  Eight  pre- 
operculomandibular  pores,  four  emanating  from 
dentary,  one  from  anguloarticular,  and  three  from 
preopercle.  Interorbital  and  occipital  (supratem- 
poral)  pores  absent.  Body  lateral  line  of  two 
branches:  mediolateral,  originating  in  pectoral 
axil  on  vertical  through  pectoral  base,  complete 
to  tail  tip,  and  ventral,  originating  just  posterior 
to  fourth  postorbital  pore,  coursing  gradually 
across  body  to  just  above  anus,  then  running  to 
tail  tip  above  anal  fin  base. 

Vertebrae  symmetrical,  no  geographical  vari- 
ation exhibited  numerically.  Caudal  fin  rays  vari- 
able, with  one  or  two  epural,  4-6  upper  hypural, 
and  4-5  lower  hypural  rays.  Last  anal  ray  asso- 
ciated with  second  preural  vertebra,  last  dorsal 
ray  associated  with  fourth  preural  vertebra.  Dor- 
sal fin  origin  associated  with  vertebrae  3-5,  with 
no  free  pterygiophores.  All  dorsal  elements  soft 
rays,  first  one  not  segmented,  but  bilaterally  di- 
vided at  least  near  base.  Anal  fin  origin  associ- 
ated with  antepenultimate  to  ultimate  precaudal 
vertebrae,  with  3-5  pterygiophores  inserted  an- 
terior to  haemal  spine  of  first  caudal  vertebra. 

Gill  rakers  short,  roughly  triangular  in  adults, 
more  slender,  but  not  pointed,  in  smallest  spec- 
imens; adults  with  3-5  epibranchial  rakers  and 
usually  1 1-1 3  on  ceratobranchial;  raker  denticles 
absent.  Branchiostegal  rays  six:  four  articulating 
with  ceratohyal  and  two  with  epihyal.  Two  nub- 
like  pyloric  caeca,  not  appreciably  longer  in 
smallest  specimens  than  largest.  Pseudobranch 
filaments  relatively  long,  about  one  eye  diameter 
in  length  (or  more  in  smallest  specimens),  num- 
bering 3-5. 

Fresh  specimens  uniformly  dark  brown,  head 
and  pectoral  fin  darker,  almost  black.  Lining  of 
orobranchial  chamber  dark  brown  or  black.  Mar- 
gins of  vertical  fins  and  peritoneum  black. 

One  gravid  female  known  (CAS  60524,  350 
mm  SL),  with  80  ova  5.2-7.2  mm  in  diameter 
(X  =  5.8  mm).  Stomach  contents  of  several  spec- 


ANDERSON  AND  PEDEN:  NORTH  PACIFIC  PACHYCARA 


91 


imens  consisted  of  amphipods,  isopods,  poly- 
chaetes,  and  trap  bait. 

DISTRIBUTION.— Off  the  Queen  Charlotte  Is- 
lands, British  Columbia,  south  to  ofFGuadalupe 
Island,  Mexico,  and  in  the  Gulf  of  California  in 
1 ,829-3,2 1 9  m  over  brown  and  green  mud  bot- 
toms. 

ETYMOLOGY.— From  the  Greek  yv^vba  (naked) 
and  Iviov  (nape)  in  reference  to  the  species'  scale- 
less  head. 

Pachycara  lepinium  n.  sp. 

(Figure  6) 

Lycodes  sp.  Hubbs  et  al.,  1979:14. 
Lycenchelys  "D."  Pearcy  et  ah,  1982:387. 

HOLOTYPE.-USNM  280120  (male,  465  mm  SL);  British 
Columbia,  W  of  Tasu  Sound,  Queen  Charlotte  Islands; 
52°38.0'N,  1 32°05.8'W;  trap,  2,744  m;  TALAPUS  set  8;  4  Feb. 
1980. 

PARATYPES.- British  Columbia:  BCPM  980-121  (1);  same 
as  holotype.  BCPM  980-98  (1);  W  of  Tasu  Sound;  trap,  2,889 
m;  EASTWARD  HO  set  6;  Aug.  1979.  NMC  86-0445  (1);  W 
of  Tasu  Sound;  trap,  2,744  m;  EASTWARD  HO  set  5;  16 
Aug.  1979.  USNM  221 1 14  (1);  W  of  Tasu  Sound;  trap,  2,889 
m;  EASTWARD  HO  set  12;  22  Aug.  1979.  Cascadia  Abyssal 
Plain:  CAS  60526  (1);  44°23.0-25.4'  N,  125°34.0-35.0'W;  trawl, 
2,878m;  YAQUINA  sta.  CP- 1 F,  BMT  4 1 0;  4  Aug.  1 974.  CAS 
60527  (1);  45°46.9^5.7'N,  126°35.5'W;  trawl,  2,665  m;  YA- 
QUINA sta.  CP-2A,  BMT  259;  15  Feb.  1971.  CAS  60528  (1); 
44°41. 7-43. 6'N,  125°33.5-32.5'W;  trawl,  2,825  m;  YA- 
QUINA sta.  CP-1E,  BMT  120;  6  Oct.  1969.  CAS  60529  (1); 
45°53.1-51.8'N,  126°39.0-42.5'W;  trawl,  2,713  m;  YA- 
QUINA sta.  CP-2A,  BMT  154;  16  Jan.  1970.  CAS  60530  (1); 
44°29.6-45.2'N,  126°29.0-35.9'W;  trawl,  2,818  m;  YA- 
QUINA sta.  CP-2D,  OTB  90;  23  Oct.  1965.  OSUO  2563  (1); 
45°09.3-10.5'N,  125°38.3-38.0'W;  trawl,  2,669  m;  YA- 
QUINA sta.  CP- 1C,  BMT  93;  15-16July  1 969.  LACM  443 15-1 
(1);  44°39.7-39.0'N,  126°39.6-44.8'W;  trawl,  2,816  m;  CA- 
YUSE  sta.  CP-2E,  BMT  253;  30  Sept.  1970.  California:  SIO 
66-50  (1);  off  Cape  Mendocino;  40°34.6-34.7'N,  125°51.4- 
51.5'W;  trap,  2,970  m;  ALEXANDER  AGASSIZ;  21-22  May 
1966.  SIO  60-475  (1);  off  Farallon  Isl.;  trawl,  ca.  1,800  m;  10 
Nov.  1960.  SIO  59-288  (1);  San  Clemente  Basin;  32°39.0'N, 
1 18°09.3'W;  trap,  2,027  m;  23-24  July  1959.  SIO  59-289  (3); 
San  Clemente  Basin;  32°39.0'N,  118°09.3'W;  trap,  2,023  m; 
23-24  July  1959.  SIO  65-452  (4);  San  Clemente  Basin; 
32°38.7'N,  1 18°08.4'W;  trap,  2,008  m;  ALEXANDER  AGAS- 
SIZ; 25  Sept.  1965.  LACM  321 14-2  (1);  San  Clemente  Basin; 
32°38.0-37.5'N,  1 18°09.0-06.4'W;  trap,  1,829-1,939  m;  VE- 
LERO  IV  sta.  12521;  10  Dec.  1968.  Mexico:  SIO  59-365  (5); 
off  Cabo  Colnet;  31°02.7'N,  116°59.0'W;  trap,  2,140  m;  27- 
28  Oct.  1959. 

ADDITIONAL  MATERIAL. -SIO  59-364  (1);  off  Cabo  Colnet, 
Mexico;  31°00.5'N,  1 18°06.0'W;  trap,  1,728  m;  27  Oct.  1959. 
SIO  59-366  (1);  off  Cabo  Colnet,  Mexico;  31°02.7'N, 
116°59.3'W;  trap,  2,140  m;  27-28  Oct.  1959. 

DIAGNOSIS.— A  species  of  Pachycara  as  de- 
nned by  Anderson  (1984)  distinguished  by  the 
following  combination  of  characters:  pelvic  fins 


present,  their  length  1 1.5-17.3%  HL;  mediolat- 
eral  branch  of  lateral  line  originating  posterior 
to  pectoral  fin  margin;  scales  present  on  nape, 
extending  to  interorbital  region;  pectoral  fin  length 
72. 5-89.9%  HL;  vertebrae  23-26  +  80-94=  105- 
120;  head  length  13.8-15.9%  SL;  anal  fin  origin 
associated  with  vertebrae  23-26. 

DESCRIPTION.  — Counts  and  measurements 
presented  in  Table  1  were  compiled  from  all 
known  specimens,  221-597  mm  SL  (including 
one  gravid  female).  The  following  description  is 
based  on  six  adult  males,  eight  adult  females, 
and  1 6  juveniles  of  both  sexes.  Head  large,  ovoid, 
wider  in  adults  than  juveniles.  Body  relatively 
short,  deep,  broader  in  cross  section  in  adults 
than  juveniles.  Tail  laterally  compressed,  more 
so  posteriorly,  tapering  gradually  to  tip.  Skin  firm, 
not  gelatinous,  but  somewhat  pliable  around 
head,  especially  near  mouth  and  cheeks,  due  to 
subdermal  lipid  layer.  Scales  minute,  cycloid, 
imbedded,  covering  body,  tail,  and  vertical  fins 
almost  to  their  margins;  scales  on  nape  extend 
anteriorly  to  vertical  one  eye  diameter  anterior 
to  dorsalmost  edge  of  gill  opening  in  smallest 
specimen,  to  interorbital  region  in  adults.  Scales 
present  in  pectoral  axil  and  on  base,  extending 
on  outer  surface  of  fin  to  half  or  more  its  length; 
scales  present  on  cheeks.  Eye  circular,  relatively 
larger  in  proportion  to  head  in  smallest  speci- 
mens than  largest.  Gill  slit  long,  extending  ven- 
trally  almost  to  pelvic  fin  origin;  slight,  rounded, 
posteriorly-directed  lobe  formed  at  dorsal  mar- 
gin of  operculum.  Single  pair  of  nostrils  at  snout 
tip,  nasal  tube  very  short,  not  reaching  upper  lip. 
Pectoral  fin  large,  its  origin  just  below  body  mid- 
line;  posterior  margin  almost  evenly  rounded; 
lowermost  5-6  rays  shorter  and  thicker  than  oth- 
ers, their  tips  somewhat  excised;  pectoral  rays 
usually  17-18  (Table  2). 

Mouth  subterminal  in  largest  adults,  terminal 
in  young;  upper  jaw  longer  in  relation  to  head  in 
adult  males  than  comparably  sized  adult  females; 
upper  jaw  48.8-52.0%  HL  (X  =  50.3,  n  =  5)  in 
males  over  380  mm  SL,  41.7-46.4%  HL  (X  = 
44.8%,  n  =  7)  in  females  over  380  mm  SL.  Mi- 
nute, epidermal  prickles  on  anterior  portions  of 
lips  and  chin.  Palatine  membrane  (oral  valve) 
well  developed,  reaching  anterior  edge  of  vomer. 
Teeth  in  jaws  and  palate  small,  conical;  dentition 
not  sexually  dimorphic.  Jaw  teeth  in  two  irreg- 
ular rows  anteriorly  in  smallest  specimens,  in  4- 
5  anterior  rows  in  adults,  blending  into  single 
posterior  row  in  all.  Vomerine  teeth  in  irregular 


92 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  3 


patch;  palatine  teeth  in  one  or  two  (largest  spec- 
imens) irregular  rows. 

Cephalic  lateralis  pore  system  reduced,  nu- 
merical variation  present  only  in  pores  of  post- 
orbital  canal.  Postorbital  pores  one,  three,  and 
four  present,  emanating  from  frontal,  pterotic 
and  lateral  extrascapular,  respectively;  pores  one 
and  four  only  in  two  specimens;  pores  one  and 
four  on  one  side,  one,  three,  and  four  on  the  other 
in  two  other  specimens.  Two  pairs  of  nasal  pores, 
one  set  mesial  to  nasal  tube,  the  other  set  pos- 
teromesially.  Six  suborbital  pores,  all  emanating 
from  ventral  ramus  of  suborbital  bone  chain. 
Eight  preoperculomandibular  pores,  four  ema- 
nating from  dentary,  one  from  anguloarticular, 
and  three  from  preopercle.  Interorbital  and  oc- 
cipital (supratemporal)  pores  absent.  Body  lat- 
eral line  of  two  branches:  mediolateral,  origi- 
nating at  or  just  posterior  to  pectoral  fin  margin, 
complete  to  tail  tip,  and  ventral,  originating  just 
posterior  to  fourth  postorbital  pore,  coursing 
gradually  across  body  to  just  above  anus,  then 
running  just  above  anal  fin  to  tail  tip. 

Vertebrae  symmetrical.  Individuals  from  ba- 
sins of  southern  California  and  northern  Baja 
California  with  significantly  more  vertebrae  than 
those  from  the  north  (specimens  from  British 
Columbia  to  Cape  Mendocino  with  105-110 
vertebrae  [n  =  13,  *  =  107.2,  SD  =  1.58],  spec- 
imens from  the  Farallon  Islands  to  Baja  Cali- 
fornia with  1 1 6-1 20  vertebrae  [n  =  1 6,  x  =  1 1 7.2, 
SD  =  0.49]).  Caudal  fin  rays  variable,  with  two 
epural,  3-5  upper  hypural  and  3-5  lower  hypural 
rays.  Last  anal  ray  associated  with  second  preural 
vertebra,  last  dorsal  ray  associated  with  third  or 
fourth  preural  vertebrae.  Dorsal  fin  origin  asso- 
ciated with  vertebrae  4-5,  with  no  free  pteryg- 
iophores.  All  dorsal  fin  elements  soft  rays,  first 
1-2  not  segmented,  but  bilaterally  divided  at  least 
near  base.  Anal  fin  origin  associated  with  ante- 
penultimate to  ultimate  precaudal  vertebra,  with 
2-5  pterygiophores  inserted  anterior  to  haemal 
spine  of  first  caudal  vertebra. 

Gill  rakers  short,  roughly  triangular,  relatively 
longer  on  dorsal  portion  of  ceratobranchial;  larg- 
est adults  with  four  epibranchial  rakers  and  1 3- 
15  on  ceratobranchial;  raker  denticles  absent. 
Branchiostegal  rays  six:  four  articulating  with 
ceratohyal  and  two  with  epihyal.  Two  nublike 
pyloric  caeca,  not  appreciably  longer  in  smallest 
specimens.  Pseudobranch  filaments  relatively 
long,  about  one  eye  diameter  in  length  in  both 
adults  and  juveniles,  numbering  2-5. 


Fresh  specimens  uniformly  dark  brown,  head 
and  pectoral  fin  darker,  almost  black.  Lining  of 
orobranchial  chamber  dark  brown  or  black.  Mar- 
gins of  vertical  fins  and  peritoneum  black. 

One  gravid  female,  412  mm  SL  (SIO  59-364, 
ovary  somewhat  damaged),  had  ova  averaging 
4.3  mm  in  diameter.  Stomach  contents  of  several 
specimens  included  polychaetes,  bivalves,  and 
amphipods. 

DISTRIBUTION.— Off  the  Queen  Charlotte  Is- 
lands, British  Columbia,  south  to  ofTGuadalupe 
Island,  Mexico,  at  depths  of  1 ,728-2,970  m  over 
brown  and  green  mud  bottoms.  Often  taken  in 
traps  and  trawls  with  P.  gymninium,  but  this 
species  not  yet  known  from  the  Gulf  of  Califor- 
nia. 

ETYMOLOGY.— From  the  Greek  XCTTIS  (scale)  and 
Iviov  (nape)  in  reference  to  the  species'  scaly  head. 

DISCUSSION 

The  distribution  of  P.  bulbiceps  may  at  first 
glance  appear  anomalous,  in  that  the  species  is 
widely  distributed  in  the  North  Atlantic,  but  oth- 
erwise restricted  to  the  northeastern  Pacific  con- 
tinental rise  and  adjacent  abyss.  Anderson  (1988) 
provided  evidence  that  the  pan-American,  trop- 
ical, abyssal  fish  fauna  was  sundered  by  tectonic 
uplifting  in  the  region  of  the  Panamanian  isth- 
mus at  least  by  the  mid-Miocene.  The  present 
Caribbean  deep-water  ichthyofauna  is  very  dif- 
ferent from  that  of  the  eastern  tropical  Pacific 
(Garman  1899;  Anderson  et  al.  1986).  Thus,  the 
Pacific  and  Atlantic  stocks  of  P.  bulbiceps  would 
have  had  ample  time  to  speciate  if  there  were  no 
South  Atlantic-South  Pacific  populations  linking 
the  northern  groups.  Because  abyssal  trawl  sam- 
ples from  the  southwestern  Atlantic  and  south- 
eastern Pacific  are  generally  lacking  for  most  re- 
gions, isolation  of  the  northern  populations 
cannot  be  evaluated.  Based  on  the  paleogeog- 
raphy  of  the  American  tropics  cited  by  Anderson 
(1988),  the  Atlantic  and  Pacific  populations 
probably  are  not  isolated  from  each  other,  and 
future  abyssal  collections  off  Chile,  Argentina, 
Brazil,  and  probably  southern  Africa  should  pro- 
duce specimens  of  P.  bulbiceps.  Among  the  char- 
acters presented  in  Table  1 ,  as  well  as  traditional 
data  analysis  used  by  us  before  (Peden  and  An- 
derson 1978;  Peden  1981;  Anderson  1982),  in- 
cluding features  of  the  axial  skeleton,  gill  arches, 
dentition,  squamation,  and  lateral  lines,  no  nu- 
merical or  morphological  differences  were  de- 


ANDERSON  AND  PEDEN:  NORTH  PACIFIC  PACHYCARA 


93 


tected  between  the  Atlantic  and  Pacific  popula- 
tions, thus  isolation  seems  improbable.  This  is 
in  contrast  to  North  Atlantic  and  North  Pacific 
populations  of  the  grenadier  Coryphaenoides  ar- 
matus,  which  exhibited  minor  electrophoretic  but 
significant  biometric  differences,  resulting  in 
Wilson  and  Waples'  (1984)  placement  of  these 
forms  as  separate  subspecies.  Similarly,  Small 
(1981)  found  specific  differences  in  the  cosmo- 
politan bathyal  genus  Antimora  and  recognized 
an  endemic  North  Pacific  species  (A.  microlepis), 
while  a  second  species  (A.  rostratd)  was  shown 
to  be  circumglobal  but  excluded  from  the  North 
Pacific. 

The  geographic  variation  noted  for  P.  lepi- 
nium,  but  not  found  in  P.  gymninium,  is  difficult 
to  explain,  but  may  relate  to  differing  environ- 
mental conditions  that  affect  vertebral  number 
in  the  basins  of  the  Southern  California  Bight. 
Still,  no  variation  was  noted  in  the  eight  indi- 
viduals of  P.  gymninium  from  this  area,  or  in  P. 
bulbiceps  along  its  entire  range.  If  the  distribu- 
tions of  these  species  are  broader  than  present 
samples  reflect  (and  they  probably  are),  signifi- 
cant geographic  variation  may  be  found  when 
adequate  material  is  analyzed  along  the  ranges 
of  all  species,  as  was  found  for  the  eelpouts  Ly- 
codapus  mandibularis  (Peden  1979)  and  Lycodes 
diapterus  (Peden  1981).  The  lack  of  variation 
noted  for  P.  gymninium  may  be  due  to  either  the 
present  poor  sample  size  or  is  a  valid  specific 
character,  as  found  in  another  eelpout  genus, 
Gymnelus  (Anderson  1982). 

ACKNOWLEDGMENTS 

For  assistance  with  specimen  loans  and  data 
management  we  thank  Karsten  Hartel,  Susan 
Jewett,  Shelly  Johnson,  Robert  Lavenberg,  Nigel 
Merrett,  William  Pearcy,  Richard  Rosenblatt, 
Jeffrey  Seigel,  and  David  Stein.  Ray  Scarsbrook 
collected  the  British  Columbia  material,  includ- 
ing holotypes  of  both  new  species  and  generously 
donated  them  to  the  British  Columbia  Provincial 
Museum.  Patricia  Drukker-Brammall  rendered 
Figures  1,  5,  and  6.  Craig  R.  Smith  generously 
provided  Figure  3. 

LITERATURE  CITED 

ANDERSON,  M.  E.  1982.  Revision  of  the  fish  genera  Gymnelus 
Reinhardt  and  Gymnelopsis  Soldatov  (Zoarcidae),  with  two 
new  species  and  comparative  osteology  of  Gymnelus  viridis. 
Natl.  Mus.  Nat.  Sci.,  Publ.  Zool.  (17):  1-76. 

.    1 984.   On  the  anatomy  and  phytogeny  of  the  Zoar- 


cidae (Teleostei:  Perciformes).  Ph.D.  Dissertation,  College 
of  William  and  Mary,  Williamsburg,  Virginia.  254  pp. 
.    1988.    Studies  on  the  Zoarcidae  (Teleostei:  Perci- 


formes) of  the  southern  hemisphere.  I.  The  Antarctic  and 
subantarctic  regions.  Antarct.  Res.  Sen,  Amer.  Geophys. 
Union,  Biol.  Antarct.  Seas  19  (in  press). 

ANDERSON,  M.  E.,  R.  E.  CRABTREE,  H.  J.  CARTER,  K.  J.  SULAK, 
AND  M.  D.  RICHARDSON.  1986.  Distribution  of  demersal 
fishes  of  the  Caribbean  Sea  found  below  2,000  meters.  Bull. 
Mar.  Sci.  37(3):794-807. 

ANDRIASHEV,  A.  P.  1 954.  Fishes  of  the  northern  seas  of  the 
USSR.  Zool.  Inst,  Acad.  Sci.,  Contr.  Fauna  USSR  53:1- 
566. 

.  1973.  Zoarcidae.  In  Checklist  of  the  fishes  of  the 

northeastern  Atlantic  and  of  the  Mediterranean.  J.-C.  Hu- 
reau  and  T.  Monod,  eds.  Unesco,  Paris,  1:540-547. 

.  1 986.  Zoarcidae  (including  Lycodidae).  In  Fishes  of 

the  north-eastern  Atlantic  and  the  Mediterranean.  P.  J.  P. 
Whitehead,  M.-L.  Bauchot,  J.-C.  Hureau,  J.  Nielsen,  and  E. 
Tortonese,  eds.  Unesco,  Paris,  3:1 130-1 150. 

FOWLER,  H.  W.  1936.  The  marine  fishes  of  West  Africa.  Bull. 
Amer.  Mus.  Nat.  Hist.  70(2):607-1493. 

GARMAN,  S.  1899.  Reports  on  an  exploration  off  the  west 
coasts  of  Mexico,  Central  and  South  America,  and  off  the 
Galapagos  Islands,  in  charge  of  Alexander  Agassiz,  by  the 
U.S.  Fish  Commission  steamer  "Albatross"  during  1891, 
Lieut.-Commander  Z.  L.  Tanner,  U.S.N.,  commanding. 
XXVI.  The  fishes.  Mem.  Mus.  Comp.  Zool.  Harv.  Coll.  24: 
1-431. 

GOSZTONYI,  A.  E.  1977.  Results  of  the  research  cruises  of 
FRV  "Walther  Herwig"  to  South  America.  XLVIII.  Revi- 
sion of  the  South  American  Zoarcidae  (Osteichthyes:  Blen- 
nioidei),  with  the  description  of  three  new  genera  and  five 
new  species.  Arch.  FischWiss.  27(3):  19 1-249. 

HUBBS,  C.  L.,  W.  I.  FOLLETT,  AND  L.  J.  DEMPSTER.     1 979.    List 

of  the  fishes  of  California.  Occ.  Papers  Calif.  Acad.  Sci.  (133): 
1-51. 

LEVITON,  A.  E.,  R.  H.  GIBBS,  JR.,  E.  HEAL,  AND  C.  E.  DAWSON. 
1985.  Standards  in  herpetology  and  ichthyology:  Part  I. 
Standard  symbolic  codes  for  institutional  resource  collec- 
tions in  herpetology  and  ichthyology.  Copeia  1985(3):802- 
832. 

MARKLE,  D.  F.  AND  G.  R.  SEDBERRY.  1978.  A  second  spec- 
imen of  the  deep-sea  fish,  Pachycara  obesa,  with  a  discussion 
of  its  classification  and  a  checklist  of  other  Zoarcidae  off 
Virginia.  Copeia  1978(l):22-25. 

MCALLISTER,  D.  E.  AND  E.  I.  S.  REES.  1964.  A  revision  of 
the  eelpout  genus  Melanostigma  with  a  new  genus  and  with 
comments  on  Maynea.  Bull.  Natl.  Mus.  Canada  199:85- 
109. 

MERRETT,  N.  R.  AND  N.  B.  MARSHALL.  1981.  Observations 
on  the  ecology  of  deep-sea  bottom-living  fishes  collected  off 
northwest  Africa  (08°-27°).  Prog.  Oceanogr.  9:185-244. 

PAPPENHEIM,  P.  1912.  Die  Fische  der  Deutschen  Siidpolar- 
Expedition  1901-1903.  I.  Die  Fische  der  Antarktis  und 
Subantarktis.  Deutsche  Sudpolar  Exped.  13:160-182. 

PEARCY,  W.  G.,  D.  L.  STEIN,  AND  R.  S.  CARNEY.  1982.  The 
deep-sea  benthic  fish  fauna  of  the  northeastern  Pacific  Ocean 
on  Cascadia  and  Tufts  Abyssal  Plains  and  adjoining  conti- 
nental slopes.  Biol.  Oceanogr.  l(4):375-428. 

PEDEN,  A.  E.  1979.  Meristic  variation  of  Lycodapus  man- 
dibularis (Pisces:  Zoarcidae)  and  oceanic  upwelling  on  the 
west  coast  of  North  America.  J.  Fish.  Res.  Bd.  Canada  36(1): 
69-76. 

.    1981.   Meristic  variation  of  four  fish  species  exhib- 


94 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  3 


iting  lowest  median  counts  in  Georgia  Strait,  British  Colum- 
bia. Can.  J.  Zool.  59(4):679-683. 

PEDEN,  A.  E.  AND  M.  E.  ANDERSON.  1978.  A  systematic  re- 
view of  the  fish  genus  Lycodapus  (Zoarcidae)  with  descrip- 
tions of  two  new  species.  Can.  J.  Zool.  56(9):  1925-1 961. 

SMALL,  G.  J.  1981.  A  review  of  the  bathyal  fish  genus  An- 
timora  (Moridae:  Gadiformes).  Proc.  Calif.  Acad.  Sci.  42(13): 
341-348. 

WILSON,  R.  R.,  JR.  AND  R.  S.  WAPLES.  1984.  Electrophoretic 
and  biometric  variability  in  the  abyssal  grenadier  Cory- 


phaenoides  armatus  of  the  western  North  Atlantic,  eastern 

South  Pacific  and  eastern  North  Pacific  oceans.  Mar.  Biol. 

80:227-237. 
ZUGMAYER,  E.     191  la.    Diagnoses  des  poissons  nouveaux 

provenant  des  campagnes  du  yacht  "Princesse-Alice"  ( 1 90 1 

a  1910).  Bol.  Inst.  Oceanogr.,  Monaco  (193):1-14. 
.    1 9 1 1  b.   Poissons  provenant  des  campagnes  du  yacht 

"Princesse-Alice."  Result.  Camp.  Sci.  Prince  Albert  I  35:1- 

174. 


CALIFORNIA  ACADEMY  OF  SCIENCES 

Golden  Gate  Park 
San  Francisco,  California  94 1 1 8 


PROCEEDINGS 

OF  THE 

CALIFORNIA  ACADEMY  OF  SCIENCES 


Vol.  46,  No.  4,  pp.  95-136,  22  figs.,  4  tables.  March  9,  1989 


THE  HARVESTMEN  FAMILY  PHALANGODIDAE. 

1.  THE  NEW  GENUS  CALICINA,  WITH  NOTES  ON  SITALCINA 

(OPILIONES:  LANIATORES) 

By 
Darrell  Ubick 

Department  of  Entomology,  California  Academy  of  Sciences,  Golden  Gate  Park, 
San  Francisco,  California  94118-4599 

and 
Thomas  S.  Briggs 

Department  of  Entomology,  California  Academy  of  Sciences,  Golden  Gate  Park, 
San  Francisco,  California  94118-4599 


ABSTRACT:  The  Californian  phalangodid  genus  Sitalcina  Banks  is  shown  to  be  polyphyletic,  and  is  restricted 
to  include  only  those  species  closely  related  to  the  type  species  Sitalces  californica  Banks.  A  new  genus, 
Calicina,  described  to  accommodate  most  of  the  remaining  species,  is  defined  by  the  presence  of  a  tarsal  spur 
on  the  male  palpus,  penis  glans  with  apically  directed  stylus,  ovipositor  with  double  row  of  apical  setae,  and 
ovipositor  cuticle  with  microspines.  At  least  the  first  of  these  characters  is  a  synapomorphy. 

Of  the  25  species  of  Calicina  recognized,  18  are  transferred  from  Sitalcina:  breva  Briggs  (elevated  from 
subspecies),  cloughensis  Briggs  and  Horn,  digita  Briggs  and  Horn,  ensata  Briggs,  kaweahensis  Briggs  and 
Horn,  keenea  Briggs,  macula  Briggs,  mariposa  Briggs,  minor  Briggs  and  Horn,  morroensis  Briggs,  palapraeputia 
Briggs,  piedra  Briggs,  polina  Briggs,  sequoia  Briggs  and  Horn,  serpentinea  Briggs  and  Horn,  sierra  Briggs  and 
Horn,  topanga  Briggs,  and  yosemitensis  Briggs.  The  remaining  species  are  new:  arida,  basalta,  conifera, 
diminua,  dimorphica,  galena,  and  mesaensis. 

Based  on  genitalic  characters,  primarily  glans  structure,  four  species  groups  and  nine  subgroups  are  rec- 
ognized. Their  relationships  are  hypothesized  using  cladistic  analysis  and  vicariance  biogeography.  The 
distribution  of  Calicina  is  strongly  concordant  with  that  of  the  slender  salamander,  Batrachoseps,  and  with 
the  presumed  distributions  of  exotic  terranes. 

Interspecific  variation  in  the  somatic  morphology  of  Calicina  appears  to  be  the  result  of  paedomorphosis 
and  troglobism.  These  modifications,  present  in  most  species,  are  probably  adaptations  to  xeric  environments 
and,  possibly,  competition  with  other  phalangodids. 

Received  October  7,  1987.  Accepted  May  11,  1988 

„  „  Phylogeny 1 2 1 

TABLE  OF  CONTENTS  „.  ,  ~~ 

Biogeography 125 

Abstract 95  Natural  History 130 

Introduction 96  Ecology 131 

Materials  and  Methods 96  Acknowledgments 135 

Taxonomy 9  8  Literature  Cited 1 3  5 

[95] 


96 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


INTRODUCTION 

A  major  problem  in  the  systematics  of  har- 
vestmen  is  the  apparent  scarcity  of  reliable  ge- 
neric characters.  This  stems  largely  from  the 
classic  studies  of  Roewer  (e.g.,  1 923)  which  over- 
emphasized the  importance  of  easily  visible  so- 
matic characters,  such  as  variation  in  the  tarsal 
count,  in  denning  genera.  This  practice  has  re- 
sulted in  a  confusing  proliferation  of  small  and 
monotypic  genera.  For  example,  the  1 7  known 
species  of  the  eastern  Nearctic  Phalangodidae  are 
currently  assigned  to  eight  genera  (Goodnight  and 
Goodnight  1942,  1967). 

In  sharp  contrast  to  this  general  trend  is  the 
treatment  of  the  phalangodids  of  California:  only 
two  genera  are  recognized  for  an  unusually  rich 
fauna  of  39  described  species.  One  of  these, 
Banksula  Roewer,  includes  strictly  cavernico- 
lous  species  having  uniformly  high  tarsal  counts 
(4-6-5-6)  and  the  other,  Sitalcina  Banks,  con- 
tains epigean  species  with  lower  counts  (3-5-5-5 
or  less).  Recent  studies  (Briggs  1974;  Briggs  and 
Ubick  1981)  have  demonstrated  that  Banksula 
is  well  denned  by  both  somatic  and  genitalic 
characters:  the  palpal  femur  is  armed  with  dorsal 
spines  and  the  penis  has  a  bifurcate  ventral  plate. 
On  the  other  hand,  Briggs  (1968)  found  no  de- 
rived character  uniting  the  29  species  of  Sital- 
cina. He  even  commented  on  the  heterogeneity 
of  the  genus,  listing  interspecific  differences  in 
size,  tarsal  count,  presence  or  absence  of  eyes, 
and  type  of  sexually  dimorphic  structures.  He 
tentatively  arranged  the  species  into  three  groups: 

(1)  those  with  sexually  dimorphic  trochanters, 

(2)  those  with  sexually  dimorphic  palps,  and  (3) 
those  lacking  sexual  dimorphisms.  However,  in 
the  absence  of  additional  characters  and  because 
of  the  overall  somatic  similarity  of  the  species, 
he  did  not  question  the  validity  of  the  genus. 

Our  present  investigations  of  the  genitalic 
morphology  of  Sitalcina  indicate  that  the  genus 
is  indeed  polypliyletic.  The  strongest  evidence 
comes  from  the  morphology  of  the  penis  where, 
on  the  basis  of  the  glans  structure,  two  groups  of 
species  are  evident.  The  first  is  characterized  by 
a  folded,  caudally  directed  glans  that  swings  open 
180°  along  a  dorsoapical  arc  during  expansion 
(Fig.  Ib).  The  second  is  characterized  by  an  api- 
cally  directed  glans  that  expands  by  a  rectilinear, 
telescoping  movement  out  of  the  shaft  (Fig.  6f, 
g).  Such  striking  genitalic  differences  are  com- 


parable to  those  used  to  distinguish  families 
(Martens  1986). 

It  is  now  evident  that  Sitalcina  must  be  re- 
stricted to  include  only  those  species  closely  re- 
lated to  the  type  species,  Sitalces  californica 
Banks.  This  monophyletic  group,  corresponding 
to  group  "1"  of  Briggs  (1968),  is  defined  by  the 
characters  listed  in  Table  1.  In  addition  to  the 
folding  glans,  all  Sitalcina  males  have  trochan- 
teral  spurs  on  leg  IV  (Fig.  la).  The  ovipositor  of 
Sitalcina  is  distinctive  in  various  features,  of 
which  the  imbricate  cuticle  sculpturing  is  pos- 
sibly synapomorphic. 

The  nine  species  that  now  comprise  Sitalcina 
(along  with  three  others  still  unplaced  that  ap- 
parently belong  to  other  genera)  will  be  treated 
in  a  future  paper.  The  purpose  of  this  study  is  to 
describe  the  new  genus  represented  by  the  re- 
maining species,  for  which  we  propose  the  name 
Calicina.  The  species  ofCalicina  are  revised  and 
their  morphological  and  spatial  relationships  hy- 
pothesized using  cladistics  and  vicariance  bio- 
geography.  Analyses  of  natural  history  and  ecol- 
ogy are  provided. 

MATERIALS  AND  METHODS 

The  800  specimens  examined  during  the  course 
of  this  study  were  collected  almost  exclusively 
by  Briggs  and  associates  over  the  past  two  de- 
cades. Specimens  from  early  collections  were 
originally  stored  in  Oudemans'  Fluid,  which  gave 
excellent  initial  preservation  (Briggs  1968:3). 
However,  it  is  now  evident  that  prolonged  stor- 
age in  this  solution  (about  10  years  or  more) 
discolors  the  specimens,  results  in  deterioration 
of  the  internal  tissues,  and  makes  genitalic  ex- 
traction difficult.  By  comparison,  specimens  pre- 
served in  75%  to  85%  ethanol  for  much  longer 
periods  do  not  show  signs  of  such  damage. 

The  genitalia  can  usually  be  extracted  by  sim- 
ply opening  the  operculum  and  squeezing  the 
abdomen  with  forceps.  When  this  fails  it  is  nec- 
essary to  push  out  the  genitalia  carefully  with  a 
probe  inserted  into  the  anal  region  or,  as  a  last 
resort,  to  dissect  it  out  of  the  body  cavity.  Males 
often  require  additional  treatment  with  KOH  in 
order  to  fully  expand  the  glans,  which  normally 
lies  retracted  within  the  shaft.  For  this,  entire 
specimens  are  soaked  in  cold  1 0%  KOH  solution 
for  about  two  minutes  and  then  transferred  to  a 
water  bath  where,  aided  by  gentle  squeezing  of 
the  abdomen,  maximum  expansion  can  occur. 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALICINA 


97 


TABLE  1 .    COMPARATIVE  MORPHOLOGY  OF  SITALCINA  AND  CALICINA. 


Sitalcina 

Calicina 

Total  length 
Eyes 
Tarsal  count 

1.50-2.40  mm 
present 
3-5-5-5 

0.77-1.67  mm 
present  or  absent 
3-5-5-5  to  3-3-4-4 

Palp 

Patella 
Tarsus  ($) 

3  spine-bearing  tubercles 
unmodified 

2  spine-bearing  tubercles 
with  spur  (usually) 

Leg  IV 

Trochanter  (6) 

with  spur 

unmodified 

Penis 

Glans 

folding 

telescoping 

Ovipositor1 

Shape 
Cuticle 

bent 
imbricate2 

straight 
with  microspines 

Apical  setae 

Shape 
Length 
Number 
Arrangement 

apically  hooked3 
long—  cross  at  center 
12-20 
not  grouped4 

evenly  bent 
shorter 
12,  rarely  14  or  16 
in  triads 

Subapical  setae 

Number 

none 

1-3  pairs,  rarely  none 

1  Species  of  Sitalcina  examined:  californica,  borregoensis,  madera,  scapula,  and  sura. 

2  In  S.  californica. 

3  Except  S.  scapula. 

4  Except  5.  sura. 


Unless  otherwise  indicated,  the  specimens  ex- 
amined during  this  study,  including  all  primary 
types,  are  deposited  at  the  California  Academy 
of  Sciences.  Additional  specimens  are  from  the 
collections  of  the  American  Museum  of  Natural 
History  (AMNH)  and  the  University  of  Califor- 
nia at  Berkeley  (UCB). 

Complete  species  descriptions  are  given  for  new 
species;  additional  descriptions  for  those  treated 
by  Briggs  (1968).  All  specimen  measurements  are 
in  millimeters. 

ABBREVATIONS  FOR  COLLECTORS 

AGO -Andy  G.  Grubbs 
AK- Albert  Kurz 
AKSJ- Albert  K.  S.  Jung 
AL— Albert  Lee 
BL— Benton  Leong 
BM— Barbara  Martin 
CF-C.  Fox 
DB-D.  Broussard 
OCR— D.  Craig  Rudolph 
DR- David  Rentz 
DU-Darrell  Ubick 


EVI— E.  Vanlngen 
GAM-G.  A.  Marsh 
GL— Galen  Leung 
GT- Gordon  Tang 
JG— Joseph  Gee 
JRH-J.  R.  Heifer 
JSB-J.  S.  Buckett 
KH- Kevin  Horn 
KK-K.  Kay 
LMS-L.  M.  Smith 
MG— Michael  Gardner 
MM— M.  McEachern 
MW— Michael  Wong 
NB-N.  Boice 
RL— Robert  Lem 
ROS— Robert  O.  Schuster 
SK-S.  Kubota 
SO -Stefan  Ong 
SW- Steve  Winterath 
TO— Toshiro  Ohsumi 
TSB— Thomas  S.  Briggs 
VFL— Vincent  F.  Lee 
WCR- Warren  C.  Rauscher 
WES -Warren  E.  Savary 
WI- Wilton  I  vie 


98 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


FIGURE  1.  Sitalcina  californica.  (a)  Trochanter  IV  of  male,  lateral  view  (180x).  (b)  Expanded  glans  penis,  sublateral  view 
(240  x ).  Arrow  indicates  movement  of  glans  during  expansion,  (c)  Ovipositor,  lateral  view  ( 1 80  x ).  (d)  Ovipositor,  close-up  of 
lateral  surface  showing  imbricate  cuticle  (l,800x).  (e)  Ovipositor,  apical  view  (300 x). 


TAXONOMY 

Family  PHALANGODIDAE  Simon 
Subfamily  PHALANGODINAE  Simon 

Calicina  Ubick  and  Briggs,  new  genus 

DIAGNOSIS.— Males  of  Calicina  are  unique 
among  the  Nearctic  phalangodid  genera  in  hav- 
ing a  telescoping  penis  glans  (as  opposed  to  a 
folding  one).  Additionally,  most  males  of  Cali- 
cina have  a  dorsal  spur  on  the  palpal  tarsus,  which 
is  not  known  in  other  phalangodids.  Females  are 
distinct  from  those  of  Sitalcina,  and  possibly  oth- 
er genera,  in  having  ovipositors  with  a  double 
setal  fringe  and  a  cuticle  with  microspines.  Most 
species  of  Calicina  may  be  distinguished  from 


other  phalangodid  genera  by  their  reduced  tarsal 
counts  of  less  than  3-5-5-5. 

TYPE  SPECIES.  — Sitalcina  mariposa  Briggs, 
1968. 

ETYMOLOGY.— The  generic  name  is  a  contrac- 
tion of  California  Sitalcina  and  is  feminine  in 
gender. 

DESCRIPTION.— Body  length  0.77-1.67.  Color 
ranges  from  pale  yellow  to  orange  to  reddish- 
brown.  Body  cuticle  minutely  and  densely  tu- 
berculate,  with  scales,  setae,  and  nipplelike  tu- 
bercles. Scute  with  prominent  ocular  tubercle  at 
anterior  margin;  varying  in  size,  shape,  and  de- 
gree of  rugosity  (see  Briggs  1968,  fig.  45-57,  59, 
62,  63,  65,  66).  Eyes  well  developed  (Fig.  3a)  or 
reduced,  lacking  retina  and  occasionally  cornea 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALICINA 


99 


FIGURE  2.    Calicina  mariposa,  lateral  view  of  male  (left  appendages  omitted).  AT  =  anterior  tubercles,  OT  =  ocular  tubercle, 
TS  =  tarsal  spur  of  palpus,  GO  =  genital  operculum,  P  =  penis,  legs  numbered. 


(=lens)  (Fig.  3b).  Carapace  with  two  rows  of  an- 
terior tubercles  numbering  from  0  to  4  or  rarely 
to  8  pairs  (Fig.  2,  3a,  b).  Ozopore  well  developed, 
on  anteriolateral  margin,  with  distinct  posterior 
channel.  Tergites  I-V  fused,  VI-VIII  free,  with 
median,  transverse  row  of  tubercles  interspersed 
with  short  setae.  Venter  textured  similarly  as  scute 
but  more  densely  setose;  seven  free  sternites  with 
median,  transverse  rows  of  tubercles.  Genital 
operculum  between  coxae  IV,  apically  truncate 
in  males,  rounded  in  females.  Coxae  I  and  II  with 
endites. 

Leg  sculpturing  consisting  of  minutely  tuber- 
culate  scales;  calcanei  and  tarsi  smooth.  Leg  II 
longest,  leg  length  formula  (longest  to  shortest) 
II,IV,I,III.  Tarsal  counts  3-5-5-5,  3-4-4-5,  or  3-4- 
4-4;  rarely  3-5-4-5  or  3-3-4-4.  Tarsal  claws  single 
on  legs  I  and  II;  paired  on  legs  III  and  IV.  Ju- 
veniles with  hind  claws  on  onychium,  arolium 
between  paired  claws.  Palpi  and  chelicerae  dis- 
tally  setose,  with  smooth  cuticle  of  appressed 
scales.  Palpi  robust,  ventrally  armed  with  spine- 
bearing  tubercles  (femur  with  three  ectobasally 
and  one  mesodistally;  patella  with  1  pair;  tibia 
and  tarsus  with  2  pairs  each)  (Fig.  3c).  Male  pal- 
pal tarsus  with  a  well  developed  dorsal  spur  (Fig. 
4a),  with  vestigial  spur  (Fig.  4b),  or  without  spur; 
occasionally  with  enlarged  mesal  spine-bearing 
tubercle  (Fig.  4c,  d). 

Penis  without  muscles,  composed  of  basal  sac, 
shaft,  and  apical  glans.  Basal  sac  inflatable.  Shaft 
(=truncus)  cylindrical,  sclerotized;  length  3.5-7 
x  width;  apicoventral  part  (= ventral  plate) 
rounded  or  attenuated,  set  with  setae.  Glans  tele- 
scoping, with  basal  segment  and  stylus,  rarely 
with  additional  (middle)  segment  (palapraepu- 
tia).  Collar  lobes  (=1Titillator  of  Martens  1986) 
on  basal  segment  (mariposa  and  digita  groups), 
on  middle  segment  (palapraeputia),  or  appar- 
ently absent  (serpentinea  group).  Stylus  apical 


and  variable  in  shape.  Accessory  structures  either 
thin,  scalelike  parastyli  (=7Konduktor  of  Martens 
1986)  (digita  group)  or  robust,  lobed  dorsal  pro- 
cess (serpentinea  group). 

Ovipositor  short,  straight,  and  cylindrical  when 
expanded;  length  1.5-2.0  x  width;  divided  into 
basal  and  distal  segments.  Lateral  and/or  apical 
surface  covered  with  minute,  pointed  tubercles 
(microspines).  Apex  with  2  or  4  lobes,  occasion- 
ally with  dorsal  folds;  with  setal  fringe  of  1 2  api- 
cal setae  (10  in  diminua)  grouped  in  triads  and 
1-4  pairs  of  subapical  setae  (most  species)  or  with 
14  apical  setae  only  (palapraeputia  and  serpen- 
tinea subgroup);  occasionally  with  a  pair  of  apical 
teeth  (Fig.  5). 

DISTRIBUTION  AND  SPECIES.— The  genus  Cali- 
cina is  found  only  in  the  central  Sierra  Nevada 
and  the  Coast  Ranges  of  California.  The  25  known 
species  represent  four  species  groups  and  nine 
subgroups. 

Key  to  the  Calicina  Species  Groups 

and  Subgroups 
Based  on  Characters  of  the  Male  Genitalia 

1 .  Basal  glans  segment  with  a  pair  of  collar 
lobes  (Fig.  16a-d) 2 

-  Basal  glans  segment  without  collar  lobes 
(Fig.  1 6e-h) 7 

2.  Parastyli  absent  (Fig.  6,  lOa-d) 3 

-  Parastyli  present  (Fig.  7-9,  lOe-h)  (digita 
group) 4 

3.  Collar  lobes  simple  (Fig.  6) 

mariposa  group  and  subgroup 

-  Collar  lobes  bilobed  (Fig.  lOa-d) 

arida  subgroup,  in  part  (doughensis) 

4.  Parastyli  ventral  (Fig.  7) digita  subgroup 

-  Parastyli  lateral  or  dorsal 5 

5.  Parastyli  lateral,  broad  and  ornate  (Fig.  8) 

kaweahensis  subgroup 


100 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


FIGURE  3.  Calicina,  morphology,  (a)  C.  mariposa,  male  paratopotype,  anterior  part  of  scute  showing  ocular  tubercle  with 
large  cornea,  a  row  of  paracular  anterior  tubercles  (AT),  and  the  ozopore  (O)  (230  x ).  (b)  C.  serpentinea,  female,  same  view 
showing  reduced  number  of  anterior  tubercles  and  degenerate  cornea  (C)  (230  x ).  (c)  C.  mariposa,  male  paratopotype,  chelicerae 
and  palpi  showing  tarsal  spurs  (TS)  (80  x ). 


-  Parastyli  dorsal,  narrow 6 

6.  Parastyli  separate;  stylus  straight;  ventral 

plate  acuminate  (Fig.  9)     topanga  subgroup 

-  Parastyli  fused;  stylus  sinuous;  ventral  plate 
unmodified  (Fig.  lOe-h) 

arida  subgroup,  in  part  (aridd) 


Middle  segment  of  glans  completely  sur- 
rounding stylus;  collar  lobes  hornlike  (Fig. 

1 1) palapraeputia  group  and  subgroup 

Middle  segment  of  glans  forming  the  dorsal 
process  that  is  distinct  from  stylus;  lobes 
variable  (Fig.  1 6g,  h)  (serpentinea  group)  8 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALICINA 


101 


FIGURE  4.  Calicina,  male  palpal  tarsi,  (a)  C.  mariposa,  showing  large  tarsal  spur  (310x).  (b)  C.  mesaensis,  showing  reduced 
tarsal  spur  (3 10  x ).  (c)  C.  minor,  showing  enlarged  mesobasal  spine-bearing  tubercle  (X)  (470  x ).  (d)  C.  ensata,  showing  enlarged 
mesoapical  spine-bearing  tubercle  (X)  (310x). 


102 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


FIGURE  5.  Calicina,  ovipositors,  (a)  C.  mariposa,  lateral  view  (270  x).  (b)  C.  mariposa,  close-up  of  lateral  surface  showing 
microspines  (2,700x).  (c)  C.  mariposa,  apical  view  (390 x).  (d)  C.  topanga,  lateral  view  showing  subapical  setae  (S)  (230 x).  (e) 
C.  palapraeputia,  ventral  view  showing  apical  tooth  (T)  (170x).  (f)  C.  palapraeputia,  apical  view  showing  microspines  and 
anterior  margin  of  genital  operculum  (230  x).  (g)  C.  serpentinea,  lateral  view  showing  apical  tooth  (310x).  (h)  C.  serpentinea, 
apical  view  showing  microspines  (390 x). 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALICINA 


103 


104 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CAL1CINA 


105 


8.  Dorsal  process  with  apical  bifurcation  (Fig. 

1 2) minor  subgroup 

-  Dorsal  process  with  apical  lobes 9 

9.  Dorsal  process  with  5  apical  lobes;  ventral 
plate  with  at  most  4  pairs  of  lateral  setae 
(Fig.  13,  14) serpent inea  subgroup 

-  Dorsal  process  with  at  most  4  apical  lobes; 
ventral  plate  with  5  to  8  pairs  of  lateral 
setae  (Fig.  1 5) sequoia  subgroup 

THE  MARIPOSA  SPECIES  GROUP 
(The  mariposa  Subgroup) 

DIAGNOSIS.— Members  of  the  mariposa  group 
may  be  recognized  by  their  simple  genitalia.  The 
glans  consists  of  a  stylus  and  a  basal  segment 
bearing  collar  lobes  and  lacks  additional  sclerites. 
The  ovipositor  contains  the  full  complement  of 
12  apical  and  3  pairs  of  subapical  setae.  The 
dorsal  grooves  on  the  collar  lobes  are  synapo- 
morphic  for  the  group.  The  three  species  be- 
longing to  this  group  are  all  summer-active  forest 
dwellers. 

DISTRIBUTION.— The  mariposa  group  is  found 
only  in  the  central  Sierra  Nevada  in  Yosemite 
National  Park  at  elevations  of  1,200-2,000  m. 

Calicina  mariposa  (Briggs,  1968), 
new  combination 

(Fig.  2,  3a,  c,  4a,  5a-c,  6a,  b,  16a,  b) 
Sitalcina  mariposa  Briggs,  1968:31. 

DIAGNOSIS.— This  species  differs  from  other 
members  of  the  group  in  having  a  high  tarsal 
count  of  3-5-5-5;  males  lack  an  apical  notch  on 
the  ventral  plate  of  the  penis. 

ADDITIONAL  DESCRIPTION.— Carapace  with  6- 
8  pairs  of  anterior  tubercles.  Penis  large;  ventral 
plate  apically  entire  and  with  8  pairs  of  marginal 
setae;  collar  lobes  pointed  and  pocketlike;  stylus 
long  and  sinuous.  Ovipositor  with  1 2  apical  setae 
and  3  pairs  of  lateral,  subapical  setae;  micro- 
spines  on  apical  and  lateral  surfaces. 

TYPE  SPECIMENS.— Male  holotype,  female  allotype,  and  three 
paratypes  (one  male,  two  females),  under  logs  in  a  giant  sequoia 
grove  at  Mariposa  Big  Trees,  Yosemite  National  Park,  Mari- 
posa County,  28  July  1967,  TSB  and  KH. 


SPECIMENS  EXAMINED.— Four  males,  12  females. 

NEW  RECORDS.- MARIPOSA  COUNTY:  Mariposa  Big 
Trees,  26  July  1984  (TSB,  SK,  GT);  TUOLUMNE  COUNTY: 
Yosemite  National  Park:  Tioga  Pass  Road,  June  1970  (TSB); 
9.2  km  E  Crane  Flat  Junction,  27  July  1984  (TSB,  SK,  GT); 
Wawona  Camp,  17  Sep.  1941  (WI;  AMNH  collection). 

NATURAL  HISTORY. —Known  from  sequoia  and 
yellow  pine  forests;  collected  from  June  to  Sep- 
tember. 


Calicina  yosemitensis  (Briggs,  1968), 
new  combination 

(Fig.  6e-g) 

Sitalcina  yosemitensis  Briggs,  1968:31. 

DIAGNOSIS.— Males  of  this  species  have  a  glans 
with  a  uniquely  modified  stylus  tip;  females  may 
be  separated  from  other  group  members  in  hav- 
ing an  ovipositor  with  only  1  pair  of  subapical 
setae. 

ADDITIONAL  DESCRIPTION.— Carapace  with  4- 
8  pairs  of  anterior  tubercles.  Penis  medium-sized; 
ventral  plate  with  apical  notch  and  7  pairs  of 
lateral  setae;  collar  lobes  small,  pointed,  apically 
grooved;  stylus  tip  with  lateral  expansions.  Ovi- 
positor as  in  mariposa  but  with  only  1  pair  of 
subapical  setae. 

TYPE  SPECIMENS.  — Male  holotype,  female  allotype,  and  three 
paratypes  (two  males,  one  female),  under  fallen  bark  in  conif- 
erous forest  5.0  mi  (8.0  km)  E  Crane  Flat  Ranger  Station, 
Yosemite  National  Park,  Tuolumne  County,  28  July  1967, 
TSB  and  AL. 

SPECIMENS  EXAMINED.— Five  males,  three  females. 

NEW  RECORD. -TUOLUMNE  COUNTY:  Yosemite  Na- 
tional Park:  3.8  mi  (6.1  km)  NE  Crane  Flat  Ranger  Station, 
21  July  1968  (TSB,  MW,  GL). 


Calicina  conifera  Ubick  and  Briggs, 
new  species 

(Fig.  6c,  d) 

DIAGNOSIS.— This  species  differs  from  others 
in  the  mariposa  group  by  its  small  size,  pale  col- 
oration, and  absence  of  retina. 

ETYMOLOGY.— The  specific  name  refers  to  the 
coniferous  biome  in  which  this  species  lives. 

DESCRIPTION.— Color  yellowish  orange.  Car- 


FIOURE  6.  Male  genitalia  of  the  mariposa  group,  (a,  b)  C.  mariposa  (Crane  Flat  Junction),  ventral  and  lateral  views,  (c,  d) 
C.  conifera  (holotype),  ventral  and  lateral  views,  (e-g)  C.  yosemitensis.  (e,  f)  Dorsal  and  lateral  views  showing  retracted  glans 
(paratopotype).  (g)  Dorsolateral  view  showing  expanded  glans,  arrow  indicates  movement  of  glans  during  expansion  (3.8  mi  NE 
Crane  Flat).  Note:  Unless  otherwise  indicated,  the  scale  bar  for  these  and  all  subsequent  genitalic  drawings  equals  0.30  mm, 
and  all  drawings  are  of  fully  expanded  glandes. 


106 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


apace  with  3  or  4  pairs  of  anterior  tubercles. 
Retina  absent,  lens  present.  Tarsal  count  3-4-4-5. 

Male:  Total  body  length,  1.20.  Scute  length, 
0.80.  Scute  width,  0.88.  Eye  tubercle  length,  0. 1 2. 
Eye  tubercle  width,  0.16.  Leg  II  length,  2.13. 
Palpal  tarsus  with  dorsal  spur.  Penis  small;  ven- 
tral plate  apically  notched  and  with  7  pairs  of 
marginal  setae;  collar  lobes  small,  rounded  and 
apically  grooved;  stylus  uniformly  tapering  and 
sinuous. 

Female:  Total  body  length,  1 .00.  Scute  length, 
0.63.  Scute  width,  0.7 1 .  Eye  tubercle  length,  0. 1 3. 
Eye  tubercle  width,  0. 1 1 .  Ovipositor  as  in  mar- 
iposa  but  with  two  pairs  of  subapical  setae. 

TYPE  SPECIMENS.— Male  holotype  and  two  female  para  types, 
under  fallen  bark  in  a  mixed  coniferous  forest  at  1.8  km  E 
Crane  Flat  Junction,  Tuolumne  County,  29  July  1984,  TSB, 
SK,  and  GT. 

SPECIMENS  EXAMINED.— Only  the  type  series. 

THE  DIGITA  SPECIES  GROUP 

DIAGNOSIS.— The  parastyli  on  the  glans  penis 
is  the  presumed  synapomorphy  for  this  group. 

The  digita  Subgroup 

DIAGNOSIS.— The  digita  subgroup  is  distin- 
guished from  all  others  by  the  unique  male  gen- 
italia:  the  ventral  plate  is  arrow-shaped  and  the 
glans  bears  a  pair  of  slender,  ventral  parastyli. 
Males  have  a  large  dorsal  spur  on  the  palpal  tar- 
sus. 

DISTRIBUTION.— Middle  elevation  Sierra  Ne- 
vada from  Amador  south  to  Tulare  counties. 


Calicina  digita  (Briggs  and  Horn,  1967), 
new  combination 

(Fig.  7a-d,  16c,  d) 

Sitalcina  digitus  Briggs  and  Horn,  1967:51.  Briggs,  1968:20. 

DIAGNOSIS.  — This  species  is  easily  distin- 
guished from  the  other  in  its  subgroup  by  a  higher 
tarsal  count  of  3-5-5-5. 

ADDITIONAL  DESCRIPTION.— Carapace  with  6 
or  7  pairs  of  anterior  tubercles.  Glans  with  a  pair 
of  short,  scalelike  parastyli  and  long,  sinuous  sty- 
lus. Ventral  plate  with  7  pairs  of  setae.  Ovipositor 
similar  to  that  ofmariposa  but  with  lower  density 
of  microspines  and  with  2  or  3  pairs  of  subapical 
setae. 

TYPE  SPECIMENS.— Male  holotype,  female  allotype,  and  20 
paratypes  (12  males,  eight  females),  under  fallen  bark  at  2.2 


mi  (3.5  km)  S  Giant  Forest,  Sequoia  National  Park,  Tulare 
County,  15  May  1966,  TSB  and  KH. 

SPECIMENS  EXAMINED.— Twenty-seven  males,  22  females. 

NEW  RECORDS. -MARIPOSA  COUNTY:  3  mi  (5  km)  E 
Briceburg,  6  Feb.  1 954  (ROS;  AMNH  collection);  Merced  Riv- 
er, N  Fork  near  Main  Fork,  3 1  Mar.  1 983  (TSB,  SO).  FRESNO 
COUNTY:  Saturday  Cave,  near  Boyden  Cave,  15  May  1966 
(TSB);  Big  Creek,  7  Aug.  1956  (ROS;  UCB  collection).  TU- 
LARE COUNTY:  Sequoia  National  Park:  0.9  mi  (1.4  km)  S 
Giant  Forest  Village,  17  Aug.  1984  (TSB,  VFL,  DU);  4  mi  (6.4 
km)  N  Hospital  Campground,  13  Feb.  1967  (JSB,  MG). 

NATURAL  HISTORY.— Found  in  a  variety  of 
biomes,  including  forests  (giant  sequoia-yellow 
pine  forest,  oak  woodland)  and  grassland,  and 
collected  throughout  the  year.  There  is  one  col- 
lection from  a  cave. 

REMARKS.— The  disjunct  populations  do  not 
appear  to  differ  in  reproductive  or  somatic  char- 
acters. 

Calicina  sierra  (Briggs  and  Horn,  1967), 
new  combination 

(Fig.  7e-g) 

Sitalcina  sierra  Briggs  and  Horn,  1967:49. 

Sitalcina  sierra  sierra  Briggs  and  Horn.  Briggs,  1968:20. 

DIAGNOSIS. — Calicina  sierra,  a  paedomorphic 
species,  is  most  readily  distinguished  from  digita 
by  its  low  tarsal  count  of  3-4-4-4. 

ADDITIONAL  DESCRIPTION.— Carapace  with  3 
or  4  pairs  of  anterior  tubercles.  Penis  similar  to, 
but  smaller  than,  that  of  digita.  Ovipositor  as  in 
mariposa. 

JUVENILES.— One  of  the  two  juveniles  is  an  ear- 
ly instar  (from  Scat  Cave)  having  a  tarsal  count 
of  1-1-2-2,  the  other  is  a  penultimate  (McLean's 
Cave  area,  2  Nov.  1975)  just  prior  to  moulting. 
Both  instars  lack  anterior  tubercles.  The  late  in- 
star  has  several  robust  spines  on  each  of  the  free 
tergites;  the  early  instar  has  only  one  pair  of  such 
spines  restricted  to  tergite  VII.  The  eyes  of  both 
instars  are  well  developed,  as  in  adults. 

TYPE  SPECIMENS.— Male  holotype,  female  allotype,  and  five 
paratypes  (three  males,  two  females),  6.2  mi  (10.0  km)  N  Mer- 
cer Caverns,  Sheep  Ranch  Road,  Calaveras  County,  26  Mar. 
1966,  TSB  and  KH. 

SPECIMENS  EXAMINED.  — Forty-three  males,  30  females,  two 
juveniles. 

NEW  RECORDS.-CALAVERAS  COUNTY:  Music  Hall  Cave, 
7.2  km  ESE  Angel's  Camp,  18  May  1977  (AGG,  SW,  MM, 
NB);  Carlow's  Bat  Cavern,  7  km  SE  Angel's  Camp,  25  May 
1977  (DB,  SW,  AGG);  Scat  Cave,  4  mi  (6.4  km)  W  Columbia, 
26  Mar.  1979  (BM,  SW);  unnamed  cave,  Peruvian  Gulch,  2 
mi  (3.2  km)  NW  Columbia,  10  May  1980  (DU);  Speleogen 
Cave,  7  km  SE  Angel's  Camp,  20  Apr.  1980  (TSB,  DU). 
TUOLUMNE  COUNTY:  Lower  van  Trump  Mine,  Fox  Gulch, 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALICINA 


107 


FIGURE  7.  Male  genitalia  of  the  digita  subgroup,  (a-d)  C.  digita.  (a,  b)  Lateral  and  subventral  views  showing  retracted  glans 
(Merced  River),  (c,  d)  Lateral  view  of  expanded  glans  and  dorsal  view  of  left  collar  lobe  (Giant  Forest),  (e-g)  C.  sierra  (McLean's 
Cave  area),  ventral,  dorsal,  and  lateral  views. 


23  Sep.  1975  (TSB,  RL);  Porcupine  Skull  Cave,  2  mi  (3.2  km) 
N  Columbia,  11  Feb.  1979  (OCR,  SW);  2.5  mi  (4  km)  NW 
Columbia,  2  Feb.  1979  (OCR,  SW);  S  Fork  Stanislaus  River 
at  Marble  Quarry  Road,  5  Mar.  1981  (DU);  2.5  mi  (4  km)  N 
Columbia,  McLean's  Cave  area,  2  Nov.  1975  (TSB,  RL),  2 
Apr.  1979  (OCR,  BM,  SW). 

DISTRIBUTION.— Known  from  Amador,  Ca- 
laveras,  and  Tuolumne  counties. 

NATURAL  HISTORY.— Found  in  digger  pine-oak 
forests  during  the  winter  and  spring  months  (No- 
vember-April). C.  sierra  is  also  known  from 
caves,  where  it  is  active  throughout  the  year, 


although  it  has  never  been  collected  in  associa- 
tion with  the  sympatric,  cavernicolous  species  of 
Banksula.  In  an  extensive  survey  of  the  Cala- 
veras  Formation  caves,  Banksula  was  recorded 
from  more  than  30  caves  and  C.  sierra  from  six 
different  ones  (Briggs  and  Ubick  1981). 

The  kaweahensis  Subgroup 

DIAGNOSIS.— Males  of  the  kaweahensis  sub- 
group are  unique  in  having  a  penis  with  ornate, 
lateral  parastyli. 


108 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


FIGURE  8.    Male  genitalia  of  the  kaweahensis  subgroup,  (a-d)  C.  galena,  (a)  Sublateral  view  showing  retracted  glans  (holotype). 
(b-d)  Lateral,  dorsal,  and  ventral  views  (paratopotype).  (e-g)  C.  kaweahensis  (Rocky  Hill),  dorsal,  ventral,  and  lateral  views. 


DISTRIBUTION.  — Known  only  from  Tulare 
County. 


Calicina  kaweahensis  (Briggs  and  Horn,  1966), 
new  combination 

(Fig.  8e-g) 

Sitalcina  kaweahensis  Briggs  and  Horn,  1966:266.  Briggs,  1968: 

22. 

DIAGNOSIS.  — This  species  may  be  distin- 
guished by  the  combination  of  large  size  (body 
length  greater  than  1 .4  mm)  and  a  tarsal  count 
of  3-5-4-5.  Males  have  a  glans  with  uniquely 
modified  parastyli. 

ADDITIONAL  DESCRIPTION.— Carapace  with  4 
pairs  of  anterior  tubercles.  Penis  with  collar  lobes 
in  basal  position;  parastyli  dorsally  fused  and 


apically  attenuated;  stylus  straight  and  basally 
enlarged;  ventral  plate  with  8  pairs  of  marginal 
setae.  Ovipositor  as  in  mariposa  but  with  2  pairs 
of  subapical  setae  and  with  a  pair  of  large  apical 
teeth;  microspines  on  entire  distal  segment  but 
sparse  on  the  ventral  surface. 

TYPE  SPECIMENS.  — Female  holotype,  male  allotype,  and  two 
paratypes  (damaged),  under  granitic  rocks  at  0.3  mi  (0.5  km) 
N  Junction  of  Woodlake-Lemoncove  Road  on  road  to  Badger, 
Tulare  County,  26  Nov.  1965,  TSB  and  VFL. 

SPECIMENS  EXAMINED.— Seventeen  males,  11  females,  four 
specimens  of  undetermined  sex  (damaged). 

NEW  RECORDS. -TULARE  COUNTY:  0.3  mi  (0.5  km)  N 
junction  Woodlake-Lemoncove  Road,  18  Dec.  1965  (KH); 
near  Lake  Kaweah  Offices  on  Hwy  198,  14  May  1966  (TSB); 
near  Kaweah  Dam  on  Hwy  198,  26  Nov.  1965  (VFL);  2.1  mi 
(3.4  km)  E  Rocky  Hill  Town,  19  Dec.  1966  (TSB,  KH);  1.4 
mi  (2.3  km)  E  Rocky  Hill,  22  Jan.  1967  (TSB,  KH,  AKSJ); 
Yokohl  Valley,  near  Lindsay,  Feb.  1971  (TSB);  Three  Rivers, 
Nov.  1967  (TSB). 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALICINA 


109 


NATURAL  HISTORY.— Known  from  oak  wood- 
land biomes  and  collected  throughout  the  winter 
and  spring  months  (November-May). 

Calicina  galena  Ubick  and  Briggs, 
new  species 

(Fig.  8a-d) 

DIAGNOSIS.— Males  of  this  species  differ  from 
all  others  by  the  ornately  scrolled  parastyli  and 
the  unique  transverse  rows  of  setae  on  the  ventral 
plate. 

ETYMOLOGY.— The  specific  name  refers  to  the 
type  locality  and  is  a  name  in  apposition. 

DESCRIPTION.— Carapace  with  1  pair  of  ante- 
rior tubercles.  Eyes  present.  Tarsal  count  3-5-5-5 
or  3-5-4-5. 

Male:  Total  body  length,  1.17.  Scute  length, 
0.79.  Scute  width,  0.72.  Eye  tubercle  length,  0. 1 7. 
Eye  tubercle  width,  0.12.  Leg  II  length,  1.94. 
Color  yellowish  orange.  Palpal  tarsus  without 
dorsal  spur.  Penis  with  ventral  plate  bearing  6 
pairs  of  setae  arranged  in  two  transverse  rows; 
collar  lobes  reduced  to  small  flaps;  parastyli  in- 
tricately scrolled;  stylus  short  and  spinelike. 

Female:  Total  body  length:  1.01.  Scute  length, 
0.69.  Scute  width,  0.78.  Eye  tubercle  length,  0. 1 3. 
Eye  tubercle  width,  0.12.  Color  orange.  Ovipos- 
itor as  in  mariposa. 

VARIATION.— Some  specimens  from  the  John- 
sondale  localities  have  the  basal  tarsomeres  of 
tarsus  III  fused,  giving  a  tarsal  count  of  3-5-4-5. 

JUVENILES.— The  single  known  juvenile  is  very 
pale,  has  a  tarsal  count  of  3-5-4-5,  and  has  well 
developed  eyes.  As  in  adults,  one  pair  of  anterior 
tubercles  is  present,  and  the  abdominal  tergites 
lack  robust  setae. 

TYPE  SPECIMENS.— Male  holotype  and  11  paratypes  (two 
males,  eight  females,  one  juvenile),  Galena  Creek,  Mountain 
Home  State  Forest,  Tulare  County,  4  Sep.  1967,  TSB.  Ten 
paratypes  (seven  males,  three  females),  2.25  mi  (3.6  km)  W 
Johnsondale,  Tulare  County,  1,600  m,  2  July  1988,  TSB  and 
DU.  Nineteen  paratypes  (12  males,  seven  females),  Double 
Bunk  Meadows,  3.5  air  mi  (5.6  km)  SW  Johnsondale,  Tulare 
County,  1,850  m,  3  July  1988,  TSB  and  DU. 

SPECIMENS  EXAMINED.— Only  the  type  series. 

NATURAL  HISTORY.— The  original  series  was 
collected  beneath  granite  rocks  in  a  giant  sequoia 
forest.  The  recent  collections  are  from  mixed  co- 
niferous forests;  specimens  were  collected  be- 
neath and  within  decaying  logs,  large  slabs  of 
fallen  bark,  and  dense  mats  of  duff.  The  popu- 


lation at  3.6  km  W  Johnsondale  is  sympatric 
with  a  species  of  Sitalcina. 

The  topanga  Subgroup 

DIAGNOSIS.— This  subgroup  is  distinguished 
from  all  others  by  several  unique  male  genital 
characters:  apex  of  ventral  plate  attenuated  and 
armed  with  two  pairs  of  strong  setae;  stylus 
straight  and  elongated,  somewhat  flattened; 
parastyli  short,  hornlike  in  shape  and  dorsally 
situated;  and  collar  lobes  small  and  covered  with 
minute  tubercles. 

DISTRIBUTION.— Members  of  this  subgroup 
occur  in  the  southern  Coast  Ranges  and  Te- 
hachapi  Mountains,  the  southernmost  distribu- 
tion of  the  genus. 

Calicina  topanga  (Briggs,  1968), 
new  combination 

(Fig.  5d,  9c,  d) 

Sitalcina  topanga  Briggs,  1968:25. 

DIAGNOSIS.  — This  species  may  be  distin- 
guished from  others  in  the  subgroup  by  its  high 
tarsal  count  of  3-5-5-5. 

ADDITIONAL  DESCRIPTION.— Carapace  with  2 
or  3  pairs  of  anterior  tubercles.  Penis  with  ventral 
plate  attenuated  and  bearing  7  pairs  of  setae,  of 
which  four  apical  ones  are  enlarged;  collar  lobes 
earlike  lateral  projections  with  microspines; 
parastyli  dorsal,  short  and  hornlike;  stylus  long 
and  straight,  somewhat  flattened.  Ovipositor  as 
in  mariposa  but  with  2-4  pairs  of  subapical  setae 
(one  specimen  with  9  setae). 

TYPE  SPECIMENS.— Male  holotype,  female  allotype,  and  four 
paratypes  (two  males,  two  females),  under  sandstone  in  a  dense 
oak  forest  at  3.3  mi  (5.3  km)  N  Topanga  Beach  in  Topanga 
Canyon,  Los  Angeles  County,  7  Apr.  1966,  TSB,  KH,  and 
VFL. 

SPECIMENS  EXAMINED.— Thirty-eight  males,  32  females,  one 
specimen  of  undetermined  sex. 

NEW  RECORDS.  -SANTA  BARBARA  COUNTY:  Santa  Ynez 
Mountains,  Cold  Springs,  19  Dec.  1967  (TSB,  AKSJ).  Santa 
Cruz  Island:  canyon  N  UCSB  Field  Station,  1 9  Dec.  1 967  (TSB, 
KH,  AKSJ),  21  Dec.  1967  (TSB);  0.1  mi  (0.2  km)  SW  UCSB 
Field  Station,  20  Dec.  1967  (AKSJ);  Canada  del  Puerto,  Pris- 
oners Harbor,  21  Dec.  1967  (KH,  AKSJ);  Raven's  Wood  Can- 
yon, 22  Dec.  1967  (TSB);  Cristi  Pines,  19-22  Dec.  1967  (TSB, 
KH,  AKSJ). 

NATURAL  HISTORY.— Known  from  both  closed 
and  open  forests  (broadleaf  evergreen,  digger  pine- 
oak,  and  oak  woodland),  and  collected  during 
the  winter  and  spring  months  (December-April). 


110 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


FIGURE  9.    Male  genitalia  of  the  topanga  subgroup,  (a,  b)  C.  morroensis  (Madonna  Inn),  ventral  and  subdorsal  views,  (c,  d) 
C.  topanga  (Santa  Cruz  Island),  subventral  and  lateral  views,  (e-g)  C.  keenea  (holotype),  lateral,  ventral,  and  dorsal  views. 


REMARKS:  The  mainland  and  island  popula- 
tions of  topanga  appear  similar  in  both  genitalic 
and  somatic  characters. 

Calicina  keenea  (Briggs,  1968), 
new  combination 

(Fig.  9e,  f) 

Sitalcina  keenea  Briggs,  1968:25. 

DIAGNOSIS.— This  species  is  distinguished  from 
others  in  its  group  by  its  tarsal  count,  3-4-4-5. 

ADDITIONAL  DESCRIPTION.— Carapace  with  2 
or  3  pairs  of  anterior  tubercles.  Penis  similar  to 
but  smaller  than  that  of  topanga;  ventral  plate 
with  6  pairs  of  lateral  setae.  Ovipositor  as  in 
mariposa  but  with  4  pairs  of  subapical  setae  and 
with  a  lower  density  of  microspines. 


TYPE  SPECIMENS.  —  Male  holotype  and  female  allotype,  under 
granite  in  moist  oak  grassland  at  3.2  mi  (5.1  km)  NW  Keene, 
Kern  County,  19  Mar.  1967,  VFL. 

SPECIMENS  EXAMINED.— Only  the  type  series. 

Calicina  morroensis  (Briggs,  1 968), 
new  combination 

(Fig.  9a,  b) 

Sitalcina  morroensis  Briggs,  1968:26. 

DIAGNOSIS.— This  paedomorphic  species  dif- 
fers from  others  in  the  group  by  the  combination 
of  low  tarsal  count  of  3-4-4-4  and  degenerate  eyes 
(lacking  retina  and  sometimes  cornea). 

ADDITIONAL  DESCRIPTION.— Carapace  with  1 
pair  of  anterior  tubercles.  Penis  typical  for  the 
group;  ventral  plate  with  strong  apical  setae  and 
6  pairs  of  lateral  setae;  stylus  broad  and  flat, 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALICINA 


111 


FIGURE  10.  Male  genitalia  of  the  arida  subgroup,  (a-d)  C.  cloughensis.  (a-c)  Lateral,  ventral,  and  dorsal  views  (topotype). 
(d)  Dorsal  view  of  slide-mounted  specimen  showing  retracted  glans  (holotype).  (e-h)  C.  arida.  (e-g)  Lateral,  dorsal,  and  ventral 
views  (holotype).  (h)  Sublateral  view  showing  retracted  glans  (paratopotype). 


parastyli  very  small.  Ovipositor  similar  to  that  TYPE  SPECIMENS. -Male  holotype  and  female  allotype,  under 

K-          .           ;           •,-,-/•                       j        j  serpentine  in  marshy  grassland  at  0.8  mi  (1.3  km)  N  Morro 

of  manposa  but  with  3  pairs  of  strongly  reduced  Bay  San  Luis  Obispo  County  2  Mar  1 967>  VFL. 

Subapical  setae  and  with  Sparse  distribution  of  SPECIMENS  EXAMINED. -Twenty-nine  males,  25  females. 

microspines.  NEW  RECORDS. -SAN  LUIS  OBISPO  COUNTY:  San  Luis 


112 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


Obispo,  behind  Madonna  Inn,  22  Mar.  1970  (TSB,  GL);  2.4 
mi  (3.9  km)  NW  San  Simeon  in  serpentine  grassland,  13  Feb. 
1987  (TSB,  VFL). 


The  arida  Subgroup 

DIAGNOSIS.— Members  of  this  subgroup  are 
distinguished  from  all  other  species  by  their 
unique  glans  penis:  the  stylus  is  strongly  sinuous 
and  the  collar  lobes  are  greatly  enlarged.  Males 
also  lack  spurs  on  their  palpal  tarsi  but  have 
enlarged  mesobasal,  spine-bearing  tubercles. 

DISTRIBUTION.— Known  from  a  single  locality 
in  San  Benito  County  and  a  cave  in  Tulare  Coun- 
ty. 


Calicina  arida  Ubick  and  Briggs, 
new  species 
(Fig.  10e-h) 

DIAGNOSIS.  — Males  of  this  paedomorphic 
species  are  unique  in  the  genus  in  having  the 
parastyli  represented  by  a  single  rod. 

ETYMOLOGY. —The  specific  name  is  a  Latin  ad- 
jective meaning  dry,  and  refers  to  the  climatic 
conditions  in  the  vicinity  of  the  type  locality. 

DESCRIPTION.— Color  yellowish  orange.  Car- 
apace with  1  pair  of  anterior  tubercles.  Retina 
absent,  cornea  present.  Tarsal  count  3-4-4-4. 

Male:  Total  body  length,  0.80.  Scute  length, 
0.57.  Scute  width,  0.56.  Eye  tubercle  length,  0.10. 
Eye  tubercle  width,  0.13.  Leg  II  length,  1.65. 
Palpal  tarsus  without  dorsal  spur  but  with  en- 
larged mesobasal  spine-bearing  tubercle.  Penis 
relatively  stout;  ventral  plate  with  slight  apical 
attenuation  and  6  pairs  of  setae;  collar  lobes 
greatly  enlarged  and  apically  directed;  parastyli 
represented  by  a  single,  dorsal  rod;  stylus  strong- 
ly sinuous  and  dorsally  directed. 

Female:  Total  body  length,  0.73.  Scute  length, 
0.50.  Scute  width,  0.53.  Eye  tubercle  length,  0.06. 
Eye  tubercle  width,  0.08.  Ovipositor  with  1 2  api- 
cal setae,  with  3  pairs  of  subapical  setae  and  with 
a  pair  of  short  apical  teeth;  microspines  sparsely 
distributed  apically  and  laterally. 


TYPE  SPECIMENS.  — Male  holotype  and  18  paratypes  (nine 
males,  nine  females),  under  serpentine  rocks  in  an  oak  gully 
at  Panoche  Road,  12.7  mi  (20.4  km)  SE  Paicines,  San  Benito 
County,  1  Dec.  1984,  TSB,  VFL,  and  DU.  One  male  parato- 
potype,  1  Dec.  1971,  KH.  Two  paratypes  (male,  female)  de- 
posited at  AMNH  collection. 

SPECIMENS  EXAMINED.  — Eleven  males,  10  females. 


Calicina  cloughensis  (Briggs  and  Horn,  1967), 
new  combination 

(Fig.  lOa-d) 

Sitalcina  cloughensis  Briggs  and  Horn,  1967:52.  Briggs,  1968: 
23. 

DIAGNOSIS.— This  species  is  the  only  troglo- 
bitic  Calicina  and  is  recognized  by  its  pale  col- 
oration, complete  loss  of  eyes  (cornea  as  well  as 
retina),  and  a  high  tarsal  count  (3-5-5-5).  Fur- 
thermore, males  of  cloughensis  have  bilobed  col- 
lar lobes  and  a  unique  stylus  bearing  a  median 
enlargement. 

ADDITIONAL  DESCRIPTION.— Carapace  with  1 
pair  of  anterior  tubercles.  Penis  with  ventral  plate 
bearing  6  pairs  of  lateral  setae;  collar  lobes  di- 
vided into  apical  and  basal  portions;  stylus  long 
and  sinuous,  with  crescent-shaped  enlargement 
at  midlength,  and  with  apical  extension.  Ovi- 
positor with  1 2  apical  setae  (nine  in  one  speci- 
men) and  2  or  3  pairs  of  subapical  setae;  mi- 
crospines sparsely  distributed  on  lateral  surface. 

JUVENILES.— Two  of  the  paratypes  are  early  in- 
star  juveniles  having  tarsal  counts  of  1  -2-2-2  and 
completely  lacking  eyes.  Both  lack  anterior  tu- 
bercles but  have  robust  spines  on  the  free  tergites 
(adults  have  normal  setae). 

TYPE  SPECIMENS.  — Male  holotype,  female  allotype,  and  six 
paratypes  (male,  three  females,  two  juveniles),  in  Ladder  Room 
of  Clough  Cave  at  South  Fork  Ranger  Station  1,070  m  (3,500 
ft),  Sequoia  National  Park,  Tulare  County,  14  May  1966,  TSB, 
VFL,  and  KH. 

SPECIMENS  EXAMINED.  — Four  males,  nine  females,  two  ju- 
veniles. 

NEW  RECORDS. -TULARE  COUNTY:  Clough  Cave,  1 8  Apr. 
1979  (SW,  EVI),  31  Aug.  1986  (TSB,  KH,  AKSJ). 

REMARKS.— The  stylus  of  the  holotype  male 
lacks  the  apical  extension  found  in  other  speci- 
mens. 

THE  PALAPRAEPUTIA  SPECIES  GROUP 

(The  palapraeputia  Subgroup) 

DIAGNOSIS.— The  single  species  representing 
this  group  has  unique  male  genitalia:  glans  of 
penis  three  segmented  with  middle  segment 
bearing  a  pair  of  hornlike  collar  lobes. 

DISTRIBUTION.— Found  only  in  Fresno  Coun- 
ty. 

Calicina  palapraeputia  (Briggs,  1968), 
new  combination 

(Fig.  5e,  f,  11,  16e,  f) 

Sitalcina  palapraeputia  Briggs,  1968:29. 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALICINA 


113 


DIAGNOSIS.— Same  as  for  group. 

ADDITIONAL  DESCRIPTION.— Carapace  with  2 
or  3  pairs  of  anterior  tubercles.  Genital  opercu- 
lum  with  concave  anterior  margin.  Penis  with 
ventral  plate  truncate,  apically  elongated  and 
armed  with  7  pairs  of  setae;  glans  with  two  col- 
lars; collar  lobes  raptorial;  stylus  stout.  Ovipos- 
itor with  short  distal  segment,  with  7  pairs  of 
apical  setae,  and  with  a  pair  of  large  apical  teeth; 
microspines  present  only  at  apex. 

TYPE  SPECIMENS.  — Male  holotype,  female  allotype,  and  six 
paratypes  (three  males,  three  females),  under  serpentine  rocks 
in  open  grassland  at  7  mi  (1 1  km)  NE  Piedra,  Fresno  County, 
21  Jan.  1967,  TSB  and  KH. 

SPECIMENS  EXAMINED.— Fifteen  males,  10  females. 

NEW  RECORDS. -FRESNO  COUNTY:  10.2  mi  (16.4  km) 
SW  Trimmer,  27  Jan.  1968  (KH,  JG);  8.7  mi  (14  km)  SW 
Trimmer,  27  Jan.  1968  (TSB,  KH);  5.3  mi  (8.5  km)  SW  Trim- 
mer, 27  Jan.  1968  (TSB). 

NATURAL  HISTORY.— Found  beneath  serpen- 
tine and  granite  rocks  in  oak  woodland  and  grass- 
land biomes. 

THE  SERPENTINEA  SPECIES  GROUP 

DIAGNOSIS.— Males  of  this  group  have  a  dorsal 
process  on  the  glans  penis.  Females  differ  from 
those  in  the  digita  and  mariposa  groups  in  having 
an  ovipositor  with  only  apical  microspines. 

The  minor  Subgroup 

DIAGNOSIS.— The  morphology  of  the  penis, 
dorsal  process  with  ventrally  directed  basal  lobes 
but  no  apical  lobes,  is  diagnostic  for  the  subgroup. 

DISTRIBUTION.— Known  only  from  San  Mateo 
County. 


FIGURE  1 1 .  Male  genitalia  of  C.  palapraeputia.  (a,  b)  Lat- 
eral and  dorsal  views  showing  partially  expanded  glans  (5.3 
mi  SW  Trimmer),  (c)  Ventral  view  showing  fully  expanded 
glans  (10.2  mi  SW  Trimmer). 


Calicina  minor  (Briggs  and  Horn,  1 966), 
new  combination 

(Fig.  4c,  12) 

Sitalcina  minor  Briggs  and  Horn,  1966:263.  Briggs,  1968:27. 

DIAGNOSIS.— This  paedomorphic  species  is 
distinguished  from  all  others  by  its  unique  glans 
penis:  basal  segment  with  transverse  rows  of  small 
tubercles  and  dorsal  process  lacking  apical  lobes 
but  with  prominent  basal  lobes. 

ADDITIONAL  DESCRIPTION.— Carapace  with  at 
most  1  pair  of  anterior  tubercles.  Male  palpal 
tarsus  with  enlarged  mesobasal  spine-bearing  tu- 
bercle instead  of  spur.  Penis  with  ventral  plate 
apically  produced  and  bearing  7  pairs  of  setae; 


dorsal  process  apically  bifurcate,  with  ventrally 
directed  basal  lobes;  stylus  straight  and  tapering, 
basally  thickened.  Ovipositor  with  6  pairs  of  api- 
cal and  3  pairs  of  subapical  setae;  microspines 
restricted  to  apex. 

TYPE  SPECIMENS.  — Female  holotype,  male  allotype,  and  four 
paratypes  (one  male,  three  females),  under  serpentine  rocks  on 
grassland  hillside  at  0.75  mi  (1.2  km)  N  Crystal  Springs  Dam 
on  County  Road  No.  14,  San  Mateo  County,  23  Jan.  1966, 
TSB  and  KH. 

SPECIMENS  EXAMINED.— Twenty  males,  23  females. 

NEW  RECORDS. -SAN  MATEO  COUNTY:  Edgewood  Park, 
serpentine  grassland  on  N  half  of  park,  2  Jan.  1 983  (TSB,  VFL, 
DU);  6  Apr.  1986  (TSB,  TO);  2  Jan.  1987  (TSB,  VFL,  DU); 
24  Dec.  1987  (TSB,  VFL,  WES,  DU). 


114 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


FIGURE  12.    Male  genitalia  of  C.  minor  (Edgewood  Park),  (a)  Ventral  view  of  penis  shaft,  (b-d)  Ventral,  dorsal,  and  lateral 
views. 


DISTRIBUTION.  — This  species  is  presently 
known  only  from  Edgewood  Park.  No  specimens 
have  been  collected  at  the  type  locality  (despite 
repeated  attempts)  since  the  completion  of  the 
adjacent  highway.  Also,  the  series  of  collections 
from  Santa  Clara  County  listed  by  Briggs  and 
Horn  (1966)  represent  an  undescribed  species  be- 
longing to  a  related  genus. 

NATURAL  HISTORY.— All  collections  are  from 
serpentine  grassland. 


The  serpentined  Subgroup 

DIAGNOSIS. —This  subgroup,  the  largest  in  Cal- 
icina,  includes  those  species  whose  males  have 
a  glans  with  a  ventral  stylus  and  a  dorsal  process 
bearing  five  apical  lobes.  The  palpal  spur,  when 
present,  is  small  or  vestigial. 

DISTRIBUTION.— The  subgroup  has  a  disjunct 
distribution  with  two  populations  in  the  lower 
Sierra  Nevada  foothills  and  a  third  along  the 
central  and  northern  Coast  Ranges. 


Key  to  Males  of  the  Species  of  the 
serpentined  Subgroup 

1 .  Glans  with  a  pair  of  basal  lobes  (Fig.  1 3a- 
c,  18) piedrd 

-  Glans  without  basal  lobes  (Fig.  1 3d-m,  1 4) 
2 

2.  Stylus  longer  than  dorsal  process  (Fig.  1 3d- 
m) 3 

-  Stylus  shorter  than  dorsal  process  (Fig.  1 4) 
5 

3.  Tarsal  count  3-4-4-5 serpentined 

-  Tarsal  count  3-4-4-4 4 

4.  Dorsal  process  with  lateral  lobes  as  long  as 
wide  (Fig.  1 3d-f ) ensata 

-  Dorsal  process  with  lateral  lobes  longer  than 
wide  (Fig.  1 31,  m) polind 

5.  Body  with  dark  maculations macula 

-  Body  concolorous 6 

6.  Dorsal  process  with  lateral  lobes  as  long  as 
wide  (Fig.  14a-f) _ 7 

-  Dorsal  process  with  lateral  lobes  longer  than 
wide  (Fig.  14J-1) 8 


FIGURE  13.  Male  genitalia  of  the  serpentined  subgroup.  1.  (a-c)  C.  piedra  (holotype),  lateral,  ventral,  and  dorsal  views,  (d- 
f)  C.  ensata  (Tollhouse),  (d,  e)  Ventrolateral  and  dorsal  views,  (f )  Dorsolateral  view  of  stylus  tip.  (g-k)  C.  serpentinea.  (g,  h) 
Dorsal  and  ventral  views  (Monte  Vista),  (i-k)  Ventrolateral,  dorsal,  and  ventral  views  (San  Felipe  Road).  (I,  m)  C.  polina 
(Novato),  dorsolateral  and  ventral  views.  (Scale  bar  equals  0.25  mm,  except  for  "f"  where  it  equals  0.10  mm.) 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALICINA 


115 


116 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


7.  Median  lobe  of  dorsal  process  apically 
notched  (Fig.  14d-f) basalta 

-  Median  lobe  of  dorsal  process  apically 
pointed  (Fig.  14a-c) breva 

8.  Tarsal  count  3-4-4-5 dimorphica 

-  Tarsal  count  3-4-4-4 mesaensis 

Calicina  piedra  (Briggs,  1968), 
new  combination 

(Fig.  13a-c) 

Sitaldna  piedra  Briggs,  1968:24. 

DIAGNOSIS.— This  species,  the  largest  of  the 
serpentinea  group,  is  easily  distinguished  by  the 
structure  of  the  male  genitalia:  dorsal  process  of 
glans  subequal  in  length  to  stylus,  with  a  pair  of 
basoventral  lobes,  and  with  small  apical  lobes. 

ADDITIONAL  DESCRIPTION.— Carapace  with  2 
or  3  pairs  of  anterior  tubercles.  Penis  with  ventral 
plate  apically  attenuated  and  with  5  pairs  of  lat- 
eral setae;  stylus  flat,  broad,  and  sinuous.  Ovi- 
positor as  in  minor  but  without  subapical  setae. 

TYPE  SPECIMENS.  — Male  holotype,  female  allotype,  and  fe- 
male paratype,  under  rocks  in  an  oak  grassland  at  1 .6  mi  (2.6 
km)  SW  Piedra,  Fresno  County,  21  Jan.  1967,  TSB,  KH,  and 
AJ. 

SPECIMENS  EXAMINED.— Only  the  type  series. 

Calicina  serpentinea  (Briggs  and  Horn,  1 966), 

new  combination 

(Fig.  3b,  5g,  h,  13g-k,  16g,  h) 

Sitaldna  serpentinea  Briggs  and  Horn,  1966:268.  Briggs,  1968: 
26. 

DIAGNOSIS.— This  species  differs  from  others 
in  its  group  by  the  combination  of  a  tarsal  count 
of  3-4-4-5  and  structure  of  the  male  genitalia 
(stylus  longer  than  the  dorsal  process  and  median 
lobe  narrow  and  elongated). 

ADDITIONAL  DESCRIPTION.— Carapace  with  2 
or  3  pairs  of  anterior  tubercles.  Eyes  variable, 
cornea  present  or  absent.  Tarsal  spur  of  male 
palpi  may  be  absent  or  extremely  small.  Penis 
with  ventral  plate  bearing  7  pairs  of  setae;  dorsal 
process  with  narrow  median  lobe  (with  or  with- 
out an  apical  notch)  and  rounded  lateral  lobes. 
Ovipositor  with  apical  microspines,  with  6-8 
pairs  of  apical  setae,  and  with  one  pair  of  apical 
teeth. 

VARIATION.  —  Calicina  serpentinea  is  poly- 
morphic with  regard  to  several  characters.  In- 
dividuals from  the  three  westernmost  localities 
(dense  forest  biomes)  lack  corneas  and  tarsal 


spurs,  which  are  present  in  the  remaining  pop- 
ulations (grassland  and  oak  woodland  biomes). 
The  female  from  Monte  Vista  lacks  the  paired 
apical  teeth  on  the  ovipositor.  In  individuals  from 
Santa  Clara  and  Contra  Costa  counties,  the  me- 
dian lobe  is  apically  entire  (Fig.  1 3g,  h),  in  others 
it  is  apically  notched  (Fig.  1 3i-k). 

JUVENILES.— The  two  known  juveniles,  from 
Bob's  Secret  Cave,  are  middle  instars  having  tar- 
sal counts  of  2-2-3-3.  They  lack  anterior  tuber- 
cles. As  in  adults  from  this  locality,  the  eyes  are 
completely  reduced  (lacking  both  retina  and  cor- 
nea) and  the  free  tergites  lack  enlarged  spines. 

TYPE  SPECIMENS.  —  Female  holotype,  male  allotype,  and  six 
paratypes  (two  males,  four  females),  under  serpentine  in  marshy 
grassland  along  Silver  Creek  Road,  5  mi  (8  km)  SE  Tully  Road, 
San  Jose,  Santa  Clara  County,  27  Feb.  1966,  TSB  and  CF. 

SPECIMENS  EXAMINED.  — Fifty-one  males,  38  females,  two  ju- 
veniles, seven  specimens  of  undetermined  sex. 

NEW  RECORDS. -SANTA  CLARA  COUNTY:  Silver  Creek 
Road,  3.6  mi  (5.8  km)  from  San  Felipe  Road,  2  Jan.  1983 
(TSB,  VFL,  DU);  Silver  Creek  at  San  Felipe  Reservoir,  27 
Nov.  1966  (TSB,  AKSJ);  Monte  Vista,  near  Permanente  Quar- 
ry, 24  Mar.  1 968  (TSB);  Metcalf  Road,  1 .8  mi  (2.9  km)  N  Hwy 
101,2  Jan.  1983  (TSB,  VFL,  DU).  SANTA  CRUZ  COUNTY: 
Cave  Gulch,  Bob's  Secret  Cave,  6  May  1984  (TSB,  VFL,  DU); 
Cave  Gulch,  1 8  June  1 984  (TSB,  DU).  SAN  BENITO  COUN- 
TY: Lone  Tree  Road,  8.6  mi  (13.8  km)  W  Fairview  Road,  27 
Dec.  1984  (TSB,  VFL,  DU). 

DISTRIBUTION.— Known  from  the  southern  Bay 
Area  from  San  Mateo  and  Contra  Costa  to  San 
Benito  counties. 

NATURAL  HISTORY.  — This  wide-ranging 
species  occurs  in  a  variety  of  habitats;  most  com- 
monly beneath  serpentine  in  grassland  and  oak 
woodland  biomes,  but  also  from  redwood, 
broadleaf  evergreen,  and  digger  pine-oak  asso- 
ciations. There  is  also  one  collection  from  a  cave. 
Most  specimens  were  taken  from  under  serpen- 
tine, but  some  were  also  collected  beneath  lime- 
stone and  sandstone  rocks,  and  fallen  bark.  While 
more  commonly  collected  during  the  rainy  sea- 
son, this  species  is  apparently  active  throughout 
the  year  in  mesic  refugia. 

Calicina  polina  (Briggs,  1968), 
new  combination 

(Fig.  131,m) 

Sitaldna  polina  Briggs,  1968:30. 

DIAGNOSIS.— A  unique  dorsal  process  with  a 
cordate  median  lobe  and  narrow  lateral  lobes 
distinguishes  males  of  this  paedomorphic  species 
from  all  other  species. 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALICINA 


117 


ADDITIONAL  DESCRIPTION.— Carapace  with  1 
or  2  pairs  of  anterior  tubercles.  Penis  with  stylus 
longer  than  dorsal  process;  ventral  plate  with  6 
pairs  of  setae.  Ovipositor  with  7  pairs  of  apical 
setae;  microspines  apical. 

TYPE  SPECIMENS.— Male  holotype,  female  allotype,  and  three 
paratypes  (two  males,  one  female),  under  serpentine  rocks  in 
a  moist  grassland  at  El  Polin  Spring,  Presidio,  San  Francisco 
County,  11  Dec.  1966,  TSB. 

SPECIMENS  EXAMINED.— Forty-six  males,  31  females,  one 
specimen  of  undetermined  sex. 

NEW  RECORDS. -SAN  FRANCISCO  COUNTY:  Presidio 
Golf  Course,  9  Nov.  1975  (TSB).  MARIN  COUNTY:  1  mi 
(1.6  km)  NW  Novato,  2  Apr.  1967  (TSB);  Marin  Dell  Ranch, 
near  Novato,  2  Jan.  1972  (TSB,  AK);  Novato,  San  Marin 
Drive,  7  Mar.  1982  (TSB,  DU),  14  Mar.  1982  (DU);  Novato, 
W  of  Burdell  Mountain,  100  m  (350  ft),  2  Jan.  1986  (TSB, 
DU);  Novato,  SW  of  Burdell  Mountain,  170  m  (550  ft),  2  Jan. 
1986  (TSB,  DU).  SONOMA  COUNTY:  2.5  mi  (4  km)  E  Shell- 
ville,  2  Apr.  1967  (TSB);  ALAMEDA  COUNTY:  Cull  Canyon 
Road,  1  km  N  Crow  Canyon  Road,  22  Jan.  1984  (TSB,  DU); 
Cull  Canyon  Road,  1.5  mi  (2.4  km)  N  Crow  Canyon  Road, 
22  Jan.  1984  (TSB,  DU);  Woolsey  Canyon,  Berkeley,  21  Dec. 

1 983  (TSB,  VFL,  DU);  Redwood  Road,  3. 1  mi  (5  km)  N  Castro 
Valley  Road,  22  Jan.  1984  (TSB,  DU);  Chabot  Reservoir  at 
Lake  Chabot  Road,  6  Apr.  1982  (TSB,  DU).  CONTRA  COS- 
TA COUNTY:  Wildcat  Canyon  Park  near  Hill  Road,  30  Jan. 

1984  (TSB),  14  July  1984  (TSB). 

DISTRIBUTION.— This  species  is  found  in  the 
northern  part  of  the  San  Francisco  Bay  Region, 
from  Sonoma  south  to  San  Francisco  and  Ala- 
meda  counties. 

NATURAL  HISTORY. —The  collections  ofpolina 
west  of  San  Francisco  Bay  (Sonoma,  Marin,  and 
San  Francisco  counties)  are  from  serpentine 
grassland,  whereas  those  from  the  East  Bay  are 
from  beneath  sandstone  in  oak  forests. 

Calicina  ensata  (Briggs,  1968), 
new  combination 

(Fig.  4d,  13d-f) 

Sitalcina  ensata  Briggs,  1968:21. 

DIAGNOSIS.  — Males  of  this  paedomorphic 
species  may  be  distinguished  from  others  in  the 
subgroup  in  having  a  stylus  which  is  longer  than 
the  dorsal  process  and  has  an  apical  prong. 

ADDITIONAL  DESCRIPTION.— Carapace  with  2 
or  3  pairs  of  anterior  tubercles.  Penis  with  ventral 
plate  bearing  7  pairs  of  lateral  setae;  dorsal  pro- 
cess with  large  lateral  lobes;  stylus  long,  tapering. 
Ovipositor  with  12  apical  setae  and  3  pairs  of 
subapical  setae;  microspines  apical. 

TYPE  SPECIMENS.  — Male  holotype,  female  allotype,  and  10 
paratypes  (eight  males,  two  females),  under  rhyolite  in  an  oak 


woodland  at  2.0  mi  (3.2  km)  NW  Trimmer,  Fresno  County, 
16  Apr.  1967,  TSB. 

SPECIMENS  EXAMINED.— Sixteen  males,  seven  females. 

NEW  RECORD.— FRESNO  COUNTY:  W  entrance  to  Watts 
Valley,  28  Jan.  1968  (TSB). 

NATURAL  HISTORY.— This  species  is  known 
from  oak  woodland  and  has  been  collected  be- 
neath granite,  rhyolite,  and  serpentine  rocks. 

Calicina  macula  (Briggs,  1968), 
new  combination 

(Fig.  14g-i) 

Sitalcina  macula  Briggs,  1968:23. 

DIAGNOSIS.— This  is  the  only  Calicina  with  a 
dark  marbled  color  pattern.  Males  of  this  pae- 
domorphic species  may  also  be  distinguished  by 
the  proportions  of  the  dorsal  process  lobes  and 
in  having  enlarged  mesoapical  spine-bearing  tu- 
bercles on  male  palpal  tarsi. 

ADDITIONAL  DESCRIPTION.— Carapace  with  2 
pairs  of  anterior  tubercles.  Penis  typical  for  the 
serpentinea  group;  dorsal  process  with  notched 
median  lobe;  ventral  plate  with  7  pairs  of  lateral 
setae.  Ovipositor  with  1 2  apical  and  3  pairs  of 
subapical  setae;  microspines  apical. 

TYPE  SPECIMENS.— Male  holotype,  female  allotype,  and  nine 
paratypes  (eight  males,  one  female),  under  serpentine  rocks  in 
an  oak  grassland  at  9  mi  (14.5  km)  SE  Academy,  Fresno  Coun- 
ty, 16  Apr.  1967,  TSB. 

SPECIMENS  EXAMINED.— Only  the  type  series. 

Calicina  dimorphica  Ubick  and  Briggs, 
new  species 

(Fig.  14m,  n) 

DIAGNOSIS.— Males  of  this  species  are  distin- 
guished from  all  others  in  the  group  in  having  a 
dorsal  process  with  a  diamond-shaped  central 
lobe  and  elongated  lateral  lobes.  Males  lack  the 
palpal  spur  but  have  an  enlarged  mesoapical 
spine-bearing  tubercle  on  the  palpal  tarsus. 

ETYMOLOGY.— The  specific  name  refers  to  the 
sexually  dimorphic  palpal  spination  character- 
istic of  this  species. 

DESCRIPTION.— Color  yellowish  orange.  Car- 
apace with  3  pairs  of  anterior  tubercles.  Eyes  well 
developed.  Tarsal  count  3-4-4-5. 

Male:  Total  body  length,  1.22.  Scute  length, 
0.86.  Scute  width,  0.90.  Eye  tubercle  length,  0. 1 7. 
Eye  tubercle  width,  0.18.  Leg  II  length,  2.07. 
Palpal  tarsus  without  dorsal  spur  but  with  en- 
larged mesoapical  spine-bearing  tubercle.  Penis 


118 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


h 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALICINA 


119 


with  ventral  plate  bearing  7  pairs  of  setae,  of 
which  2  distal  pairs  are  enlarged.  Dorsal  process 
with  long,  tapering  lateral  lobes  and  diamond- 
shaped  median  lobe. 

Female:  Total  body  length,  1 .00.  Scute  length, 
0.63.  Scute  width,  0.60.  Eye  tubercle  length,  0. 1 3. 
Eye  tubercle  width,  0.12.  Palpal  tarsus  without 
enlarged  spine-bearing  tubercle.  Ovipositor  with 
apical  microspines,  12  apical  setae,  and  1  pair 
of  subapical  setae. 

TYPE  SPECIMENS.— Male  holotype  and  three  paratypes  (two 
males,  one  female)  under  granite  in  an  oak  grassland  at  the  NE 
entrance  to  Watts  Valley,  Fresno  County,  28  Jan.  1968,  TSB. 

SPECIMENS  EXAMINED.— Only  the  type  series. 

Calicina  mesaensis  Ubick  and  Briggs, 
new  species 

(Fig.  4b,  14J-1) 

DIAGNOSIS.  — Males  of  this  paedomorphic 
species  differ  from  all  others  in  having  a  dorsal 
process  with  a  quadrate  median  lobe  and  elon- 
gate lateral  lobes. 

ETYMOLOGY. —The  specific  name  is  a  reference 
to  the  type  locality,  Table  Mountain. 

DESCRIPTION.— Color  yellow-orange.  Cara- 
pace with  2  pairs  of  anterior  tubercles.  Eyes  well 
developed.  Tarsal  count  3-4-4-4. 

Male:  Total  body  length,  1.14.  Scute  length, 
0.87.  Scute  width,  0.82.  Eye  tubercle  length,  0. 1 8. 
Eye  tubercle  width,  0.17.  Leg  II  length,  2.02. 
Palpal  tarsus  with  greatly  reduced  dorsal  spur 
(Fig.  4b).  Penis  with  7  pairs  of  setae  on  ventral 
plate.  Dorsal  process  with  quadrate  median  lobe 
and  tapering  lateral  lobes.  Stylus  with  apical 
prong. 

Female:  Total  body  length,  1.13.  Scute  length, 
0.78.  Scute  width,  0.71.  Eye  tubercle  length,  0.15. 
Eye  tubercle  width,  0. 1 3.  Ovipositor  with  apical 
microspines  and  7  pairs  of  apical  setae. 

TYPE  SPECIMENS.— Male  holotype  and  25  paratypes  (1 1  males, 
14  females),  under  basalt  rocks  in  an  oak  grassland  at  Table 
Mountain,  2.8  mi  (4.5  km)  N  Millerton  Lake  Road  on  Sky 
Harbor  Road,  Fresno  County,  31  Mar.  1985,  TSB,  TO,  WCR, 
and  DU.  Two  paratypes  (male,  female)  deposited  at  the  AMNH 
collection. 

SPECIMENS  EXAMINED.— Only  the  type  series. 


Calicina  breva  (Briggs,  1968), 
new  combination 

(Fig.  14a-c) 

Sitalcina  sierra  breva  Briggs,  1968:21. 

DIAGNOSIS.  — Males  of  this  paedomorphic 
species  may  be  distinguished  by  the  structure  of 
the  dorsal  process:  median  lobe  pentagonal,  sub- 
equal  to  the  laterals,  and  ventral  plate  with  only 
5  pairs  of  setae. 

ADDITIONAL  DESCRIPTION.— Carapace  with  1 
or  2  pairs  of  anterior  tubercles.  Penis  with  ventral 
plate  bearing  5  pairs  of  small  setae;  dorsal  process 
with  three  subequal,  cloverlike  lobes;  stylus  short 
and  straight.  Ovipositor  with  apical  microspines 
and  7  pairs  of  apical  setae. 

TYPE  SPECIMENS.— Male  holotype,  female  allotype,  and  five 
paratypes  (three  males,  two  females),  under  basalt  rocks  in  a 
grassland  at  1  mi  (1 .6  km)  S  Knight's  Ferry,  Stanislaus  County, 
11  Apr.  1967,  TSB. 

SPECIMENS  EXAMINED.— Only  the  type  series. 


Calicina  basalta  Ubick  and  Briggs, 
new  species 
(Fig.  I4d-f) 

DIAGNOSIS.— This  paedomorphic  species  is 
most  closely  related  to  breva  from  which  it  may 
be  distinguished  by  the  notched  median  lobe  of 
the  dorsal  process  of  the  penis. 

ETYMOLOGY.— The  specific  name  refers  to  the 
microhabitat  of  this  species. 

DESCRIPTION.— Color  yellow.  Carapace  with  1 
or  2  pairs  of  anterior  tubercles.  Retina  occasion- 
ally absent.  Tarsal  count  3-4-4-4. 

Male:  Total  body  length,  1.08.  Scute  length, 
0.70.  Scute  width,  0.77.  Eye  tubercle  length,  0. 1 3. 
Eye  tubercle  width,  0.15.  Leg  II  length,  2.07. 
Palpal  tarsus  without  dorsal  spur.  Penis  similar 
to  that  of  breva  except  that  dorsal  lobe  has  notched 
median  lobe. 

Female:  Total  body  length,  0.86.  Scute  length, 
0.60.  Scute  width,  0.63.  Eye  tubercle  length,  0. 1 2. 
Eye  tubercle  width,  0. 1 2.  Ovipositor  with  apical 
microspines  and  7  pairs  of  apical  setae. 


FIGURE  14.  Male  genitalia  of  the  serpentinea  subgroup.  2.  (a-c)  C.  breva  (paratopotype),  lateral,  dorsal,  and  ventral  views, 
(d-f)  C.  basalta  (holotype),  lateral,  dorsal,  and  ventral  views,  (g-i)  C.  macula  (holotype),  sublateral,  dorsal,  and  ventral  views 
showing  retracted  glans.  (j-1)  C.  mesaensis  (holotype),  lateral,  dorsal  and  ventral  views,  (m,  n)  C.  dimorphica  (paratopotype), 
lateral  and  ventral  views. 


120 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


FIGURE  1 5.  Male  genitalia  of  the  sequoia  subgroup,  (a-e)  C.  diminua  (paratopotype).  (a,  b)  Dorsal  and  lateral  views,  (c-e) 
Ventral,  dorsal,  and  lateral  views,  (f-p)  C.  sequoia,  (f)  Ventral  view  of  penis,  truncus  (Little  River),  (g-j)  Lateral  to  ventral  views 
(Little  River),  (k,  1)  Dorsal  and  ventral  views  (Pine  Ridge  Road),  (m)  Lateral  view  showing  retracted  glans  (Mendocino).  (n-p) 
Lateral,  dorsal,  and  ventral  views  (Usal  Creek).  (Scale  bar  equals  0.25  mm,  except  for  "c-e"  where  it  equals  0.10  mm). 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALICINA 


121 


TYPE  SPECIMENS.  —  Male  holotype  and  five  paratypes  (three 
males,  two  females),  under  basalt  rocks  in  a  grassland  area  8 
mi  (12.9  km)  SW  Jamestown,  Tuolumne  County,  10  Feb.  1968, 
TSB.  Six  paratypes  (three  males,  three  females)  taken  in  similar 
conditions  at  1 3  mi  (2 1  km)  E  Oakdale,  Stanislaus  County,  1 0 
Feb.  1968,  TSB  and  BL. 

SPECIMENS  EXAMINED.— Seven  males,  five  females. 

The  sequoia  Subgroup 

DIAGNOSIS.— The  presence  of  at  least  5  pairs 
of  regularly  spaced  lateral  setae  on  the  ventral 
plate  and  the  extremely  complex  dorsal  process 
are  diagnostic  for  this  subgroup. 

DISTRIBUTION.— Known  from  the  northern 
Coast  Ranges  of  California. 

Calicina  diminua  Ubick  and  Briggs, 
new  species 

(Fig.  15a-e) 

DIAGNOSIS.— This  paedomorphic  species,  the 
smallest  Calicina,  is  most  easily  distinguished  by 
the  elaborate  structure  of  the  glans  penis. 

ETYMOLOGY. —The  specific  name  is  a  reference 
to  the  small  size  of  this  species. 

DESCRIPTION.— Color  yellow.  Carapace  with- 
out anterior  tubercles.  Eyes,  both  retina  and  lens, 
absent.  Tarsal  count  3-4-4-4. 

Male:  Total  body  length,  0.77.  Scute  length, 
0.49.  Scute  width,  0.53.  Eye  tubercle  length,  0.08. 
Eye  tubercle  width,  0.09.  Leg  II  length  1 .4 1 .  Pal- 
pal tarsus  with  dorsal  spur.  Penis  small;  ventral 
plate  with  5  pairs  of  lateral  setae  and  two  triads 
of  ventral  setae;  dorsal  process  ornate;  stylus  short 
and  spinelike. 

Female:  Total  body  length,  0.60.  Scute  length, 
0.46.  Scute  width,  0.46.  Eye  tubercle  length,  0.06. 
Eye  tubercle  width,  0.07.  Ovipositor  with  only 
10  apical  setae  and  without  subapical  setae  or 
microspines. 

TYPE  SPECIMENS.  — Male  holotype  and  11  paratypes  (four 
males,  seven  females),  under  serpentine  on  a  grassland  hillside 
at  Novato,  San  Marin  Drive,  Marin  County,  2  Jan.  1986,  TSB 
and  DU.  Five  paratypes,  0.5  km  SW  of  type  locality,  7  Dec. 
1985  (male)  and  28  Apr.  1968  (three  males,  one  female),  TSB. 
Two  paratypes  (male,  female)  deposited  at  the  AMNH  collec- 
tion. 

SPECIMENS  EXAMINED.— Nine  males,  eight  females. 

Calicina  sequoia  (Briggs  and  Horn,  1966), 
new  combination 

(Fig.  15f-p) 

Sitalcina  sequoia  Briggs  and  Horn,  1966:267.  Briggs,  1968:28. 

DIAGNOSIS.— In  addition  to  their  distinctive 
male  genitalia  (dorsal  process  large  with  1  or  2 


pairs  of  lobes  and  ventral  plate  with  6  to  8  pairs 
of  lateral  setae)  these  small  phalangodids  have 
the  most  reduced  tarsal  count  of  any  Calicina 
(3-4-4-4  or  less),  but  unlike  most  other  paedo- 
morphic species,  have  well  developed  eyes. 

ADDITIONAL  DESCRIPTION.— Tarsal  count  usu- 
ally 3-3-4-4,  occasionally  3-4-4-4.  Carapace  with 
1  pair  of  anterior  tubercles.  Penis  with  ventral 
plate  bearing  6-8  pairs  of  lateral  setae  and  2  or 

3  pairs  of  ventral  setae;  dorsal  process  with  two 
pairs  of  lobes  (the  basal  pair  sometimes  missing) 
and  an  apical  portion  capable  of  hyperextension; 
stylus  ventral,  spinelike.  Ovipositor  with  apical 
microspines,  6  pairs  of  apical  setae,  2  pairs  of 
subapical  setae,  and  1  pair  of  apical  teeth. 

VARIATION.— The  five  specimens  from  the 
southernmost  localities  (Pine  Ridge  Road  and 
Talmage)  have  the  higher  tarsal  count  of  3-4-4- 

4  and  lack  dorsobasal  lobes  on  the  dorsal  process 
(Fig.  1 5k,  1)  but,  otherwise,  do  not  differ  in  gen- 
italic  features. 

JUVENILES.— The  only  known  juvenile,  from 
Mendocino,  is  an  early  instar  having  a  tarsal  count 
of  1-1-2-2.  As  in  adults,  the  eyes  are  well  de- 
veloped. However,  the  anterior  tubercles  are  ab- 
sent and  the  free  tergites  are  armed  with  robust 
spines  which  are  absent  in  adults. 

TYPE  SPECIMENS.  — Female  holotype,  male  allotype,  and  six 
paratypes,  under  rocks  and  logs  in  a  Douglas  fir  forest  at  2.3 
mi  (3.7  km)  S  Piercy,  Mendocino  County,  13  Mar.  1966,  TSB 
andKH. 

SPECIMENS  EXAMINED.— Twenty- two  males,  24  females,  one 
juvenile,  seven  specimens  of  undetermined  sex. 

NEW  RECORDS. -MENDOCINO  COUNTY:  2.3  mi  (3.7  km) 

5  Piercy,  17-18  June  1966  (KH,  VFL);  Mendocino,  4  May 
1963  (OCR);  5  mi  (8  km)  S  Usal  Creek,  17  Apr.  1976  (TSB); 
Pine  Ridge  Road,  1.6  mi  (2.6  km)  S  Low  Gap  Road,  W  of 
Ukiah,  10  Feb.  1985  (TSB);  Casper,  3  Aug.  1957  (JRH,  GAM; 
AMNH  collection);  Talmage,  29  July  1959  (LMS,  ROS;  UCB 
collection). 

NATURAL  HISTORY.— This  species  is  known 
from  redwood  and  Douglas  fir  forests  and  has 
been  collected  beneath  both  logs  and  rocks.  Six 
samples  contained  individuals  of  "Sitalcina" 
cockerelli  Goodnight  and  Goodnight. 


PHYLOGENY 
MONOPHYLY  OF  CALICINA 

Calicina  is  believed  to  be  monophyletic  on  the 
basis  of  its  unique  palpal  tarsus  and  possible  syn- 
apomorphies  in  the  ovipositor  and  glans  mor- 
phologies. Among  the  Nearctic  phalangodids, 
only  the  species  of  Calicina  have  a  telescoping 


122 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


FIGURE  16.  Calicina,  glans  types  (a,  c,  e,  g  =  lateral  views; 
b,  d,  f,  h  =  dorsal  views),  (a,  b)  C.  mariposa,  ST  =  stylus,  CL 
=  collar  lobe,  (c,  d)  C.  digita,  PS  =  parastylus.  (e,  f)  C.  pala- 
praeputia,  MS  =  middle  segment,  (g,  h)  C.  serpentined,  DP  = 
dorsal  process. 


glans  (Fig.  1 6).  Representatives  of  all  other  nom- 
inal genera  were  examined  and  found  to  have  a 
folding  glans.  At  this  time  it  is  not  possible  to 
determine  which  of  these  character  states  is  de- 
rived. Outgroup  comparison  cannot  be  used  since 
the  sister  group  of  these  (presumably  closely  re- 
lated) genera  is  not  known.  It  may  be  argued  that 
the  folding  glans,  because  of  its  widespread  oc- 
currence, is  plesiomorphic.  However,  the  tele- 
scoping glans  is  functionally  simpler,  whereas  the 
folding  glans  is  usually  associated  with  other  de- 
rived genital  characters,  such  as  the  bifurcate 
ventral  plate  ofBanksula  Roewer,  Texella  Good- 
night and  Goodnight,  and  all  Appalachian  gen- 
era. 

The  tarsal  spur  is  not  known  in  any  other  pha- 
langodid  genus  and  is,  clearly,  a  synapomorphy 
for  Calicina  (Fig.  17,  component  1).  However, 
several  species  that  we  include  in  Calicina  lack 
the  spur,  which  we  interpret  to  be  the  result  of 
a  character  reversal.  The  tarsal  spur  is  most 
strongly  developed  in  the  two  least  derived  taxa 
of  Calicina  (the  mariposa  group  and  the  digita 
subgroup)  but  is  reduced  in  size  or  absent  in  most 


of  the  remaining  species.  Of  the  1 1  species  that 
completely  lack  the  spur,  most  are  closely  related 
(based  on  genitalic  characters)  to  species  having 
distinct,  but  small,  spurs  (the  seven  species  be- 
longing to  the  kaweahensis,  serpentinea,  and  se- 
quoia subgroups).  Likewise,  on  the  basis  of  their 
glans  structure,  the  two  species  in  the  arida 
subgroup  and  the  monotypic  minor  subgroup  are 
clearly  representatives  of  the  digita  and  serpen- 
tinea species  groups,  respectively.  This  leaves  only 
one  unassociated  species,  palapraeputia,  that 
lacks  the  spur.  However,  despite  the  unique  glans 
structure  of  this  species,  its  generic  placement  is 
not  questioned  because,  in  addition  to  having  a 
telescoping  glans,  its  ovipositor  closely  resembles 
that  of  some  Calicina  species  in  microspine  dis- 
tribution and  setal  arrangement  (compare  Fig. 
5e,  f  and  5g,  h). 

The  ovipositor  of  Calicina  differs  from  that  of 
Sitalcina  in  several  characters  (Table  1).  Of  these, 
microspines  may  be  derived  because,  with  few 
exceptions,  they  do  not  occur  in  other  Califor- 
nian  phalangodid  species.  Also,  two  setal  series 
(apical  and  subapical)  present  in  most  Calicina 
species  have  not  been  observed  in  Sitalcina  nor 
in  other  Californian  phalangodid  genera.  Further 
investigation  will  be  necessary  to  determine  the 
polarities  of  these  character  states  and  their  uni- 
versality. 

SISTER  GROUP 

There  are  two  possible  sister  groups  of  Cali- 
cina. The  first  consists  of  the  genera  which  have 
a  telescoping  glans.  Of  the  European  phalango- 
dids,  apparently  only  Ptychosoma  Soerensen  has 
this  type  of  glans  (see  figures  in  Brignoli  1968). 
Two  additional  species  with  telescoping  glans 
have  recently  been  described  in  Scotolemon  Lu- 
cas (espanoli  Rambla,  1973  and  balearicus  Ram- 
bla,  1977)  but,  based  on  the  published  genital 
illustrations,  appear  to  be  congeneric  with  Pty- 
chosoma. Although  similar  to  Calicina  in  general 
body  structure,  Ptychosoma  differs  in  having  a 
greater  number  of  palpal  spine-bearing  tubercles, 
in  having  a  strongly  modified  ventral  plate,  and 
in  lacking  sexually  dimorphic  structures. 

A  telescoping  glans  is  likewise  found  in  some 
Southeast  Asian  genera.  Buparellus  Roewer  and 
Bupares  Thorell  have  a  glans  structure  quite  sim- 
ilar to  some  Calicina  species,  but  they  differ  in 
having  more  strongly  armed  palpi  (see  figures  in 
Suzuki  1985).  The  Japanese  Parabeloniscus  Su- 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALICINA 


123 


TABLE  2.    LIST  OF  CHARACTERS  AND  THEIR  PRESUMED  POLARITIES. 


Character 

Plesiomorphic 

Apomorphic 

1. 

male  palpal  tarsus 

unmodified 

with  spur 

2. 

parastylus 

absent 

present 

3. 

stylus  shape 

sinuous 

straight 

4. 

parastylus  insertion 

ventral 

(a)  lateral,  (b)  dorsal 

5. 

microspine  distribution 

distal  half  of  ovipositor 

restricted  to  apex 

6. 

glans  segments 

two 

three 

7. 

dorsal  process  (DP) 

absent 

present 

8. 

DP  apex 

bifurcate 

lobate 

9. 

basoventral  lobes  of  DP 

present 

absent 

10. 

median  lobe  length 

shorter  than  paramedian  lobes 

longer  than  paramedian  lobes 

11. 

stylus  length 

subequal  to  DP 

(a)  longer  than  DP 

(b)  shorter  than  DP 

12. 

lateral  lobe  shape 

length  =  width 

length  >  width 

zuki  shares  with  Calicina  a  similar  glans  and 
ventral  plate  spination  but  has  increased  palpal 
spination  and  a  sexually  dimorphic  third  tergite 
(elongated  in  males)  (see  figures  in  Suzuki  1 973). 
Although  placed  in  the  Phalangodinae,  these  gen- 
era also  lack  the  characteristic  eye  tubercle  and 
may  belong  elsewhere. 

Finally,  the  genitalia  of  Assamiidae  and  Bian- 
tidae  strongly  resemble  those  of  some  Calicina 
species  (see  figures  in  Martens  1977,  1986). 
Members  of  these  families,  however,  have  highly 
modified  palpi  (segments  are  elongated  and  at- 
tenuated and  spination  is  reduced)  and  generally 
lack  well  defined  eye  tubercles. 

The  relationship  of  Calicina  to  the  above  taxa 
depends  on  the  phylogenetic  significance  of  the 
telescoping  glans.  If  this  glans  structure  is  de- 
rived, then  these  and  additional  taxa  may  ac- 
tually be  closely  related.  On  the  other  hand,  if 
the  telescoping  state  is  plesiomorphic,  then  dif- 
ferences in  somatic  characters  assume  greater 
importance.  In  this  case,  the  somatically  similar 
Californian  phalangodids  become  the  second 
group  of  potential  candidates  for  the  Calicina 
sister  group.  Of  these,  the  most  likely  possibility 
is  a  recently  discovered  (and  undescribed)  phal- 
angodid  genus  that  has  a  glans  that  unfolds  and 
telescopes  during  expansion.  This  relationship 
will  be  explored  in  a  future  study. 

INTRAGENERIC  RELATIONSHIPS 

The  proposed  relationships  among  the  Cali- 
cina species  are  based  primarily  on  the  male  gen- 
italia. The  characters  used,  along  with  their  pre- 
sumed polarities,  are  given  in  Table  2  and  the 


resulting  cladograms  are  presented  in  Figures  1 7 
and  18. 

As  described  earlier,  four  types  of  glans  occur 
in  Calicina.  These  appear  to  define  the  major 
lineages  that  are  here  recognized  as  species  groups. 
Each  group  is  defined  by  unique  structures  or 
character  states,  interpreted  as  autapomorphies: 
the  mariposa  group  by  dorsally  grooved  collar 
lobes,  the  digita  group  by  parastyli,  the  pala- 
praeputia  group  by  an  additional  glans  segment, 
and  the  serpentinea  group  by  a  dorsal  process 
(Fig.  16). 

The  mariposa  group  contains  species  with  the 
simplest  glans  morphology  (Fig.  6,  16a,  b).  This 
glans  contains  two  segments:  the  basal  bears  a 
pair  of  collar  lobes  and  the  apical  consists  of  a 
sinuous  stylus.  This  structural  simplicity  is  as- 
sumed to  be  plesiomorphic.  The  presence  of  ple- 
siomorphic somatic  characters  in  this  group  is  at 
least  consistent  with  this  interpretation.  First, 
Calicina  mariposa  has  a  tarsal  count  of  3-5-5-5. 
With  the  exception  of  a  few  species  in  the  digita 
group,  all  other  Calicina  species  have  lower 
counts.  High  tarsal  counts,  3-5-5-5  or  higher, 
occur  in  virtually  all  other  Holarctic  phalango- 
dids. Second,  the  males  of  all  species  in  the  mar- 
iposa group  have  well  developed  tarsal  spurs. 
This  state,  as  argued  previously,  is  presumed  to 
be  plesiomorphic  within  the  genus. 

The  glans  in  the  digita  group  is  similar  to  that 
in  the  mariposa  group,  but  contains  an  additional 
pair  of  sclerites,  the  parastyli  (PS)  (Fig.  1 6c,  d, 
component  2).  The  digita  subgroup  has  ventral, 
scalelike  parastyli  and  a  sagittate  ventral  plate 
(Fig.  7).  It  is  the  most  generalized  element  of  the 
group  because  it  retains  the  high  tarsal  count, 


124 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


Calicina  subgroups 

mar     dig     kaw     top      ari       pal      min     ser    seq 
A(R) 


an 
R 
A(R) 
R 
Ab 
P 
P 
P 


Calicina  serpentinea  subgroup 
pie     dim    mes    mac     bre      bas     ens      pol     ser 


A(R)   A(R) 

P        P 
A(R)      A 


9. 
10. 
11. 
12. 


P  =  plesiomorphy 

A  =  apomorphy 
a,b  =  alternate  apomorphic  states 

R  =  character  reversal 
(R)  =   reversal  in  some  species 


C.  piedra 


FIGURE  17.  Cladogram  with  data  matrix  showing  the  re- 
lationships of  Calicina  subgroups  using  characters  given  in 
Table  2.  Abbreviations:  mar  =  mariposa,  dig  =  digita,  kaw  = 
kaweahensis,  top  =  topanga,  ari  =  arida,  pal  =  palapraeputia, 
min  =  minor,  ser  =  serpentinea,  and  seq  =  sequoia. 


well  developed  tarsal  spurs,  and  a  mariposa-like 
sinuous  stylus.  The  stylus  of  the  kaweahensis  and 
topanga  subgroups  is  straight,  and  considered  to 
be  derived  (component  3).  In  the  former,  the  PS 
are  broad,  ornate,  and  laterally  inserted  (Fig.  8, 
component  4a).  In  the  topanga  subgroup  the  PS 
are  rodlike  and  inserted  dorsolaterally  to  the  sty- 
lus (Fig.  9).  The  two  species  of  the  arida  subgroup 
have  an  unusual  glans  morphology  for  their  group 
(Fig.  10).  In  arida  the  PS  are  represented  by  a 
single  dorsal  rod,  whereas  in  doughensis  they  are 
totally  absent.  Despite  the  absence  of  PS,  dough- 
ensis is  undoubtedly  a  member  of  this  subgroup 
because  of  its  close  relationship  with  arida.  (See 
the  diagnosis  of  the  subgroup  for  synapomor- 
phies.)  Using  the  digita  subgroup  as  an  outgroup, 
the  dorsal  position  of  the  PS  in  the  arida  and 
topanga  subgroups  is  synapomorphic  (compo- 
nent 4b). 
Calicina  palapraeputia,  the  sole  representative 


FIGURE  18.  Cladogram  with  data  matrix  showing  the  re- 
lationships of  species  in  the  serpentinea  subgroup  using  char- 
acters given  in  Table  2.  The  drawing  is  the  glans,  ventral  view, 
of  C.  piedra.  Abbreviations:  pie  =  piedra,  dim  =  dimorphica, 
mes  =  mesaensis,  mac  =  macula,  bre  =  breva,  bas  =  basalta, 
ens  =  ensata,  pol  =  polina,  ser  =  serpentinea,  DP  =  dorsal 
process,  ML  =  median  lobe,  PML  =  paramedian  lobe,  LL  = 
lateral  lobe,  BL  =  basal  lobe,  and  ST  =  stylus. 


of  its  group,  differs  from  all  other  Calicina  species 
in  having  a  three-segmented  glans  (Fig.  11,  1 6e, 
f ).  If  the  stylus  and  basal  segment  of  this  glans 
are  homologous  to  comparable  parts  of  the  mar- 
iposa glans,  the  middle  segment  is  left  unasso- 
ciated.  One  possibility  is  that  this  segment  rep- 
resents a  highly  modified  PS  that  enveloped  and 
fused  around  the  stylus,  as  appears  to  occur  in 
kaweahensis  (Fig.  8e-g).  Alternatively,  the  mid- 
dle segment  could  have  developed  from  similarly 
modified  collar  lobes  from  an  arida-like  ancestor 
(Fig.  lOe-h).  Either  of  these  mechanisms  would 
probably  be  accompanied  by  the  fusion  of  these 
structures  (PS  or  collar  lobes)  on  the  middle  seg- 
ment, but  no  evidence  of  this  has  been  detected. 
Furthermore,  both  interpretations  require  the  in- 
dependent acquisition  of  the  middle  segment 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALICINA 


125 


lobes  in  palapraeputia.  A  more  parsimonious  al- 
ternative (and  the  one  we  accept)  is  that  the  lobes 
in  palapraeputia  are  actually  collar  lobes,  ho- 
mologous to  those  found  in  the  mariposa  and 
iigita  groups.  In  this  case  the  palapraeputia  glans 
:ould  be  derived  from  a  mariposa  type  through 
the  development  of  an  additional  collar  on  the 
3asal  segment  of  the  latter. 

In  the  serpentinea  group  the  glans  has  a  de- 
*ived,  lobe-bearing  structure,  the  dorsal  process 
'Fig.  1 6g,  h,  component  7).  In  most  species  (the 
serpentinea  subgroup)  the  dorsal  process  (DP) 
sears  five  apical  lobes  (Fig.  13,  14).  The  remain- 
ng  three  species  all  have  unique  DP  morphol- 
Dgies.  Calicina  minor,  the  sole  representative  of 
ts  subgroup,  has  the  simplest  DP  bearing  a  pair 
:>f  dorsobasal  lobes  and  an  apical  bifurcation  in- 
stead of  lobes  (Fig.  1 2).  Of  the  members  of  the 
sequoia  subgroup,  the  DP  of  diminua  is  divided 
.nto  two  lateral,  extremely  ornate,  hemispheres; 
whereas  in  sequoia  it  contains  a  pair  of  basal  and 
ipical  lobes  in  addition  to  a  complex  apical  re- 
>ion  (Fig.  1 5).  Despite  fundamental  differences 
n  the  glans  structure,  the  last  two  species  seem 
:o  be  related  as  they  share  a  unique  ventral  plate 
spination.  The  presence  of  apical  lobes  on  the 
DP  appears  to  be  a  synapomorphy  uniting  all 
species  of  the  serpentinea  group  except  minor 
component  8). 

The  species  of  the  serpentinea  subgroup  all  have 
•ather  similar  glans  morphologies,  with  the  ex- 
option  of  piedra.  The  piedra  glans  differs  from 
;he  others  in  four  characters:  (1)  DP  with  basal 
obes,  (2)  paramedian  lobes  longer  than  median 
obe,  (3)  lateral  lobes  wider  than  long,  and  (4) 
stylus  subequal  to  DP  in  length  (Fig.  1 3a-c,  1 8). 
Because  basal  lobes  are  found  in  the  other 
subgroups,  their  presence  is  a  synapomorphy  for 
;he  entire  serpentinea  group,  but  a  plesiomorphy 
it  the  subgroup  level.  The  implication  that  piedra 
,s  the  most  generalized  species  in  the  subgroup 
s  compatible  with  its  relatively  higher  tarsal  count 
3-4-4-5).  The  remaining  species  in  this  subgroup 
ire  united  in  lacking  basal  lobes  (component  9) 
ind  in  having  relatively  larger  median  lobes 
[component  10).  They  are  subdivided  into  two 
groups  based  on  stylus  length.  Using  piedra  as 
the  outgroup,  stylus  lengths  either  longer  (com- 
ponent 1  la)  or  shorter  (component  1  Ib)  than  the 
DP  appear  to  be  synapomorphic.  Finally,  elon- 
gated lateral  lobes  (component  12)  are  synapo- 
morphic for  some  species. 

Because  of  the  serpentinea  group's  unique  glans 


morphology,  its  relationship  to  the  other  groups 
would  appear  remote.  However,  its  ovipositor 
structure  implies  a  close  relationship  with  pa- 
lapraeputia. In  both  groups  the  distribution  of 
microspines  is  restricted  to  the  apical  surface  of 
the  ovipositor  which,  using  mariposa  as  the  out- 
group,  is  synapomorphic  (component  5).  This 
relationship  suggests  that  the  glans  morphologies 
in  the  two  groups  are  likewise  closely  related; 
namely,  that  the  dorsal  process  (DP)  is  homol- 
ogous to  the  palapraeputia  middle  glans  segment 
(component  6).  The  DP  could  conceivably  have 
developed  from  the  middle  segment  by  the  en- 
largement of  the  stylus  opening  along  the  ven- 
troapical  surface.  In  fact,  the  basal  lobes  of  the 
DP,  evident  in  members  of  all  subgroups,  may 
well  be  vestiges  of  the  ventral  surface  of  the  mid- 
dle segment. 

Two  alternative  explanations  for  the  origin  of 
the  DP  are  possible.  First,  the  DP  might  repre- 
sent highly  derived  parastyli.  In  kaweahensis,  for 
example,  the  parastyli  are  greatly  enlarged,  par- 
tially envelop  the  stylus,  and  are  fused  dorsally 
(Fig.  8e-g).  Further  enlargement  could  result  in 
a  DP.  However,  this  interpretation  requires  the 
concomitant  loss  of  collar  lobes  in  the  serpenti- 
nea group.  Second,  the  DP  might  be  homologous 
to  collar  lobes.  In  arida  the  collar  lobes  are  large 
and  apically  produced  (Fig.  lOe-h)  and  could 
conceivably  form  a  DP  through  fusion  and  ad- 
ditional enlargement.  However,  this  interpreta- 
tion requires  a  character  reversal  of  parastyli  in 
the  serpentinea  group.  Furthermore,  both  of  these 
explanations  imply  a  sister  group  relationship 
between  the  serpentinea  group  and  either  ka- 
weahensis or  arida  which  is  not  supported  by  any 
other  known  character. 


BlOGEOGRAPHY 

As  is  to  be  expected  of  cryptozoic  organisms 
having  low  dispersal  potential,  the  species  of  Cal- 
icina are  allopatric  or  parapatric  (see  Fig.  19). 
The  four  known  instances  of  sympatry,  between 
palapraeputia  and  ensata,  diminua  and  polina, 
digita  and  kaweahensis,  and  digita  and  clough- 
ensis,  are  all  between  members  of  different  groups 
or  subgroups.  Furthermore,  these  cases  of  sym- 
patry may  not  actually  involve  interspecific  in- 
teraction. In  the  last  case,  for  example,  clough- 
ensis  is  troglobitic  and  digita  epigean.  Also, 
though  diminua  and  polina  live  in  close  prox- 


126 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


sequoia- 


diminu 


polina 


minor- 


sierra 


serpentinea 


arida 


morroensis 


topanga 


\ 


conif era 

yosemitensis 

mariposa 


\ 


dimorphica 
palapraeputia 


digita 
cloughensis 

galena 


FIGURE  19.     Map  of  central  California  showing  the  distribution  of  species  of  Calicina.  A  black  dot  indicates  the  localities  of 
paedomorphic  species,  a  circle  those  of  non-paedomorphics.  Overlapping  symbols  denote  sympatry. 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALICINA 


127 


FIGURE  20.  Distribution  of  the  subgroups  ofCalicina  and  a  cladogram  of  their  relationships.  A  dotted  line  connects  disjunct 
elements  and  a  dashed  line  indicates  the  presumed  boundaries  of  the  Sierran  exotic  terranes  (from  Hendrickson  1 986).  Abbre- 
viations: K  =  kaweahensis,  M  =  minor,  and  P  =  palapraeputia. 


128 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


dimorphica 


O   ensata 


"7"  piedra 


FIGURE  2 1 .  Map  of  central  California  showing  the  distribution  of  the  species  of  the  serpentined  subgroup  with  a  cladogram 
of  their  relationships.  The  'x'  indicates  the  presumed  time  of  the  subgroup's  disjunction  by  the  Central  Valley.  Abbreviation: 
m  =  macula. 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALICINA 


129 


imity,  they  have  never  been  collected  in  the  same 
rock  outcrops. 

An  area  cladogram  of  the  groups  and  subgroups 
is  given  in  Figure  20.  Two  gradients  are  apparent. 
An  altitudinal  gradient  exists  between  the  mar- 
iposa,  digita,  and  serpentined  groups,  which  oc- 
cupy progressively  lower  elevations.  A  latitudi- 
nal gradient  is  also  evident:  the  digita  group  occurs 
primarily  along  the  southern  mountains  and 
coast,  whereas  the  serpentinea  group  occupies  the 
central  and  northern  parts  of  the  state. 

The  distribution  pattern  of  the  subgroups  of 
Calicina  is  compatible  with  an  origin  through 
vicariance.  The  initial  speciation  events  are  like- 
ly to  have  taken  place  in  the  Sierra.  The  Sierran 
component  of  Calicina  contains  representatives 
of  all  four  groups,  including  the  most  generalized 
members  (mariposa  group,  digita  subgroup,  and 
piedrd).  The  initial  speciations  may  have  resulted 
from  two  barriers  (given  the  basal  trichotomy) 
located  between  the  ranges  of  the  mariposa,  dig- 
ita, and  serpentinea  groups. 

It  is  worth  noting  that  these  presumed  barriers 
closely  correspond  to  the  proposed  boundaries 
of  exotic  terranes  given  by  Hendrickson  (1986) 
and  indicated  in  Figure  20.  The  coastal  Calicina 
distributions  also  correspond  closely  to  the  maps 
of  exotic  terranes.  The  species  topanga  and  mor- 
roensis  each  occupy  separate  terranes,  and  the 
coastal  distribution  of  the  serpentinea  group  (with 
the  exception  of  the  two  westernmost  serpentinea 
localities)  is  located  within  the  Franciscan  com- 
plex of  terranes. 

A  few  taxa  exhibit  disjunct  distributions  that, 
in  conjunction  with  our  knowledge  of  geologic 
history,  may  be  used  for  estimating  their  mini- 
mum ages.  Thepolina  distribution  surrounds  the 
San  Francisco  Bay,  suggesting  that  the  species 
predates  the  formation  of  the  bay.  Similarly,  to- 
panga occurs  on  Santa  Cruz  Island  as  well  as  the 
adjacent  mainland.  If  the  currently  disjunct  dis- 
tribution resulted  from  the  formation  of  the  San- 
ta Barbara  Channel,  then  the  species  must  be 
rather  old.  Recent  estimates  for  a  land  connec- 
tion to  these  islands  are  much  older  (although 
unspecified)  than  the  previous  model  of  a  Pleis- 
tocene land  bridge  (Wenner  and  Johnson  1980). 
It  may  be  more  plausible,  however,  that  the  San- 
ta Cruz  Island  population  represents  a  more  re- 
cent colonization  from  the  mainland.  Rafting  is 
a  commonly  evoked  mechanism  (for  example, 
by  Yanev  (1982)  for  slender  salamanders)  and  a 
good  possibility  for  the  forest-dwelling  topanga. 


The  Central  Valley  is  currently  an  obvious  and 
impenetrable  barrier  for  Calicina  (owing  to  the 
absence  of  favorable  mesic  habitats)  and  divides 
the  distributions  of  two  subgroups,  arida  and 
serpentinea.  The  Sierran  uplift,  which  formed  the 
Central  Valley,  is  thought  to  have  begun  in  the 
early  Pleistocene.  A  barrier  could  eventually  have 
resulted  from  the  drying  effects  of  glacial  retreats, 
as  proposed  for  the  populations  of  Ortholasma 
levipes  Shear  and  Gruber  (1983:12).  Our  inter- 
pretation of  the  serpentinea  subgroup  relation- 
ships indicates  that  the  Central  Valley  could  have 
been  responsible  for  the  disjunction  of  ensata 
from  serpentinea-polina  (indicated  by  an  "x"  in 
Fig.  21). 

Another  method  of  estimating  the  time  of  cla- 
dogenic  events  in  Calicina  is  to  use  the  findings 
for  another  organism  from  the  same  ancestral 
biota.  A  good  example  is  the  slender  salamander, 
Batrachoseps.  It  appears  very  likely  that  Batra- 
choseps  and  Calicina  experienced  a  comparable 
evolutionary  history  for  three  reasons:  their  sim- 
ilar habitat  preferences,  congruent  distributions, 
and  compatible  phylogenies. 

The  ecological  requirements  of  Batrachoseps, 
as  described  by  Yanev  (1980)  are  remarkably 
similar  to  those  of  Calicina.  Both  genera  include 
species  restricted  to  forests  or  oak  woodlands. 
Both  are  specialized  for  subterranean  life,  appear 
at  the  surface  only  during  favorable  periods,  and 
are  able  to  survive  in  small  isolates  of  suitable 
habitat.  Studies  on  Batrachoseps  indicate  that 
individuals  are  extremely  sedentary  and  have  a 
very  low  dispersal  potential,  as  is  probably  the 
case  for  Calicina. 

The  phylogeny  of  Batrachoseps  consists  of  three 
main  lineages.  The  basal  branch,  currently  rep- 
resented by  two  disjunct  isolates,  is  thought  to 
have  formerly  ranged  along  the  eastern  Sierra 
Nevada,  close  to  that  of  the  present  distribution 
of  the  mariposa  group.  The  second  branch  is 
represented  by  B.  attenuatus,  which  has  a  coastal 
component  that  is  almost  identical  to  that  of  the 
serpentinea  group  and  a  Sierran  component  in- 
tersecting the  distribution  of  breva  and  basalta. 
The  third  branch  contains  the  remaining  four 
species  whose  combined  distribution  is  very  sim- 
ilar to  that  of  the  digita  group. 

On  the  basis  of  electrophoretic  and  immuno- 
logical  studies,  Yanev  (1980)  proposed  an  origin 
for  Batrachoseps  in  the  early  Eocene  (ca.  50  mya). 
However,  Hendrickson  (1986)  criticized  these 
values  as  being  far  too  recent  and  suggested  that 


130 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


TABLE  3.    FIELD  DATA  OF  CALICINA. 


Taxa 

N' 

Biome 

Habitat 

Season2 

mariposa  gp 

10 

sequoia-yellow  pine 

snags,  logs 

summer 

digita 

9 

sequoia-yellow  pine,  oak  wood,  grass 

logs,  gra,  sla 

all  year 

digita  (cave) 
sierra 
sierra  (cave) 

1 
6 
8 

sequoia-yellow  pine 
digger  pine-oak 
digger  pine-oak 

cave 
lim,  sch,  sla 
caves 

spr3 
win-spr 
all  year 

galena 
kaweah 

3 

12 

sequoia-yellow  pine 
oak  wood 

gra,  logs 
gra 

summer 
win-spr 

topanga 

9 

broadleaf  evergreen,  digger  pine-oak, 
oak  wood 

logs,  bas,  snd,  bre 

win-spr3 

kee 
mor 

1 

4 

oak  wood 
grass 

gra 
srp 

win-spr 
win-spr 

clough 
arida 

3 

2 

oak  wood 
oak  wood 

cave 
srp 

srp-sum3 
win-spr 

pala 

5 

oak  wood 

gra,  srp 

win-spr 

minor 

4 

grass 

srp 

win-spr 

serp  sbgp" 
serp 

serp  (cave) 

29 
12 

1 

oak  wood,  grass 
redwood-broadleaf  evergreen,  digger 
pine-oak,  oak  wood,  grass 
redwood 

spr,  bas,  gra,  snd,  rhy 
logs,  srp,  lim,  snd 

cave 

win-spr 
all  year 

spr3 

diminua 
sequoia 

3 
11 

grass 
redwood-Douglas  fir 

srp 
logs,  rocks 

win-spr 
all  year 

Abbreviations:  N  =  number  of  collection  samples;  gp  =  group;  sbgp  =  subgroup;  kaweah  =  kaweahensis;  kee  =  keenea;  mor 
=  morroensis;  clough  =  cloughensis\  pala  =  palapraeputia\  serp  =  serpentinea;  wood  =  woodland;  grass  =  grassland;  bas  =  basalt; 
bre  =  breccia;  gra  =  granite;  lim  =  limestone;  rhy  =  rhyolite;  sch  =  schist;  sla  =  slate;  snd  =  sandstone;  srp  =  serpentine;  spr 
=  spring;  sum  =  summer;  win  =  winter. 

'  Roughly  3/4  of  all  samples  include  biome  and  habitat  data. 

2  Time  of  the  year  when  adults  are  active. 

3  Probably  active  all  year. 

4  Excluding  serpentinea. 


the  isolation  of  Batrachoseps  must  have  taken 
place  between  the  late  Jurassic  and  the  early  Cre- 
taceous. This  conclusion  was  based  on  the  cor- 
relation between  cladistic  branching  and  geologic 
events,  especially  the  history  of  exotic  terranes. 
The  geologic  history  of  the  Californian  ter- 
ranes is  still  largely  unresolved.  Most  of  the  ter- 
ranes are  believed  to  have  distant  southern 
origins,  presumably  in  the  vicinity  of  southern 
North  America.  Their  northward  displacement 
as  isolated  units  is  thought  to  have  commenced 
in  the  mid-Mesozoic.  Subsequent  accretions  to 
the  North  American  Plate  range  from  late  Ju- 
rassic to  Cretaceous  for  the  Sierran  terranes,  to 
early  Eocene  for  the  southern  coastal  terranes, 
and  to  late  Miocene  for  the  western  elements  of 
the  Franciscan  complex.  If  the  present  distri- 
bution pattern  of  Calicina  is  the  result  of  isola- 
tion on  terranes,  then  a  possible  scenario  is  that 
the  original  breakup  of  the  Sierran  terranes  iso- 
lated the  ancestral  populations  of  Calicina,  which 
eventually  evolved  into  distinct  species.  Follow- 


ing accretion  to  their  present  position,  some  dis- 
persal from  the  terranes  would  be  necessary  to 
account  for  the  present  distribution,  especially 
for  the  south  Sierran  members  of  the  serpentinea 
group.  The  later  arrival  of  the  coastal  terranes 
suggests  widespread  dispersal  from  the  Sierran 
populations. 

NATURAL  HISTORY 

Information  on  the  natural  history  of  Calicina 
comes  from  our  field  observations  and  the  data 
associated  with  some  1 30  collection  samples.  This 
information,  summarized  in  Table  3,  is  dis- 
cussed below. 

HABITAT.— Species  of  Calicina  occur  in  mesic 
habitats  (in  conditions  of  total  darkness  and  high 
humidity)  but  are  apparently  absent  from  those 
areas  that  are  periodically  inundated  or  situated 
in  saturated  soils.  Most  species  are  found  exclu- 
sively beneath  medium  to  large-sized  rocks  that 
are  in  contact  with  the  soil  and  undisturbed.  Oth- 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALICINA 


131 


er  species  also  occur  beneath  decomposing  logs. 
One  group  of  species  (the  mariposa  group)  is 
found  exclusively  under  logs  and,  more  com- 
monly, beneath  and  among  fragments  of  fallen 
bark  that  surround  large  snags.  Four  species  have 
been  recorded  from  caves.  One  of  these  (clough- 
ensis)  is  a  troglobite,  the  others  are  troglophiles 
with  predominantly  surface  populations.  These 
habitat  preferences  of  Calicina  are  similar  to  those 
previously  recorded  for  all  Californian  phalan- 
godids  (Briggs  1968)  except  that  Calicina,  unlike 
some  species  of  Sitalcina,  has  never  been  col- 
lected in  leaf  litter. 

Species  of  Calicina  are  associated  with  many 
rock  types,  the  most  frequent  being  serpentine. 
Of  the  64  collection  samples  that  indicate  specific 
rock  associations,  28  are  from  serpentine.  Other 
frequently  recorded  rocks  are  granite  (13  sam- 
ples), basalt  (seven),  sandstone  (seven),  and  lime- 
stone (four,  excluding  cave  collections).  The  Coast 
Range  species  have  been  recorded  primarily  from 
serpentine  and  sandstone,  the  Sierran  ones  from 
granite  and  basalt. 

BIOME.— The  species  of  Calicina  are  known 
from  a  variety  of  biomes  (see  also  Fig.  22).  Some 
are  apparently  restricted  to  dense  forests.  Mem- 
bers of  the  mariposa  group  and  galena  are  known 
only  from  high  elevations  (1,200-2,000  m)  in 
primary  stands  of  giant  sequoia-yellow  pine  as- 
sociation. Calicina  sequoia  has  been  collected  only 
in  redwood-Douglas  fir  association.  Three  ad- 
ditional species  have  been  collected  in  dense  for- 
ests. Calicina  digita  is  recorded  from  giant  se- 
quoia-yellow pine,  topanga  and  serpentinea  from 
broadleaf  evergreen,  the  latter  also  in  association 
with  redwood.  However,  these  three  species  are 
also  known  from  more  open  forests  (digger  pine- 
oak  and  oak  woodland),  as  well  as  from  grass- 
land. All  remaining  species  are  known  only  from 
oak  woodland  and  grassland  biomes. 

ACTIVITY.— The  period  of  adult  surface  activ- 
ity varies  between  the  species.  Three  patterns  are 
evident.  (1)  Species  active  during  the  rainy  sea- 
son (winter  to  spring).  These  species  (14)  live  in 
grassland  and  oak- woodland  biomes  (one  species 
from  digger  pine-oak).  (2)  Species  active  during 
the  summer.  Species  of  the  mariposa  group  and 
galena,  collected  in  the  summer  months,  are 
known  only  from  giant  sequoia-yellow  pine  for- 
ests. (3)  Species  active  throughout  the  year.  The 
residents  of  dense,  low-elevation  forests  (sequoia 
and  forest  populations  of  serpentinea,  digita,  and 
possibly  topanga)  have  a  year-long  adult  activity 


period.  Likewise,  the  cavernicolous  populations 
of  sierra  have  been  collected  throughout  the  year. 
It  is  expected  that  the  other  cavernicolous  Cal- 
icina also  have  year-round  activity. 

LIFE  CYCLE.— Almost  all  individuals  of  Cali- 
cina encountered  in  the  field  and  in  collections 
are  adults.  Despite  repeated  efforts  to  collect  ju- 
veniles, only  three  subadults  are  known  from 
epigean  habitats  (five  additional  juveniles  are 
known  from  caves).  This  absence  of  juveniles  at 
the  surface  suggests  that  development  in  Calicina 
takes  place  within  the  soil.  The  appearance  of 
adults  of  species  from  drier  environments  co- 
incides with  the  onset  of  the  rainy  season.  Nor- 
mally, only  one  or  a  few  individuals  are  present 
on  each  rock  or  log  undersurface.  However,  in 
several  species  we  have  observed  aggregations  of 
1 0  or  more  individuals,  often  containing  at  least 
one  mating  pair. 

COHABITANTS.— The  habitat  of  Calicina  is 
densely  populated  by  a  rich  assortment  of  or- 
ganisms. The  most  commonly  observed  insects 
are  campodeid  diplurans  and  collembolans.  Ob- 
servations suggest  that  collembolans  are  the  like- 
ly prey  ofBanksula  (Briggs  and  Ubick  1 98 1)  and 
probably  also  of  Calicina. 

Several  laniatorid  opilionids  live  sympatri- 
cally  with  Calicina.  Zuma  Goodnight  and  Good- 
night is  known  from  dense  coniferous  forests  at 
higher  elevations  of  the  Sierra  Nevada.  Sitalcina 
is  found  in  forests  of  both  coastal  and  Sierran 
regions.  "Sitalcina"  cockerelli  Goodnight  and 
Goodnight  lives  in  the  northern  coastal  conif- 
erous forests.  Banksula  is  restricted  to  caves  of 
the  central  Sierra  foothills,  but  has  never  been 
collected  sympatrically  with  Calicina. 

In  the  central  Coast  Range  and  Sierran  foothill 
regions,  numerous  spiders  are  repeatedly  col- 
lected with  Calicina.  Some  of  the  more  common 
(and  distinctive)  are  leptonetids  (Archoleptone- 
td),  telemids  (Usofild),  oonopids  (Orchestina  and 
Scaphielld),  amaurobioids  (Blabomma,  Calym- 
maria,  Pimus,  and  Titiotus)  and  many  gnapho- 
sids.  Interestingly,  these  taxa,  with  the  exception 
of  the  oonopids,  gnaphosids,  and  Pimus,  all  have 
local  cavernicolous  representatives. 

CONSERVATION.— Several  Calicina  species  are 
known  only  from  single  localities  and  may  be 
facing  extinction  as  a  result  of  habitat  destruc- 
tion. An  extreme  example  is  arida  whose  known 
range  consists  of  several  square  meters  in  a  small 
ravine,  surrounded  by  extremely  xeric  or  dis- 
turbed environments.  Several  additional  species 


132 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


TABLE  4.  THE  RELATIONSHIP  BETWEEN  PAEDOMORPHIC  AND 
TROGLOBITIC  CHARACTERS  (TARSAL  COUNT  AND  EYE  Loss)  AND 
XERIC  ENVIRONMENTS  (OAK-GRASSLAND  BIOMES)  IN  CALICINA. 


Calicina  species 

Tarsal 
count 

Eyesight 

Biome 

mariposa  gp. 

mariposa  sbgp. 

mariposa 

3-5-5-5 

yosemitensis 

3-4-4-5 

conifera 

3-4-4-5 

blind 

digita  gp. 

digita  sbgp. 

digita 

3-5-5-5 

xeric 

(most) 

sierra 

3-4-4-4 

xeric 

kaweahensis  sbgp. 

kaweahensis 

3-5-4-5 

xeric 

galena 

3-5-5-5 

(3-5-4-5) 

topanga  sbgp. 

topanga 

3-5-5-5 

blind 

(some) 

keenea 

3-4-4-5 

xeric 

morroensis 

3-4-4-4 

blind 

xeric 

arida  sbgp. 

cloughensis 

3-5-5-5 

blind 

arida 

3-4-4-4 

blind 

xeric 

palapraeputia  gp. 

palapraeputia  sbgp. 

palapraeputia 

3-4-4-5 

xeric 

serpentinea  gp. 

minor  sbgp. 

minor 

3-4-4-4 

blind 

xeric 

serpentinea  sbgp. 

piedra 

3-4-4-5 

xeric 

ensata 

3-4-4-4 

xeric 

serpentinea 

3-4-4-5 

blind 

xeric 

(most) 

polina 

3-4-4-4 

blind 

xeric 

macula 

3-4-4-4 

blind 

xeric 

(some) 

mesaensis 

3-4-4-4 

xeric 

dimorphica 

3-4-4-5 

xeric 

breva 

3-4-4-4 

blind 

xeric 

basalta 

3-4-4-4 

blind 

xeric 

(some) 

sequoia  sbgp. 

diminua 

3-4-4-4 

blind 

xeric 

sequoia 

3-3-4-4 

(3-4-4-4) 

are  known  from  single  localities:  basalta,  clough- 
ensis, conifera,  diminua,  dimorphica,  keenea, 
macula,  mesaensis,  minor,  and  piedra.  Conser- 
vation programs  will  be  necessary  if  these  or- 
ganisms and  their  habitats  are  to  be  protected. 


ECOLOGY 

Calicina  exhibits  considerable  size-related  in- 
terspecific variation.  Larger  species  have  rela- 
tively higher  tarsal  counts,  darker  pigmentation, 
greater  number  of  anterior  tubercles,  larger  tarsal 
spurs  (of  male  palpi),  and  well  developed  eyes. 
These  structures  are  reduced  in  size  or  absent  in 
their  smaller  relatives.  The  possibility  that  one 
of  these  character  clusters  is  synapomorphic  is 
incompatible  with  our  cladistic  analysis,  as  nei- 
ther group  of  species  represents  a  clade.  How- 
ever, there  appears  to  be  a  correlation  between 
reduced  size  and  xeric  biomes  (see  Table  4  and 
Fig.  22).  This  section  will  explore  the  hypothesis 
that  the  character  transformations  result  from 
two  phenomena,  paedomorphosis  and  troglob- 
ism,  and  represent  adaptations  to  xeric  condi- 
tions. 

PAEDOMORPHOSIS.— The  direction  of  the  so- 
matic character  transformation  can  be  inferred 
by  comparison  with  the  ancestral  state.  As  ar- 
gued in  the  phylogeny  section,  the  probable  out- 
group  for  the  remaining  species  of  Calicina 
is  mariposa,  the  species  with  the  simplest 
(=plesiomorphic)  glans  morphology.  The  forest- 
dwelling  mariposa  is  a  large,  strongly  pigmented 
species,  which  has  a  high  tarsal  count  (3-5-5-5), 
the  greatest  number  of  anterior  tubercles  (seven 
to  eight  pairs),  a  large  tarsal  spur,  and  well  de- 
veloped eyes.  That  the  character  states  associated 
with  large  size  are  plesiomorphic  is  further  sup- 
ported by  the  condition  in  other  Phalangodidae. 
With  the  exception  of  troglobitic  species,  vir- 
tually all  remaining  Nearctic  phalangodids  are 
large,  strongly  pigmented  forest  dwellers  with  high 
tarsal  counts  and  well  developed  eyes.  It  is,  thus, 
reasonable  to  suppose  that  the  reduction  in  size 
and  structure  is  derived. 

Ontogenetic  transformations  were  ascertained 
indirectly  by  examining  the  juvenile  morpholo- 
gy. A  total  of  eight  specimens,  representing  five 
species,  were  examined  (attempts  to  collect  ad- 
ditional juveniles  were  not  successful).  All  ju- 
veniles are  white,  lacking  any  of  the  orange  pig- 
mentation found  in  adults,  all  lack  tarsal  spurs, 
and  all,  with  the  exception  of  one  penultimate 
instar  (galena),  lack  anterior  tubercles.  Based  on 
relative  size,  these  specimens  represent  early  (four 
individuals),  middle  (two),  and  penultimate  in- 
stars  (two).  The  only  observable  correlation  with 
growth,  other  than  size,  is  an  increase  in  tarsal 
count.  The  earliest  instars  have  tarsal  counts  of 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALICINA 


133 


4.0-1 


3.0- 


O 

:    Pi 


2.5- 


2.0- 


1.5 


TC  <  3-4-U-U 


si 

O 


•  €    O 


br 


pa 


O 


•  O    ka 


TC>  3-4-U-5 


:.     dn 


me 

C      / 


0.8 


0.9 


1.0 


1.1 


1.2 


1.3 


1.4 


1.5 


1.6 


1.7 


TOTAL  BODY  LENGTH   (MM) 

FIGURE  22.  A  plot  showing  the  distribution  of  paedomorphic  and  troglobitic  characters  in  Calicina.  The  Total  Body  Length 
and  Leg  II  Length  values  are  from  the  measurements  of  male  holotypes  (male  allotypes  of  digita,  kaweahensis,  minor,  and 
serpentined).  A  circle  indicates  species  having  well  developed  eyes,  a  black  dot  denotes  blind  species,  and  a  mixed  symbol 
indicates  the  species  with  partial  retinal  loss.  The  vertical  line  separates  those  species  with  tarsal  counts  of  3-4-4-4  or  less 
(paedomorphics)  from  those  with  higher  counts.  The  forest-dwelling  species  are  encircled  by  a  solid  line,  those  from  oak  woodland 
and  grassland  by  a  dotted  line,  and  those  living  in  both  densely  forested  and  grassland  biomes  by  a  dashed  line.  Abbreviations: 
a  =  arida,  ba  =  basalta,  br  =  breva,  cl  =  cloughensis,  co  =  conifera,  dg  =  digita,  dn  =  diminua,  dr  =  dimorphica,  e  =  ensata, 
g  =  galena,  ka  =  kaweahensis,  ke  =  keenea,  me  =  macula,  me  =  mesaensis,  mi  =  minor,  mo  =  morroensis,  mr  =  mariposa, 
pa  =  palapraeputia,  pi  =  piedra,  po  =  polina,  si  =  sierra,  sq  =  sequoia,  sr  =  serpentinea,  t  =  topanga,  and  y  =  yosemitensis. 


134 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


1  - 1  -2-2  (sierra  and  sequoia)  and  1  -2-2-2  (dough- 
ensis); later  instars  (serpentined)  have  2-2-3-3; 
and  the  oldest  (galena),  3-5-4-5.  Eyes  are  present 
in  species  with  eyed  adults  (sierra,  galena,  and 
sequoia),  absent  in  the  others  (doughensis  and 
serpentined).  Despite  the  small  sample  size,  it 
seems  probable  that  these  observations  apply  to 
all  Calidna,  given  the  generally  conservative  na- 
ture of  juvenile  characters. 

Using  mariposa  as  the  outgroup,  the  character 
states  found  in  the  small  species  of  Calidna  are 
derived.  However,  a  comparison  with  their  pre- 
sumed ontogeny  indicates  that  they  are  plesio- 
morphic.  The  inescapable  conclusion  is  that  the 
small  species  are  derived  through  the  retention 
of  juvenile  characters;  namely,  that  they  are  pae- 
domorphic.  The  other  possible  alternative  is  that 
the  juvenile  states  are  in  reality  character  re- 
versals; i.e.,  that  penultimate  instars  have  adult 
somatic  morphology.  However,  this  latter  pos- 
sibility is  clearly  less  parsimonious  and  not  sup- 
ported by  the  morphology  of  the  two  penulti- 
mates examined. 

Paedomorphosis  is  a  widespread  phenomenon 
in  Calidna.  With  the  exception  of  three  species 
(mariposa,  digita,  and  topangd),  all  species  have 
some  degree  of  paedomorphic  modification. 
However,  only  the  1 2  species  having  the  lowest 
tarsal  counts  (3-4-4-4;  3-3-4-4  in  sequoia)  and 
the  high  correlation  to  other  characters  (see  Fig. 
22)  are  here  considered  to  be  significantly  pae- 
domorphic. Given  our  phylogenetic  interpreta- 
tions, paedomorphic  species  are  present  in  all 
subgroups  except  mariposa,  kaweahensis,  and 
palapraeputia,  and  must  have  evolved  indepen- 
dently at  least  nine  times.  Most  of  the  paedo- 
morphics  (nine)  belong  to  the  serpentinea  group; 
the  remainder  belong  to  the  digita  group,  where 
they  occupy  the  periphery  of  the  group's  distri- 
bution (see  Fig.  1 9). 

Paedomorphosis  has  been  well  documented  in 
Opiliones.  In  her  review,  Rambla  (1980)  listed 
many  examples  of  the  phenomenon  in  Caddidae, 
Ischyropsalidae,  Nemastomatidae,  and  Penta- 
nychidae.  Briggs  (1986),  in  addition  to  pointing 
out  the  presence  of  paedomorphic  Calidna  (as 
Sitaldnd)  species,  included  Triaenonychidae 
among  the  examples. 

TROGLOBISM.— Although  only  one  species  of 
Calidna  (doughensis)  is  an  obligate  cavernicole, 
troglobitic  characters  are  found  in  many  species 
of  Calidna,  including  most  paedomorphics.  The 
most  obvious  character  is  eye  loss.  Eight  of  the 
paedomorphic  species  lack  retinae,  along  with 


three  non-paedomorphics  (excluding  doughen- 
sis). However,  eye  loss  does  not  appear  to  be  a 
paedomorphic  character,  as  juveniles  of  eyed 
species  have  well  developed  eyes  (for  example, 
the  early  instars  of  sierra  and  sequoia).  Addi- 
tional character  states  of  paedomorphics,  such 
as  pale  coloration  and  small  body  size,  could  also 
be  troglobitic  adaptations.  Another  common 
troglobitic  character  is  increased  appendage 
length.  Preliminary  results  (Fig.  22)  suggest  that 
the  blind  species  of  Calidna  have  slightly  longer 
leg  II  lengths  than  comparable-sized  eyed  ones. 

SELECTIVE  PRESSURES.— In  his  exhaustive  study 
of  heterochrony,  Gould  (1977)  differentiates  two 
distinct  phenomena  included  in  paedomorpho- 
sis:  progenesis  and  neoteny.  Progenesis  evolves 
(through  r  selection)  in  harsh,  unstable,  density- 
independent  situations  that  favor  rapid  matu- 
ration. This  is  achieved  by  shortening  the  life 
cycle  through  premature  maturation,  and  results 
in  adults  juvenilized  in  both  size  and  shape.  Neo- 
teny, on  the  other  hand,  evolves  (through  K  se- 
lection) in  stable,  density-dependent  situations 
that  favor  maintenance.  Here  the  selection  is  not 
for  early  maturation  but  for  adaptive  larval  char- 
acters. Adult  neotenics  are,  therefore,  juvenilized 
in  shape  but  not  size. 

The  small  size  of  paedomorphic  Calidna 
strongly  suggests  progenesis.  As  mentioned  ear- 
lier, all  paedomorphic  species  (except  sequoia) 
reside  exclusively  in  oak  woodland  and  grassland 
biomes,  whereas  non-paedomorphics  occur  pre- 
dominantly in  dense  forests.  The  former  biomes 
are  much  more  xeric  and  experience  severe  dry 
seasons,  which  may  favor  both  the  small  size  and 
shorter  life  cycle  of  progenetics.  The  presence  of 
troglobitic  characters  in  paedomorphic  Calidna 
indicates  a  prolonged  subterranean  existence, 
which  would  clearly  favor  a  reduction  in  size. 
Also,  a  shorter  life  cycle  would  be  an  advantage 
in  harsh  environments  having  short  growing  sea- 
sons. This  was  postulated  by  Shear  (1975)  for 
another  progenetic  harvestman,  Caddo  pepper- 
ella  Shear. 

A  somewhat  weaker  argument  may  be  made 
that  the  paedomorphosis  in  Calidna  is  the  result 
of  neoteny.  First,  it  is  theoretically  possible 
(though  perhaps  less  parsimonious)  that  the  re- 
duction in  size  and  juvenilization  are  indepen- 
dently derived;  the  former  resulting  from  pro- 
portioned dwarfism,  the  latter  from  neoteny. 
Second,  the  presence  of  troglobitic  characters  in 
paedomorphic  Calidna  suggests  cavelike  selec- 
tive pressures.  Studies  cited  by  Culver  (1982) 


UBICK  AND  BRIGGS:  NEW  PHALANGODID  GENUS,  CALIC1NA 


135 


indicated  that  certain  cave  salamanders  become 
neotenic  in  response  to  a  low  food  supply.  If  the 
subterranean  habitat  of  paedomorphic  Calicina 
is  cavelike  in  these  respects,  such  as  constant 
conditions  and  resource  scarcity,  then  K  selec- 
tion and  neoteny  emerge  as  possibilities. 

Neoteny  is  also  a  possibility  for  the  three  larg- 
est species  of  Calicina:  kaweahensis,  palaprae- 
putia,  and  piedra.  In  contrast  to  the  forest-dwell- 
ing habits  of  other  large  species,  these  species 
inhabit  grassland  biomes.  They  exhibit  some  re- 
duction in  the  tarsal  count,  most  pronounced  in 
the  latter  two  (which  also  lack  tarsal  spurs  on  the 
male  palpi).  Large  size  in  combination  with  ju- 
venile structures  suggests  neoteny. 

Two  situations  suggest  a  relationship  between 
competition  and  paedomorphosis.  First,  Cali- 
cina sequoia  is  unique  among  the  paedomorphic 
species.  On  the  basis  of  its  low  tarsal  count  (most 
populations  have  3-3-4-4),  it  is  the  most  pae- 
domorphic species.  However,  it  has  well  devel- 
oped eyes,  lives  in  dense  forests,  and  has  a  year- 
round  activity  period.  Paedomorphosis  here  does 
not  seem  to  be  a  response  to  xeric  environment. 
Interestingly,  this  species  is  fully  sympatric  with 
another  phalangodid,  "  Sitalcina"  cockerelli.  The 
two  species  not  only  occupy  the  same  biomes, 
but  share  identical  habitats  (we  have  six  samples 
containing  both  species).  They  differ  most  strik- 
ingly in  size:  cockerelli  has  a  body  length  two  to 
three  times  that  of  sequoia.  It  seems  conceivable 
that  paedomorphosis  could  have  evolved  in  se- 
quoia as  a  means  of  reducing  competition  for 
similar  prey. 

Second,  competition  may  also  be  involved  in 
the  evolution  of  paedomorphosis  in  other  Cali- 
cina species.  The  Coast  Range  species,  from  po- 
lina  south  to  morroensis,  are  closely  sympatric 
with  species  of  Sitalcina.  However,  Sitalcina 
species  are  non-paedomorphic  forest  dwellers, 
whereas  the  coastal  Calicina  species  are  predom- 
inantly paedomorphics  from  oak  woodland  and 
grassland  biomes  (except  for  some  populations 
of  serpentined).  Since  the  original  biomes  for 
Calicina  were  most  probably  dense  forests,  it  is 
likely  that  the  coastal  Calicina  species  also  re- 
sided in  forests.  Perhaps  Calicina  was  excluded 
from  the  more  favorable  environments  by  Si- 
talcina. 

ACKNOWLEDGMENTS 

For  their  assistance  with  both  field  and  labo- 
ratory work  we  thank  Vincent  F.  Lee,  Kevin  Horn, 


Albert  K.  S.  Jung,  along  with  several  other  for- 
mer Galileo  High  School  students.  Specimens  of 
Calicina  and  related  phalangodids  were  kindly 
loaned  by  Norman  I.  Platnick,  James  C.  Co- 
kendolpher,  and  William  A.  Shear.  Special  thanks 
go  to  Mary  Ann  Tenorio  for  producing  the  scan- 
ning electron  micrographs;  she  and  Leo  Andres 
gave  invaluable  assistance  in  artistic  matters. 
William  A.  Shear  kindly  made  available  numer- 
ous detailed  drawings  of  related  phalangodids, 
and  Lynne  R.  Parenti  provided  an  essential  ref- 
erence and  additional  information  on  biogeog- 
raphy.  Finally,  we  thank  James  C.  Cokendol- 
pher,  Daphne  G.  Fautin,  Willis  J.  Gertsch,  David 
H.  Kavanaugh,  Vincent  F.  Lee,  Norman  I.  Plat- 
nick,  Wojciech  J.  Pulawski,  William  A.  Shear, 
and  others  for  critically  reading  our  manuscript 
and  offering  many  useful  suggestions. 

LITERATURE  CITED 

BRIGGS,  T.  S.  1968.  Phalangids  of  the  laniatorid  genus  Si- 
talcina  (Phalangodidae:  Opiliones).  Proc.  Calif.  Acad.  Sci. 
(ser.  4)36:1-32. 

.  1974.  Phalangodidae  from  caves  in  the  Sierra  Ne- 
vada (California)  with  a  redescription  of  the  type  genus 
(Opiliones:  Phalangodidae).  Occas.  Pap.  Calif.  Acad.  Sci. 
108:1-15. 

.    1986.    Neotenic  morphology  in  Pacific  Coast  opil- 


ionids.  Proc.  Ninth  International  Congress  of  Arachnology, 
Panama,  p.  303.  (Abstr.) 

BRIGGS,  T.  S.  AND  K.  HOM.  1966.  Five  new  species  of  Pha- 
langodidae from  California  (Opiliones).  Pan-Pac.  Entomol. 
42:262-269. 

.  1967.  New  Phalangodidae  from  the  Sierra  Nevada 

Mountains  (Opiliones).  Pan-Pac.  Entomol.  43:48-52. 

BRIGGS,  T.  S.  AND  D.  UBICK.  1981.  Studies  on  cave  har- 
vestmen  of  the  central  Sierra  Nevada  with  descriptions  of 
new  species  ofBanksula.  Proc.  Calif.  Acad.  Sci.  42:3 1 5-322. 

BRIGNOLI,  P.  M.  1968.  Note  su  Sironidae,  Phalangodidae  e 
Trogulidae  italiani,  cavernicoli  ed  endogei  (Opiliones).  Fragm. 
Entomol.  5:259-293. 

CULVER,  D.  C.  1982.  Cave  life.  Evolution  and  ecology.  Har- 
vard University  Press,  Massachusetts.  185  pp. 

GOODNIGHT,  C.  J.  AND  M.  L.  GOODNIGHT.  1942.  New  Pha- 
langodidae (Phalangida)  from  the  United  States.  Amer.  Mus. 
Novit.  1188:1-18. 

.  1967.  Opilionids  from  Texas  caves  (Opiliones,  Pha- 
langodidae). Amer.  Mus.  Novit.  2301:1-8. 

GOULD,  S.  J.  1 977.  Ontogeny  and  phylogeny.  Harvard  Uni- 
versity Press,  Massachusetts.  501  pp. 

HENDRICKSON,  D.  A.  1986.  Congruence  of  bolitoglossine  bio- 
geography  and  phylogeny  with  geologic  history:  paleotrans- 
port  on  displaced  suspect  terranes?  Cladistics  2:1 13-129. 

MARTENS,  J.  1977.  Opiliones  aus  dem  Nepal-Himalaya.  III. 
Oncopodidae,  Phalangodidae,  Assamiidae  (Arachnida). 
Senckenb.  Biol.  (1976)  57:295-340. 

.  1986.  Die  Grossgliederung  der  Opiliones  und  die 

Evolution  der  Ordnung  (Arachnida).  Proc.  Tenth  Interna- 
tional Congress  Arachnol.,  Jaca,  1986,  1:289-310. 

RAMBLA,  M.   1973.  ContribucionalconocimientodelosOpil- 


136 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  4 


iones  de  la  fauna  iberica.  Estudio  de  los  subordenes  Lani- 
atores  y  Palpatores  (pars.).  Summary  of  Ph.D.  Thesis.  Uni- 
versity of  Barcelona.  2 1  pp. 
.    1977.    Un  nuevo  Scotolemon  cavernicola  de  la  isla 


de  Mallorca  (Arachnida,  Opiliones,  Phalangodidae).  Speleon 
23:7-13. 

1980.    Neoteny  in  Opiliones.  Proc.  Eighth  Interna- 


tional Congress  Arachnol.,  Vienna,  1980:489^92. 

ROEWER,  C.-F.  1923.  Die  Weberknechte  der  Erde.  Verlag 
von  Gustav  Fischer,  Jena.  IV,  1116  pp. 

SHEAR,  W.  A.  1975.  The  opilionid  family  Caddidae  in  North 
America,  with  notes  on  species  from  other  regions  (Opil- 
iones, Palpatores,  Caddoidea).  J.  Arachnol.  (1974)  2:65-88. 

SHEAR,  W.  A.  AND  J.  GRUBER.  1983.  The  opilionid  subfamily 
Ortholasmatinae  (Opiliones,  Troguloidea,  Nemastomati- 
dae).  Amer.  Mus.  Novit.  2757:1-65. 


SUZUKI,  S.  1973.  Opiliones  from  the  South-west  Islands,  Ja- 
pan. J.  Sci.  Hiroshima  Univ.,  Ser.  B,  Div.  1  (Zool.)  24:205- 
279. 

.  1985.  A  synopsis  of  the  Opiliones  of  Thailand 

(Arachnida).  I.  Cyphophthalmi  and  Laniatores.  Steenstrupia 
11:69-110. 

WENNER,  A.  M.  AND  D.  L.  JOHNSON.  1980.  Land  vertebrates 
on  the  California  Channel  Islands:  sweepstakes  or  bridges? 
Pp.  497-530  in  The  California  islands:  proceedings  of  a  mul- 
tidisciplinary  symposium.  D.  Power,  ed.  Santa  Barbara  Mu- 
seum of  Natural  History,  Santa  Barbara. 

YANEV,  K.  P.  1980.  Biogeography  and  distribution  of  three 
parapatric  salamander  species  in  coastal  and  borderland  Cal- 
ifornia. Pp.  531-550  in  The  California  islands:  proceedings 
of  a  multidisciplinary  symposium.  D.  Power,  ed.  Santa  Bar- 
bara Museum  of  Natural  History,  Santa  Barbara. 


CALIFORNIA  ACADEMY  OF  SCIENCES 

Golden  Gate  Park 
San  Francisco,  California  94 1 1 8 


PROCEEDINGS 

OF  THE 

CALIFORNIA  ACADEMY  OF  SCIENCES 


Vol.  46,  No.  5,  pp.  137-150,  6  figs. 


March  9,  1989 


FIVE  NEW  BERRY-FRUITED  SPECIES  OF 
TROPICAL  AMERICAN  MELASTOMATACEAE 

By 
Frank  Almeda 

Department  of  Botany,  California  Academy  of  Sciences,  Golden  Gate  Park, 
San  Francisco,  California  94118-9961 


ABSTRACT:  Five  species  of  berry-fruited  Melastomataceae  are  described  as  new:  Blakea  fuchsioides  from 
western  Panama;  Clidemia  hammelii  from  Costa  Rica,  Panama,  and  western  Colombia;  Miconia  calocoma 
and  M.  dissitiflora  from  Costa  Rica;  and  Tococa  croatii  from  eastern  Panama.  Predaceous  mites  of  the  genus 
Ololaelaps  are  recorded  as  residents  in  the  foliar  pocket  domatia  of  C.  hammelii.  This  is  the  first  report  of 
an  association  between  a  mite  and  a  species  of  Melastomataceae.  Discussions,  distributional  notes,  and 
diagnostic  illustrations  are  provided  for  each  of  the  species. 

Received  June  22,  1988.  Accepted  September  14,  1988. 


INTRODUCTION 

Ongoing  field  work  in  little-explored  areas  of 
Central  America  and  the  study  of  numerous  col- 
lections sent  for  identification  continue  to  yield 
many  new  taxa  among  the  berry-fruited  genera 
of  Melastomataceae.  In  this  paper  I  describe  new 
species  in  Blakea,  Clidemia,  Miconia,  and  To- 
coca. The  taxonomic  notes  and  discussions  in- 
cluded here  emphasize  the  diagnostic  characters 
that  make  each  of  the  new  species  unusual  or 
anomalous  additions  to  their  respective  genera. 

Blakea  fuchsioides  Almeda,  sp.  nov. 

Figure  1 

TYPE:  —  PANAMA.  Chiriqui:  La  Fortuna  area,  ca.  7  mi  N 
of  Los  Planes  de  Hornito  in  forest  along  small  draw,  elev.  3,600 
ft  (1,097  m),  26  Aug.  1983,  Hammel  &  Kress  13473  (holotype: 
CAS!;  isotype:  DUKE). 

Frutex  epiphyticus.  Ramuli  sicut  petioli  folia  subtus  inflo- 
rescentia  hypanthiaque  modice  pilis  castaneis  laevibus  (0.5-) 
l-1.5(-2.5)  mm  longis  armati.  Petioli  0.8-2.4  cm  longi,  lam- 
ina 3-8. 9  x  2-4.7cmelliptico-ovatavelovataapiceacuminata 
basi  rotundata  vel  subcordata,  5-nervata,  membranacea  vel 


subcoriacea  et  integra  vel  obscure  dentata.  Flores  6-meri  in 
quoque  nodo  superiore  singuli  vel  bini,  pedicellis  3.2-1 1.5  cm 
longis,  bracteae  elliptico-lanceolatae;  bracteae  exteriores  3-4.5 
x  1  cm  acutae  liberae;  bracteae  interiores  2.2-3.5  x  0.9-1.2 
cm  acutae  liberae.  Hypanthium  (ad  torum)  5-7  mm  longum 
extus  strigillosum;  calycis  tubus  5  mm  longus,  lobis  4-6  mm 
longis.  Petala  19-23  x  14-15  mm  elliptico-ovata  apicerotun- 
dato.  Filamenta  5-7  mm  longa;  antherae  2-3.5  x  2  mm  ob- 
longae  lateraliter  non  cohaerentes  apicaliter  biporosae;  con- 
nectivum  dorsaliter  ca.  0.5  mm  supra  thecarum  basim 
inconspicue  tuberculatum.  Stylus  1.8-3.2  cm;  ovarium  6-lo- 
culare  omnino  inferum  apice  glabro  (cono  et  collo  non  evo- 
luto). 

Pendent  viny  epiphytes  or  hemiepiphytes  ad- 
hering to  the  bark  of  host  trees  by  adventitious 
roots.  Older  cauline  internodes  glabrate,  terete, 
becoming  striate  or  cracked  in  age.  Distal  in- 
dument  of  smooth  rusty  brown  hairs  (0.5-)1- 
1.5(-2.5)  mm  long.  Leaves  of  a  pair  slightly  un- 
equal in  size;  petioles  0.8-2.4  cm  long;  blades 
membranaceous  to  subcoriaceous,  3-8.9  cm  long 
and  2-4.7  cm  wide,  elliptic-ovate  to  ovate,  acu- 
minate at  apex,  broadly  rounded  to  subcordate 
at  base,  margin  typically  entire  but  varying  to 


[137] 


138 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  5 


FIGURE  1 .  Blakea  fuchsioides  Almeda.  A,  habit,  x  ca.  %;  B,  representative  leaf  (lower  surface),  x  ca.  %;  C,  enlargement  of 
lower  leaf  surface  showing  pubescence  and  elevated  median  nerve;  D,  representative  flower  with  petals  and  stamens  removed, 
x%;  E,  outer  floral  bract  (adaxial  surface),  x2;  F,  petal,  x4;  G,  stamens,  lateral  view  (left)  and  ventral  view  (right),  xca.  15; 
H,  seeds,  xca.  15.  (A-F  from  the  holotype;  G  from  Hammel  2261.) 


remotely  dentate,  5 -nerved  below,  the  outermost 
pair  of  subparallel  primaries  often  inconspicuous 
and  concealed  by  the  revolute  margins  on  drying, 
the  transverse  secondaries  essentially  parallel  on 
the  upper  surface  and  diverging  from  the  ele- 


vated primaries  at  nearly  right  angles,  moder- 
ately to  densely  strigillose  to  velutinous  above 
with  widely  to  antrorsely  spreading  smooth  hairs, 
(0. 5-)  1-1. 5  mm  long,  but  often  varying  to  gla- 
brate  at  maturity  with  the  pubescence  persisting 


ALMEDA:  FIVE  NEW  MELASTOMATACEAE 


139 


to  varying  degrees  in  the  furrows  created  by  the 
impressed  primary  nerves,  moderately  to  dense- 
ly hirsute  below  with  smooth  rusty  brown  hairs 
1-2.5  mm  long.  Flowers  6-merous,  pendent,  sol- 
itary, or  rarely  geminate  in  the  axils  of  distal 
branches;  pedicels  3.2-11.5  cm  long,  densely 
covered  with  minutely  roughened  brown  spread- 
ing hairs  1  mm  long.  Floral  bracts  red,  foliaceous, 
sessile,  and  entire,  moderately  to  densely  covered 
with  roughened  spreading  hairs  0.5-1  mm  long 
on  both  surfaces;  outer  bracts  3-5 -nerved,  free 
to  the  base,  3-4.5  x  1  cm,  lanceolate,  apex  acute, 
base  truncate  to  rounded;  inner  bracts  7-14- 
nerved,  free  to  the  base,  2.2-3.5  x  0.9-1.2  cm, 
lanceolate,  apex  acute,  base  truncate  to  rounded. 
Hypanthia  (at  anthesis)  campanulate,  5-7  mm 
long  to  the  torus  (vascular  ring),  moderately  to 
densely  strigillose  with  obscurely  roughened  hairs 
0.5  mm  long.  Calyx  tube  ca.  5  mm  long,  flaring 
and  flangelike;  calyx  lobes  copiously  pubescent 
with  minutely  roughened  brown  hairs  0.25-0.5 
mm  long,  deltoid  basally  between  sinuses  but 
abruptly  tapered  into  linear-oblong,  mostly  en- 
tire segments  4-6  mm  long  and  1  mm  wide.  Pet- 
als glabrous,  pink  (fide  McPherson  7733)  or 
magenta  (fide  Hammel  &  Kress  13473),  elliptic- 
ovate,  19-23  mm  long,  14-15  mm  wide,  con- 
nivent  and  somewhat  concave  adaxially  with 
slightly  involute  margins  when  expanded,  apex 
rounded.  Stamens  12,  isomorphic,  free  and 
forming  a  pendent  circle  around  the  style;  fila- 
ments complanate,  5-7  mm  long;  anthers  linear- 
oblong  to  ovoid-oblong,  truncate  at  apex,  2-3.5 
mm  long,  2  mm  wide,  each  anther  tipped  with 
two  confluent  pores;  connective  slightly  thick- 
ened dorsally  and  dilated  basally  about  0.5  mm 
above  base  of  anther  thecae  into  a  bluntly  round- 
ed deflexed  caudiform  appendage  about  0.5  mm 
long.  Ovary  inferior,  6-celled,  glabrous  at  apex 
but  not  expanded  into  a  cone  or  collar.  Style 
straight,  glabrous,  1.8-3.2  cm  long,  typically 
exserted  beyond  the  petals  and  sometimes  ex- 
ceeding the  subtending  pair  of  floral  bracts;  stig- 
ma truncate  to  slightly  rounded.  Berry  globose, 
about  1  cm  long  and  1  cm  in  diameter.  Seeds 
numerous,  clavate  to  narrowly  pyriform  or 
bluntly  deltoid,  1(-1.5)  mm  long,  beige  with  a 
smooth  glossy  testa  and  a  prominent  lateral  raphe. 

ADDITIONAL  SPECIMENS  EXAMINED.  — PANAMA.  Bocas  del 
Toro:  along  trail  on  divide  separating  Chiriqui  and  Bocas  del 
Toro,  ca.  08°45'N,  82°15'W,  22  Oct.  1985,  McPherson  7199 
(CAS);  Fortuna  Dam  region  along  continental  divide  trail, 


08°45'04"N,  82°15'W,  23  Dec.  1986,  McPherson  &  Aranda 
10062  (CAS).  Bocas  del  Toro/Chiriqui  Border:  above  Fortuna 
Dam,  ca.  08°45'N,  82°15'W,  3  Dec.  1985,  McPherson  7733 
(CAS).  Chiriqui:  road  to  Fortuna  Dam  site  N  of  Gualaca,  22.7 
mi  [36.8  km]  beyond  bridge  over  Rio  Esti,  22  Nov.  1979, 
Antonio  2770  (CAS);  La  Fortuna  hydroelectric  project,  on  ridge 
behind  camp,  22  Mar.  1978,  Hammel  2261  (MO);  trail  west 
from  Fortuna  Dam  camp  to  La  Fortuna,  08°43'N,  82°14'W, 
23  Mar.  1985,  Hampshire  &  Whitefoord  903  (BM);  2  km  S  of 
Fortuna  Lake,  trail  east,  08°43'N,  82°14'W,  24  Mar.  1985, 
Hampshire  &  Whitefoord  912  (BM);  campamento  Bijao  en 
Fortuna,  20  Mar.  1976,  Mendoza  et  al.  249  (PMA,  US);  entre 
alto  de  Guayabo  y  la  linea  divisoria  continental,  25  Sep.  1976, 
Correa  et  al.  2837  (PMA,  US). 


DISTRIBUTION.— Known  only  from  the  wet 
cloud  forests  of  the  Fortuna  Dam  region  in  west- 
ern Panama  along  the  Chiriqui/Bocas  del  Toro 
border  at  1,097-1,400  m. 

PHENOLOGY.— Flowering  specimens  have  been 
collected  from  August  through  December;  the 
only  available  fruiting  collection  was  made  in 
March. 

Labels  on  collections  of  this  species  describe 
it  as  a  climbing  or  pendent  epiphyte.  I  have  not 
studied  this  species  in  the  field  but  its  climbing 
habit,  adventitious  roots,  and  pendent  inflores- 
cences lead  me  to  suspect  that  it  will  ultimately 
be  shown  to  be  a  secondary  hemiepiphyte  like 
Miconia  arboricola  Almeda  (Almeda  1 984).  Sec- 
ondary hemiepiphytes  are  vine-like  plants  that 
germinate  terrestrially,  ascend  nearby  trees  by 
adventitious  roots,  and  later  become  epiphytic 
by  losing  root  contact  with  the  ground  (Putz  and 
Holbrook  1986). 

In  addition  to  its  copious  cover  of  brown 
smooth  hairs  on  distal  branchlets,  petioles,  lower 
leaf  surfaces,  and  floral  bracts,  B.  fuchsioides  is 
distinguished  by  an  extraordinary  combination 
of  floral  characters.  As  emphasized  by  the  spe- 
cific epithet,  the  flowers  of  this  species  are  rem- 
iniscent of  Fuchsia  (Onagraceae)  in  posture  and 
coloration.  In  B.  fuchsioides  ihe  flowers  are  borne 
on  flexible  pendent  pedicels  exposing  floral  or- 
gans in  a  way  that  favors  hovering  pollinators. 
The  connivent  petals  form  a  broad  tube-like  con- 
formation, and  the  floral  parts  are  vividly  col- 
ored. The  hypanthia,  calyx  lobes,  and  decussate 
floral  bracts  that  closely  subtend  the  flowers  are 
bright  red,  whereas  the  petals  are  reportedly  pink 
or  magenta. 

If  future  field  observations  confirm  that  flow- 
ers of  this  species  produce  nectar,  we  can  predict 
that  this  species  is  probably  bird  pollinated  be- 


140 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  5 


cause  it  exhibits  so  many  of  the  features  asso- 
ciated with  ornithophily  (Percival  1965;  Faegri 
and  van  der  Fiji  1971;  Proctor  and  Yeo  1973). 

This  species  and  other  undescribed  Central 
American  species  currently  under  study  (Alme- 
da,  in  prep.)  exhibit  anther  characters  that  weak- 
en the  tenuous  distinction  between  Blakea  and 
the  closely  related  segregate  genus  Topobea.  The 
traditional  differences  between  these  genera  are 
based  on  anther  shape  and  the  relative  size  and 
proximity  of  the  apical  pores.  In  Blakea,  the  an- 
thers are  oval,  oblong,  or  elliptic  and  blunt  or 
broadly  rounded  at  the  summit  with  two  typi- 
cally well-separated  (often  minute)  apical  pores. 
In  Topobea,  the  anthers  are  linear-oblong  to  lan- 
ceolate or  subulate  with  the  dorsally  inclined  api- 
cal pores  approximate  and  often  confluent.  An- 
ther size  and  shape  in  B.  fuchsioides  agree  with 
the  syndrome  typically  found  in  Blakea.  The 
broad  confluent  pores,  however,  dictate  assign- 
ment of  this  species  to  Topobea.  The  truncate 
position  of  the  anther  pores  is  characteristic  of 
neither  genus  and  underscores  the  difficulty  of 
assigning  unusual  new  species  to  either  of  the 
genera  as  they  are  currently  denned.  Until  a 
counterargument  can  be  advanced  for  the  con- 
tinual recognition  of  Topobea  based  on  some- 
thing other  than  arbitrary  characters,  it  seems 
appropriate  to  follow  Baillon  (1879)  and  Mac- 
bride  ( 1 94 1 )  in  submerging  Topobea  in  the  older 
and  much  larger  genus,  Blakea. 

Because  B.  fuchsioides  is  one  of  the  most  un- 
usual species  yet  to  be  described  in  Blakea,  it  is 
not  surprising  that  its  relationships  are  obscure. 
A  peculiar  feature  which  has  not  been  reported 
in  other  species  of  Blakea  or  Topobea  is  the  del- 
icate, veil-like  membrane  that  covers  the  broad 
anther  pores  prior  to  anthesis.  These  veil-like 
coverings  may  well  function  to  preclude  exces- 
sive intrafloral  pollen  dispersion  prior  to  petal 
expansion  on  the  pendent  flowers. 

Clidemia  hammelii  Almeda,  sp.  nov. 

Figure  2 

TYPE. -COSTA  RICA.  Heredia:  Finca  La  Selva,  OTS  [Or- 
ganization for  Tropical  Studies]  Field  Station  on  Rio  Puerto 
Viejo  just  E  of  its  junction  with  Rio  Sarapiqui,  elev.  about  100 
m.  Slopes  along  Q.  El  Salto,  2,900  m  S,  2  Sep.  1980,  Hammel 
9682  (holotype:  CAS!;  isotype:  DUKE). 

Frutex  vel  arbuscula  1-5  m.  Ramuli  teretes  primum  sicut 
folia  novella  inflorescentia  sparse  vel  modice  setosi  pilis  2-3.5 
mm  longis  et  modice  vel  sparse  stellulato-furfuracei  demum 
glabrati.  Petioli  1-4  cm  longi;  lamina  1 1.5-29  x  6.1-14.5  cm 


elliptica  apice  acuminata  basi  acuta  vel  cuneata,  5-7-nervata 
vel  5-7-plinervata,  nervi  in  axillis  acarodomatiis  instructi.  In- 
florescentia primum  terminalis  demum  lateralis  pauciramosa 
plus  minusve  deflexa;  flores  5-meri,  pedicellis  (ad  anthesim) 
1-2.5  mm  longis,  bracteolis  1.5-2.5  mm  longis  subulatis  per- 
sistentibus.  Hypanthium  (ad  torum)  1 .5-2.5  mm  longum;  calyx 
primum  in  cono  apiculato  clausus  demum  in  lobos  irregulares 
persistentes  ruptus,  dentibus  exterioribus  0.5-1  mm  eminen- 
tibus.  Petala  3-3.5  x  1-1.5  mm  oblonga-rotundato  glabra. 
Stamina  isomorphica  vel  paulo  anisomorphica  glabra;  fila- 
menta  1.5-2  mm  longa;  antherarum  thecae  1-1.25  x  0.5  vel 
1.25-1.75  x  0.5  mm  oblongae,  poro  paulo  dorsaliter  inclinato; 
connectivum  nee  prolongatum  nee  appendiculatum.  Stylus  5- 
6  mm  glaber  in  ovarii  collo  0.5  mm  immersus;  ovarium 
5-loculare  omnino  inferum  apice  modice  glanduloso-setuloso. 

Shrubs  or  small  trees  1-5  m  tall.  Internodes 
terete,  glabrate  at  maturity;  distal  branchlets, 
vegetative  buds,  and  inflorescences  beset  with  a 
sparse  to  moderate  covering  of  smooth,  spread- 
ing hairs  (2-3.5  mm)  that  is  typically  intermixed 
with  inconspicuous,  early  deciduous,  asperous- 
headed  hairs  and  underlain  by  a  moderate  to 
dense  understory  of  stellulate-furfuraceous  or 
short  asperous-headed  hairs.  Leaves  of  a  pair 
typically  somewhat  unequal  in  size;  petioles  1- 
4  cm  long;  blades  membranaceous,  1 1.5-29  cm 
long  and  6.1-14.5  cm  wide,  elliptic,  apex  long- 
acuminate,  base  acute  to  cuneate,  margin  entire 
to  inconspicuously  crenulate,  5-7-nerved  or  5- 
7-plinerved  abaxially  with  pocket  domatia  typ- 
ically formed  in  the  angle  between  the  median 
nerve  and  each  of  the  two  proximal  lateral  nerves, 
moderately  strigose  to  subhirsute  above  with 
smooth  hairs  mostly  1-2  mm  long,  moderately 
hirsute  below  with  a  mixture  of  smooth  hairs  (1- 
2.5  mm  long)  and  minute  glandular  hairs  essen- 
tially restricted  to  the  elevated  primary  and  higher 
order  venation.  Inflorescence  a  pseudolateral 
modified  dichasium,  3-6  cm  long,  divaricately 
branched  from  the  base;  bracts  of  the  rachis  nodes 
paired,  narrowly  lanceolate  to  subulate,  3-3.5 
mm  long,  1  mm  wide,  sparingly  stellulate-fur- 
furaceous to  glabrate;  bracteoles  sessile,  persis- 
tent, paired  but  sometimes  fused  into  an  incon- 
spicuous nodal  collar,  narrowly  lanceolate  to 
subulate,  1.5-2.5  mm  long,  up  to  0.5  mm  wide, 
essentially  glabrous  and  entire  but  terminating 
in  a  solitary  hair.  Pedicels  1-2.5  mm  long,  spar- 
ingly to  moderately  stellulate-furfuraceous  and 
basally  encircled  by  a  deciduous,  tufted  ring  of 
barbellate  or  asperous-headed  hairs.  Hypanthia 
(at  anthesis)  campanulate,  1.5-2.5  mm  long  to 
the  torus  (vascular  ring),  moderately  to  sparsely 
covered  with  spreading  smooth  hairs  0.5-2  mm 


ALMEDA:  FIVE  NEW  MELASTOMATACEAE 


141 


FIGURE  2.  Clidemia  hammelii  Almeda.  A,  habit,  x%;  B,  representative  flower  with  stamens  removed  (left)  and  floral  bud 
showing  rupturing  calyx  (right),  x  9;  C,  petal,  x  10;  D,  larger  (antepetalous)  stamens,  dorsal  view  (left)  and  3/4  lateral  view  (right), 
x  ca.  1 0;  E,  smaller  (antesepalous)  stamens,  %  lateral  view  (left)  and  dorsal  view  (right),  x  ca.  1 2;  F,  mature  berry,  x  3;  G,  seeds, 
x  16.  (A  from  Folsom  5864;  B-E  from  de  Nevers  3962;  F,  G  from  Hammel  12668.) 


long  and  a  sparse  understory  of  sessile  stellulate- 
furfuraceous  hairs.  Calyx  closed  in  bud  and 
crowned  by  an  apiculum  0.5  mm  long  but  rup- 
turing irregularly  at  anthesis  into  3-5  hyaline, 
persistent,  rounded  lobes  0.5  mm  long;  calyx  teeth 
5,  subulate,  0.5-1  mm  long.  Petals  5,  glabrous 
and  reflexed,  white  or  reportedly  pale  pink,  ob- 
long, rounded  and  often  bluntly  erose  apically, 


otherwise  entire,  3-3.5  mm  long,  1-1.5  mm  wide. 
Stamens  10,  isomorphic  or  alternately  subiso- 
morphic  with  larger  stamens  inserted  on  gla- 
brous torus  opposite  petals  and  smaller  stamens 
inserted  opposite  calyx  lobes;  filaments  glabrous, 
complanate  and  constricted  distally,  1.5-2  mm 
long;  anthers  1-1.5  (-1.75)  mm  long  and  0.5  mm 
wide,  yellow,  linear-oblong,  deeply  channeled 


142 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  5 


ventrally  between  thecae,  truncate  to  slightly 
emarginate  distally  with  a  circular,  somewhat 
dorsally  inclined  pore;  connective  simple,  slight- 
ly thickened  dorsally  but  not  dilated  or  prolonged 
below  thecae.  Ovary  inferior,  5 -celled,  fluted  and 
distended  apically  into  a  glandular-puberulent, 
lobulate,  stylar  collar  mostly  0.5  mm  high  that 
becomes  depressed  and  bowl-like  on  maturing 
berries.  Style  straight,  glabrous,  5-6  mm  long; 
stigma  truncate.  Berry  reportedly  dark  purple  at 
maturity,  globose,  6-7  mm  long  and  5-6  mm  in 
diameter.  Seeds  numerous,  galeiform  to  deltoid, 
0.5  mm  long,  white  or  beige,  irregularly  angulate 
with  a  densely  papillate  testa  and  a  lateral  flat- 
tened or  somewhat  convex  raphe. 

ADDITIONAL  SPECIMENS  EXAMINED.— COSTA  RICA.  Car- 
tago:  Valle  Escondido,  30  Mar.  1966,  Schnell  627  (US).  He- 
redia:  Finca  La  Selva,  OTS  Field  Station  on  Rio  Puerto  Viejo, 
just  E  of  its  junction  with  Rio  Sarapiqui,  21  Jul.  1981,  Hammel 
11026  (DUKE);  forest  N  of  road  to  Puerto  Viejo  at  Chilamate, 
22  May  1 982,  Hammel  12423  (CAS,  CR);  hills  near  Chilamate 
along  road  between  Puerto  Viejo  and  La  Virgen,  30  May  1 982, 
Hammel  12668  (CAS);  Chilamate  de  Sarapiqui,  S  end  of  Cer- 
ros  Sardinal,  N  side  of  Rio  Sarapiqui,  10°27.5'N,  84°4'W,  3 
Jul.  1985,  Hammel  &  Grayum  14103  (CAS);  Finca  El  Bejuco, 
S  base  of  Cerros  Sardinal,  Chilamate  de  Sarapiqui,  10°27'N, 
84°04'W,  2  Jun.  1985,  Grayum  &  Jacobs  5355  (CAS);  between 
Rio  Peje  and  Rio  Sardinalito,  Atlantic  slope  of  Volcan  Barva, 
10°18.5'N,  84°04'W,  8  Apr.  1986,  Grayum  6910  (CAS-2 
sheets);  Tirimbina,  9  Jun.  1971,  Proctor  32244  (MO).  PAN- 
AMA. Bocas  del  Toro:  between  Fortuna  and  Chiriqui  Grande, 
8.5  mi  [13.7  km]  N  of  bridge  over  the  Fortuna  Lake,  08°46'N, 
82°14'W,  10  Mar.  1985,  Croat  60151  (CAS);  Fortuna  to  Chi- 
riqui, ca.  3  km  N  of  continental  divide,  08°48'N,  82°1 1'W,  12 
Feb.  1986,  Hammel  &  McPherson  14466  (CAS);  Fortuna  Dam 
region,  near  Chiriqui  Grande,  08°45'N,82°1 5' W,  18  Jan.  1986, 
McPherson  807 9  (CAS);  8.5  road-miles  [13.7  km]  from  bridge 
near  Fortuna  Dam  on  road  towards  Chiriqui  Grande,  08°50'N, 
82°15'W,  10  Mar.  1985,  McPherson  6760  (CAS);  road  to  Chi- 
riqui Grande,  6  km  N  of  continental  divide,  08°50'N,  82°07' W, 
2  Aug.  1984,  Todzia  et  al.  2558  (CAS);  Fortuna  Dam  area  to 
Chiriqui  Grande,  2  Aug.  1984,  D'Arcy  et  al.  15996  (CAS); 
above  Chiriqui  Grande  ca.  2  mi  E  on  a  side  road  ca.  10  road- 
miles  below  continental  divide,  08°55'N,  82°10'W,  11  Mar. 
1988,  Almeda  et  al.  6082  (CAS,  MO,  PMA,  US).  Code:  2  mi 
N  of  Cerro  Pilon,  16  Mar.  1973,  Liesner  726  (CAS,  MO); 
between  La  Junta  and  Limon,  5  hour  walk  N  of  Alto  Calvario, 
1 1  Oct.  1977,  Folsom  5864  (CAS);  Caribbean  side  of  divide 
at  El  Cope,  08°45'N,  80°35'W,  4  Feb.  1983,  Hamilton  &  Da- 
vidse  2776  (CAS);  road  from  La  Pintada  to  Coclesito,  08°45'N, 
80°30'W,7  Feb.  1983,  Hamilton  &Davidse  28 15  (CAS).  Colon: 
9  mi  [14.6  km]  S  of  Portobello,  17  Jul.  1970,  Croat  11372  (F, 
US);  Rio  Iguanita  and  inland  to  2  km,  7  Feb.  1981,  D'Arcy 
14615  (CAS).  Darien:  vicinity  of  gold  mining  camp  of  T.  Kit- 
tredge  on  headwaters  of  Rio  Tuquesa  ca.  2  air  km  from  con- 
tinental divide,  26  Aug.  1974,  Croat  27210  (F);  forested  ridges 
and  valleys  at  Mamey,  6  Mar.  1982,  Whitefoord  &  Eddy  374 
(BM,  MO).  Panama:  13  km  N  of  El  Llano  on  road  to  Carti, 
13  Feb.  1973,  Busey  374  (US);  El  Llano-Carti  road,  14.8  km 
N  of  Panamanian  Hwy.,  27  Jan.  1977,  Folsom  et  al.  1479 


(MO).  San  Bias:  Nusagandi,  trail  from  camp  NW  to  a  quebrada, 
09°19'N,  78°15'W,  31  Jul.  1984,  de  Nevers  &  de  Leon  3595 
(CAS);  Nusagandi,  El  Llano-Carti  road,  09°19'N,  78°15'W,  10 
Aug.  1984,  de  Nevers  &  Gonzalez  3655  (CAS);  Llano-Carti 
road,  19  km  from  Interamerican  Hwy.,  09°19'N,  78°55'W,  4 
Sep.  1 984,  de  Nevers  &  Herman  3831  (CAS);  Llano-Carti  road, 
20  km  from  Interamerican  Hwy.,  09°19'N,  78°55'W,  30  Sep. 
1 984,  de  Nevers  et  al.  3962  (CAS).  COLOMBIA.  Choco:  forest 
NW  of  Alto  Curiche,  20  May  1967,  Duke  &  Idrobo  11240 
(US). 

DISTRIBUTION.— Currently  known  from  river 
banks,  shaded  sites,  and  light  gaps  in  rain  forests 
from  the  Puerto  Viejo  region  of  northeastern 
Costa  Rica  south  through  Panama  to  an  area 
northwest  of  Alto  Curiche  in  western  Colombia, 
from  sea  level  to  900  m. 

PHENOLOGY.— Flowering  and  fruiting  occur 
sporadically  throughout  the  year. 

Although  variable  in  foliar  size  and  the  degree 
to  which  the  elevated  primaries  diverge  from  the 
median  nerve  on  lower  foliar  surfaces,  C.  ham- 
melii  is  readily  denned  by  the  following  set  of 
characters:  a  divaricately  branched  pseudolateral 
inflorescence;  an  irregularly  rupturing  apiculate 
calyx;  a  fluted,  glandular-puberulent  stylar  col- 
lar; and  seeds  that  are  galeiform  to  angularly  del- 
toid in  outline  with  a  densely  papillate  testa.  There 
also  appears  to  be  some  intrafloral  variation  in 
staminal  size  among  the  three  flowering  collec- 
tions available  for  study.  A  collection  from  Pan- 
ama (de  Nevers  et  al.  3962)  has  stamens  that  are 
alternately  somewhat  unequal  in  size  (see  Fig.  2), 
whereas  Hammel  9682  from  Costa  Rica  and 
Todzia  et  al.  2558  from  Panama  have  stamens 
that  are  clearly  isomorphic.  This  slight  variation 
may  prove  to  be  taxonomically  inconsequential 
but  warrants  further  observation  as  additional 
flowering  collections  are  made. 

Although  C.  hammelii  appears  to  have  no  close 
allies  among  described  congeners,  it  bears  a  su- 
perficial vegetative  resemblance  to  C.  reitziana 
Cogn.  &  Gleason  and  C.  costaricensis  Cogn.  & 
Gleason.  Both  differ  from  C.  hammelii  in  having 
an  adaxially  pubescent  torus,  regularly  devel- 
oped (nonrupturing)  calyx  lobes,  and  smooth 
seeds. 

A  distinctive,  but  not  always  conspicuous,  fea- 
ture of  C.  hammelii  is  the  occurrence  of  small 
funnel-shaped  pockets  or  chambers  in  the  angles 
between  the  median  nerve  and  the  base  of  each 
proximal  lateral  nerve  (see  Fig.  3).  The  function 
of  these  structures  in  any  kind  of  symbiotic  re- 
lationship is  unclear.  Of  the  numerous  collec- 
tions cited  above,  label  information  on  only  one 


ALMEDA:  FIVE  NEW  MELASTOMATACEAE 


143 


B 


FIGURE  3.  Clidemia  hammelii  Almeda.  A,  representative  leaf  (abaxial  surface),  xca.  '/»;  B,  enlargement  of  lower  leaf  surface 
showing  pocket  domatia,  xca.  5;  C,  the  mite,  Ololaelaps  sp.,  dorsal  view  (above)  and  ventral  view  (below),  xca.  20.  (A,  B  from 
Almeda  et  al.  6082.) 


from  Panama  (de  Never s  &  Herman  3831)  makes 
note  of  "ants  in  the  leaves."  During  recent  field 
work  I  examined  several  individuals  of  another 
Panamanian  population  (Almeda  et  al.  6082)  and 
found  mites  of  the  genus  Ololaelaps  (Gamasida, 
family  Laelapidae)  in  the  domatia  of  at  least  three 
individuals  (see  Fig.  3).  No  other  insects  or  ar- 
thropods, however,  were  encountered  in  the  33 
domatia  examined.  According  to  G.  W.  Krantz 
(in  litt.,  16  June  1988)  a  domatium  is  an  odd 
habitat  for  Ololaelaps  because  most  are  de- 
scribed from  moss  and  litter  substrates,  or  from 
the  nests  of  small  mammals.  Because  mites  of 
this  genus  are  thought  to  be  predaceous,  Krantz 
suggests  that  the  mites  may  feed  on  resident 
nematodes,  or  they  may  simply  use  the  domatia 
as  shelters. 

In  his  review  of  plant  domatia,  Jacobs  (1966) 
concludes  that  there  is  no  evidence  to  demon- 
strate a  biological  relationship  between  domatia- 
bearing  plants  and  mites.  Benson  (1985),  how- 
ever, notes  that  scavenging  mites  are  fairly 


common  opportunistic  invaders  of  plant  cavi- 
ties. He  suggests  that  pocket  domatia,  often  re- 
ferred to  as  acarodomatia,  probably  function  as 
feeding  sites  for  sucking  homopterans  (i.e.,  the 
coccids  or  scale  insects,  and  the  membracids  or 
treehoppers).  This  ant-coccid  linkage  is  intrigu- 
ing when  one  considers  that  homopterans  can 
attract  ants  with  their  honeydew  secretions  and 
gain  protection  from  other  predaceous  insects. 
Charles  Turner  (pers.  comm.,  26  May  1988),  who 
is  currently  working  on  mite  ecology,  proposes 
a  facultative  mutualistic  interaction  to  explain 
the  plant-mite  association.  He  suggests  that  the 
pocket  domatia  may  serve  as  egg  nurseries  and 
help  to  shelter  the  mites  from  predators.  The 
mites,  in  turn,  benefit  the  plant  by  feeding  on 
fungal  spores  and  the  eggs  of  predaceous  insects. 
The  plant-mite  relationship  thus  remains  un- 
solved, requiring  more  sophisticated  field-ori- 
ented study  than  it  has  received  so  far. 

This  species  is  named  for  Barry  E.  Hammel, 
a  student  of  neotropical  Clusiaceae  whose  col- 


144 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  5 


lecting  efforts  in  Costa  Rica  and  Panama  have 
added  many  new  and  interesting  plants  to  the 
flora  of  Mesoamerica. 

Miconia  calocoma  Almeda,  sp.  nov. 

Figure  4 

TYPE. -COSTA  RICA.  Heredia:  Finca  La  Selva,  OTS  Field 
Station  on  Rio  Puerto  Viejo  just  E  of  its  junction  with  Rio 
Sarapiqui,  southeast  corner,  elev.  100m,  17  Apr.  \9&l,Folsom 
9776  (holotype:  CAS!;  isotype:  DUKE). 

Frutex  vel  arbuscula  3-5  m.  Ramuli  teretes  sicut  petioli 
foliorum  subtus  venae  primariae  inflorescentia  hypanthiaque 
modice  pilis  stellatis  induti.  Petioli  4-10  mm  longi;  lamina 
(4.2-)8.5-19.5  x  (2.3-)5-8.5  cm  elliptica  vel  ovato-elliptica 
apice  acuminata  basi  rotundato-obtusa,  5(-7)-plinervata, 
membranacea  et  undulata  vel  undulato-dentata.  Inflorescentia 
3-7  cm  longa  multiflora;  flores  4-meri,  pedicellis  (ad  anthesim) 
0.5  mm  longis,  bracteolis  2-3  mm  longis  anguste  oblongis 
persistentibus.  Hypanthium  (ad  torum)  2  mm  longum;  calyx 
primum  in  cono  apiculato  clausus  demum  in  lobos  regulares 
persistentes  ruptus,  dentibus  exterioribus  1-1.5  mm  eminen- 
tibus.  Petala  3-4  x  1.5-2.5  mm  obovato-oblongaglabra.  Stam- 
ina isomorphica  glabra;  filamenta  1.5-2  mm  longa;  anthera- 
rum  thecae  1-1.5  x  0.5  mm  angustae  oblongae,  poro  paulo 
ventraliter  inclinato;  connectivum  nee  prolongatum  nee  ap- 
pendiculatum.  Stylus  5-10  mm  glaber;  ovarium  4-loculare  et 
'A  inferum  glaber. 

Shrubs  or  small  trees  3-5  m  tall.  Older  cauline 
internodes  terete  and  essentially  glabrous.  Distal 
branchlets,  petioles,  and  inflorescences  densely 
covered  with  a  mixture  of  rusty  brown  sessile- 
stellate  and  stipitate-stellate  hairs.  Leaves  of  a 
pair  typically  unequal  in  size;  petioles  4-10  mm 
long;  blades  membranaceous,  (4.2-)8.5-19.5  cm 
long  and  (2.3-)5-8.5  cm  wide,  elliptic  to  elliptic- 
ovate,  apex  acuminate,  base  typically  obtuse  to 
broadly  rounded  but  sometimes  varying  to 
slightly  oblique  at  base  and  then  somewhat  de- 
current  on  petiole,  margin  undulate  to  undulate- 
dentate;  5(-7)-plinerved  with  the  inner  pairs  of 
subparallel  primary  nerves  elevated  and  diverg- 
ing from  median  nerve  in  opposite,  subopposite, 
or  irregularly  alternate  fashion  at  successive  points 
above  the  blade  base,  sparingly  stellate  pubescent 
to  glabrous  above  at  maturity,  copiously  stellate 
pubescent  on  elevated  primary  nerves  below  with 
a  sparser  hair  covering  on  prominulous  network 
of  transverse  secondary  and  higher  order  veins. 
Inflorescence  terminal  but  sometimes  appearing 
pseudolateral  because  of  elongation  of  axillary 
shoots,  paniculiform  with  ultimate  branchlets 
terminating  in  multiflowered,  congested  glomer- 
ules;  bracts  of  rachis  nodes  paired,  linear-oblong, 
2-5  mm  long,  0.5-1  mm  wide,  glabrous  above 
and  copiously  stellate  below;  bracteoles  persis- 


tent, 2-3  per  pedicel,  sessile,  linear-oblong,  at- 
tenuate apically,  0.5-1.5  mm  long,  0.25-0.5  mm 
wide,  margin  entire  with  a  solitary  apical  hair, 
glabrous  above  and  sparingly  to  moderately  stel- 
late below.  Pedicels  copiously  stellate  pubescent, 
0.5  mm  long  but  inconspicuous  and  concealed 
by  congested  glomerules.  Hypanthia  (at  anthesis) 
campanulate,  2  mm  long  to  the  torus  (vascular 
ring),  densely  stellate  pubescent  throughout.  Ca- 
lyx closed  in  bud  and  crowned  by  an  apiculum 
0.25  mm  long  but  rupturing  at  anthesis  into  2- 
4  persistent  deltoid  to  semicircular  hyaline  lobes 
mostly  1  mm  long  and  1-1.5  mm  wide  basally; 
exterior  calyx  teeth  4,  linear-oblong,  1-1.5  x  0.5 
mm,  copiously  stellate  and  widely  spreading  to 
recurved  on  fruiting  hypanthia.  Petals  4,  gla- 
brous, white,  obovate  to  oblong-obovate,  widely 
spreading  to  reflexed,  rounded  apically,  3-4  mm 
long  and  1.5-2.5  mm  wide  distally.  Stamens  8, 
isomorphic;  filaments  glabrous,  complanate, 
constricted  and  incurved  distally,  1.5-2  mm  long; 
anthers  1-1.5  mm  long  and  0.5  mm  wide,  yellow, 
linear-oblong,  laterally  compressed  and  deeply 
channeled  ventrally  between  thecae,  truncate  to 
slightly  emarginate  distally  with  a  somewhat 
ventrally  inclined  terminal  pore;  connective 
thickened  dorsally  but  not  dilated  or  prolonged 
below  the  filament  insertion.  Ovary  ca.  %  infe- 
rior, 4-celled,  and  glabrous.  Style  erect,  often 
somewhat  curved  distally,  glabrous,  5-10  mm 
long  and  conspicuously  overtopping  the  sur- 
rounding stamens;  stigma  truncate  to  capitellate. 
Berry  reportedly  blue-black  to  purple  at  matu- 
rity, globose,  6-8  mm  long  and  6-8  mm  in  di- 
ameter. Seeds  numerous,  obovoid  to  pyriform, 
1  mm  long,  white  or  beige,  somewhat  angulate 
with  a  densely  tuberculate  testa  and  a  lateral  flat- 
tened raphe  on  the  opposing  face. 

ADDITIONAL  SPECIMENS  EXAMINED.— COSTA  RICA.  Here- 
dia: S  of  Puerto  Viejo,  2  km  S  of  Magsasay  Penal  Colony,  W 
of  road,  5  Feb.  1983,  Garwood  et  al.  1125  (BM,  MO);  Finca 
La  Selva,  OTS  Field  Station  on  Rio  Puerto  Viejo,  just  E  of  its 
junction  with  Rio  Sarapiqui,  3,000  m  line,  South  Boundary 
along  Central  Trail,  2  May  1981,  Folsom  9990  (CAS);  Finca 
La  Selva,  OTS  Field  Station,  South  Boundary,  1 ,900  m  E,  6 
Aug.  1 980,  Hammel  9452  (DUKE);  Finca  La  Selva,  OTS  Field 
Station,  South  Boundary,  1,600  m  E,  27  Sep.  1980,  Hammel 
9872  (DUKE);  Finca  La  Selva,  OTS  Field  Station,  South 
Boundary  on  slope  just  W  of  Q.  Esquina,  1 5  Jul.  1 982,  Hammel 
&  Trainer  13181  (CAS).  Limon:  Cerro  Coronel,  E  of  Laguna 
Danto,  10°41'N,  83°38'W,  16-23  Jan.  1986,  Stevens  23730 
(MO). 

DISTRIBUTION.— A  little-collected  rain  forest 
species  currently  known  from  northeastern  Costa 


ALMEDA:  FIVE  NEW  MELASTOMATACEAE 


145 


FIGURE  4.  Miconia  calocoma  Almeda.  A,  habit,  x  %;  B,  representative  leaf  (lower  surface),  x  %;  C,  enlargement  of  lower 
leaf  surface  showing  stellate  pubescence;  D,  representative  flower,  x!2;  E,  petal,  xlO;  F,  stamens,  %  lateral  view  (left)  and 
ventral  view  (right),  x  10;  G,  mature  berry,  x4;  H,  seeds,  x  15.  (A-F  from  the  holotype;  G,  H  from  Hammel  13181.) 


Rica  in  an  area  extending  from  Cerro  Coronel 
just  south  of  Barra  del  Colorado  southwest  to  the 
OTS  La  Selva  Field  Station  and  vicinity  at  20- 
200m. 

PHENOLOGY.— Flowering  and  fruiting  speci- 
mens have  been  collected  from  January  through 
May  and  July  through  September,  respectively. 

Miconia  calocoma  is  distinguished  by  its  rusty 
stellate  pubescence,  4-merous  flowers,  irregular- 


ly rupturing  apiculate  calyx,  unappendaged  an- 
thers, 4-celled  ovary,  and  densely  tuberculate, 
obovoid  to  pyriform  seeds. 

Placement  of  this  species  into  Cogniaux's 
(1891)  sectional  classification  depends  on  wheth- 
er one  chooses  to  emphasize  characters  of  the 
calyx  or  stamens.  In  having  oblong  unappen- 
daged stamens  with  ventrally  inclined  apical 
pores,  M.  calocoma  would  appear  to  belong  to 


146 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  5 


section  Amblyarrhena.  Within  this  section,  how- 
ever, no  particular  species  or  group  of  species 
can  be  singled  out  as  close  relatives.  The  irreg- 
ularly rupturing  apiculate  calyx  is  characteristic 
of  section  Laceraria,  but  M.  calocoma  lacks  the 
dorsally  appendiculate  anthers  that  are  typical  of 
that  section.  Within  this  latter  section,  M.  cal- 
ocoma is  most  similar  to  M.  rupticalyx  Wurdack 
of  Venezuela  and  M.  wagneri  Macbride  of  Bo- 
livia and  Peru.  These  two  allopatric  taxa  share 
several  diagnostic  features  with  M.  calocoma  in- 
cluding the  stellate  pubescence,  rupturing  apic- 
ulate calyx,  and  4-merous  flowers.  Characters  that 
consistently  separate  M.  rupticalyx  from  M.  cal- 
ocoma include  the  larger  (12-28  x  9-19  cm) 
ovate,  5-7-nerved  leaves  with  rounded  serrulate 
margins,  reniform  bracteoles,  dorsally  inclined 
anther  pores,  and  2(-3)-celled  ovary.  Miconia 
wagneri  is  readily  separated  from  M.  calocoma 
by  many  of  these  same  features  but  it  differs  most 
notably  in  having  larger  (15-25  x  8-12  cm),  5- 
7-nerved  leaves  that  are  entire  or  remotely  cal- 
lose-denticulate  and  anther  connectives  that  are 
dilated  at  the  filament  insertion  into  dorso-basal 
spurs. 

The  epithet  for  this  species  is  derived  from  the 
Greek  word  kalos,  beautiful,  and  the  Latin  word 
coma  (home  in  Greek),  hair  tuft,  in  reference  to 
the  attractive  covering  of  rusty  stellate  hairs  on 
branchlets,  leaves,  and  inflorescences. 

Miconia  dissitiflora  Almeda,  sp.  nov. 

Figure  5 

TYPE.— COSTA  RICA.  Puntarenas:  above  Golfito  along  road 
to  television  tower,  elev.  50-500  m,  16  Jul.  1977,  Almeda  et 
al.  3093  (holotype:  CAS!;  isotypes:  BM!,  BR!,  CR!,  DUKE!, 
F!,  MO!,  NY!,  US!). 

Frutex  1-3  m.  Ramuli  teretes  sicut  foliorum  subtus  venae 
primariae  inflorescentia  hypanthiaque  sparse  vel  modice  glan- 
duloso-furfuracei.  Folia  subsessilia  vel  petiolis  l-5(-9)  mm 
longis;  lamina  6.5-22  x  3.1-10cm,ellipticavelelliptico-ovata 
apice  acuminata  vel  attenuata  basi  paulo  cordata  vel  rotundata, 
5(-7)-plinervata,  membranacea  et  undulata  vel  subintegra.  In- 
florescentia 7-20.5  cm  longa  multiflora;  flores  5-meri,  pedi- 
cellis  (ad  anthesim)  1.5-2.5  mm  longis;  bracteolis  0.5  mm 
longis  subulatis.  Hypanthium  (ad  torum)  1.5  mm  longum;  ca- 
lyx primum  in  cono  apiculato  clausus  demum  in  lobos  irre- 
gulares  persistentes  ruptus,  dentibus  exterioribus  0.5  mm  em- 
inentibus.  Petala  3.5  x  1.5-2  mm  oblongo-lanceolata  glabra. 
Stamina  paulo  anisomorphica  glabra;  filamenta  1.5-2  mm  lon- 
ga; antherarum  thecae  2  x  0.5  vel  1.5  x  0.5  mm  angustae 
oblongae,  poro  paulo  dorsaliter  inclinato;  connectivum  nee 
prolongatum  nee  appendiculatum.  Stylus  2.5-3  mm  glaber; 
ovarium  5-loculare  omnino  inferum  apice  modice  puberulo. 

Sparingly  branched  shrubs  1-3  m  tall.  Older 


cauline  internodes  terete  and  glabrous.  Distal 
branchlets  and  inflorescences  moderately  to 
sparingly  stellate-furfuraceous.  Leaves  of  a  pair 
somewhat  unequal  in  size,  typically  subsessile 
and  clasping  or  sometimes  with  petioles  l-5(-9) 
mm  long;  blades  membranaceous  to  subcoria- 
ceous  at  maturity,  6.5-22  cm  long  and  3.1-10 
cm  wide,  elliptic  to  elliptic-ovate,  apex  acumi- 
nate to  attenuate,  base  rounded  to  subcordate 
but  sometimes  varying  to  slightly  oblique,  mar- 
gin bluntly  undulate-dentate  to  subentire,  5(-7)- 
plinerved  below  with  inner  pairs  of  subparallel 
primary  nerves  elevated  and  diverging  from  me- 
dian nerve  in  opposite  fashion  at  successive  points 
above  blade  base,  glabrous  above,  glabrous  to 
sparingly  stellate  on  and  adjacent  to  elevated  pri- 
maries below.  Inflorescence  a  terminal,  divari- 
cately branched  paniculiform  dichasium  7-20.5 
x  7-30  cm;  bracts  of  the  rachis  nodes  paired, 
lance-triangular,  0.5-4  mm  long,  0.5-1  mm  wide, 
essentially  glabrous  throughout  or  sparingly  stel- 
late pubescent  below;  bracteoles  paired,  gla- 
brous, sessile  and  evidently  fused  into  a  short 
nodal  collar  usually  evident  as  an  elevated  in- 
terpetiolar  ridge,  lance-triangular  to  subulate,  0.5 
mm  long,  0.5  mm  wide  at  base,  margin  entire. 
Pedicels  sparingly  stellate-furfuraceous  to  gla- 
brous, 1.5-2.5  mm  long.  Hypanthia  (at  an  thesis) 
campanulate,  1.5  mm  long  to  torus  (vascular  ring), 
glabrous  to  sparingly  stellulate  puberulent  ba- 
sally.  Calyx  closed  in  bud  and  crowned  by  an 
apiculum  ca.  0.5  mm  long  but  rupturing  irreg- 
ularly at  anthesis  into  2-5  persistent  hyaline  lobes 
1-1.5  mm  long;  exterior  calyx  teeth  5,  subulate, 
0.5  mm  long,  glabrous,  erect  to  antrorsely 
spreading.  Petals  5,  glabrous,  white,  oblong-lan- 
ceolate, acute  to  retuse  apically,  3.5  mm  long  and 
1.5-2  mm  wide,  margin  entire.  Stamens  10,  al- 
ternately unequal  with  the  larger  stamens  in- 
serted on  the  torus  opposite  the  petals  and  the 
smaller  ones  inserted  opposite  the  calyx  lobes; 
filaments  glabrous,  complanate,  somewhat  con- 
stricted and  geniculate  distally,  1.5-2  mm  long; 
anthers  1.5-2  mm  long  and  0.5  mm  wide,  yellow, 
linear-oblong,  truncate  distally  with  a  somewhat 
dorsally  inclined  terminal  pore;  connective 
slightly  thickened  dorsally  but  not  dilated  or  pro- 
longed below  the  filament  insertion.  Ovary  in- 
ferior, 5 -celled,  minutely  and  caducously  puber- 
ulent at  the  summit.  Style  straight,  glabrous,  2.5- 
3  mm  long;  stigma  truncate.  Berry  globose,  2.5- 
3.5  mm  long  and  ca.  2.5  mm  in  diameter.  Seeds 
numerous,  white,  galeiform,  0.5  mm  long  with 


ALMEDA:  FIVE  NEW  MELASTOMATACEAE 


147 


H 


FIGURE  5.  Miconia  dissitiflora  Almeda.  A,  habit,  x%;  B,  enlargement  of  inflorescence  node  showing  bracts  and  stellate- 
furfuraceous  pubescence;  C,  floral  bud  showing  rupturing  calyx  and  apiculum,  x  1 2;  D,  flower  with  petals  and  stamens  removed, 
x  1 2;  E,  petal,  x  9;  F,  larger  (antepetalous)  stamens,  dorsal  view  (left)  and  lateral  view  (right),  x  6;  G,  smaller  (antesepalous) 
stamens,  lateral  view  (left)  and  dorsal  view  (right),  xca.  6;  H,  mature  berry,  xca.  10;  I,  seeds,  xca.  18.  (A-G  from  the  holotype; 
H,  I  from  Liesner  2013.) 


a  densely  muricate  or  verrucose  testa  on  the  con- 
vex side  and  a  lateral  raphe  on  the  opposing  face. 

ADDITIONAL  SPECIMENS  EXAMINED.— COSTA  RICA.  Pun- 
tarenas:  main  ridge  and  NE  slopes  of  Fila  de  Cal,  between  San 
Vito  and  Ciudad  Neily,  08°41'N,  82°56.5'W,  13  Sep.  1985, 
Graytim  et  al.  6034  (CAS);  NE  slopes  of  Fila  de  Cal  between 
San  Vito  and  Ciudad  Neily,  08°41'N,  82°56.5'W,  12Jul.  1985, 
Hammel  &  Grayum  14187  (CAS);  Rincon  de  Osa,  Rio  Agua 


Buena  to  ca.  4  km  above  it,  10  Feb.  1974,  Liesner  20 13  (CAS); 
Corcovado  National  Park,  trail  from  base  of  hills  to  Los  Chiles, 
9  Jul.  1977,  Liesner  3071  (CAS,  CR);  trail  from  radio  tower 
on  ridge  above  Golfito,  opposite  docks,  08°39'N,  83°10'W,  25 
Jul.  1977,  Webster  22045  (CAS). 

DISTRIBUTION.— Known  only  from  the  Golfo 
Dulce  region  of  southern  Costa  Rica  in  an  area 
extending  from  Los  Chiles  in  Corcovado  Na- 


148 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  5 


tional  Park  eastward  to  Fila  de  Cal  between  San 
Vito  and  Ciudad  Nelly  at  20-620  m. 

PHENOLOGY.— Flowering  specimens  have  been 
collected  in  July  and  September;  the  only  fruiting 
specimen  was  collected  in  July. 

Miconia  dissitiflora  is  a  well-defined  species, 
readily  distinguished  by  its  sparse  stellate- 
furfuraceous  indument,  divaricately  branched 
inflorescence,  irregularly  rupturing  calyx,  unap- 
pendaged  anisomorphic  stamens,  and  oblong- 
lanceolate  petals.  It  is  also  unusual  and  evidently 
unique  among  New  World  berry-fruited  mela- 
stomes  in  having  the  larger  stamens  inserted  on 
the  torus  opposite  the  petals.  Among  neotropical 
melastomes  this  condition  is  known  only  in  the 
capsular-fruited  genus  Monochaetum. 

The  problem  of  placing  M.  dissitiflora  into  a 
section  parallels  the  ambiguous  situation  de- 
scribed above  for  M.  calocoma.  An  emphasis  on 
staminal  characters  would  dictate  placement  into 
section  Amblyarrhena,  whereas  characters  of  the 
calyx  suggest  a  relationship  with  the  species  of 
section  Laceraria.  The  new  species  bears  no  strong 
resemblance  to  any  of  the  described  taxa  cur- 
rently placed  in  section  Amblyarrhena.  A  search 
for  morphologically  similar  species  in  section 
Laceraria  has  only  served  to  highlight  the  am- 
biguities involved  when  assessing  relationships 
based  on  different  character  sets.  Miconia  cen- 
trodesma  of  section  Laceraria  shares  an  irregu- 
larly rupturing  calyx  and  a  sparse  stellate-fur- 
furaceous  indument  with  M.  dissitiflora,  but  it 
differs  in  a  number  of  diagnostic  reproductive 
characters.  It  has  4-merous  flowers,  isomorphic 
stamens,  dorsally  appendiculate  anthers,  a  2-3- 
celled  ovary,  and  narrow  angular-oblong  tuber- 
culate  seeds.  The  seeds  of  M.  centrodesma  are 
unusual  in  having  a  dilated  testa  at  the  distal  end 
that  forms  a  foot  or  spur  composed  of  enlarged 
cells  that  collapse  on  drying.  This  "foot,"  which 
can  be  observed  only  in  fresh  or  hydrated  ma- 
terial, imparts  an  overall  shape  reminiscent  of  a 
miniature  boomerang.  Thus,  the  seeds  of  M  cen- 
trodesma differ  dramatically  from  the  muricate 
or  verrucose  galeiform  seeds  of  M.  dissitiflora. 
These  considerable  differences  suggest  that  the 
relationship  between  M.  centrodesma  and  M. 
dissitiflora  is  not  a  particularly  close  one.  Con- 
sequently, M.  dissitiflora  is  probably  best  re- 
garded as  an  isolated  species  not  readily  placed 
within  the  sectional  classification  proposed  by 
Cogniaux(1891). 

The  epithet  chosen  for  this  species  is  derived 
from  the  Latin  word  dissitus,  remote  or  lying 


apart,  referring  to  the  well-spaced  position  of 
flowers  on  the  divaricately  branched  inflores- 
cence. It  is  this  character,  together  with  the  small 
apiculate  buds  and  plinerved  leaves,  that  give  M. 
dissitiflora  an  aspect  reminiscent  of  Conostegia 
cinnamomea  (Beurl.)  Wurdack.  The  latter  differs 
in  having  isomorphic  stamens,  a  calyptriform, 
circumscissile  calyx,  and  oblong,  deltoid  seeds 
that  are  somewhat  angulate  on  the  convex  face. 

Tococa  croatii  Almeda,  sp.  nov. 

Figure  6 

TYPE.— PANAMA.  Darien:  Serrania  de  Pirre,  9-10  km  due 
N  of  Alto  de  Nique,  elev.  1,520-1,560  m,  27  Jul.  1976,  Croat 
37873  (holotype:  CAS!;  isotype:  MO!). 

Frutex  0.5-15  m.  Ramuli  primum  rotundato-quadrangulati 
demum  teretes  sicut  foliorum  venae  primariae  et  secundariae 
sparse  vel  modice  setosi  pilis  laevibus  laxis  paulo  retrorsis  0.5- 
2mmlongis.  Petioli2.5-12.5cmlongi;lamina  12-19.5  x  8.3- 
1 5.3  cm  ovata  apice  acuminata  basi  rotundata  vel  cordata,  7(- 
9)-nervata,  membranacea  et  crenulata  ciliata.  Inflorescentia 
10-18  cm  long  multiflora;  flores  5-meri,  pedicellis  (ad  anthe- 
sim)  1-4  mm  longis,  bracteolis  1  mm  longis  subulatis  persis- 
tentibus.  Hypanthium  (ad  torum)  2-2.5  mm  longum;  calyx 
primum  in  cono  apiculato  clausus  demum  in  lobos  regulares 
persistentes  ruptus,  dentibus  exterioribus  0.25-0.5  mm  emi- 
nentibus.  Petala  5-6  x  3.5-5  mm  obovata  glabra.  Stamina 
isomorphica  glabra;  filamenta  2.5-3.5  mm  longa;  antherarum 
thecae  1.5-2  x  1  mm  oblongae  inter  se  cohaerentes  poro  0.5 
mm  diam.;  connectivum  dorsaliter  ad  basim  dente  ca.  1-1.5 
mm  longo  descendenti  armatum.  Stylus  5.5-8.5  mm;  ovarium 
5-loculare  omnino  inferum  apice  sparsissime  pilis  0.5-1  mm 
longis  glandulosis  armato. 

Shrubs  0.5-1.5  m  tall.  Distal  branchlets 
subquadrangular,  sparsely  to  moderately  beset 
with  a  caducous  cover  of  straight  spreading  and/ 
or  somewhat  crisped  hairs  mostly  1-2  mm  long. 
Leaves  of  a  pair  equal  to  slightly  unequal  in  size; 
petioles  2.5-12.5  cm  long;  blades  thin  and  mem- 
branaceous,  12-19.5  cm  long  and  8.3-15.3  cm 
wide,  broadly  ovate  to  cordate,  apex  acuminate, 
base  broadly  rounded  to  cordate,  margin  ciliate- 
crenulate,  7(-9)-nerved  below  with  all  primary 
nerves  diverging  from  a  common  point  at  base 
of  blade,  sparingly  to  moderately  setulose  above 
with  smooth  appressed  to  somewhat  spreading 
hairs  mostly  0.5-2  mm  long,  moderately  fine  se- 
tulose on  and  between  primary  nerves  below  with 
tardily  caducous,  simple,  and  glandular  hairs  0.5- 
1  mm  long.  Inflorescence  a  terminal  multiflow- 
ered  panicle  10-18  cm  long  with  subverticillate 
branching  at  secondary  nodes;  rachis  moderately 
to  copiously  covered  with  spreading  glandular 
hairs  (0.5-1  mm  long)  and  a  ground  layer  of 
sessile  glandular  hairs;  bracts  of  rachis  nodes 
paired,  linear-oblong,  2.5-3  mm  long,  0.5  mm 


ALMEDA:  FIVE  NEW  MELASTOMATACEAE 


149 


FIGURE  6.  Tococa  croatii  Almeda.  A,  habit,  x'/i;  B,  flower  with  petals  and  staminal  ring  removed,  x4;  C,  petal,  x5;  D, 
staminal  ring  as  seen  from  above  (left)  and  lateral  view  (right),  xca.  4;  E,  seeds,  xca.  25.  (A-D  from  the  holotype;  E  from 
Gentry  &  Mori  13817.) 


wide,  sparsely  glandular  setulose  below  inter- 
mixed with  or  replaced  by  an  inconspicuous 
ground  layer  of  sessile  glandular  hairs;  bracteoles 
paired,  sessile,  and  persistent,  subulate  to  seti- 
form,  1  mm  long,  essentially  glabrous  or  with  a 
sparse  cover  of  sessile  glandular  hairs.  Pedicels 
1-4  mm  long,  beset  with  minute  glandular  hairs 
up  to  0.5  mm  long.  Hypanthia  (at  anthesis)  cam- 
panulate,  2-2.5  mm  long  to  torus  (vascular  ring), 
moderately  covered  with  spreading  smooth  glan- 
dular hairs  0.5-1  mm  long  intermixed  with  or 
replaced  by  a  ground  layer  of  sessile  dot-like 
glandular  hairs.  Calyx  closed  in  bud,  crowned  by 
an  apiculum  0.5  mm  long  but  rupturing  regularly 
at  anthesis  into  5  triangular  hyaline  lobes  2x2 
mm;  exterior  calyx  teeth  5,  fused  to  calyx  lobes 
for  much  of  their  length  with  the  free  subulate 


portions  projecting  0.25-0.5  mm  from  the  distal 
portion  of  each  calyx  lobe.  Petals  5,  glabrous, 
white  (fide  Croat  37873),  pink  (fide  Hartman 
12458),  or  bluish-purple  (fide  Hartman  12404), 
obovate,  rounded  apically,  5-6  mm  long  and  3.5- 
5  mm  wide,  margin  entire.  Stamens  10,  iso- 
morphic;  filaments  glabrous,  complanate,  2.5- 
3.5  mm  long;  anthers  1.5-2  mm  long,  1  mm 
wide,  laterally  compressed  and  coherent  in  a  ring, 
apex  truncate  to  somewhat  emarginate  on  ven- 
tral side;  connective  thickened  dorsally  and  pro- 
longed at  base  into  a  retrorse  triangular  appen- 
dage 1-1.5  mm  long.  Ovary  inferior,  5 -celled, 
sparingly  glandular-setose  at  summit  surround- 
ing style  base.  Style  straight,  minutely  papillate 
distally,  5.5-8.5  mm  long  and  conspicuously 
overtopping  surrounding  stamen  ring;  stigma 


150 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  5 


truncate  to  capitellate.  Berry  reportedly  blue  at 
maturity,  globose,  5-6  mm  long,  5-6  mm  in  di- 
ameter. Seeds  numerous,  ovoid,  white,  0.5  mm 
long  with  a  shallowly  scrobiculate  testa. 

ADDITIONAL  SPECIMENS  EXAMINED.  — PANAMA.  Darien: 
Cerro  Mali,  vicinity  of  base  camp,  1,400  m,  near  Colombian 
border,  23  Jan.  1975,  Gentry  &  Mori  13817  (F,  US);  SW  ridge 
leading  to  Alturas  de  Nique  on  border  with  Colombia,  1 , 1 00- 
1,200  m,  30  Dec.  1980,  Hartman  12404  (CAS);  Alturas  de 
Nique  and  ridge  leading  SW,  1,250-1,500  m,  31  Dec.  1980, 
Hartman  12458  (CAS). 

DISTRIBUTION.— A  little-collected  species  of 
eastern  Panama  currently  known  from  Cerro  Mali 
in  the  Serrania  del  Darien  and  the  region  of  Al- 
turas de  Nique  in  the  Serrania  de  Pirre  in  south- 
ern Darien  province  near  the  Panama/Colombia 
frontier  at  1,100-1,560  m. 

PHENOLOGY.— Flowering  specimens  have  been 
collected  in  July  and  December;  the  only  known 
fruiting  specimen  was  collected  in  January. 

Tococa  croatii  is  denned  by  its  caducous  cover 
of  straight  or  crisped  cauline  hairs  (1-2  mm  long), 
regularly  rupturing  hyaline  calyx,  and  shallowly 
scrobiculate,  ovoid  seeds.  The  most  extraordi- 
nary feature  of  this  species  is  the  morphology  of 
the  anther  thecae.  These  are  distally  truncate  to 
emarginate  with  a  comparatively  broad  terminal 
pore,  laterally  coherent  in  a  ring,  and  beset  dor- 
sally  with  deflexed  triangular  appendages.  Among 
described  species  of  Tococa,  T.  croatii  shares  its 
habit,  inflorescence  architecture,  staminal  de- 
tails, and  fruit  morphology  only  with  T.  sym- 
phyandra  (Triana)  Cogn.  of  Colombia  and  Ec- 
uador. The  consistent  differences  between  these 
species  involve  foliage  characters  and  pubes- 
cence details  of  the  distal  internodes,  hypanthia 
and  ovaries.  Tococa  symphyandra  has  long  sub- 
retrorse-setose  hairs  (0.5-1.5  cm  long)  on  young 
branchlets  and  petioles,  didymous  formicaria 
about  1  cm  long  that  are  immersed  in  the  blade 
base,  and  essentially  glabrous  hypanthia  and 
ovaries.  The  derived  staminal  characters  that 
unite  T.  croatii  and  T.  symphyandra  as  a  dis- 
tinctive species  pair  are  also  the  features  that 
make  them  anomalous  members  of  the  genus. 
These  staminal  characters  would  also  make  them 
equally  anomalous  if  included  in  Miconia  or  any 
other  genus  of  the  Miconieae.  Consequently,  the 
precise  relationships  and  best  generic  disposition 
for  these  two  species  must  await  detailed  mono- 
graphic studies  of  both  Miconia  and  Tococa. 

This  species  is  named  for  Thomas  B.  Croat,  a 
student  of  neotropical  Araceae  who  has  contrib- 


uted significantly  to  our  knowledge  of  the  neo- 
tropical flora  through  his  many  fine  collections 
and  publications. 

ACKNOWLEDGMENTS 

The  research  presented  here  was  supported,  in 
part,  by  U.S.  National  Science  Foundation  Grants 
DEB  76-83040,  DEB  78-25620,  and  BSR 
8614880  (Flora  Mesoamerica).  I  am  grateful  to 
the  following  individuals  who  contributed  to  this 
study  by  assisting  me  in  the  field  or  collecting 
specimens  at  my  request:  Greg  de  Nevers,  Thom- 
as F.  Daniel,  Barry  E.  Hammel,  Gordon  Mc- 
Pherson,  and  Robert  L.  Wilbur.  I  thank  the  Mu- 
seo  Nacional  de  Costa  Rica,  the  Organization  for 
Tropical  Studies,  and  the  Missouri  Botanical 
Garden  for  logistical  support.  I  also  thank  John 
J.  Wurdack  for  comments  on  the  manuscript;  G. 
W.  Krantz  (Dept.  of  Entomology,  Oregon  State 
University)  for  identifying  the  mites;  Charles 
Turner  (U.S.  Department  of  Agriculture-ARS, 
Albany,  California)  for  information  on  plant- 
mite  associations;  and  the  curators  and  staffs  of 
the  following  herbaria  who  either  made  their  fa- 
cilities available  for  on-site  study  or  provided 
loans  and/or  gifts  of  important  collections:  BM, 
BR,  C,  CR,  DUKE,  F,  G,  K,  MEXU,  MO,  NY, 
P,  PMA,  US.  The  illustrations  are  the  work 
of  Terry  Bell  (Fig.  1,  3-6)  and  Ellen  del  Valle 
(Fig.  3). 

LITERATURE  CITED 

ALMEDA,  F.  1 984.  New  and  noteworthy  additions  to  the  Me- 
lastomataceae  of  Panama.  Proc.  Calif.  Acad.  Sci.  43(17): 
269-282. 

BAILLON,  H.  E.  1879.  Melastomacees.  Hist.  PI.  7:1-65.  (En- 
glish transl.  7:1-65.  1881.) 

BENSON,  W.  W.  1985.  Amazon  ant-plants.  Pp.  239-266  in 
Amazonia.  G.  T.  Prance  and  T.  E.  Lovejoy,  eds.  Pergamon 
Press,  Oxford,  England. 

COGNIAUX,  C.  A.  1891.  Melastomaceae.  Pp.  1-1256  in 
Monographiae  phanerogamarum,  Vol.  7.  A.  de  Candolle  and 
C.  de  Candolle,  eds.  G.  Masson,  Paris. 

FAEGRI,  K.  AND  L.  VAN  DER  PUL.  1971.  The  principles  of 
pollination  ecology,  2nd  ed.  Pergamon  Press,  Oxford,  En- 
gland. 

JACOBS,  M.  1966.  On  domatia— the  viewpoints  and  some 
facts.  I,  II,  III.  Proc.  Koninkl.  Nederl.  Akad.  Wetensch.  C69: 
275-316. 

MACBRIDE,  J.  F.  1 94 1 .  Melastomataceae.  In  Flora  of  Peru. 
Field  Mus.  Nat.  Hist.,  Bot.  Sen  13:249-521. 

PERCIVAL,  M.  1965.  Floral  biology.  Pergamon  Press,  Oxford, 
England. 

PROCTOR,  M.  AND  P.  YEO.  1973.  The  pollination  of  flowers. 
Williams  Collins  Sons  and  Co.  Ltd.,  London,  England. 

PUTZ,  F.  E.  ANDN.  M.  HOLBROOK.  1986.  Notes  on  the  natural 
history  of  hemiepiphytes.  Selbyana  9(l):61-69. 


PROCEEDINGS 

OF  THE 

CALIFORNIA  ACADEMY  OF  SCIENCES 


Vol.  46,  No.  6,  pp.  151-178,  13  figs.,  2  tables.  August  24,  1989 


SYSTEMATIC  REVISION  AND  DESCRIPTION  OF  NEW 

SPECIES  OF  SUCKERMOUTH  CATFISHES 

(CHILOGLANIS,  MOCHOKIDAE) 

FROM  CAMEROUN 

By 
Tyson  R.  Roberts 

California  Academy  of  Sciences,  San  Francisco,  California  94118 


ABSTRACT:  Ten  species  of  Chiloglanis  occur  in  Cameroun:  C.  batesii  Boulenger,  1904  (all  or  almost  all  river 
basins);  C.  benuensis  Daget  and  Stauch,  1963  (Niger);  C.  cameronensis  Boulenger,  1904  (Lokunje,  Kribi, 
Kienke,  Ntem);  C,  disneyi  Trewavas,  1974  (Mungo  and  Cross);  C.  harbinger  new  species  (Lokunje);  C.  niger 
new  species  (Niger);  C.  polypogon  new  species  (Cross);  C.  reticulatus  new  species  (Congo);  C.  sanagaensis 
new  species  (Sanaga);  and  C.  voltae  Daget  and  Stauch,  1963  (Niger). 

Chiloglanis  disneyi,  C.  harbinger,  C.  niger,  C.  polypogon,  and  C.  sanagaensis  are  known  only  from  Cameroun; 
C.  benuensis  occurs  in  the  River  Benue  in  Cameroun  and  Nigeria.  Chiloglanis  reticulatus  also  occurs  in  the 
Congo  basin  in  northwestern  Zaire.  Chiloglanis  batesii  (including  its  synonym  C.  micropogon  Poll,  1952) 
occurs  in  suitable  habitats  throughout  the  Niger  and  Congo  basins,  and  is  the  only  species  of  Chiloglanis 
known  from  the  Chad  basin.  Chiloglanis  cameronensis  is  the  only  species  known  in  the  Ogooue  basin  (Gabon). 
Chiloglanis  voltae  is  also  known  from  the  Benue  in  Nigeria  and  the  Volta  basin  (Ghana  and  Burkina  Faso). 

In  all  of  the  major  river  systems  of  Cameroun  two  or  even  three  species  of  Chiloglanis  occur  sympatrically. 
Sympatric  species  usually  differ  strongly  in  sucker  morphology,  number  and  length  of  barbels,  dentition, 
coloration,  adult  size,  and  other  characteristics.  Secondary  sexual  dimorphism  is  marked  in  some  species, 
apparently  absent  in  others. 

Received  September  22,  1987.  Accepted  January  9,  1989. 

uensis  Daget  and  Stauch,  1963;  C.  cameronensis 

INTRODUCTION  „     .  innj,   ^  j-        •  T  im/i   /~ 

Boulenger,  1904;  C.  disneyi  Trewavas,  1974;  C. 

Chiloglanis  belongs  to  the  endemic  African  micropogon  Poll,  1952;  and  C.  voltae  Daget  and 

catfish  family  Mochokidae.  Adults  range  in  size  Stauch,  1963.  For  a  review  of  the  literature  on 

from  about  18  to  100  mm  standard  length.  The  species  of  Chiloglanis  and  their  distribution  see 

sucker  or  oral  disc  is  formed  by  the  enlarged  Daget  etal.  (1986).  Two  recent  papers  have  dealt 

upper  and  lower  lips.  The  maxillary  barbels  lie  with  Chiloglanis  from  parts  of  Cameroun:  the 

free  at  the  sides  of  the  upper  lip  but  the  man-  Camerounian  portion  of  the  Benue  (Niger  basin) 

dibular  or  mental  (chin)  barbels  are  largely  in-  by  Daget  and  Stauch  (1963),  and  the  Mungo  and 

corporated  into  the  lower  lip.  Meme,  small  coastal  streams  in  western  Cam- 

A  total  of  34  species  of  Chiloglanis  has  been  eroun,  by  Trewavas  (1974).  The  results  of  these 

recognized,  of  which  six  have  been  reported  from  studies  and  my  interpretations  of  them  are  in- 

Cameroun:  C.  batesii  Boulenger,  1 904;  C.  ben-  corporated  in  the  present  paper. 

[151] 


152 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  6 


Most  species  of  Chiloglanis  are  represented  in 
museum  collections  by  only  a  few  specimens, 
often  in  poor  condition.  In  1980  I  travelled  ex- 
tensively in  Cameroun,  collecting  in  the  Niger, 
Ndian,  Cross,  Sanaga,  and  Congo  basins.  Ad- 
ditional collections  were  made  in  Cameroun  in 
1987-88.  Specimens  of  Chiloglanis  were  col- 
lected readily  with  a  nylon  flyscreen  pushnet  in 
rocky  substrate,  gravel  bars,  and  sand  riffles.  This 
resulted  in  discovery  of  four  undescribed  species. 
An  additional  new  species,  from  the  Lokunje  ba- 
sin, was  discovered  in  the  fish  collection  of  the 
California  Academy  of  Sciences.  I  visited  the 
Lokunje  in  December  1987  but  was  unable  to 
collect  additional  specimens.  Five  of  the  six 
species  previously  reported  from  Cameroun  are 
recognized  as  valid.  After  study  of  over  40  lots 
and  nearly  2,000  specimens  referable  to  C. 
batesii  and  C.  micropogon,  including  type  spec- 
imens, I  regard  C.  micropogon  as  a  junior  syn- 
onym of  C.  batesii.  Chiloglanis  batesii  is  perhaps 
the  most  widely  distributed  species  of  Chilo- 
glanis. It  occurs  in  suitable  habitat  in  all  or  al- 
most all  river  basins  of  Cameroun,  throughout 
the  Niger  and  Zaire  basins,  and  is  the  only  species 
of  the  genus  in  the  Chad  basin;  it  is  absent  west 
of  the  Niger  basin.  Chiloglanis  disneyi,  hitherto 
known  only  from  the  Mungo,  is  reported  from 
the  Cross  basin.  The  new  species  described  from 
Cameroun  are:  C.  niger  (Niger  basin);  C.  poly- 
pogon  (Cross);  C.  sanagaensis  (Sanaga);  C.  har- 
binger (Lokunje);  and  C.  reticulatus  (Congo).  The 
first  four  are  known  only  from  Cameroun.  The 
fifth  is  also  present  in  the  Congo  basin  in  Zaire. 

Little  is  known  about  the  biology  of  Chilo- 
glanis or  of  any  other  mochokids.  At  least  some 
species  of  Chiloglanis  have  surprisingly  large  eggs. 
Ovarian  eggs  of  1.7  and  1.9  mm  were  reported 
in  C.  batesii  of  28  and  35.5  mm,  respectively,  by 
Trewavas  (1974:363).  A  37.2  mm  specimen  from 
Kivu  has  ovarian  eggs  of  2.0  mm.  The  total  num- 
ber of  eggs  in  this  specimen  is  at  least  a  dozen, 
but  no  more  than  20  (all  but  one  or  two  have 
been  left  in  situ).  The  holotype  of  C.  reticulatus 
is  a  gravid  female  with  about  40  eggs  to  1 .8  mm. 
In  C.  batesii  and  in  C.  reticulatus,  ripe  ovaries 
contain  several  size  classes  of  eggs.  Sexually  ma- 
ture males  of  all  species  have  lobate  testes,  which 
in  very  ripe  individuals  may  cause  abdominal 
expansion  comparable  to  that  caused  by  enlarged 
ovaries  of  gravid  females.  In  several  species  males 
have  an  elongate  genital  papilla,  the  functional 


significance  of  which  is  unknown.  Males  of  a  few 
species  have  the  humeral  process  of  the  pectoral 
girdle  greatly  enlarged,  but  the  significance  of  this 
is  also  unknown.  Chiloglanis  feed  mainly  on  in- 
sects and  attached  algae.  The  species  differ  strik- 
ingly in  dentition  and  other  morphological  fea- 
tures related  to  feeding,  but  these  differences  have 
not  been  correlated  with  particular  foods. 

STUDY  MATERIAL 

Material  examined  is  deposited  in  the  follow- 
ing institutions  (standard  symbolic  codes  from 
Leviton  et  al.  1985):  BMNH,  British  Museum  of 
Natural  History;  CAS  and  CAS-SU,  California 
Academy  of  Sciences;  FMNH,  Field  Museum  of 
Natural  History  (Chicago);  MCZ,  Museum  of 
Comparative  Zoology;  MHNG,  Museum  d'His- 
toire  Naturelle,  Geneve;  MNHN,  Museum  Na- 
tional d'Histoire  Naturelle;  MRAC,  Musee  Roy- 
al de  FAfrique  Centrale  (Tervuren);  UMMZ, 
Museum  of  Zoology,  University  of  Michigan;  and 
USNM,  Smithsonian  Institution. 

Camerounian  and  non-Camerounian  material 
of  species  of  Chiloglanis  found  in  Cameroun  is 
listed  in  the  systematic  account  of  each  species. 

The  abbreviation  "R.",  always  placed  in  front 
of  the  proper  name,  stands  for  River,  riviere,  Rio 
(Spanish),  and  rio  (Portuguese).  I  have  adopted 
this  practice  when  writing  on  African  freshwater 
fishes  because  the  same  river  is  often  shared  by 
two  or  more  English-,  French-,  Spanish-,  or  Por- 
tuguese-speaking countries;  authors  in  French 
usually  refer  to  any  river,  regardless  of  what 
country  it  is  in,  as  "riviere,"  while  authors  in 
English  use  "River";  and  although  in  English  the 
word  "River"  may  precede  or  succeed  the  proper 
name,  in  the  other  languages  the  word  for  river 
always  precedes  the  proper  name. 

The  following  comparative  material  of  non- 
Camerounian  species  of  Chiloglanis  and  of  the 
suckermouthed  mochokid  genera  Atopochilus  and 
Euchilichthys  has  been  examined: 

Chiloglanis  brevibarbis  Boulenger,  1902:  MHNG  2000.68-82, 
15:  29.3-57.7  mm,  Kenya,  R.  Athi  at  Kabaa. 

Chiloglanis  congicus  Boulenger,  1920:  CAS(SU)  66706,  51.6 
mm,  Congo  basin,  probably  Stanley  Pool;  CAS  44887,  2: 
58.0-59.5  mm,  aquarium  specimens;  CAS  60812,  60.7  mm, 
Congo  basin,  R.  Inkisi  below  Chute  de  Sanga. 

Chiloglanis  deckenii  Peters,  1868:  MHNG  1443.79-83,  5:  25.8- 
33.4  mm,  Kenya,  R.  Kerio  near  Tambach. 

Chiloglanis  neumanni  Boulenger,  1911:  UMMZ  199975, 
199988,  and  200183,  19:  19.5-40.8  mm,  Zambia,  Luapula- 
Congo  basin. 


ROBERTS:  CAMEROUNIAN  SUCKERMOUTH  CATFISHES 


153 


Chiloglanis  niloticusBoulenger,  1900:MNHN  1907-230,39.4 
mm,  R.  Nile  at  Arko,  Sudan  (syntype). 

Chiloglanis  occidental  Pe\\egrin,  1933:  UMMZ  182016,  24.0 
mm,  tributary  of  R.  Senegal;  CAS(SU)  62926,  56:  19.9-40.5 
mm,  Ghana,  R.  Ankwao;  CAS(SU)  62927, 4:  30.7^6.4  mm, 
Ghana,  Opon  Valley;  CAS(SU)  62928,  56.7  mm,  Ghana,  R. 
Ahiri  near  Wasa  Akropong;  MCZ  48063,  25.6  mm,  Ghana, 
Pra  basin,  Twifu  Prasu;  V.  G.  F.  Smith  collection,  uncat., 
26.4  mm,  Nigeria,  Niger  basin,  R.  Lere,  Jos  Plateau;  BMNH 
1935.5.29:52,  18.8  mm,  Nigeria,  Niger  basin,  R.  Kaduna. 

Chiloglanis  polyodon  Norman,  1932:  BMNH  1 90 1 . 1 2.26,  Sier- 
ra Leone,  headwaters  of  R.  Bagbwe  (holotype). 

Chiloglanis  sardinhai  Ladiges  and  Voelker,  1961:  USNM 
265443,  5:  28.1-37.1  mm,  Angola,  Benguela,  Caibambo. 

Atopochilus  savorgnani  Sauvage,  1878:  CAS(SU)  15715,  Cam- 
eroun,  R.  Ntem,  Nyabessan. 

Atopochilus  sp.  undet.  a:  CAS(SU)  64793,  98.2  mm,  Congo 
basin,  Stanley  Pool. 

Atopochilus  sp.  undet.  b:  CAS  60815,  31.5  mm,  Congo  basin, 
R.  Lufu. 

Euchilichthys  dybowskii  (Vai\\anl,  1892):  MNHN  1892-81,  34.0 
mm,  Congo  basin,  Oubanghi  (syntype);  MCZ  48356, 5: 40.8- 
63.2  mm,  Congo  basin,  R.  Mbomou  at  Gozobangui. 

Euchilichthys  royauxi  Boulenger,  1902:  CAS(SU)  64612,  2: 
89.6-175  mm,  Congo  basin,  rapids  of  lower  Congo  at  Kin- 
suka. 

Characters  Useful  in  Distinguishing 
Species  of  Chiloglanis 

Apart  from  Trewavas  (1974),  almost  no  au- 
thors have  given  adequate  discussion  or  defini- 
tions of  the  characters  used  to  define  species  of 
Chiloglanis.  Some  easily  observed  features  of  the 
sucker  and  dentition  have  not  been  utilized,  and 
discussion  of  variation  with  age  or  sex  is  gen- 
erally lacking.  Observation  of  previously  unob- 
served or  unreported  features,  such  as  tubercles 
and  taste  buds,  is  facilitated  by  scanning  electron 
microscopy.  Some  useful  characters,  none  pre- 
viously utilized,  can  be  observed  in  radiographs. 

SUCKER  AND  BARBELS.— All  species  of  Chi- 
loglanis have  a  well-developed  sucker  or  oral  disc 
(Fig.  1),  made  of  the  greatly  expanded  and  united 
upper  and  lower  lips.  The  lower  lip  incorporates 
much  or  almost  all  of  the  mental  (mandibular  or 
chin)  barbels.  The  most  obvious  differences  in 
the  sucker  of  various  species  involves  the  mental 
barbels.  In  some  species  they  project  consider- 
ably from  the  posterior  border  of  the  lower  lip, 
in  others  they  project  moderately,  and  in  still 
others  slightly  or  not  at  all.  Some  species  have 
only  two  prominent  mental  barbels  on  each  side 
of  the  lower  lip.  These  correspond  to  the  primary 
or  primitive  pair  of  mental  barbels  found  in  al- 
most all  groups  of  catfishes.  In  other  species  there 
are  up  to  five  auxiliary  mental  barbels  on  each 


side  of  the  lower  lip.  In  most  species  with  aux- 
iliary mental  barbels,  these  occur  inside  the  outer 
primary  mental  barbel,  but  one  species  (C.  poly- 
pogori)  usually  has  an  auxiliary  mental  barbel 
external  to  the  outer  primary  mental  barbel. 
Camerounian  species  of  Chiloglanis  with  mental 
barbels  projecting  considerably  from  the  lower 
lip  (i.e.,  with  "large"  mental  barbels)  are  C.  ben- 
uensis,  C.  polypogon,  C.  reticulatus,  and  C.  vol- 
tae.  Chiloglanis  reticulatus  and  C.  voltae  have 
only  primary  mental  barbels,  while  C.  benuensis 
and  C.  polypogon  have  well-developed  auxiliary 
mental  barbels.  Chiloglanis  voltae  is  unique  in 
having  very  elongate  outer  mental  barbels  and 
very  short  inner  ones.  Species  with  mental  bar- 
bels projecting  very  little  ("minute"  mental  bar- 
bels) are  C.  batesii,  C.  disneyi,  C.  harbinger,  and 
C.  niger.  Chiloglanis  cameronensis  and  C.  sa- 
nagaensis  have  moderately  projecting  mental 
barbels. 

Other  obvious  differences  between  species  in- 
volve size  and  shape  of  the  sucker.  In  some  species 
it  is  transversely  oval  (i.e.,  broader  than  long) 
and  relatively  (proportionately)  small  compared 
to  the  size  of  the  fish,  e.g.,  in  C.  benuensis,  C. 
polypogon,  and  C.  reticulatus.  In  others  it  is  rel- 
atively round,  e.g.,  C.  disneyi,  C.  harbinger,  and 
C.  niger.  The  last  two  species  have  perhaps  the 
largest  suckers  of  any  species  of  Chiloglanis.  The 
papillae  on  the  sucker  also  differ  among  species. 
In  C.  benuensis,  C.  polypogon,  and  C.  reticulatus 
they  are  especially  large,  particularly  on  the  me- 
dian part  of  the  lower  lip.  In  C.  voltae  the  papillae 
are  greatly  reduced  or  absent.  In  C.  sanagaensis 
they  are  extremely  numerous,  evenly  distributed 
and  of  nearly  the  same  size  over  virtually  the 
entire  sucker.  In  all  species  the  two  halves  of  the 
lower  lip  are  separated  distally  by  a  median  cleft. 
In  most  species  the  cleft  is  very  deep,  so  that 
most  of  the  median  papillate  area  of  the  lower 
lip  is  also  cleft,  but  in  C.  niger  the  cleft  is  shal- 
lower and  the  median  papillate  area  undivided. 
The  boundary  between  upper  and  lower  lips  is 
relatively  well  defined  in  some  species,  e.g.,  C. 
benuensis,  C.  polypogon,  and  C.  reticulatus,  but 
is  indistinguishable  in  others,  e.g.,  C.  niger.  The 
upper  lip  of  C.  disneyi  extends  posteriorly  under 
the  lower  and  is  broadly  fused  to  it,  but  retains 
a  free  posterior  margin  where  it  has  joined  the 
lower  lip. 

Each  half  of  the  lower  lip  has  two  medullary 
cartilages.  Difficult  to  see  in  some  species  or  spec- 


154 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  6 


FIGURE  1 .  Suckers  of  Camerounian  species  of  Chiloglanis.  a,  C.  reticulatus,  31.3  mm  paratype  (Congo  basin);  b,  C.  polypogon, 
25.0  mm  paratype  (Cross  basin);  c,  C.  sanagaensis,  28.6  mm  paratype  (Sanaga  basin);  d,  C.  batesii,  21.7  mm  (Sanaga  basin);  e, 
C.  disneyi,  23.0  mm  (Cross  basin);/  C.  niger,  27.8  mm  paratype  (Niger  basin). 


imens,  in  others  they  are  large,  naturally  reddish 
or  purplish,  and  readily  observed  with  trans- 
mitted light.  The  distal  portion  of  each  medullar 
cartilage  typically  is  trifurcate;  occasionally  a  car- 
tilage may  have  four  branches  (see  Trewavas 


1 974:360-36 1 ,  fig.  5a).  The  branches  of  the  med- 
ullary cartilages  are  longest  in  species  with  large 
barbels.  In  C.  reticulatus  and  C.  voltae,  which 
have  no  auxiliary  mental  barbels,  the  medullary 
cartilages  are  nevertheless  tri-  or  quadrifurcate. 


ROBERTS:  CAMEROUNIAN  SUCKERMOUTH  CATFISHES 


155 


The  large  central  branch  of  each  medullary  car- 
tilage lies  in  the  middle  of  a  primary  mental  bar- 
bel and  extends  almost  to  its  tip.  The  lateral 
branches,  much  shorter  than  the  central  one,  ex- 
tend along  the  lateral  margins  of  the  portions  of 
the  disc  projecting  posteriorly  between  the  bases 
of  the  barbels.  In  C.  polypogon,  in  which  the 
portions  of  the  disc  between  the  primary  mental 
barbels  are  greatly  prolonged  as  auxiliary  mental 
barbels,  a  lateral  branch  of  the  medullary  carti- 
lage lies  in  the  center  of  each  auxiliary  barbel. 

DENTITION.— The  dentition  of  Chiloglanis,  like 
that  of  most  mochokids,  is  highly  specialized, 
and  its  description  has  given  rise  to  much  con- 
fusion. As  in  Synodontis,  the  mochokid  genus 
with  the  largest  number  of  species,  the  teeth  are 
S-shaped.  An  exception  is  provided  by  C.  voltae, 
in  which  the  teeth  are  almost  straight.  The  mor- 
phology of  the  teeth  has  been  described  and  il- 
lustrated for  many  species  of  Synodontis  (Poll 
1971),  but  this  has  not  been  done  for  Chiloglanis. 
The  premaxillary  (upper  jaw)  teeth  are  of  two 
sizes,  large  and  small.  The  large  teeth  are  ar- 
ranged in  two  oval  patches,  one  on  each  pre- 
maxillary, separated  by  a  toothless  median  gap. 
The  teeth  in  these  patches  form  somewhat  irreg- 
ular rows  that  nevertheless  can  be  counted;  the 
number  varies  from  two  to  eight.  Most  species 
have  three  to  five  rows,  C.  batesii  and  C.  disneyi 
three  or  four,  C.  harbinger  to  seven,  C.  niger  to 
seven  or  eight;  C.  voltae  only  two.  In  addition  to 
the  oval  patches  of  large  teeth,  the  upper  jaw 
bears  numerous  small  teeth  forming  a  shallowly 
V-shaped  series  across  the  angle  of  the  upper  jaw 
where  the  flat  external  surface  bearing  the  oval 
patches  curves  sharply  upwards  toward  the  roof 
of  the  mouth.  Some  of  these  teeth  are  firmly 
implanted  on  the  premaxillary  bones,  while  oth- 
ers lie  more  or  less  loose  in  the  membrane  be- 
tween the  premaxillary  bones.  The  innermost  or 
most  posterior  row  of  teeth  in  this  series  may  be 
somewhat  longer  and  more  slender,  forming  a 
row  of  ciliiform  teeth  (sensu  Trewavas  1974:359 
et  seq.).  This  tooth  row,  while  it  varies  in  prom- 
inence and  degree  of  differentiation  from  the  rest 
of  the  small  premaxillary  teeth,  is  present  in  most 
species  of  Chiloglanis.  According  to  Trewavas 
(1974:358-359)  the  disposition  of  small  pre- 
maxillary teeth  differs  in  C.  batesii  and  C.  mi- 
cropogon:  "in  C.  batesii  the  attached  and  free 
teeth  compose  a  uniform  group,  but  in  C.  mi- 
cropogon  the  median  teeth  are  still  smaller  and 


sections  show  that  the  membrane  in  which  they 
lie  forms  a  minute  pouch  opening  backwards, 
the  teeth  at  the  edge  of  the  pouch  forming  a 
transverse  series  across  the  midline.  Examined 
with  a  dissecting  microscope,  these  latter  appear 
as  a  row  of  minute  ciliiform  teeth  on  a  small 
membranous  curtain."  This  lengthy  quote  pro- 
vides the  best  account  of  a  peculiar  feature  of  the 
dentition  of  Chiloglanis,  and  purportedly  a  basis 
for  distinguishing  C.  batesii  and  C.  micropogon. 
My  observations  suggest  that  a  complete  row  of 
ciliiform  teeth  is  characteristic  of  most  species 
of  Chiloglanis  except  C.  voltae.  In  many  freshly 
preserved  specimens  identified  by  me  as  C. 
batesii,  the  ciliiform  teeth  are  highly  visible,  and 
the  membranous  pouch  described  by  Trewavas 
is  fully  evident.  In  other  specimens,  including 
some  collected  by  Bates  at  Efulen  (type  locality 
of  C.  batesii),  they  appear  at  first  glance  to  be 
absent.  However,  in  at  least  some  of  these  spec- 
imens the  delicate  membrane  bearing  the  cili- 
iform teeth  is  shrunken  and  collapsed  against  the 
roof  of  the  mouth,  so  that  the  teeth  no  longer 
project  normally  downwards.  In  this  connection 
it  may  be  noted  that  a  small  specimen  from  the 
R.  Lobi  in  Cameroun,  originally  identified  as  C. 
batesii  by  Boulenger,  was  re-identified  as  C.  mi- 
cropogon by  Trewavas  (1974:361),  and  that  part 
of  the  type  series  of  C.  batesii  was  collected  in 
tributaries  of  the  Lobi.  It  may  also  be  noted  that 
in  another  species,  C.  brevibarbis,  Trewavas 
( 1 974:364)  reported  ciliiform  teeth  present  in  only 
one  specimen  among  several  she  examined. 

The  mandibular  or  lower  jaw  teeth,  arrayed  in 
a  relatively  small  area  near  the  symphysis  of  the 
lower  jaw  in  all  species  of  Chiloglanis  except  C. 
voltae,  usually  comprise  an  internal  row  of  func- 
tional teeth  and  an  external  row  (rarely  two  or 
three  external  rows)  of  replacement  teeth  (Fig. 
2).  In  most  species  the  teeth  in  the  replacement 
row  erupt  nearly  simultaneously,  so  that  de- 
pending upon  the  stage  in  the  replacement  cycle, 
teeth  of  the  lower  jaw  are  either  in  only  a  single 
row  (the  functional  row)  or  in  two  rows  (func- 
tional and  replacement).  This  simple  situation 
may  be  complicated  when  the  row  of  functional 
teeth  has  been  shed  or  only  partially  shed,  and 
only  part  of  the  replacement  tooth  row  has  moved 
into  functional  position.  In  some  species,  espe- 
cially those  with  numerous  teeth  such  as  C.  dis- 
neyi, C.  harbinger,  and  C.  niger,  the  replacement 
process  tends  to  be  very  regular,  and  it  seems 


156 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  6 


FIGURE  2.  Mandibular  teeth  of  Camerounian  species  of  Chiloglanis.  a,  C.  reticulatus,  31.3  mm  paratype;  b,  C.  polypogon, 
25.0  mm  paratype;  c,  C.  sanagaemis,  28.6  mm  paratype;  d,  C.  batesii,  21.7  mm;  e,  C.  disneyi,  23.0  mm;/  C.  niger,  27.8  mm 
paratype. 


that  an  old  row  of  functional  teeth  is  shed  nearly  replaced  by  corresponding  teeth  from  the  re- 

simultaneously  or  as  a  unit,  to  be  replaced  at  placement  row  before  the  more  lateral  functional 

once  by  a  complete  row  of  replacement  teeth.  In  teeth  are  shed.  The  sequence  of  tooth  replace- 

others,  e.g.,  C.  polypogon  and  C.  congicus,  it  seems  ment  has  not  been  investigated  in  detail, 

that  those  near  the  symphysis  fall  out  first,  to  be  In  Chiloglanis  generally,  allowing  for  individ- 


ROBERTS:  CAMEROUNIAN  SUCKERMOUTH  CATFISHES 


157 


ual  variation  due  to  loss  or  injury  of  teeth  and 
irregularities  in  replacement,  there  are  basically 
two  rows  of  mandibular  teeth,  an  internal  func- 
tional row  and  an  external  replacement  row;  the 
number  of  teeth  of  the  left  and  right  side  of  the 
lower  jaw  is  usually  equal.  The  number  of  teeth 
in  the  replacement  row  usually  corresponds  to 
that  in  the  functional  row.  There  is  some  increase 
in  number  of  teeth  with  growth,  but  throughout 
the  size  range  of  specimens  examined  for  this 
report,  the  ontogenetic  increase  does  not  amount 
to  more  than  four  to  six  functional  mandibular 
teeth  even  in  species  with  the  highest  tooth  counts. 
In  numerous  specimens  of  C.  batesii,  the  func- 
tional tooth  row  has  6  teeth  on  the  left  side  and 
6  on  the  right,  and  the  replacement  tooth  row 
has  the  same  count.  Thus,  the  basic  number  of 
mandibular  teeth  in  this  species  usually  is  6  +  6. 
Although  some  specimens  of  C.  batesii  have  few- 
er than  6  +  6  (usually  5  +  5  or  4+4)  I  have  not 
observed  any  with  more  than  6  +  6. 

Species  of  Chiloglanis  previously  reported  to 
have  numerous  mandibular  teeth  are  C.  congi- 
cus,  C.  disneyi,  C.  lamottei  Daget,  1948,  C.  mar- 
lieri  Poll,  1952,  and  C.  polyodon.  These  were 
discussed  in  connection  with  the  description  of 
C.  disneyi  by  Trewavas  (1974:365-367).  Chi- 
loglanis congicus  has  mandibular  teeth  tightly 
bunched  together  ("en  bouquet").  The  teeth  near 
the  symphysis  (middle  of  the  bouquet)  have  their 
crowns  projecting  more  than  those  at  the  sides. 
Trewavas  gave  the  mandibular  tooth  count  as  1 8 
but  did  not  explain  how  the  count  was  made.  In 
four  specimens  of  C.  congicus  I  find  the  total 
number  of  mandibular  teeth  is  12,  16,  16,  and 
1 9.  In  the  first  three  specimens  there  are  basically 
two  rows  of  teeth  (one  functional,  one  replace- 
ment) with  4+4  or  5  +  5  teeth  in  each  row.  The 
fourth  specimen  has  two  or  three  rows  of  teeth, 
one  functional  and  one  or  two  replacement,  with 
the  rows  deranged  and  some  teeth  missing.  On 
each  side  of  any  of  the  rows  the  maximum  num- 
ber of  teeth  is  4.  Chiloglanis  lamottei,  known 
only  from  the  holotype  from  the  R.  Cavally,  has 
a  single  row  of  1 4  mandibular  teeth  (Daget  1 948), 
presumably  7  +  7.  Chiloglanis  polyodon,  known 
only  from  the  holotype  from  the  R.  Bagbwe,  was 
originally  described  as  having  mandibular  teeth 
"arranged  in  two  parallel  rows,  each  consisting 
of  1 5  or  16  teeth,  those  of  the  inner  row  larger 
and  more  curved  than  those  of  the  outer  (Nor- 
man 1932:1 84).  Some  40  years  later  the  holotype 
was  re-examined  and  found  to  have  1 5  inner  and 
14  outer  mandibular  teeth  bunched  together" 


TABLE  1 .    Mandibular  tooth  counts  in  species  of  Chilogla- 
nis. 


C.  polypogon 
C.  reticulatus 
C.  congicus 
C.  sanagaensis 
C  benuensis 
C.  cameronensis 
C.  batesii 
C.  lamottei 
C.  polyodon 
C.  disneyi 
C.  niger 
C.  marlieri 
C.  harbinger 


3  +  3  or  4+4 

3  +  3  to  5  +  5? 

4+4  (occasionally  5  +  5?) 

4+4  or  5  +  5  (rarely  6+6) 

3  +  3  to  6+6 

4+4  to  8  +  8 

4+4  to  6  +  6 

7  +  7 

7  +  7  to  8  +  8 

8  +  8  to  10+10 
8+8  to  10+10 
13+ 13.  (or  14+14?) 
13+13to  15+15 


(Trewavas  1974:367).  Both  accounts  are  cited 
here  because  of  the  possibility  that  in  the  inter- 
vening period  a  tooth  was  removed  from  each 
row,  possibly  by  Norman  himself.  The  basic 
mandibular  tooth  count  for  C.  polyodon  presum- 
ably is  7  +  7  or  8  +  8.  Chiloglanis  marlieri  is  known 
only  from  the  holotype  collected  in  the  R.  Nda- 
kirwa,  a  tributary  of  the  R.  Lowa,  west  of  Lake 
Kivu.  As  originally  illustrated  this  specimen  has 
one  complete  mandibular  tooth  row  of  1 5  + 1 5 
teeth,  plus  an  incomplete  internal  row  with  7 
teeth  on  the  extreme  left  side,  8  on  the  extreme 
right,  and  none  in  the  large  intervening  space 
(Poll  1952:227,  fig.  2).  Trewavas  (1974)  re-ex- 
amined this  specimen  and  reported  the  mandib- 
ular teeth  as  "one  row  of  26,  a  widely  interrupted 
row  of  6  +  7  and  2  incompletely  erupted  teeth  at 
the  midline."  Thus  the  basic  tooth  count  pre- 
sumably is  13  +  13  or  14+14.  Mandibular  tooth 
counts  for  various  species  of  Chiloglanis  are  pre- 
sented in  Table  1 . 

TASTE  BUDS.— Taste  buds  (Figs.  3,  4)  were  re- 
ported (not  figured)  on  the  papillae  or  "bosses  of 
thickened  epithelium"  on  the  oral  disc  of  C. 
batesii  and  C.  micropogon  (regarded  here  as  a 
synonym  of  C.  batesii)  by  Trewavas  (1974).  I 
have  observed  taste  buds  with  scanning  electron 
microscopy  in  one  or  two  specimens  of  five 
species  from  Cameroun,  and  failed  to  find  them 
in  specimens  of  a  sixth  (Figs.  3,  4).  The  following 
remarks  are  presented  as  generalizations  on  the 
taste  buds  of  these  six  species,  but  the  limited 
number  of  specimens  on  which  they  are  based 
should  be  kept  in  mind.  The  buds  are  largest  and 
most  abundant  in  C.  polypogon  and  C.  reticu- 
latus. In  these  species  they  are  distributed  fairly 
uniformly  on  the  papillae  and  barbels  of  the  low- 
er lip.  In  C.  reticulatus  the  buds  are  about  50  /*m 
diameter,  and  the  sensory  field  (with  numerous 


158 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  6 


W^i -•      V  '•".    ' 

FIGURE  3.    Taste  buds  in  Camerounian  species  of  Chiloglanis.  a,  b,  C.  reticulatus,  31.3  mm;  c,  d,  C.  polypogon,  25.0  mm;  e, 
f,  C.  sanagaensis,  28.6  mm.  Horizontal  field  widths:  a,  c,  e  (top  row)  =  450  nm\  b,  d,  /(bottom  row)  =  45  /*m. 


minute  pores  for  the  sensillae)  of  each  bud  is 
about  1 8  nm  diameter.  The  sensory  field  projects 
from  the  summit  or  apex  of  the  taste  bud  (Fig. 
36).  The  density  of  the  taste  buds,  calculated 
from  Figure  3b,  is  about  240  per  square  mm;  this 
is  about  the  same  as  that  reported  for  taste  buds 
in  the  mouth  and  pharynx  of  the  common  Eur- 
asian loach  Cobitis  taenia  Linnaeus,  1758  by  Ja- 
kubowski  (1983:850).  The  taste  buds  and  sen- 
sory fields  of  C.  polypogon  are  of  about  the  same 
size,  but  the  sensory  field  tends  to  be  sunken  into 
the  middle  of  the  bud  rather  than  elevated  (Fig. 
3d).  A  similar  condition  is  reported  for  the  ex- 
ternal taste  buds  of  Cobitis  taenia  by  Jakubowski 
(1983:851,  fig.  ID)  but  has  not  been  observed  in 
other  Chiloglanis.  Chiloglanis  disneyi  has  rela- 
tively large,  numerous  buds  on  the  lower  lip.  The 
buds  are  up  to  22  ^m  diameter  and  the  sensory 
fields  to  7  or  8  /j.m  diameter  (Fig.  4d).  Slightly 
larger  buds  occur  in  C.  batesii  (Fig.  46).  Extensive 
viewing  of  the  sucker  of  C.  niger  with  SEM  re- 


vealed no  large  or  well  formed  taste  buds.  Only 
a  few  small  structures  tentatively  identified  as 
taste  buds  were  observed  on  papillae  on  the  lower 
lip  (Fig.  4f).  The  largest  bud  is  1 2  nm  diameter 
with  a  sensory  field  of  about  8  /urn  diameter.  The 
sensory  field  has  fewer  pores  than  observed  in 
other  species  of  Chiloglanis.  Taste  buds  appar- 
ently are  absent  in  C.  sanagaensis.  I  could  not 
find  any  on  the  sucker  of  a  28.6  mm  specimen. 
Although  the  oral  papillae  of  this  specimen  have 
raised  areas  similar  to  those  where  taste  buds 
occur  in  other  species,  their  summit  is  occupied 
only  by  densely  convoluted  microvillar  ridges, 
without  any  sensillar  pores.  The  structures  su- 
perficially most  similar  to  taste  buds  observed 
in  C.  sanagaensis  are  illustrated  in  Figure  3e,  f). 
The  species  of  Chiloglanis  with  the  largest  and 
most  numerous  taste  buds,  C.  polypogon  and  C. 
reticulatus,  have  suckers  with  large  barbels  and 
large  papillae.  Taste  buds  of  intermediate  size 
and  abundance  occur  in  C.  batesii  and  C.  disneyi, 


ROBERTS:  CAMEROUNIAN  SUCKERMOUTH  CATFISHES 


159 


FIGURE  4.    Taste  buds  in  Camerounian  species  of  Chiloglanis  (cont.).  a,  b,  C.  batesii,  21.7  mm;  c,  d,  C.  disneyi,  23.0  mm;  e, 
f,  C.  niger,  27.8  mm.  Horizontal  field  widths:  a,  c,  e  (top  row)  =  450  nm;  b,  ^./(bottom  row)  =  45  nm. 


species  with  greatly  reduced  barbels  and  rela- 
tively small  papillae.  Chiloglanis  niger,  with  a 
very  large  sucker,  minute  barbels,  and  moder- 
ately large  papillae,  has  taste  buds  few  in  number 
and  poorly  developed.  Chiloglanis  sanagaensis, 
with  moderately  elongate  barbels  and  numerous 
well-developed  papillae,  apparently  lacks  taste 
buds.  Finally,  it  should  be  noted  that  taste  buds 
are  not  limited  to  the  sucker  in  Chiloglanis.  Rel- 
atively large  buds  occur  on  the  dorsal  surface  of 
the  head  in  C.  reticulatus  (Fig.  5e,  f). 

TUBERCLES.— Tubercles,  composed  of  keratin- 
ized  and  sometimes  also  greatly  enlarged  super- 
ficial epidermal  cells,  occur  on  the  head  and  body 
of  all  or  almost  all  species  of  Chiloglanis.  In  con- 
trast to  the  situation  in  most  cyprinoid  fishes, 
the  tuberculation  apparently  is  not  sexually  di- 
morphic. Among  Camerounian  species  it  is 
particularly  well  developed  in  C.  reticulatus,  C. 
polypogon,  and  C.  batesii.  Although  less  evident 
because  they  are  fewer  and  further  apart,  well-de- 


veloped tubercles  are  present  also  in  C.  disneyi, 
C.  niger,  and  C.  sanagaensis.  The  Camerounian 
(and  Zairean)  species  with  the  most  striking  tu- 
berculation is  C.  reticulatus,  in  which  large  spec- 
imens of  both  sexes  have  long  tuberculated  ridges 
forming  a  network  or  reticulum  covering  the  dor- 
sal and  lateral  surfaces  of  the  head  and  much  of 
the  body.  The  ridges  are  especially  evident  on 
the  head.  They  are  up  to  10  cells  wide;  the  crest 
of  a  ridge  is  regularly  tuberculate,  each  tubercle 
consisting  of  about  a  dozen  enlarged  epidermal 
cells  (Fig.  5d).  The  surface  of  the  non-tuberculate 
polygonal  cells  forming  the  ridge  is  not  smooth, 
the  center  of  each  cell  being  occupied  by  a  low- 
lying  roughened  protuberance  evidently  equiv- 
alent to  an  unculus  (for  review  of  unculi  see  Rob- 
erts 1982).  The  epidermal  surfaces  between  the 
ridges,  in  contrast,  consist  of  slightly  smaller,  non- 
unculiferous  (and  therefore  relatively  smooth- 
surfaced)  polygonal  cells,  widely  scattered  taste 
buds,  and  even  more  widely  scattered  pores.  The 


160 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  6 


FIGURE  5.  Epidermal  features  on  head  of  Chiloglanis  reticulatus.  a,  dorsal  view  of  head  of  31.3  mm  paratype,  with  elongate 
tuberculate  ridges  well  developed  (but  not  connecting  up  to  form  a  network  as  in  larger  specimens);  b-e,  dorsal  surface  of  head 
posterior  to  eye  of  41.6  mm  holotype,  with  well-developed  network  of  tuberculated  ridges  and  isolated  taste  buds;/  enlarged 
view  of  taste  bud  in  e.  Horizontal  field  widths:  a,  7.5  mm;  b,  1  mm;  c,  750  ^m;  d,  360  ^m;  e,  1 10  /j.m;  and/  1 1  /am. 


pores,  two  of  which  are  seen  in  Figure  5b,  c,  lie 
in  shallow  depressions  about  50  nm  diameter 
and  are  themselves  about  20  pm  diameter.  The 
taste  buds,  elevated  above  the  general  epidermal 
surface,  are  also  about  20  /um  diameter  (Fig.  5/). 
The  gross  aspects  of  all  of  these  features  (but  not 
their  cellular  nature)  can  be  observed  under  the 
dissecting  microscope.  Although  no  other  species 
of  Chiloglanis  observed  by  me  has  tuberculated 
ridges  forming  such  an  extensive  and  well-de- 
fined network  as  C.  reticulatus,  more  or  less  elon- 
gate tuberculated  ridges  are  present  on  some  larg- 
er specimens  of  C.  batesii,  C.  occidentalis,  and 
C.  polypogon. 

Numerous  round  or  oval  tubercles,  comprising 
some  60  to  80  cells  and  about  40  MHI  diameter, 
occur  on  the  dorsal  surface  of  the  head  of  C. 
polypogon  (Fig.  6a,  b).  Similar  tubercles  were 
observed  in  C.  batesii  from  Ngudi  (Fig.  6e).  Some 


larger  C.  batesii  have  tuberculated  ridges,  but 
these  were  not  observed  with  SEM.  In  C.  poly- 
pogon the  cells  comprising  a  tubercle  are  uncu- 
liferous,  whereas  in  the  specimens  of  C.  batesii 
observed  they  are  not.  In  a  28.6  mm  C.  sana- 
gaensis,  prominent  tubercles  were  observed  on 
the  dorsal  surface  of  the  snout.  These  have  a 
round  base  30  to  35  nm  diameter,  and  a  rounded 
or  pointed  apex  to  20  to  30  nm  high.  They  appear 
to  have  a  perimeteral  dehision  ring  (Fig.  6d).  The 
head  of  this  specimen  exhibits  many  shallow  cra- 
ter-like depressions,  tentatively  identified  as  scars 
from  tubercles  that  had  been  shed.  Of  the  species 
of  Chiloglanis  examined  with  SEM,  the  weakest 
tuberculation  is  in  C.  niger.  Tubercles  on  its  head 
are  widely  separated,  up  to  40  /urn  diameter,  but 
not  elevated,  consisting  of  only  a  single  layer  of 
thin,  apparently  individually  dehiscent  polygo- 
nal epidermal  cells,  with  irregular  projections 


ROBERTS:  CAMEROUNIAN  SUCKERMOUTH  CATFISHES 


161 


FIGURE  6.  Tubercles  on  head  of  Camerounian  species  of  Chiloglanis.  a,  b,  C.  polypogon,  25.0  mm;  c,  d,  C.  sanagaensis,  28.6 
mm;  e,  C.  batesii,  25.7  mm  (R.  Ngudi);/  C.  niger,  27.8  mm.  Horizontal  field  widths:  a,  250  nm;  b,  c,  115  nm;  d,  105  nm;  e, 
450  /urn;  and/  105  nm. 


(unculi?)  arising  from  their  centers  or  cell  mar- 
gins (Fig.  6/). 

FIN  RAYS.— Counts  of  fin  rays  are  of  limited 
usefulness  in  distinguishing  species  of  Chilogla- 
nis. Most  useful  are  dorsal  and  pectoral  fin  ray 
counts.  Most  Camerounian  species  usually  have 
5  soft  dorsal  fin  rays,  but  C.  batesii,  C.  niger,  and 
C.  voltae  usually  have  6.  However,  occasional 
specimens  of  the  former  species  have  only  5  and 
occasional  specimens  of  the  latter  species  have 
6.  Chiloglanis  harbinger  has  4  or  5.  Many  species 
have  7  or  8  soft  pectoral  fin  rays;  no  species  has 
been  observed  with  fewer  than  7.  Chiloglanis 
disneyi  has  8  or  9,  C.  harbinger  and  C.  niger  9 
or  10. 

In  many  specimens  the  number  of  anal  fin  rays 
is  difficult  to  determine.  I  have  not  obtained  ad- 
equate data  to  evaluate  anal  fin  ray  counts  for 
distinguishing  species.  Camerounian  species  have 
a  total  of  7  to  12  anal  fin  rays,  but  much  of  the 
variation  is  in  the  number  of  anterior  simple 


rays,  which  range  from  2  to  5.  Especially  in  spec- 
imens with  more  simple  rays,  the  anteriormost 
ones  tend  to  be  very  small,  sometimes  impos- 
sible to  count  accurately  in  whole  specimens  or 
even  in  radiographs.  The  number  of  branched 
anal  fin  rays  is  usually  7  (range  5  to  8).  All  species 
of  Chiloglanis  examined  have  7  pelvic  fin  rays 
and  8/9  principal  caudal  fin  rays. 

COLORATION.— The  basic  elements  of  color 
pattern— lightly  and  darkly  pigmented  areas— 
are  remarkably  constant  in  many  species  of  Chi- 
loglanis. This  can  be  appreciated  readily  by  com- 
paring the  photographs  of  various  Camerounian 
species  in  this  paper.  The  main  feature  of  the 
color  pattern  is  three  broad  dark  vertical  bands 
on  the  dorsolateral  body  surface.  Although  the 
outline  of  the  bands  is  irregular,  closer  exami- 
nation reveals  the  same  or  similar  irregularities 
in  the  bands  of  quite  different  species.  This  is 
especially  true  for  the  very  irregularly  shaped 
second  band,  lying  between  the  adipose  and  anal 


162 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  6 


fins.  This  band  has  an  anteroventral  projection 
where  its  ventral  portion  is  more  or  less  exten- 
sively broken  or  invaded  by  a  depigmented  area 
arising  between  the  base  of  the  pelvic  and  anal 
fins  and  extending  dorsoposteriorly.  The  posi- 
tion and  shape  of  this  pale  area,  and  the  resulting 
configuration  of  the  dark  vertical  band  between 
the  adipose  and  anal  fins,  is  very  evident  in  the 
photos  of  C.  reticulata,  C.  polypogon,  C.  sana- 
gaensis,  and  C.  disneyi.  It  is  also  visible  in  the 
photos  of  C.  niger  and  C.  batesii,  although  in 
these  specimens  the  basic  pattern  is  obscured  by 
their  overall  duskiness.  The  barred  color  pattern 
on  the  caudal  peduncle  and  caudal  fin  is  similar 
in  many  species.  Some  species,  including  C.  poly- 
pogon and  C.  reticulatus,  have  small,  round,  pale 
spots  scattered  on  the  dark  vertical  bands  of  the 
body.  At  first  glance  these  spots  appear  irregular, 
but  their  number  and  position  is  similar  in  var- 
ious species.  The  peculiar  conservatism  of  the 
irregularly  banded  or  barred  color  pattern  of 
Chiloglanis  is  paralleled  by  a  number  of  other 
groups  of  catfishes,  notably  the  Southeast  Asian 
bagrid  genus  Leiocassis. 

Most  species  of  Chiloglanis  are  dull  and  cryp- 
tically or  disruptively  colored  in  life.  Live  Chi- 
loglanis niger  are  uniformly  dark,  almost  black. 
This  is  perhaps  also  true  in  C.  harbinger  for  which 
coloration  has  been  observed  only  in  preserved 
specimens.  The  most  brilliantly  colored  species 
probably  is  C.  disneyi,  which  has  well  defined 
bright  yellow  and  black  bands.  Sexual  dichro- 
matism  is  unknown. 

A  few  species  have  distinctive  color  features. 
Nearly  all  species  have  dark  vertical  bars  on  the 
dorsolateral  part  of  the  body  that  end  more  or 
less  abruptly  at  about  the  level  of  the  paired  fins. 
In  C.  harbinger  the  large  anterior  bar  extends 
partially  onto  the  abdomen,  the  melanophores 
becoming  fewer  and  wider  apart  towards  the  ven- 
tral midline.  In  C.  cameronensis  this  bar  often 
continues  without  diminished  intensity  right 
across  the  abdomen,  which  is  therefore  black. 
This  is  especially  so  in  samples  of  this  species 
from  the  Ntem  and  Ogooue  basins.  In  most  spec- 
imens of  C.  niloticus  and  its  close  relative  (or 
conspecific)  C.  occidentalis,  the  upper  and  lower 
caudal  fin  lobes  each  bear  a  dark  horizontal  stripe. 
Although  C.  occidentalis  is  widely  distributed  in 
western  Africa,  including  the  Niger  basin,  it  has 
not  been  found  in  Cameroun.  None  of  the  species 
of  Chiloglanis  known  from  Cameroun  have  cau- 


dal fin  lobes  with  horizontal  stripes.  The  dorsal 
surface  of  the  snout  of  C.  batesii  has  a  triradiate 
pattern.  This  consists  of  a  dark  or  dusky  median 
mark  at  the  snout  tip  separated  by  pale  inter- 
spaces from  an  anterolateral  dark  or  dusky  mark 
on  each  side  of  the  snout.  This  pattern,  observed 
in  specimens  from  throughout  the  extensive  range 
of  C.  batesii,  is  absent  or  poorly  developed  in 
other  species  of  Chiloglanis.  Chiloglanis  reticu- 
latus, occurring  sympatrically  with  C.  batesii  in 
the  Congo  basin,  often  has  a  biradiate  pattern  on 
the  snout.  This  is  due  to  a  dark  or  dusky  an- 
terolateral mark  on  each  side  of  the  snout  as  in 
C.  batesii,  but  the  large  intervening  area  includ- 
ing the  snout  tip  is  pale. 

SEXUAL  DIMORPHISM.  — Testes  of  maturing 
Chiloglanis  have  numerous  lobes.  In  extremely 
ripe  males  the  lobes  may  be  so  enlarged  as  to 
cause  abdominal  expansion  comparable  to  that 
of  gravid  females.  In  many  species  males  have 
an  elongate  genital  papilla.  Secondary  sexual  di- 
morphism occurs  in  several  species.  Often  this 
involves  enlargement  of  the  anal  fin,  and  some- 
times also  enlargement  and  modification  of  the 
shape  of  the  caudal  fin.  In  males  of  C.  macrop- 
terus  Poll  and  Stewart,  1975,  from  the  R.  Luongo 
in  Zambia  (Congo  basin),  the  anal  and  caudal 
fins  are  both  enlarged;  the  caudal  fin,  moderately 
forked  in  females,  is  triangular  in  large  males, 
with  the  distal  part  of  the  triangle  formed  by 
elongation  of  the  middle  caudal  fin  rays.  Very 
similar  sexual  dimorphism  of  the  anal  and  caudal 
fins  occurs  in  a  West  African  species,  C.  normani 
from  the  R.  Cavally,  Cote  d'lvoire.  The  only 
Camerounian  species  in  which  sexual  dimor- 
phism of  the  caudal  fin  has  been  observed  is  C. 
batesii.  In  this  species  the  caudal  fin  is  very  deep- 
ly forked,  with  the  upper  lobe  considerably  en- 
larged and  pointed  or  even  filamentous  in  mature 
males.  Such  modification  of  the  upper  caudal  fin 
lobe  of  male  Chiloglanis  has  been  reported  pre- 
viously only  in  C.  micropogon,  here  considered 
conspecific  with  C.  batesii.  In  addition  to  the 
enlarged  and  pointed  upper  caudal  fin  lobe,  ma- 
ture males  of  C.  batesii  have  lobate  testes,  an 
elongate  genital  papillae,  enlarged  anal  fin,  and 
one  other  remarkable  feature. 

Radiographs  of  Chiloglanis  from  Cameroun 
and  other  countries  reveals  several  species  with 
a  striking  sexual  dimorphism  hitherto  unre- 
ported.  In  these  species  males  have  the  humeral 
process  of  the  pectoral  fin  girdle  more  or  less 


ROBERTS:  CAMEROUNIAN  SUCKERMOUTH  CATFISHES 


163 


TABLE  2.    Vertebral  counts  in  species  of  the  suckermouth 
mochokid  genera  Chiloglanis,  Atopochilus,  and  Euchilichthys. 


TABLE  2.    Continued. 


Abdom- 

Abdom- 

inal 

Caudal 

Total 

inal        Caudal                Total 

C.  batesii 

A.  sp.  undet.  a 

Mungo 

15 

16-17 

3  1(2),  32(2) 

Stanley  Pool             19             16          35(1) 

Cross 

15 

16-17 

31(1),  32(1) 

A.  sp.  undet.  b 

Ja 

14 

17-18 

3  1(3),  32(1) 

Luala 

14-15 

16-17 

30(2),  3  1(7) 

Lufu                         21              16          37(1) 

Lufu 

14-15 

15-17 

30(4),  3  1(5),  32(1) 

E.  dybowskii 

Inkisi 

14-15 

16-17 

3  1(5),  32(5) 

Oubanghi                 19             13          32(1) 

Kwango 

15-16 

16 

3  1(1),  32(5) 

Mbomou               18-20        12-14       31(2),  32(2) 

Mbomou 

14-15 

16-17 

31(2) 

Kivu 

15 

17-18 

32(1),  33(1) 

E.  royauxi 

C.  brevibarbis 

Kinsuka                23-24        17-18       40(1),  42(1) 

Athi 

15-17 

18-20 

34(3),  35(10) 

C.  cameronensis 

greatly  enlarged.  The  enlargement  is  manifested 

Lokunje 

15 

15-16 

30(1),  31(1) 

externally  as  a  large,  skin-covered  convex  boss 

Mvila 
Ntem 

14-15 
15 

16-17 
16 

31(4) 
31(2) 

just  above  the  pectoral  fin.  It  is  most  pronounced 

Diala 

15 

16-17 

•*  *  v**/ 
3  1(2),  32(1) 

in  C.  batesii,  but  also  occurs  in  C.  occidentalis 

Ivindo 

14-15 

16 

30(2),  3  1(2) 

and  C.  polypogon.  It  apparently  does  not  occur 

C.  congicus 

in  C.  brevibarbis,  C.  cameronensis,  C.  congicus, 

Congo 

16-17 

15-16 

32(3),  33(1) 

C.  disneyi,  C.  niger,  C.  reticulatus,  C.  sanagaen- 

C. deckenii 

sis,  or  C.  voltae,  for  all  of  which  radiographs  as 

Kerio 

16-18 

15-17 

33(4)  34(2) 

well  as  specimens  have  been  examined. 

"»*"*\r/J    •/~\*  V 

VERTEBRAL  COUNTS.—  Vertebral  counts,  until 

C.  disneyi 

recently  seldom  employed  in  systematic  studies 

Mungo 
Cross 

15 
15-16 

15-16 

15-17 

30(1),  31(1) 
31(6)  32(3) 

of  catfishes,  can  be  very  useful  in  distinguishing 

•*  *  \U/J    */A*\-// 

species.  Smaller  or  less  elongate  species  of  cat- 

C. harbinger 

fishes  often  have  fewer  vertebrae  than  their  larger 

Lokunje 

15-16 

15-16 

30(1),  3  1(9) 

or  more  elongate  close  relatives.  Thus,  the  sisorid 

C.  niger 

catfish  Bagarius  bagarius  Hamilton-Buchanan, 

Menchum 

15 

16 

3KD 

1822,  is  a  small  species  superficially  similar  to 

C.  occidentalis 

the  very  large  species  Bagarius  yarrelli  Sykes, 

Ankwao 

14-15 

16-18 

3  1(1),  32(3) 

1841,  but  differs  from  it  in  having  only  38-42 

Pra 

14-15 

16-18 

3  1(4),  32(2) 

(mode  39)  vertebrae  instead  of  40-45  (mode  43) 

C.  polyodon 

(Roberts  1  983:437).  Vertebral  counts  are  of  com- 

Bagbwe 

15 

16 

3KD 

parable  utility  in  distinguishing  species  of  the 

C.  polypogon 

bagrid  genera  Leiocassis  and  Mystus  and  of  the 

Cross 

14-15 

16-18 

30(1),  32(10),  32(8) 

pangasiid  or  schilbeid  genus  Pangasius  (Roberts, 

C.  reticulatus 

in  press).  Vertebral  counts  are  determined  readi- 

Congo 

14-15 

16-19 

3  1(5),  32(4),  33(5) 

ly  from  radiographs  of  Chiloglanis.  In  some  cat- 

fish groups,  due  to  variation  in  the  length  and 

C.  sanagaensis 
Sanaga 

14-16 

16-17 

31(1),  32(4),  33(1) 

partial  or  total  fusion  of  the  anteriormost  (We- 
berian)  vertebral  centra  with  each  other,  and  oc- 

C. sardinhai 

casional  fusion  of  additional  more-posterior  ver- 

Caimbambo 

14-15 

15-17 

30(1),  3  1(4) 

tebrae  with  them,  it  is  difficult  to  obtain  reliable 

C.  voltae 

counts  of  abdominal  vertebrae.  In  Chiloglanis, 

Oti 

13-14 

16-17 

29(2),  30(5),  31(2) 

however,  the  four  Weberian  vertebrae  have  cen- 

Lakdo 

13-14 

16-18 

30(3),  3  1(1) 

tra  of  about  the  same  size  that  remain  separate; 

A.  savorgnani 

thus,  they  often  may  be  counted  directly  from 

Ntem 

17-18 

14-15 

32(5),  33(3) 

radiographs.  Furthermore,  the  fourth  vertebra  is 

164 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  6 


readily  identifiable  by  the  large  ossa  suspensoria 
and  the  fifth  vertebra  by  a  large  but  slender  pair 
of  normally  shaped  ribs.  In  practice  I  commence 
a  count  of  the  vertebral  column  with  the  fifth 
vertebra  and  proceed  backwards.  Abdominal 
vertebrae  are  defined  as  all  those  lying  anterior 
to  the  anal  fin  and  its  pterygiophores.  If  the  an- 
teriormost  anal  fin  pterygiophore  approximates 
or  even  contacts  the  hemal  spine  of  a  vertebra 
but  fails  to  lie  in  front  of  it,  the  vertebra  is  count- 
ed as  abdominal.  The  hypural  complex  is  count- 
ed as  one  vertebra.  This  method  gives  repeatable 
counts  for  most  catfishes  and  many  other  fish 
groups.  Vertebral  counts  from  a  radiographic 
survey  of  Chiloglanis  and  other  mochokid  genera 
with  sucker  mouths  are  presented  in  Table  2. 

The  counts  are  of  little  use  in  distinguishing 
most  species  of  Chiloglanis,  all  of  which  are  more 
or  less  short  bodied  and  none  very  large.  Many 
species  have  around  30  or  31  to  33  vertebrae. 
The  smallest  species,  C.  voltae,  has  29-3 1 .  This 
is  close  to  the  lowest  number  of  vertebrae  found 
in  other  groups  of  catfishes  and  in  ostariophysan 
fishes  generally.  Somewhat  higher  counts  (33-35) 
occur  in  species  from  East  Africa  including  C. 
deckeniiand  C.  brevibarbis.  Of  particular  interest 
are  the  vertebral  counts  of  the  genera  Atopochilus 
and  Euchilichthys,  in  which  the  species  generally 
are  far  larger  (to  600  mm)  than  Chiloglanis  and 
have  up  to  42  vertebrae.  In  addition  there  is  a 
shift  in  the  ratio  of  abdominal  to  caudal  verte- 
brae. In  Chiloglanis  the  ratio  is  close  to  unity, 
n:n  or  n:n+ 1  or  n+2,  while  in  Atopochilus  and 
Euchilichthys  the  abdominal  vertebrae  are  more 
numerous  than  the  caudal  vertebrae,  so  the  ratio 
is  n:n— 2  to  n— 7.  This  increase  in  abdominal 
vertebrae  appears  related  to  an  elongate  abdo- 
men accommodating  an  extremely  convoluted 
gut. 

SYSTEMATIC  ACCOUNT 
Chiloglanis  Peters,  1868 

Chiloglanis  Peters,  1868:  599  (type  species  Chi- 
loglanis deckenii  Peters,  1868,  by  monotypy). 

Chiloglanis  (gender  masculine)  differs  from  all 
other  genera  of  the  endemic  African  catfish  fam- 
ily Mochokidae,  except  Atopochilus  Sauvage, 
1878  and  Euchilichthys  Boulenger,  1 900,  in  hav- 
ing upper  and  lower  lips  greatly  expanded  and 
united  to  form  a  sucker.  Some  species  of  Syno- 
dontis  have  large  lips  superficially  resembling 


those  of  Chiloglanis  but  are  readily  distinguish- 
able because  they  have  mental  barbels  free  from 
sucker  instead  of  incorporated  into  it.  In  Ato- 
pochilus and  Euchilichthys  the  mandibular  teeth 
are  very  numerous  and  spread  across  the  entire 
free  margin  of  the  lower  jaw,  whereas  in  all  species 
of  Chiloglanis  except  C.  voltae,  the  mandibular 
teeth  are  relatively  few  and  restricted  to  a  rela- 
tively small  area  near  the  symphysis  of  the  lower 
jaw.  In  Chiloglanis  the  pectoral  fin  spine  is  in- 
variably smooth,  never  serrate,  while  in  Atopo- 
chilus and  Euchilichthys  it  bears  few  to  many 
serrae.  Euchilichthys  and  Atopochilus  are  very 
poorly  represented  in  Cameroun.  Euchilichthys 
dybowskii  is  present  in  the  R.  Ngoko  (lower  R. 
Ja,  Congo  basin)  and  Atopochilus  savorgnani  in 
the  R.  Ntem. 

KEY  TO  CHILOGLANIS  OF  CAMEROUN 

1A.  Mental  barbels  moderately  to  very  long 
(Fig.  1  a-c) 2 

IB.  Mental  barbels  short,  sometimes  absent 
(Fig.  \d-  /) 7 

2 A.  No  auxiliary  mental  barbels  (Fig.  la) 3 

2B.  From  2  to  5  well  developed  auxiliary 
mental  barbels  on  each  side  of  lower  lip 
(Fig.  1  b) 4 

3A.  Sucker  with  well  developed  papillae;  large 
teeth  on  premaxilla  in  several  rows;  man- 
dibular teeth  well  developed;  adults  to 
over  40  mm  (Congo  basin)  ...  C.  reticulatus 

3B.  Sucker  without  papillae;  large  teeth  on 
premaxilla  in  two  rows;  mandibular  teeth 
in  two  rows;  adults  to  27  mm  (Niger  ba- 
sin)   C.  voltae 

4A.  Mental  barbels  very  large;  sucker  trans- 
versely oval,  middle  of  lower  lip  with 
greatly  enlarged  papillae  (Fig.  la,  b) 5 

4B.  Mental  barbels  moderately  large;  sucker 
round,  or  longer  than  broad,  entire  sur- 
face covered  with  papillae  of  similar  size 
(Fig.  1  c) 6 

5A.  Accessory  mental  barbels  no  more  than 
3  on  each  side  of  lower  lip;  mandibular 

teeth  4+4  to  7  +  7  (Niger  basin) 

C.  benuensis 

5B.  Accessory  mental  barbels  from  3  to  5  on 
each  side  of  lower  lip;  mandibular  teeth 
3  +  3  or  4+4  (Cross  basin) C.  polypogon 

6A.  Body  relatively  elongate;  abdomen  not 
greatly  expanded,  more  or  less  exten- 
sively darkly  pigmented;  mandibular  teeth 


ROBERTS:  CAMEROUNIAN  SUCKERMOUTH  CATFISHES 


165 


4 +4  to  8 + 8  (coastal  basins  east  of  Sanaga 
basin) C.  cameronensis 

6B.  Body  stubby;  abdomen  considerably  ex- 
panded, pale;  mandibular  teeth  usually 
4+4  or  5  +  5,  rarely  6+6  (Sanaga  basin) 
C.  sanagaensis 

7  A.  Mandibular  teeth  (except  in  very  small 
specimens)  8  +  8  or  more;  sexual  dimor- 
phism absent  or  weak 8 

7B.  Mandibular  teeth  3  +  3  to  6  +  6;  sexual  di- 
morphism marked,  larger  males  with  en- 
larged anal  fin,  enlarged  and  pointed  up- 
per lobe  of  caudal  fin,  and  greatly  enlarged 
and  expanded  humeral  process  of  pec- 
toral girdle  (in  all  or  nearly  all  river  ba- 
sins)   C.  batesii 

8 A.  Sucker  relatively  huge;  large  premaxillary 
teeth  in  seven  or  eight  rows;  top  and  sides 
of  body  extensively  covered  with  mela- 
nophores,  so  it  appears  almost  uniformly 
dark;  pectoral  fin  soft  rays  9  or  1 0;  caudal 
peduncle  moderately  to  very  slender 9 

8B.  Sucker  relatively  small;  large  premaxil- 
lary teeth  in  three  or  four  rows;  top  and 
sides  of  body  with  very  sharply  contrast- 
ing dark  vertical  bands  alternating  with 
pale  areas;  pectoral  fin  soft  rays  8  or  9; 

caudal  peduncle  relatively  deep 

C.  disneyi 

9 A.  Mandibular  teeth  8+8  to  10+10;  dorsal 
fin  soft  rays  6;  adipose  fin  short  (Niger 
basin) C.  niger 

9B.  Mandibular  teeth  13+13  to  15  +  15;  dor- 
sal fin  soft  rays  4  or  5;  adipose  fin  long 
(Lokunje  basin) C.  harbinger 

Chiloglanis  reticulatus  new  species 

(Figs,  la,  2,  3a,  b,  5,  7) 

HOLOTYPE.— CAS  60786,  41.6  mm  gravid  female,  Camer- 
oun,  Congo  basin,  R.  Mwamedjwel,  a  very  small  stream  2-3 
km  W  of  Yokadouma,  Tyson  R.  Roberts,  15  Jan.  1980. 

PARATYPES.— CAS  (SU)  47471, 36.3  mm,  Cameroun,  Congo 
basin,  R.  Momjepom,  Yokadouma,  A.  I.  Good,  20  Mar.  1936; 
CAS  60787, 3 1 .0  mm,  Zaire,  Congo  basin,  R.  Lubazi,  tributary 
of  R.  Luala,  3  km  S  of  Kibunzi,  Tyson  R.  Roberts  and  Donald 
J.Stewart,  17  Aug.  1973;  CAS  60788, 10: 19.1-38.1  mm,  Zaire, 
Congo  basin,  R.  Lufu,  2  km  NW  of  Lufu  village,  Tyson  R. 
Roberts  and  Donald  J.  Stewart,  2  Aug.  1973  CAS  60789,  43: 
9.4-37.4  mm,  Zaire,  Congo  basin,  tributary  of  R.  Luala  26  km 
N  of  Kibunzi,  Tyson  R.  Roberts  and  Donald  J.  Stewart,  17 
and  24  Aug.  1973. 

Chiloglanis  reticulatus  is  a  relatively  small 
species,  largest  specimen  41.6  mm  standard 
length;  sucker  relatively  small,  transversely  oval, 


with  very  large  barbels;  auxiliary  mental  barbels 
absent;  large  teeth  of  premaxillary  in  2  to  4  rows; 
mandibular  teeth  in  1  or  2  rows,  very  irregularly 
arranged,  basic  count  from  about  2  +  2  to  5  +  5; 
eyes  large;  dorsal  fin  soft  rays  usually  5  (rarely 
6);  pectoral  fin  soft  rays  7  or  8;  adipose  fin  long, 
its  origin  well  in  advance  of  a  vertical  line  through 
anal  fin  origin;  caudal  peduncle  moderately  deep; 
caudal  fin  moderately  forked,  lower  lobe  longer 
than  upper.  Larger  specimens  of  both  sexes  with 
well-developed  tuberculated  epidermal  ridges  on 
head  and  body,  characteristically  forming  a  dense 
network  or  reticulum  on  dorsal  surface  of  head. 
The  largest  male,  37.4  mm,  from  tributary  of  R. 
Luala,  has  anal  fin  and  lower  lobe  of  caudal  fin 
greatly  enlarged;  humeral  process  of  pectoral  spine 
not  enlarged.  Its  genital  papilla  is  elongate,  with 
a  peculiarly  enlarged  and  swollen  tip,  so  that  it 
resembles  somewhat  the  proboscis  of  an  ele- 
phant. Sexual  dimorphism  absent  or  slight  in 
other  specimens  examined.  Holotype  in  life  with 
chocolate-brown  dark  bands  and  cream-colored 
interspaces. 

ETYMOLOGY.— The  name  reticulatus  (Latin, 
adjective)  refers  to  the  reticulum  or  network  of 
tuberculated  epidermal  ridges. 

DISTRIBUTION.— Known  only  from  the  north- 
western portion  of  the  Congo  basin  in  Cameroun 
and  Zaire. 

Chiloglanis  voltae  Daget  and  Stauch,  1963 

Chiloglanis  voltae  Daget  and  Stauch,  1963:99,  fig.  4  (type  lo- 
cality Bougouri  Ba  au  pont  de  Nabere,  Volta  basin;  also 
reported  from  R.  Benoue  a  Lakdo,  Cameroun);  Roman  1 966: 
1 54  (Volta  Noire  pres  de  Boromo). 

MATERIAL  EXAMINED.— CAMEROUN:  Niger  basin:  MNHN 
1962-1275,53:1 4.6-1 7.6  mm,  R.  Benoue  a  Lakdo  (paratypes); 
NIGERIA:  Niger  basin:  CAS  64128,  33:  16.9-27.0  mm,  R.  Be- 
nue  22  km  E  of  Makurdi;  GHANA:  Volta  basin:  MNHN  1984- 
551,  102:  15.1-26.3  mm,  R.  Oti  a  Sabari;  MNHN  1984-522, 
2:  22.9-27.2  mm,  R.  Wawa  a  Yadzo.  BURKINA  FASO:  Volta 
basin:  MNHN  1962-1280,  24.2  mm,  Bougouri  Ba  au  pont  de 
Nabere  (holotype);  MNHN  1961-591,  24:  14.6-16.9  mm,  same 
collection  as  holotype  (paratypes). 

Chiloglanis  voltae  is  the  smallest  known  species 
of  Chiloglanis,  largest  specimen  27.2  mm  stan- 
dard length.  Sucker  moderately  large,  non-pa- 
pillose; outer  mental  barbel  elongate,  inner  short; 
no  auxiliary  mental  barbels;  teeth  conical,  not 
strongly  S-shaped;  mandibular  teeth  in  1  or  2 
irregular  rows,  not  concentrated  near  symphysis 
but  spread  across  entire  mouth  opening;  eye  large; 
dorsal  fin  soft  rays  usually  6;  pectoral  fin  soft  rays 


166 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  6 


FIGURE  7.    Chiloglanis  reticulatus,  holotype,  gravid  female  41.6  mm  (Congo  basin). 


7  or  8;  caudal  peduncle  moderately  deep;  caudal 
fin  moderately  forked;  caudal  fin  lobes  rounded, 
lower  slightly  longer  than  upper.  Body  with  three 
dark  or  dusky  vertical  bars  separated  by  pale 
areas;  margin  of  dark  bars  defined  by  concentra- 
tion of  melanophores  not  observed  in  other 
species  of  Chiloglanis;  caudal  fin  without  well 
developed  bars  or  other  marks.  Sexual  dimor- 


phism unknown.  No  observations  available  on 
gonads  of  mature  individuals.  Total  vertebrae 
29-3 1 ,  fewer  than  in  any  other  species  of  Chi- 
loglanis  for  which  data  are  available. 

DISTRIBUTION.— Known  only  from  the  R.  Be- 
nue  (Niger  basin)  of  Cameroun  and  Nigeria  and 
the  Volta  basin.  In  the  R.  Benue  near  Makurdi, 
I  collected  C.  voltae  on  sandy  bottom,  far  from 


ROBERTS:  CAMEROUNIAN  SUCKERMOUTH  CATFISHES 


167 


FIGURE  8.    Chiloglanis  polypogon,  holotype,  immature  female  41.5  mm  (Cross  basin). 


any  other  kind  of  habitat,  in  sympatry  with  C. 
benuensis. 

Chiloglanis  benuensis  Daget  and  Stauch,  1963 

Chiloglanis  benuensis  Daget  and  Stauch,  1963:98,  fig.  3  (type 
locality  R.  Benoue  £  Lakdo,  Cameroun). 


MATERIAL  EXAMINED.— CAMEROUN:  Niger  basin:  MNHN 
1962-1273,  38.4  mm,  R.  Benoue  a  Lakdo  (holotype);  MNHN 
1962-1274,  40:  18.9-34.8  mm,  collected  with  holotype  (para- 
types);  NIGERIA:  Niger  basin:  CAS  64127,  20:  20.3-35.0  mm, 
R.  Benue  22  km  E  of  Makurdi. 

Chiloglanis  benuensis  (largest  specimen  38.4 
mm)  has  relatively  small,  transversely  oval  suck- 


168 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  6 


er  and  very  large  barbels;  up  to  three  auxiliary 
mental  barbels  on  each  side  of  lower  lip,  but  no 
auxiliary  barbel  lying  external  to  outer  primary 
mental  barbel;  eye  large;  mandibular  teeth  3  +  3 
to  5  +  5;  soft  dorsal  fin  rays  5  or  6;  pectoral  fin 
soft  rays  8;  adipose  fin  long,  its  origin  far  anterior 
to  a  vertical  line  through  anal  fin  origin;  caudal 
peduncle  moderately  slender;  caudal  fin  mod- 
erately forked,  lower  lobe  somewhat  longer  than 
upper;  sexual  dimorphism  unknown. 

DISTRIBUTION.  —  Chiloglanis  benuensis  is 
known  only  from  the  R.  Benue  in  Cameroun  and 
Nigeria.  In  the  Benue  near  Makurdi  it  was  col- 
lected sympatrically  with  C.  voltae  on  a  sandy 
bottom  in  shallow,  swift-flowing  water. 

Chiloglanis  polypogon  new  species 

(Figs.  \b,  2b,  3c,  d,  6a,  b,  8) 

HOLOTYPE.— CAS  60790, 41.5  mm  immature  female,  Cam- 
eroun, Cross  basin,  high  gradient  streams  along  road  from 
Mamfe  to  Bamenda,  88-94  km  SW  of  Bamenda,  Tyson  R. 
Roberts,  17  Mar.  1980. 

PARATYPES.-CAS  60791, 1 1: 20.2-39.7  mm,  collected  with 
holotype;  CAS  60792,  11:  23.0-26.0  mm,  Cameroun,  Cross 
basin,  riffles  of  R.  Badi  where  it  flows  into  R.  Cross  and  main- 
stream of  R.  Cross  near  Mamfe,  Tyson  R.  Roberts,  1-5  Mar. 
1980;  CAS  60793,  7:  46.4-54.9  mm,  Cameroun,  Cross  basin, 
small  stream  on  road  from  Mamfe  to  Bamenda,  98  km  from 
Bamenda  and  about  24  km  from  Mamfe,  Tyson  R.  Roberts, 
17  Mar.  1980;  MRAC  76-32-P-2213-218,  3:  35.4-38.9  mm, 
Cross  basin,  1 4  miles  [23  km]  W  of  Mamfe  towards  Eyomojok, 
H.  Stenholt  Clausen  and  Jorgen  J.  Scheel,  8  Mar.  1966. 

Chiloglanis  polypogon  is  a  moderately  large 
species,  largest  specimen  54.9  mm;  sucker  rela- 
tively small,  transversely  oval,  with  large  pa- 
pillae and  very  large  barbels;  3-5  auxiliary  men- 
tal barbels  present  on  each  side  of  lower  lip; 
majority  of  specimens  with  an  auxiliary  mental 
barbel  external  to  outer  primary  mental  barbel 
on  one  or  both  sides  of  lower  lip  (not  observed 
in  any  other  species  of  Chiloglanis);  large  pre- 
maxillary  teeth  in  3  or  4  irregular  rows;  man- 
dibular teeth  large,  3  +  3  or  4+4;  dorsal  fin  soft 
rays  usually  5  (5  in  24  specimens,  6  in  5);  pectoral 
fin  soft  rays  7  or  8;  adipose  fin  long,  its  origin 
well  in  advance  of  a  vertical  through  anal  fin 
origin;  caudal  peduncle  moderately  slender;  cau- 
dal fin  moderately  forked,  lower  lobe  longer  than 
upper;  larger  specimens  with  longitudinally  ori- 
ented tuberculated  epidermal  ridges  on  dorsal 
surface  of  head;  sexual  dimorphism  absent  or 
slight,  males  without  enlarged  humeral  process. 
In  life  pale  areas  between  dark  vertical  bars  and 
entire  ventral  surface  pinkish  or  pale  salmon, 


unlike  any  other  species  observed  by  me.  Bars 
on  body  intermediate  in  distinctness  between 
those  of  C.  disneyi  (very  distinct)  and  C.  batesii 
(poorly  defined),  species  with  which  it  occurs 
sympatrically  in  Cross  basin.  Body  with  pale 
round  spots  observed  in  C.  reticulatus  and  some 
other  species  but  not  in  C.  batesii  and  C.  ca- 
meronensis. 

Pupil  of  eye  with  flattened  upper  margin,  con- 
tractile; contraction  involves  downward  pro- 
gression of  a  silvery  membrane  or  "curtain"  from 
ventral  margin  of  exposed  portion  of  eyeball. 
During  maximum  contraction,  pupil  is  dorsally 
directed,  almost  out  of  horizontal  line  of  vision. 
A  contractile  pupil  has  not  been  observed  in  oth- 
er species  of  Chiloglanis. 

ETYMOLOGY.— The  name  polypogon  (noun  in 
apposition,  from  the  Greek  "poly,"  multiple,  and 
"pogon,"  beard)  refers  to  the  numerous  mental 
barbels. 

DISTRIBUTION.— Known  only  from  the  Cam- 
erounian  portion  of  the  Cross  basin. 

Chiloglanis  cameronensis  Boulenger,  1 904 

Chiloglanis  cameronensis  Boulenger  1904:18  (type  locality 
Efulen,  S.  Cameroon);  Boulenger  1911:482,  fig.  360;  Matthes 
1964:118  (R.  Kululu,  tributary  of  R.  Tshuapa  near  Ikela, 
Cuvette  Centrale  of  Congo  basin,  Zaire;  identification  doubt- 
ful, specimen  unavailable  for  examination);  Gery  1965:379 
(swamps  of  R.  Ivindo  near  Makokou,  Ogooue  basin,  Gabon; 
Roman  1971:134,  fig.  58  (Ntem  basin,  Rio  Muni). 

MATERIAL  EXAMINED. -CAMEROUN:  BMNH  1904.7.1:90-96, 
7:  32.5-41.2  mm,  Efulen,  Bates  collection  (syntypes);  BMNH 
1 906.5.28:68-69, 45.3  mm,  Efulen  (Bates  collection);  CAS(SU) 
15448,  42.0  mm,  Efulen,  R.  Chenge,  Kribi  basin;  CAS(SU) 
47468,  25.5  mm,  R.  Mfiande  or  R.  Seng,  Ebolowa,  Ntem  or 
Campo  basin;  CAS(SU)  1 5450,  36.3  mm,  R.  Mfiande,  Ebo- 
lowa, Ntem  basin;  CAS(SU)  15451,  4:  35.7^0.2  mm,  "Ebo- 
lowa (Mejap  II 12  mi  W)  Minya'a  R.  trib.  of  Mvila  R.";  CAS(SU) 
47470,  2:  25.6-34.2  mm,  Bikui,  Lolodorf,  Lokunje  basin; 
MHNG  757.69, 33.3  mm,  Efulen;  MNHN  1988-1 163, 3: 34.3- 
42.9  mm,  R.  Tchengue  on  road  from  Kribi  to  Ebolowa,  about 
80  km  W  of  Ebolowa,  Kienke  basin.  Rio  MUNI:  MRAC  173147- 
148, 28.7  mm,  R.  Guoro,  tributary  of  R.  Ntem.  GABON:  MRAC 
73-02-P-191 1-921,  5:  23.5-33.7  mm,  Ogooue  basin,  Makou- 
kou;  MHNG  2169.56-58,  4:  18.9-26.5  mm,  Ogooue  basin, 
marigots  apres  Ybiegn,  Makoukou;  MRAC  80-27-P-89-95,  4: 
19.6-28.2  mm,  R.  Diala. 

Chiloglanis  cameronensis  is  a  medium-sized 
species,  largest  specimen  45.3  mm;  sucker  very 
large,  longer  than  broad  (not  transversely  oval), 
extensively  covered  with  moderately  large  pa- 
pillae; barbels  short,  but  well  developed,  usually 
papillose;  up  to  3  auxiliary  mental  barbels  on 
each  side  of  lower  lip;  large  premaxillary  teeth 


ROBERTS:  CAMEROUNIAN  SUCKERMOUTH  CATFISHES 


169 


in  3  or  4  rows;  mandibular  teeth  large,  usually 
about  4+4  or  5  +  5,  but  some  specimens  with  as 
many  as  7  +  7  or  8  +  8;  eyes  large;  dorsal  fin  soft 
rays  usually  5;  pectoral  fin  soft  rays  7  or  8;  adi- 
pose fin  long,  its  origin  considerably  in  advance 
of  a  vertical  through  anal  fin  origin;  caudal  pe- 
duncle moderately  to  very  slender  (most  slender 
in  specimens  from  Ogooue);  caudal  fin  not  deeply 
forked,  lobes  rounded  or  even  truncate,  lower 
slightly  longer  than  upper.  Dark  bands  and  pale 
interspaces  on  body  fairly  distinct.  Abdomen 
more  or  less  darkly  pigmented,  almost  black  in 
some  specimens,  due  to  melanophores  contin- 
uous with  those  on  anterior  vertical  bar  of  body 
extending  across  abdomen.  This  abdominal  pig- 
mentation, absent  or  poorly  developed  in  other 
species,  highly  variable:  in  some  older  and  poorly 
preserved  specimens,  scarcely  evident;  in  some 
specimens  from  Cameroun  (including  Efulen), 
on  only  a  relatively  small  area  between  the  pelvic 
fin  bases;  in  specimens  from  Ntem  and  Ogooue, 
entire  abdomen  darkened. 

DISTRIBUTION.— Coastal  basins  of  SE  Camer- 
oun and  Rio  Muni,  including  Lokunje,  Kribi, 
Kienke,  and  Ntem  basins;  Ogooue  basin  (Ga- 
bon). Only  species  of  Chiloglanis  known  from 
Ogooue. 

Chiloglanis  sanagaensis  new  species 

(Figs.  \c,  2c,  3e,f,6c,  d,  9) 

HOLOTYPE.— CAS  60794, 26.8  mm  male,  Cameroun,  Sanaga 
basin,  gravel  bars  and  riffles  in  lower  1-2  km  of  R.  Nchit  where 
it  flows  into  R.  Mbam  (site  of  summer  palace  of  Sultan  of 
Foumban),  Tyson  R.  Roberts,  23  Mar.  1980. 

PARATYPES.— CAMEROUN:  CAS  60795,  25:  18.6-28.6  mm, 
collected  with  holotype;  MNHN  1988-1164,  11:  31.4-39.9 
mm,  Sanaga  basin,  R.  Makenene  at  bridge  on  road  to  Ndimeki 
( 1 00  m  from  Bafia-Bafoussam  route),  about  1 1 0  km  S  of  Ba- 
foussam,  Tyson  R.  Roberts,  1  Jan.  1988;  MNHN  1988-1 165, 
12:  21.1^12.8  mm,  Sanaga  basin,  R.  Mana  23  km  N  of  Bafia 
on  route  from  Bafia  to  Bafoussam,  Tyson  R.  Roberts,  1  Jan. 
1988. 

Chiloglanis  sanagaensis  is  a  medium-sized 
species,  largest  specimen  42.8  mm;  sucker  mod- 
erately large,  round,  with  numerous  papillae  of 
uniform  size  evenly  distributed  over  almost  its 
entire  surface;  barbels  short  but  well  developed, 
each  side  of  lower  lip  with  3  auxiliary  mental 
barbels;  large  premaxillary  teeth  in  3  or  4  rows; 
mandibular  teeth  usually  4+4  or  5  +  5,  rarely 
6+6;  eyes  moderately  large;  dorsal  fin  soft  rays 
5  or  6;  pectoral  fin  soft  rays  7  or  8;  adipose  fin 
low  and  very  long,  its  origin  far  in  advance  of  a 
vertical  through  anal  fin  origin;  caudal  peduncle 


slender;  caudal  fin  almost  truncate,  very  slightly 
emarginate  or  forked,  lobes  slightly  rounded, 
lower  a  bit  longer  than  upper.  Secondary  sexual 
dimorphism  not  observed. 

This  species  characterized  by  short,  stocky  or 
stubby  body  and  swollen  abdomen.  Swelling  ev- 
ident in  virtually  all  specimens,  seemingly  due 
mainly  to  exceptionally  long  and  highly  coiled 
intestine.  A  specimen  of  27.0  mm  has  intestine 
just  over  30  mm  long,  more  than  double  relative 
length  of  intestine  observed  in  C.  batesii  and  sev- 
eral other  species. 

ETYMOLOGY.— The  name  sanagaensis,  adjec- 
tive, refers  to  the  Sanaga  basin,  source  of  all 
known  specimens  of  this  species. 

DISTRIBUTION.— Known  only  from  the  Sanaga 
basin. 

Chiloglanis  batesii  Boulenger,  1 904 

(Figs.  Id,  Id,  4a,  b,  6e,  10) 

Chiloglanis  batesii  Boulenger,  1 904: 1 9  (type  locality  "Efulen" 
and  "streams  tributary  of  the  Lobi  River,  1 5  or  20  miles 
S.W.  of  Efulen"  [South  Cameroon]);  Boulenger  1911:485, 
fig.  363  (Efulen,  R.  Lobi,  R.  Ja);  Nichols  and  Griscom  1917: 
720  (lower  Congo;  Faradje;  no  description);  Trewavas  1974 
(morphology,  comparison  with  C.  micropogori). 

Chiloglanis  micropogon  Poll,  1952:228,  figs.  3,  4  (type  locality 
R.  Nzokwe,  affluent  de  la  R.  Ulindi,  Territ.  Kabare;  and  R. 
Tshinganda,  Bunyakiri,  affluent  de  la  R.  Luhoho,  West  Kivu); 
Daget  1954:307,  fig.  116  (upper  Niger);  Poll  1967:255,  fig. 
119  (R.  Luachimo,  Angola);  Trewavas  1974:361,  figs.  5,  6 
(Mungo,  Meme,  and  Lobi  basins,  Cameroun;  R.  Kaduna, 
Niger  basin,  Nigeria). 

MATERIAL  EXAMINED.— CAMEROUN:  BMNH  1904.7.127,28.2 
mm,  "15  or  20  miles  S.W.  of  Efulen  in  streams  trib.  to  R. 
Lobi"  (syntype,  a  male  with  elongate  genital  papilla,  elongate 
upper  caudal  fin  lobe,  and  enormously  expanded  humeral  pro- 
cess); BMNH  1904.7.1.97-100, 4:  27.8-33.2  mm,  Efulen,  Bates 
collection  (syntypes);  BMNH  1904.10.26:53-62, 1 1: 27.5-34.0 
mm,  R.  Ja;  BMNH  1909.4.29.97,  26.1  mm,  R.  Lobi;  CAS 
60796,  2:  36.7-39.3  mm,  high-gradient  streams  along  Mamfe- 
Bamenda  road,  16-22  km  N  of  Numba  and  88-94  km  SW  of 
Bamenda,  Cross  basin;  CAS  60797,  22.2  mm,  riffles  in  mouth 
of  R.  Badi  where  it  flows  into  R.  Cross  at  Mamfe;  CAS  60798, 
6:  19.2-26.7  mm,  R.  Nchit  where  it  flows  into  R.  Mbam, 
Sanaga  basin;  BMNH  1973.5.14:362-363,  2:  19.8-20.0  mm, 
R.  Mungo  at  Baduma;  BMNH  1973.5.14:347-354,  3:  13.7- 
24.5  mm,  R.  Blackwater,  Mungo  basin;  BMNH  1973.5.14: 
355-359,  2:  22.0-26.3  mm,  R.  Dilolo  at  Bolo,  Mungo  basin; 
BMNH  1973.5.14:364-370,2: 31.9-35.8  mm,  R.  Wowe,  Mungo 
basin;  CAS(SU)  15447,  9:  29.0-33.3  mm,  Efulen,  R.  Chenge, 
Kribi  basin;  MNHN  1988-1155,  3:  26.2-28.7  mm,  Kienke 
basin,  R.  Tchengue  on  road  from  Kribi  to  Ebolowa.  GUINEA: 
Niger  basin:  MNHN  1986-750,  85:  18.5-34.6  mm,  R.  Milo. 
MALI:  Niger  basin:  MNHN  1961-590, 14: 30. 1-35.2  mm,  Oey- 
anko,  near  Bamako.  NIGERIA:  Niger  basin:  MNHN  1988-11 54, 
4:  21.7-37.8  mm,  R.  Owena  25  km  NE  of  Ondo  on  road  to 


170 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  6 


FIGURE  9.    Chiloglanis  sanagaensis,  holotype,  male  26.8  mm  (Sanaga  basin). 


Akure;  MNHN  1988-1153,  8:  16.3-20.0  mm,  R.  Osse  near 
Uyere,  on  road  from  Benin  City  to  Akure;  BMNH  1 978. 1 2. 1 3.7- 
8,  2:  26.5-27.3  mm,  Gurara  Falls  near  Abuja;  MNHN  1977- 
14,  3:  24.1-37.6  mm,  R.  Lere,  Gindiri,  Jos  Plateau;  MNHN 
1977- 1 5,  3:  27. 1-40.0  mm,  R.  Shemankar,  tributary  of  R.  Gu, 
Jos  Plateau;  Victor.  G.  Smith  collection,  uncat.,  34.9  mm,  R. 
Assob,  Gengera  Kibo,  Jos  Plateau.  REPUBLIQUE  CENTRAFRI- 


CAINE:  Chad  basin:  MNHN  1988-1 156,  8:  21.5-27.5  mm,  R. 
Doukouma  about  65  km  W  of  Mbres;  MNHN  1988-1157, 
131:  19.3-31.8  mm,  R.  Ouaham  20  km  upriver  from  Batan- 
gafo;  MNHN  1988-1158,2:31 .3-36. 1  mm,  R.  Koumbala  be- 
low Chutes  de  Matakil;  MNHN  1988-1 159,  5:  22.2-25.7  mm, 
R.  Gounda  (near  Chad  border).  Congo  basin:  MNHN  1988- 
1 1 6 1 , 20:  1 8.2-3 1 .6  mm,  R.  Mbomou  near  Zemio;  MCZ  48360, 


ROBERTS:  CAMEROUNIAN  SUCKERMOUTH  CATFISHES 


171 


FIGURE  10.    Chiloglanis  batesii,  36.7  mm  (Sanaga  basin). 


R.  Mbomou  at  rapides  de  Gozobangui;  MNHN  1988-1160, 
41:  17.6-29.6  mm,  R.  Chinko  at  Rafai;  MNHN  1988-1162, 
4:  31.3-36.0  mm,  small  tributary  of  R.  Euwou  about  30  km 
NW  of  Alindao.  ZAIRE:  Congo  basin:  CAS  60799,  26.4  mm, 
R.  Yambi  below  Chute  de  Bidi,  2 1  km  E  of  Sundi-Mamba; 
CAS  60800,  60:  16.5-27.2  mm,  R.  Ngudi  at  Chute  Mayeux, 
23  km  NE  of  Sundi-Lutete;  CAS  60801,  139:  13.7-30.4  mm, 
R.  Lufu,  Songoloko  district;  CAS  60802,  2:  22.7-25.7  mm,  R. 


Lubazi,  tributary  to  R.  Luala  near  Kibunzi;  CAS  60803,  35: 
1 9.2-29.3  mm,  tributary  to  R.  Luala  near  Kibunzi;  CAS  60804, 
25.8  mm,  R.  Kwilu  near  Moerbeke;  60805,  29.6  mm,  R.  Sen- 
gezi,  tributary  to  R.  Inkisi-Lukusu  87  km  W  of  Kinshasa;  CAS 
60807,  2:  23.9-26.4  mm,  R.  Konzi,  tributary  to  R.  Kwango, 
at  bridge  om  road  from  Kenge  to  Masi-Manimba;  FMNH 
62918,  22.2  mm,  R.  Dungu,  Nagero;  MRAC  91480-81,  2: 
33.1-37.2  mm,  R.  Tshinganda,  tributary  of  R.  Luhoho,  Bun- 


172 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  6 


yakiri,  Kivu  (paratypes  of  C.  micropogori);  MNHN  1987-1607, 
32.0  mm,  R.  Loya  near  Mafifi. 

Chiloglanis  batesii  is  a  small  species,  largest 
known  specimen  40  mm  standard  length;  sucker 
relatively  large,  round  or  longer  than  broad  rath- 
er than  transversely  oval;  mental  barbels  greatly 
reduced;  mandibular  teeth  6  +  6  or  fewer  (often 
5  +  5,  occasionally  4 + 4);  eyes  large;  dorsal  fin  soft 
rays  usually  6  (sometimes  5);  pectoral  soft  rays 
7  or  8;  adipose  fin  high  and  posteriorly  rounded, 
but  relatively  short,  its  origin  only  slightly  in 
advance  of  a  vertical  through  anal  fin  origin;  cau- 
dal fin  deeply  forked;  caudal  peduncle  very  slen- 
der; penultimate  5-7  vertebrae,  in  juveniles  as 
well  as  adults  of  both  sexes,  with  distal  ends  of 
neural  and  hemal  spines  flared  (anteroposteriorly 
expanded).  Sexual  dimorphism  exceptionally 
pronounced,  subadult  as  well  as  adult  males  usu- 
ally with  enlarged  anal  fin,  enlarged  and  pointed 
or  filamentous  upper  caudal  fin  lobe,  and  greatly 
enlarged  humeral  spine.  Most  Chiloglanis  with 
6+6  or  fewer  mandibular  teeth  have  relatively 
small  suckers  and  long  barbels.  No  other  species 
has  a  more  slender  caudal  peduncle,  or  males 
with  such  an  elongate  upper  caudal  fin  and  en- 
larged humeral  spine  at  such  small  size.  Neural 
and  hemal  spines  in  caudal  peduncle  slender  in 
all  or  almost  all  other  species  of  Chiloglanis  ex- 
amined except  some  larger  specimens  of  C.  oc- 
cidentalis. 

Coloration  in  live  specimens  is  recorded  in  my 
fieldnotes  for  two  Camerounian  localities,  high- 
gradient  tributaries  of  the  R.  Cross  and  the  R. 
Nchit,  Sanaga  basin.  Two  large  Cross  specimens 
had  body  with  dusky,  poorly  defined  bands,  al- 
most mottled.  Pale  areas  on  body  (more  distinct- 
ly outlined  than  dusky  bands)  pale  yellowish  or 
dull  olivaceous  yellow.  Caudal  fin  with  dusky 
vertical  bar  across  base  of  lobes,  light  areas  pale 
yellow.  Dusky  markings  on  anal  fin  present  but 
less  distinct  than  in  C.  disneyi  collected  at  same 
locality.  Eye  with  small  oval  black  pupil,  iris  dull 
gray-green.  Six  specimens  from  the  Nchit  (Sa- 
naga) had  pale  areas  on  body  grayish  or  oliva- 
ceous or  pale  yellow;  alternating  dark  and  pale 
areas  on  snout  forming  in  all  six  specimens  a 
triradiate  pattern  not  evident  in  specimens  of  C. 
sanagaensis  collected  at  same  locality. 

DISTRIBUTION.  —  Chiloglanis  batesii  probably 
is  the  species  of  Chiloglanis  with  the  most  ex- 
tensive distribution.  It  is  found  throughout  the 
Niger  and  Congo  basins,  in  the  headwaters  of  the 


Chad  basin  in  Republique  Centrafricaine,  and  in 
almost  every  river  basin  in  Cameroun  except  the 
Ndian,  Nyong,  and  Ogooue.  It  occurs  sympatri- 
cally  with  six  of  the  nine  other  species  of  Chi- 
loglanis found  in  Cameroun. 

Chiloglanis  disneyi  Trewavas,  1974 
(Figs.  \e,  2e,4c,d,  11) 

Chiloglanis  disneyi  Trewavas,  1974:365,  figs.  6, 8  (type  locality 
Wowe,  Mungo  basin,  Cameroun). 

MATERIAL  EXAMINED.— CAMEROUN:  BMNH  1973:5.14.342, 
34.6  mm,  Mungo  basin,  R.  Wowe,  Disney  collection  (holo- 
type);  BMNH  1973.5.14:343-344,  2:  23.5-23.8  mm,  R. 
Dilolo,  Mungo  basin;  BMNH  1973.5.1:345,  22.5  mm,  R. 
Blackwater,  Mungo  basin;  CAS  60808,  20:  22.4-31.4  mm, 
high-gradient  streams  of  R.  Cross  along  road  from  Mamfe  to 
fBamenda. 

Chiloglanis  disneyi  apparently  is  a  small 
species,  largest  specimen  only  35  mm  (Trewavas 
1974:362);  sucker  moderately  large,  round;  pa- 
pillae of  upper  lip  with  distinctive  frond-like 
margins  (Trewavas  1974:365-366,  fig.  8);  bar- 
bels very  small;  no  auxiliary  mental  barbels;  large 
premaxillary  teeth  in  3  or  4  rows;  mandibular 
teeth  usually  in  2  rows,  each  with  8  +  8  to  10+10 
teeth;  eye  small;  dorsal  fin  soft  rays  usually  5 
(specimens  from  Cross  basin  have  5  in  18,  6  in 
2);  pectoral  fin  soft  rays  8  or  9;  caudal  peduncle 
deep;  caudal  fin  slightly  forked,  lobes  rounded, 
lower  distinctly  larger  than  upper.  Secondary 
sexual  dimorphism  apparently  absent,  but  larg- 
est specimens  available  lack  well-developed  go- 
nads. 

Live  specimens  from  Cross  basin  have  body 
with  three  well-defined,  almost  black,  vertical 
bands;  dorsal  surface  of  head  to  dorsal  fin  origin 
or  first  dark  band  on  body  yellowish  orange,  or- 
angish  brown,  or  brownish  in  different  individ- 
uals; pale  bands  on  body,  dorsal,  adipose,  and 
caudal  fins  bright  yellow  or  lemon  yellow  in  all 
individuals;  pectoral,  pelvic,  and  anal  fins  pale 
yellow;  abdomen  off-white  or  cream;  pectoral  fin 
with  faint  transverse  dusky  band  near  its  middle; 
adipose  and  anal  fin  base  with  black  spots  an- 
teriorly; caudal  fin  with  four  large  round  yellow 
spots  surrounded  by  sharply  defined  black  marks; 
eye  black.  Alternating  dark  and  pale  areas  on 
head  and  body  more  clearly  demarcated  in  C. 
disneyi  than  in  any  other  species  of  Chiloglanis 
observed  by  me  in  Cameroun. 

DISTRIBUTION.— Known  only  from  Mungo  and 
Cross  basins  in  Cameroun. 


ROBERTS:  CAMEROUNIAN  SUCKERMOUTH  CATFISHES 


173 


FIGURE  1 1 .    Chiloglanis  disneyi,  3 1 .4  mm  (Cross  basin). 


COMMENT.— I  have  compared  the  specimens 
collected  in  the  Cross  with  two  of  Trewavas's 
specimens  collected  in  the  Mungo,  and  find  they 
agree  in  most  respects.  The  sucker  is  relatively 
larger  in  the  Cross  specimens;  the  smaller  suckers 
of  the  Mungo  specimens  may  be  due  to  shrinkage 
in  preservative.  The  color  pattern  is  barely  dis- 


cernible in  the  Mungo  specimens,  which  are  in 
poor  condition. 

Chiloglanis  niger  new  species 

(Figs.  lf,2f,4e,f,6f,  12) 

HOLOTYPE.— CAS  60809,  34.4  mm,  Cameroun,  Niger  basin, 
R.  Menchum  below  high  waterfall  near  Befang,  Bamenda  high- 
lands, Tyson  R.  Roberts,  10  Mar.  1980. 


174 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  6 


FIGURE  12.     Chiloglanis  niger,  holotype,  34.4  mm  (Niger  basin). 


PARATYPES.-CAS  60810,  7:  21.7-35.3  mm,  collected  with 
holotype. 

Chiloglanis  niger  (largest  known  specimen  35.3 
mm)  has  a  huge  sucker,  rounded  and  somewhat 
longer  than  broad;  barbels  very  small;  no  aux- 
iliary mental  barbels;  large  premaxillary  teeth  in 
7  or  8  rows;  mandibular  teeth  usually  in  2  rows; 


basic  mandibular  tooth  count  8  +  8  to  10+10  in 
largest  specimens  (6  +  6  and  7  +  7  in  smallest);  eye 
very  small;  dorsal  fin  soft  rays  6;  pectoral  fin  soft 
rays  9  or  10;  adipose  fin  short,  its  origin  only 
slightly  in  advance  of  a  vertical  through  anal  fin 
origin;  caudal  peduncle  slender;  caudal  fin  mod- 
erately forked,  lobes  rounded,  equal.  No  sexual 


ROBERTS:  CAMEROUNIAN  SUCKERMOUTH  CATFISHES 


175 


FIGURE  13.    Chiloglanis  harbinger,  holotype,  immature  female  43.5  mm  (Lokunje  basin). 


dimorphism  observed  (specimens  probably  im- 
mature). 

Color  of  live  specimens  (collected  on  a  black 
coarse  sand  or  gravel  bottom)  uniformly  dark, 
almost  black,  on  entire  dorsolateral  surface  of 
head  and  body;  abdomen  pale.  Alternating  dark 
bars  and  pale  intervening  areas  on  body,  so  typ- 
ical of  species  of  Chiloglanis,  but  not  noted  in 
live  fish,  faintly  discernible  preserved  specimens. 

ETYMOLOGY.— The  name  niger  (Latin,  black, 
noun  in  apposition)  refers  to  the  Niger  basin  as 
well  as  to  the  color  in  life. 

DISTRIBUTION.  — Known  only  from  type  local- 
ity in  R.  Menchum,  Niger  basin,  Bamenda  high- 
lands of  Cameroun. 


Chiloglanis  harbinger  new  species 

(Fig.  13) 

HOLOTYPE.— CAS  6081 1,  43.5  mm  immature  female,  Cam- 
eroun, R.  Lokunje  near  Bipindi,  A.  I.  Good,  17  Feb.  1936. 

PARATYPES.-CAS(SU)  47469,  9:  34.5-44.0  mm,  collected 
with  holotype. 

Chiloglanis  harbinger  (largest  specimen  44.0 
mm)  has  a  huge  sucker,  longer  than  broad;  mi- 
nute barbels;  no  auxiliary  mental  barbels;  teeth 
very  numerous;  large  premaxillary  teeth  in  up  to 
7  rows;  mandibular  teeth  usually  in  2  rows,  basic 
mandibular  tooth  count  about  13+13  to  15+15 
(teeth  of  type  specimens  damaged  by  string  passed 
through  mouth  and  gills);  eye  very  small;  dorsal 
fin  soft  rays  usually  5  (4  in  two  specimens);  pec- 


176 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  6 


toral  fin  soft  rays  9  or  10;  adipose  fin  long  and 
low,  its  origin  well  in  advance  of  a  vertical  through 
anal  fin  origin;  caudal  peduncle  moderately  deep; 
caudal  fin  moderately  forked,  lobes  rounded, 
nearly  equal.  Males  with  a  moderately  elongate 
genital  papilla;  no  sexual  dimorphism  observed 
in  size  of  anal  fin,  shape  of  caudal  fin,  or  humeral 
process  of  pectoral  girdle. 

Color  in  life  unknown.  Preserved  specimens 
with  numerous  large  melanophores  more  or  less 
uniformly  distributed  over  entire  dorsolateral 
surface  of  body;  vertical  dark  bars  and  interven- 
ing pale  areas  on  body  (present  in  most  species 
of  Chiloglanis)  not  evident;  melanophores  ex- 
tending from  sides  of  body  onto  sides  of  abdo- 
men but  rapidly  becoming  sparser  toward  middle 
of  abdomen  (hence  very  different  from  uniformly 
distributed  melanophores  on  abdomen  in  some 
specimens  of  C.  cameronensis);  a  dark  vertical 
bar  on  base  of  caudal  fin  as  well  as  one  in  its 
middle  (very  dark  area  near  tip  of  caudal  fin  in 
lateral  photograph  of  holotype  an  artifact  due  to 
shadow  from  tip  of  caudal  fin,  which  is  broken 
and  bent  toward  camera). 

ETYMOLOGY.— The  name  harbinger  (Middle 
English,  noun  in  apposition)  is  an  allusion  to  the 
large  number  of  teeth  in  the  suckermouth  mo- 
chokid  genera  Atopochilus  and  Euchilichthys,  a 
condition  presaged  or  foreshadowed  in  this 
species  of  Chiloglanis. 

DISTRIBUTION.— Known  only  from  the  Lo- 
kunje  basin  in  southeastern  Cameroun. 

ACKNOWLEDGMENTS 

My  interest  in  Chiloglanis  was  kindled  when 
I  visited  the  late  botanist  and  ichthyologist  F.  R. 
Irvine  at  his  home  in  Achimota  shortly  after  my 
arrival  in  Ghana  in  1961  and  was  given  some 
specimens  collected  by  Ing  Zwilling  in  Nigeria. 
Permission  for  fieldwork  in  Cameroun  was 
granted  by  the  Ministry  for  Scientific  and  Tech- 
nological Research  of  the  Republic  of  Cameroun. 
Fieldwork  in  1980  was  supported  by  the  Fonds 
Leopold  III  pour  la  Conservation  et  la  Explo- 
ration de  la  Nature  and  in  1988  by  the  National 
Geographic  Society.  Fieldwork  in  Zaire  in  1973 
was  sponsored  by  the  Office  National  de  la  Re- 
cherche et  du  Developpement  of  the  Republic  of 
Zaire  and  supported  by  the  National  Geographic 
Society.  Most  of  the  photographs  are  by  Al  Cole- 
man  of  the  Museum  of  Comparative  Zoology, 
Harvard.  Scanning  electron  micrographs  are  by 


Mary  Ann  Tenorio  of  the  SEM  lab,  California 
Academy  of  Sciences.  Radiographs  were  pre- 
pared by  David  Catania  in  the  Department  of 
Ichthyology,  California  Academy  of  Sciences,  and 
Jeanine  Abel,  Laboratoire  d'Ichtyologie,  Mu- 
seum National  d'Histoire  Naturelle  (Paris).  Prof. 
Michal  Jakubowski  of  the  Department  of  Com- 
parative Anatomy  of  Jagiellonian  University 
(Cracov,  Poland)  kindly  provided  information 
concerning  taste  buds.  For  arranging  loans  of 
specimens  I  am  obliged  to  Susan  Jewett,  Karsten 
Hartel,  Douglas  Nelson,  Alwynne  Wheeler, 
Volker  Mahnert,  Dirk  Thys  van  den  Auden- 
aerde,  and  Victor  G.  F.  Smith.  The  paper  was 
written  in  the  Department  of  Ichthyology  of  the 
California  Academy  of  Sciences  and  in  the  La- 
boratoire d'Ichtyologie  of  the  Museum  National 
d'Histoire  Naturelle. 

LITERATURE  CITED 

BOULENGER,  G.  A.    1900.    On  some  little  known  African  si- 

luroid  fishes  of  the  subfamily  Doradinae.  Ann.  Mag.  Nat. 

Hist.,  ser.  7,  6:520-529. 
.    1902.   On  the  fishes  collected  by  Mr.  S.  L.  Hinde  in 

the  Kenya  district,  with  descriptions  of  four  new  species. 

Proc.  Zool.  Soc.  London  2:221-224. 

-.    1904.  Descriptions  of  new  West-African  fishes.  Ann. 


Mag.  Nat.  Hist.,  ser.  7,  14:16-20. 
.    1911.  Catalogue  of  the  freshwater  fishes  of  Africa.  2. 

Brit.  Mus.  Nat.  Hist. 
.    1920.    Poissons  recueillis  au  Congo  Beige  par  1'ex- 

pedition  du  Dr.  C.  Christy.  Ann.  Mus.  Congo  Beige,  ser.  1 , 

2:1-38,  pis.  23-25. 
DAGET,  J.    1948.    La  collection  des  poissons  d'eau  douce  de 

1'IFAN.  Cat.  Inst.  Fr.  Afr.  Noire  (Dakar).  59  pp. 
.    1954.    Les  poissons  du  Niger  superieur.  Mem.  Inst. 

Fr.  Afr.  Noire  36.  391  pp. 
DAGET,  J.,  J.-P.  GOSSE,  AND  D.  F.  E.  THYS  VAN  DEN  AUDEN- 

AERDE,  EDS.    1986.    Check-list  of  the  freshwater  fishes  of 

Africa.  2.  ISNB  (Brussels),  MRAC  (Tervuren),  ORSTOM 

(Paris). 
DAGET,  J.  AND  A.  ILTIS.   1965.  Poissons  de  Cote  d'lvoire  (eaux 

douces  et  saumatres).  Mem.  Inst.  Fr.  Afr.  Noire  74.  385  pp. 
DAGET,  J.  AND  A.  STAUCH.    1963.    Poissons  de  la  partie  ca- 

merounaise  de  bassin  de  la  Benoue.  Mem.  Inst.  Fr.  Afr. 

Noire.  68:85-107. 
GERY,  J.    1965.   Poissons  du  bassin  de  1'Ivindo.  Biol.  Gab.  1: 

375-393. 
JAKUBOWSKI,  M.   1983.  New  details  of  the  ultrastructure  (TEM, 

SEM)  of  taste  buds  in  fishes.  Zeitschr.  mikrosk.-anat.  Forsch. 

(Leipzig)  97:849-862. 
LADIGES,  W.  AND  J.  VOELKER.    1961.    Untersuchungen  iiber 

die  Fische  in  gebirgsgewassern  des  Wasserscheidenhoch- 

lands  in  Angola.  Mitt.  Hamb.  Zool.  Mus.  Inst.  59:1 17-140. 
LEVITON,  A.  E.,  R.  H.  GIBBS,  Jr.,  E.  HEAL,  AND  C.  E.  DAWSON. 

1985.    Standards  in  herpetology  and  ichthyology.  Part  I. 

Standard  symbolic  codes  for  institutional  resource  collec- 
tions in  herpetology  and  ichthyology.  Copeia  1985(3):802- 

832. 


ROBERTS:  CAMEROUNIAN  SUCKERMOUTH  CATFISHES 


177 


MATTHES,  H.    1 964.  Les  poissons  du  lac  Tumba  et  de  la  region 

d'Ikela.  Ann.  Mus.  Roy.  Afr.  Centrale  (Tervuren),  ser.  oct., 

sci.  zool.  126.  206  pp.,  6  pis. 
NICHOLS,  J.  T.  AND  L.  GRISCOM.    1917.    Freshwater  fishes  of 

the  Congo  basin.  Bull.  Amer.  Mus.  Nat.  Hist.  37:653-756. 
NORMAN,  J.  R.   1932.  A  collection  of  fishes  from  Sierra  Leone. 

Ann.  Mag.  Nat.  Hist.,  ser.  10,  10:180-185. 
PELLEGRIN,  J.   1 933.  Voyage  du  Ch.  Alluaud  et  P.  A.  Chappuis 

en  Afrique  occidental  francaise  (dec.  1930-mars  1930).  4. 

Poissons.  Arch.  Hydrobiol.  (Stuttgart)  26:101-120. 
PETERS,  W.  C.  T.    1868.  Ueber  eine  von  dem  Baron  Carl  von 

der  Decken  entdeckte  neue  Gattung  von  Welsen,  Chiloglanis 

Deckenii,  und  einige  andere  Susswasserfische  aus  Ostafrika. 

Monatsber.  Akad.  Wiss.  Berlin  1868:598-602,  pi.  2. 
POLL,  M.    1952.    Poissons  des  rivieres  de  la  region  des  lacs 

Tanganyika  et  Kivu  recueillis  par  G.  Marlier.  Rev.  Zool. 

Bot.  Afr.  46:221-236. 
.     1967.    Contribution  a  la  faune  ichthyologique  de 

1'Angola.  Diamang  Publ.  Cult.  (Lisbon)  75.  381  pp. 
.     1971.    Revision  des  Synodontis  africains  (famille 


Mochocidae).  Ann.  Mus.  Roy.  Afr.  Centr.,  ser.  oct.,  sci.  zool. 
191.497pp.,  13  pis. 

POLL,  M.  AND  D.  J.  STEWART.    1975.    Un  Mochocidae  et  un 
Kneriidae  nouveaux  de  la  riviere  Luango  (Zambia),  affluent 


du  bassin  du  Congo  (Pisces).  Rev.  Zool.  Bot.  Afr.  89:151- 

158. 
ROBERTS,  T.  R.    1982.  Unculi  (horny  projections  arising  from 

single  cells),  an  adaptive  feature  of  the  epidermis  of  ostar- 

iophysan  fishes.  Zool.  Scripta  11:55-76. 
.    1983.    Revision  of  the  south  and  southeast  Asian 

sisorid  catfish  genus  Bagarius,  with  description  of  a  new 

species  from  the  Mekong.  Copeia  1983(2):435-445. 
.  In  press.  The  freshwater  fishes  of  western  Borneo  (Ka- 
limantan Barat,  Indonesia).  Mem.  Calif.  Acad.  Sci.  14. 
ROMAN,  B.    1966.  Les  poissons  des  hauts-bassins  de  la  Volta. 

Ann.  Mus.  Roy.  Afr.  centr.,  ser.  oct.,  sci.  zool.  150.  191  pp. 
.    1971.   Peces  de  Rio  Muni  Guinea  Ecuatorial  (Aguas 

dulces  y  salobres).  Fund.  La  Salle  Cien.  Nat.  (Barcelona). 

295  pp. 
SAUVAGE,  H.  E.    1878.   Notice  sur  la  faune  ichthyologique  de 

1'Ogooue.  Bull.  Soc.  Philomath.  Paris,  ser.  3,  7:90-103. 
SMITH,  V.  G.  F.   1983.  Fish  from  fast-flowing  rivers  and  streams 

on  the  Jos  Plateau.  Nigerian  Field  47:201-21 1. 
TREWAVAS,  E.  T.    1974.  The  freshwater  fishes  of  Rivers  Mun- 

go  and  Meme  and  Lakes  Kotto,  Mboandong  and  Soden, 

West  Cameroon.  Bull.  Brit.  Mus.  Nat.  Hist.,  Zool.  26:331- 

412,  5  pis. 


CALIFORNIA  ACADEMY  OF  SCIENCES 

Golden  Gate  Park 
San  Francisco,  California  94118 


PROCEEDINGS 

OF  THE 

CALIFORNIA  ACADEMY  OF  SCIENCES 


Vol.  46,  No.  7,  pp.  179-192,  36  figs. 


December  20,  1989 


AMERICAN  DANCE  FLIES  OF  THE 

DRAPETIS  ASSIMILIS  SPECIES  GROUP 

(DIPTERA:  EMPIDIDAE) 

By 
Edward  Rogers 

Naval  Medical  Research  Institute,  Detachment  Lima,  Peru, 
APO  Miami,  Florida  34031-0008 


ABSTRACT:  Eight  American  species  are  newly  recognized  as  belonging  to  the  Drapetis  assimilis  species  group, 
including  five  new  species  described  herein:  Drapetis  arnaudi  n.  sp.  from  Florida;  D.  cerina  n.  sp.  from  New 
Mexico;  D.  destituta  n.  sp.  from  Baja  California;  Z).  Solaris  n.  sp.  from  Mexico  (Distrito  Federal);  and  D. 
tonilosa  n.  sp.  from  Baja  California.  Lectotypes  are  designated  for  D.  latipennis  Melander,  1902  and  D. 
discalis  Melander,  1918.  Drapetis  discalis  Melander  is  removed  from  Crossopalpus  Bigot  and  redescribed. 
Drapetis  infumata  Melander  is  redescribed.  A  key  to  American  species  is  provided. 

Received  November  10, 1988.  Accepted  August  24, 1989. 


INTRODUCTION 

Flies  of  the  genus  Drapetis  are  tiny  cursorial 
predators  that  are  seldom  observed  or  collected. 
Members  of  the  D.  assimilis  species  group  have 
been  found  on  tree  trunks  and  in  flowers,  and 
reared  from  litter  in  hollow  trees;  their  other  hab- 
its are  unknown. 

The  D.  assimilis  group  was  first  alluded  to  by 
Collin  (1961)  in  discussing  characters  shared  by 
the  Palearctic  species,  Drapetis  assimilis  Fallen, 
D.  simulans  Collin,  and  D.  arcuata  Loew.  Sub- 
sequently, Kovalev  (1972)  named  and  further 
characterized  the  group,  basing  it  upon  the  same 
three  species  and  D.  ingrica  Kovalev.  This  clas- 
sification was  followed  by  Chvala  (1975)  in  a 
treatment  of  Scandinavian  Tachydromiinae. 

The  last  revision  of  the  American  Drapetis 
(Melander  1918)  recognized  no  infrasubgeneric 
categories,  although  three  members  of  the  D.  as- 
similis group  were  among  the  1 1  valid  species 


then  known.  At  that  time,  distinctions  among 
Drapetis  Meigen,  Crossopalpus  Bigot  (=Eudra- 
petis  Melander),  and  Elaphropeza  Macquart  were 
poorly  understood,  with  the  result  that  for  the 
last  70  years,  the  American  D.  assimilis  group 
species  have  been  divided  between  Drapetis  and 
Crossopalpus.  The  present  paper  corrects  this  un- 
natural arrangement  by  uniting  the  group  and 
placing  it  within  Drapetis,  s.  str. 

None  of  the  D.  assimilis  group  species  have 
Holarctic  distributions,  and  despite  study,  none 
are  yet  known  from  the  Neotropical  Region. 

METHODS 

Specimens  from  the  following  collections  were 
studied.  Abbreviations  given  in  parentheses  are 
used  in  the  text  to  denote  depositories:  (AC)  Col- 
lection of  the  author;  (AMNH)  American  Mu- 
seum of  Natural  History;  (CAS)  California  Acad- 
emy of  Sciences;  (CIS)  University  of  California 


[179] 


180 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  7 


at  Berkeley;  (CNC)  Canadian  National  Collec- 
tion; (INHS)  Illinois  Natural  History  Survey; 
(MCZC)  Museum  of  Comparative  Zoology; 
(PHA)  Paul  H.  Arnaud,  Jr.,  Collection;  (UCD) 
University  of  California  at  Davis;  (UCR)  Uni- 
versity of  California  at  Riverside;  (IUCM)  Iowa 
State  University;  (UK)  University  of  Kansas; 
(USNM)  United  States  National  Museum  of 
Natural  History;  (WSU)  Washington  State  Uni- 
versity. In  listings  of  materials  examined,  all  col- 
lection data  preceding  a  specimen  apply  to  that 
specimen,  unless  data  that  follow  are  in  conflict. 

Type  specimens  of  all  American  species  were 
examined  and  compared  with  original  descrip- 
tions of  types  and  with  specimens  and  descrip- 
tions of  this  study.  Descriptions  of  new  species 
have  been  based  solely  on  the  holotype,  with 
variation  among  individuals  discussed  in  suc- 
ceeding paragraphs. 

Male  genitalia  have  been  drawn  in  ventrolat- 
eral  and  dorsolateral  views,  using  a  camera  lu- 
cida;  wings  have  been  traced  from  microscope 
slide  projections.  Terminology  of  genitalia  fol- 
lows Griffiths  (1972). 

Drapetis  Meigen 

Drapetis  Meigen,  1822:91.  Type  species:  Drapetis  exilis  Mei- 
gen, 1 822  (by  monotypy). 

A  diagnosis  of  this  genus  is  given  elsewhere 
(Rogers  1983).  Drapetis  is  not  to  be  confused 
with  its  sister  genera,  Crossopalpus  Bigot  and 
Elaphropeza  Macquart.  Drapetis  discalis  Melan- 
der  was  originally  described  in  Eudrapetis  Me- 
lander,  a  junior  synonym  of  Crossopalpus.  It,  and 
all  species  treated  here,  belong  to  Drapetis  Mei- 
gen. 

Drapetis  assimilis  Group 

DESCRIPTION.— Males:  Head  round  in  frontal 
view;  palpus  light  to  dark  brown;  antenna  uni- 
formly brown,  the  second  segment  bearing  a  dis- 
tal circlet  of  setae  with  one  ventral  seta  longer 
than  all  others  in  the  circlet,  arista  long  in  relation 
to  the  short  third  segment.  Thorax  scantily  to- 
mentose;  when  present,  propleural  tomentum  not 
reaching  anterior  spiracle;  pteropleuron  pol- 
ished; scutellar  tomentum  invading  posterior  edge 
of  mesonotum  between  calli.  Wing  lacking  mac- 
ulations,  veins  R4+5  and  M  converging  or  sub- 
parallel  distally;  second  basal  (M)  cell  long,  dis- 
tally  extending  to  below  the  junction  of  vein  R, 
with  the  costa;  rm  crossvein  attaching  just  prox- 


imad  of  center  of  M  cell;  me  crossvein  perpen- 
dicular to  Cu  or  forming  a  slightly  obtuse  prox- 
imal angle  at  junction  with  Cu.  Legs  long;  anterior 
femur  tomentose  ventrally;  center  femur  lacking 
corrugation  patterns,  ventral  surface  bearing  one 
or  more  longitudinal  rows  of  long  spines;  pos- 
terior femur  lacking  strong  anteapical  spines  on 
posterior  surface;  anterior  and  center  tibiae  to- 
mentose along  flexor  surfaces;  center  tibia  lack- 
ing a  strong  apical  tooth;  posterior  tibia  lacking 
true  extensor  bristles,  apical  lappet  short,  blunt, 
unarmed.  Abdomen  black,  tergites  lacking  in- 
volutions on  corners;  sclerotization  of  tergite  I 
interrupted  near  dorsal  midline,  other  tergites 
entire;  tergites  IV  and  V  laterally  bearing  flat- 
tened spines;  sternites  I,  II,  III,  VI,  VII,  and  VIII 
entire;  sternite  IV  divided  through  center;  scler- 
otization of  segment  VIII  narrow  and  of  uniform 
width  dorsally;  aedeagus  short,  concealed  within 
periandrium.  Females:  As  in  males,  except  all 
abdominal  sternites  entire. 

The  D.  assimilis  group  is  distinguished  from 
all  others  by  the  unique  combination  of  a  long 
projecting  seta  on  the  venter  of  the  second  an- 
tennal  segment,  a  long  basal  cell,  and  division  of 
the  male  abdominal  sternite  IV.  The  antennal 
seta  and  basal  cell  have  been  used  before  to  sep- 
arate the  D.  assimilis  group  from  other  Drapetis 
in  keys  (Kovalev  1972);  the  importance  of  the 
divided  sternite  IV  has  been  overlooked. 

With  the  exception  of  D.  latipennis  Melander, 
all  of  the  American  D.  assimilis  group  species 
have  widely  polished  orbits,  and  the  tomentum 
of  the  prothorax  fails  to  reach  the  antero ventral 
lateral  corner  of  the  humerus.  The  center  tibia 
bears  two  parallel  rows  of  short  spines  along  its 
flexor  length  in  all  American  species  except  D. 
torulosa  n.  sp.  Sternite  V  is  divided  in  a  few 
species,  as  noted  in  their  respective  descriptions. 

None  of  the  D.  assimilis  species  have  the  long 
exserted  aedeagus,  divided  sternite  HI,  wing  vein 
anomalies,  nor  tibial  deformations  that  appear 
in  various  species  allied  to  D.  divergens  Loew. 
Species  related  to  D.  dividua  Melander  have  ex- 
tensor spines  on  the  posterior  tibiae  and  a  pe- 
dunculate structure  articulating  with  the  left  peri- 
andrium, both  absent  in  the  D.  assimilis  group. 
Members  of  the  D.  naica  group  (Rogers  1983) 
have  an  apical  thorn  on  the  flexor  surface  of  the 
center  tibia,  widely  divergent  radial  and  medial 
veins,  a  short  M  cell,  and  oval  head  shape.  Rel- 
atives of  D.  exilis  Meigen  lack  the  long  ventral 


ROGERS:  AMERICAN  DRAPETIS 


181 


1 


0.1 


mm 


FIGURES  1-5.    Third  antennal  segments  of  paratype  males.  Figure  1.  Drapetis  arnaudi  n.  sp.  Figure  2.  Drapetis  destitute,  n. 
sp.  Figure  3.  Drapetis  cerina  n.  sp.  Figure  4.  Drapetis  Solaris  n.  sp.  Figure  5.  Drapetis  torulosa  n.  sp. 


seta  on  the  second  antennal  segment,  present  in 
the  D.  assimilis  and  D.  naica  groups,  and  fre- 
quently have  corrugations  on  the  center  femur. 

KEY  TO  SPECIES  OF  THE  D.  ASSIMILIS  GROUP 

This  key  is  designed  for  provisional  identifi- 
cation of  unmacerated  males.  Identifications 
should  be  confirmed  by  examination  of  cleared 
genitalia.  Females  are  seldom  identifiable. 


(Males) 


1.  Anterodorsal  surface  of  center  femur  bulg- 
ing near  proximal  quarter,  Figure  34;  center 
tibia  lacking  flexor  spines;  genitalia,  Figures 
35,  36 torulosa  n.  sp. 

-  Anterodorsal  surface  of  center  femur  shaped 
normally,  Figure  31;  center  tibia  bearing 
flexor  spines 2 

2.  Posterior  tibia  bearing  a  group  of  long, 
downcurved  setae  about  distal  third  of  ex- 
tensor surface,  Figure  28;  genitalia,  Figures 
29,  30 latipennis  Melander 

-  Posterior  tibia  bearing  short  setae  of  nearly 
uniform  length  along  extensor  surface 3 

3.  Setae  of  anterior  surface  of  center  femur 
directed  ventrally,  Figure  3 1 ;  genitalia,  Fig- 
ures 32,  33 Solaris  n.  sp. 

-  Setae  of  anterior  surface  of  center  femur  all 
directed  distally 4 

4.  Halter  yellow;  one  pair  of  vertical  bris- 
tles.  5 

-  Halter  brown;  two  or  three  pairs  of  verti- 
cals   6 

5.  Ventral  spines  of  center  femur  arranged  in 
a  single  row;  propleuron  partially  tomen- 

tose;  genitalia,  Figures  21-23 

discalis  Melander 

-  Ventral  spines  of  center  femur  arranged  in 
two  or  more  rows;  propleuron  completely 


polished;  genitalia,  Figures  15,  16 

arnaudi  n.  sp. 

6.  Gena  obscured  by  eye  in  profile;  anterior 
surface  of  center  femur  thickly  whitish  yel- 
low tomentose;  two  pairs  of  vertical  bris- 
tles; genitalia,  Figures  17,  18 cerina  n.  sp. 

-  Gena  narrowly  visible  beneath  eye  in  pro- 
file; anterior  surface  of  center  femur  pol- 
ished; three  pairs  of  verticals 7 

7.  Bearing  a  tubercle  or  a  blunt  raised  projec- 
tion near  posteroventral  proximal  third  of 
posterior  femur;  propleuron  usually  par- 
tially tomentose;  abdominal  tergites  VI  and 
VII  partially  polished;  genitalia,  Figures  24- 
27 infumata  Melander 

-  Lacking  a  tubercle  or  prominence  on  pos- 
teroventral surface  of  posterior  femur;  pro- 
pleuron completely  polished;  abdominal 
tergites  VI  and  VII  completely  tomentose; 
genitalia,  Figures  1 9,  20 destituta  n.  sp. 

Drapetis  arnaudi  n.  sp. 

(Figs.  1,  10,  15,  16) 

DIAGNOSIS.— One  pair  of  vertical  bristles;  pro- 
pleuron completely  polished;  halter  yellow;  male 
genitalia  as  in  Figures  15,  16. 

DESCRIPTION.— Male.  Length  1.1  mm.  Head 
dark  brown;  one  pair  of  verticals;  frons  width  at 
ventral  end  one-fifth  length  of  third  antennal  seg- 
ment; face  linear,  slightly  narrowed  toward  cen- 
ter; gena  entirely  visible  in  profile;  third  antennal 
segment  shaped  as  in  Figure  1 ,  arista  five  times 
its  length.  Thorax  dark  brown;  propleuron  com- 
pletely polished.  Wing  1.3  mm  long,  neuration 
as  in  Figure  10.  Legs  and  coxae  bright  yellow, 
ultimate  tarsal  articles  of  each  leg  dark  brown. 
Abdomen  laterally  bearing  tiny  flattened  spines 
on  segments  IV  and  V. 

TYPES. -Holotype:  male  (CAS  13448),  USA: 


182 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  7 


14 


FIGURES  6-14.  Wing  venation  of  males.  Figure  6.  Drapetis  cerina  n.  sp.,  paratype.  Figure  7.  Drapetis  discalis,  Brewster  Co., 
Texas.  Figure  8.  Drapetis  torulosa  n.  sp.,  paratype.  Figure  9.  Drapetis  latipennis,  Orange  Co.,  California.  Figure  10.  Drapetis 
arnaudi  n.  sp.,  paratype.  Figure  11.  Drapetis  Solaris  n.  sp.,  paratype.  Figure  12.  Drapetis  discalis,  Riverside  Co.,  California. 
Figure  13.  Drapetis  destituta  n.  sp.,  paratype.  Figure  14.  Drapetis  infumata,  Chelan  Co.,  Washington. 


ROGERS:  AMERICAN  DRAPETIS 


183 


18 


FIGURES  15-18.    Genitalia  of  male  paratypes.  Figures  15  (ventrolateral  view)  and  16  (dorsolateral  view).  Drapetis  arnaudi 
n.  sp.  Figures  1 7  (ventrolateral  view)  and  1 8  (dorsolateral  view).  Drapetis  cerina  n.  sp. 


184 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  7 


labelled,  "FLA:  Highlands  Co.,  Archbold  Biol. 
Sta.,  7.X.  1964,  P.  H.  Arnaud,  Jr."  Paratypes, 
same  locality:  2  males,  1 0.X.  1 964  (CAS  and  AC), 
1  male,  7.X.  1964  (CAS),  1  male,  12.X.1964(AQ. 

REMARKS.— The  paratypes  have  three  rows  of 
ventral  spines  on  the  center  femur,  merging  dis- 
tally  into  two  rows,  not  visible  on  the  holotype. 
The  posterior  femur  bears  a  single  row  of  poorly 
developed  ventral  setae. 

This  species  is  named  for  its  collector,  Dr.  Paul 
H.  Arnaud,  Jr. 

Drapetis  cerina  n.  sp. 

(Figs.  3,6,  17,  18) 

DIAGNOSIS.— Two  pairs  of  vertical  bristles;  gena 
obscured  by  eye  in  profile;  halter  brown;  anterior 
surface  of  center  femur  whitish  yellow  tomen- 
tose;  male  genitalia  as  in  Figures  17,  18. 

DESCRIPTION.  — Male.  Length  1.8  mm.  Head 
black;  two  pairs  of  verticals,  outer  pair  short; 
frons  width  at  ventral  end  one-seventh  length  of 
third  antennal  segment;  face  linear;  gena  ob- 
scured by  eye  in  profile;  third  antennal  segment 
shaped  as  in  Figure  3,  arista  six  times  its  length. 
Thorax  black;  one  weakly  developed  humeral 
bristle;  propleural  tomentum  present,  narrowly 
failing  to  reach  anteroventral  lateral  corner  of 
humerus;  halter  dark  brown.  Wing  2.0  mm  long, 
neuration  as  in  Figure  6.  Legs  and  coxae  yellow, 
with  a  brownish  cast  in  coxae  and  tarsi;  anterior 
surface  of  center  femur  completely  coated  with 
translucent,  waxy,  whitish  yellow  tomentum, 
ventral  surface  bearing  brown  spines  set  in  two 
irregular  rows,  spines  longest  near  centers  of  rows; 
flexor  spines  of  center  tibia  very  poorly  devel- 
oped; posterior  femur  with  one  row  of  ventral 
setae  and  four  long  anteroventral  distal  bristles 
near  knee.  Abdomen  laterally  bearing  erect  flat- 
tened spines  on  tergites  IV  and  V. 

TYPES. -Holotype:  male  (USNM  762713), 
USA:  labelled,  "Las  Vegas  HS,  14.5,  NM,  H.  S. 
Barber  Collector."  Paratype:  1  male,  USA:  New 
Mexico:  Santa  Fe  Co.,  Santa  Fe,  VII  (USNM). 

REMARKS.— The  anterior  surface  of  the  center 
femur  looks  much  as  though  it  were  smeared 
with  wax,  producing  greasy  reflections.  The  ad- 
jective cerina  refers  to  this  waxy  quality. 

Drapetis  destituta  n.  sp. 

(Figs.  2,  13,  19,20) 

DIAGNOSIS.— Three  pairs  of  strong  vertical 
bristles;  propleuron  completely  polished;  halter 


dark  brown;  abdominal  tergites  VI  and  VII  com- 
pletely tomentose;  male  genitalia  as  in  Figures 
19,20. 

DESCRIPTION.— Male.  Length  14  mm.  Head 
black;  three  verticals;  frons  width  at  ventral  end 
one-sixth  length  of  third  antennal  segment;  face 
linear;  gena  entirely  visible  in  profile;  third  an- 
tennal segment  shaped  as  in  Figure  2,  arista  four 
times  its  length.  Thorax  dark  brown;  propleuron 
completely  polished;  halter  dark  brown.  Wing 
1 .6  mm  long,  neuration  as  in  Figure  1 3.  Legs  and 
coxae  yellow,  except  tarsi  and  posterior  knees 
brown;  center  femur  bearing  two  rows  of  strong, 
dark  brown  ventral  spines,  merging  distally  into 
one  row  at  about  half  length,  continuing  along 
entire  femur;  posterior  femur  bearing  two  rows 
of  long  yellow  setae  along  entire  length,  one  row 
anteroventral,  the  other  posteroventral.  Abdo- 
men laterally  bearing  long  flattened  spines  on 
tergites  IV  and  V;  tergites  VI  and  VII  completely 
tomentose. 

TYPES. -Holotype:  male  (CAS  13451),  MEX- 
ICO: labelled,  "MEXICO:  Baja  California:  Agua 
Caliente  (San  Carlos),  18.5  km  east  of  Manea- 
dero,  6.  VII.  1 973,  Paul  H.  Arnaud,  Jr."  Paratype: 
1  male,  same  data  (PHA). 

REMARKS.— Maceration  of  the  paratype  re- 
veals that  abdominal  sternite  V  is  divided.  Rec- 
ognition of  this  species  relies  primarily  on 
genitalic  differences;  it  lacks  other  unique  char- 
acteristics. 


Drapetis  discalis  (Melander) 

(Figs.  7,  12,  21-23) 

Eudrapetis  discalis  Melander,  1918:198. 

Drapetis  (Eudrapetis)  discalis;  Melander,  1928:310. 

Drapetis  (Crossopalpus)  discalis;  Melander,  1965:477. 

DIAGNOSIS. —One  pair  of  vertical  bristles;  male 
genitalia  as  in  Figures  21-23. 

REDESCRIPTION. — The  original  description  will 
suffice  to  recognize  D.  discalis  with  the  following 
corrections  based  on  the  lectotype:  body  shining 
brown;  halter  yellowish-clear;  second,  third,  and 
fourth  sections  of  costa  proportioned  2.4:2.8:8. 1 ; 
rm  crossvein  at  five-twelfths  of  the  length  of  sec- 
ond basal  cell;  outer  two  sections  of  vein  M,+2 
proportioned  1:2. 

TYPES.— Melander  labelled  his  Washington 
specimen  "type,"  and  his  California  specimen 
"paratype,"  but  neglected  to  publish  a  designa- 
tion of  the  holotype.  Lectotype  male  (USNM), 
here  designated,  USA:  labelled,  "Wawawai  Wash 


ROGERS:  AMERICAN  DRAPETIS 


185 


22. VI."  Paralectotype  male  (USNM),  USA:  Cal- 
ifornia: Santa  Clara  Co.,  Palo  Alto. 

OTHER  MATERIAL  EXAMINED.— USA:  California:  Contra 
Costa  Co.,  1  male,  Orinda  Village,  18. VI.  1970,  E.  I.  Schlinger 
(AC);  Monterey  Co.,  3  males,  25.IX.1934,  A.  L.  Melander 
(USNM),  1  male,  8  km  SW  Greenfield,  3.V.1975,  P.  A.  Rude 
(AC),  1  male,  Soledad,  20. V.I 956,  H.  R.  Moffitt  (UCD);  Or- 
ange Co.,  1  male,  14. VI.  1929,  P.  W.  Oman  (UK);  Riverside 
Co.,  1  male,  8  km  S  of  Sage,  16.IV.  1965,  C.  A.  Toschi  (CIS), 
2  males,  Whitewater  Cyn.,  1 1  .IV.  1 965,  J.  T.  Doyen  (CIS);  San 
Bernardino  Co.,  1  male,  Loma  Linda,  13.V.1953,  R.  E.  Ryck- 
man  (USNM),  1  male,  Upper  Santa  Ana  River,  28. IX.  1953, 
A.  L.  Melander  (USNM),  1  male,  Mountain  Home  Cyn., 
27.IX.1955,  A.  L.  Melander  (USNM),  1  male,  Verdemont, 
I.V.I  946,  A.  L.  Melander  (USNM);  San  Diego  Co.,  1  male, 
4.8  km  WSW  of  Escondido,  23.111.1975,  P.  A.  Rude  (AC); 
Santa  Barbara  Co.,  1  male,  Santa  Barbara,  16. VI.  1951,  K.  W. 
Tucker  (UCD);  Santa  Clara  Co.,  2  males,  Los  Gatos,  off  Stacia 
Street,  Quercus  trunk,  30.V/23.VII.1977,  E.  J.  Rogers  (AC),  5 
males,  Stanford,  19.XI.1952,  P.  H.  Arnaud  (PHA);  Shasta  Co, 
1  male,  McArthur,  12.X.1952,  E.  I.  Schlinger  (UCD);  Ventura 
Co.,  1  male,  Saticoy,  5.V.1924,  S.  E.  Banders  (CAS).  Idaho: 
Nez  Perce  Co.,  1  male,  Lewiston  Hill,  3 I.V.I 924,  A.  L.  Me- 
lander (USNM),  1  male,  30.  VI.  1 975,  E.  J.  Rogers  (AC).  Texas: 
Brewster  Co.,  2  males,  Panther  Junction,  Big  Bend  Nat'l.  Park, 
1,067  m,  14.V.1959,  J.  F.  McAlpine  (CNC);  Gillespie  Co.,  1 
male,  Pedernales  River,  4.IV.1955,  W.  W.  Wirth  (USNM); 
San  Patricio  Co.,  1  male,  Padres  Id.  near  Port  Aransas, 
23.111.1965,  J.  G.  Chillcott  (CNC).  Utah:  Cache  Co.,  1  male, 
Blacksmith  Fork  Cyn.,  1 1. VIII.  1975,  G.  F.  Knowlton  (UCD). 
Washington:  Asotin  Co.,  1  male,  Fields'  Spring  State  Park, 
31. VII.  1971,  W.  J.  Turner  (WSU),  1  male,  27.4  km  S  of  An- 
atone,  594  m,  fls.  Lonicera,  15.IV.  1977,  W.  J.  Turner  (WSU); 
Whitman  Co.,  3  males,  Yakawawa  Cyn.,  1 1.3  km  NW  of  Col- 
ton,  762  m,  25/27. VI.  1977,  W.  J.  Turner  (WSU),  1  male,  Big 
Almota  Cyn.,  Almota,  25.VII.1973,  D.  Corredor  and  S.  Ber- 
kenkamp  (WSU).  MEXICO:  Baja  California:  1  male,  9  km  E 
of  Hamilton  Ranch  Arr.  Santo  Domingo,  23.IV.  1963,  H.  B. 
Leech  and  P.  H.  Arnaud  (CAS). 

REMARKS.— The  gena  is  usually  narrowly  vis- 
ible in  profile  (not  in  specimens  from  Brewster 
Co.,  Texas).  The  ventral  half  of  the  propleuron 
is  usually  tomentose  (completely  polished  in  all 
Texas  specimens).  Most  individuals  show  a  sin- 
gle, poorly  distinguishable  humeral  bristle.  The 
halter  may  be  yellow  or  (rarely)  brown.  Ventral 
spines  of  the  center  femur  are  arranged  in  a  single 
row,  and  the  posterior  femur  bears  a  ventral  row 
of  setae.  Abdominal  sternite  V  is  divided. 

Distinguishing  between  D.  discalis  and  D.  in- 
fumata  Melander  is  difficult.  Cerci  of  the  two 
species  vary  widely  in  shape  (Figs.  21,  22,  24, 
25),  and  Pacific  Coast  D.  discalis  occasionally 
have  brown  halters,  or  infumated  wings,  or  have 
the  me  cross  vein  located  proximad  of  the  junc- 
tion of  vein  R,  with  the  costa,  all  as  in  D.  infu- 
mata.  Difference  in  the  number  of  vertical  bris- 
tles appears  to  be  the  best  character  for  separation. 


mm 


FIGURES  1 9,  20.  Genitalia  of  paratype  male  Drapetis  des- 
tituta  n.  sp.  Figure  19.  Ventrolatral  view.  Figure  20.  Dorso- 
lateral  view. 


186 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  7 


21 


23 


FIGURES  21-23.    Genitalia  of  male  Drapetis  discalis.  Figure  21  (ventrolateral  view).  Brewster  Co.,  Texas.  Figure  22  (ventro- 
lateral  view).  Cerci,  Monterey  Co.,  California.  Figure  23  (dorsolateral  view).  Santa  Clara  Co.,  California. 


FIGURES  24-27.  Genitalia  of  male  Drapetis  infumata.  Figure  24  (ventrolateral  view).  Cerci,  Chelan  Co.,  Washington.  Figure 
25  (ventrolateral  view).  Monmouth  Co.,  New  Jersey.  Figure  26  (dorsolateral  view).  Chelan  Co.,  Washington.  Figure  27  (dor- 
solateral view).  Monmouth  Co.,  New  Jersey. 


ROGERS:  AMERICAN  DRAPETIS 


187 


25 


27 


0.1    mm 


188 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  7 


Eudrapetis  Melander  ( 1 9 1 8)  is  an  obvious  syn- 
onym of  Crossopalpus  Bigot  (1857).  Crossopal- 
pus  have  produced  genae,  a  single  pair  of  ocellar 
bristles,  and  lack  mesopleural  setae.  Like  other 
Drapetis,  D.  discalis  has  short  genae,  two  pairs 
of  ocellars,  and  hairy  mesopleurae. 

Drapetis  infumata  Melander 

(Figs.  14,  24-27) 

Drapetis  latipennis  Melander,  1 902:209  (in  part,  male  paratype 

from  Milwaukee,  Wisconsin). 
Drapetis  infumata  Melander,  1918:194. 
Drapetis  naica  Melander,  1918:195  (in  part,  male  paratype 

from  Avon,  Idaho). 

DIAGNOSIS.— Three  pairs  of  strong  vertical 
bristles;  a  gentle  prominence  or  pointed  tubercle 
near  posteroventral  proximal  third  of  posterior 
femur;  male  genitalia  as  in  Figures  24-27. 

REDESCRIPTION.— Males.  Length  1.2  to  1.5  mm. 
Head  dark  brown;  three  pairs  of  verticals;  frons 
width  at  ventral  end  one-third  to  one-fifth  length 
of  third  antennal  segment.  Thorax  black  to  brown; 
one  weak  humeral  bristle;  propleural  tomentum 
present,  not  attaining  anteroventral  lateral  cor- 
ner of  humerus  nor  anterior  spiracle;  halter  dark 
brown.  Wing  1.6  to  1.9  mm  long,  occasionally 
infumated  along  veins,  me  crossvein  located 
proximad  to  junction  of  R,  with  costa,  Figure 
14.  Legs  and  coxae  dark  yellow  to  dark  brown, 
anterior  femur  and  coxa  and  proximal  areas  of 
other  femora  lightest  colored;  center  femur  bear- 
ing a  single  row  of  ventral  spines;  posterior  femur 
bearing  one  row  of  posteroventral  setae  along 
distal  half,  one  row  of  short  to  long  setae  along 
entire  ventral  length,  and  a  posteroventral  tu- 
bercle or  node  near  proximal  third.  Abdomen 
laterally  bearing  tiny  flattened  spines  on  tergites 
IV  and  V,  closely  appressed  to  body;  abdominal 
sternite  V  usually  divided;  genitalia  as  in  Figures 
24-27. 

TYPES— Holotype  (D.  infumata):  female 
(USNM),  CANADA:  labelled,  "Nelson,  British 
Columbia."  Paratype  (D.  infumata)  female 
(USNM),  USA:  Priest  Lake,  Bonner  Co.,  Idaho 
(may  be  one  of  two  females  from  Priest  Lake, 
collected  1. VIII.  19 16,  but  lacking  a  type  label). 
Paratype  (of  D.  latipennis)  male  (AMNH  729), 
USA:  Milwaukee,  Wisconsin,  23.VI.1895,  W.  M. 
Wheeler.  Paratype  (of  D.  naica)  male,  USA: 
Avon,  Idaho  (USNM). 

OTHER  MATERIAL  EXAMINED.  — USA:  California:  El  Dorado 
Co.,  1  male,  Blodgett  Forest,  21  km  E  Georgetown, 


12. VIII.  1975,  P.  A.  Rude(AC);HumboldtCo.,  1  male,  Willow 
Creek,  12. VIII.  1948,  W.  W.  Wirth  (USNM);  Madera  Co.,  1 
male,  SE  slope  of  Green  Mtn.,  2,316  m,  20.  VIII.  1971,  H.  B. 
Leech  (CAS);  Mono  Co.,  1  male,  8  km  N  Bridgeport,  Huntoon 
Forest  Camp,  2,073  m,  22. VIII.  1966,  P.  H.  Arnaud  (AC); 
Siskiyou  Co.,  1  male,  Poker  Flat,  1,536  m,  13.VIII.1966,  H. 
B.  Leech  (CAS);  Tuolumne  Co.,  1  male,  Pinecrest,  1 1  .VIII.  1 948, 
P.  H.  Arnaud  (PHA).  Georgia:  Dade  Co.,  1  male,  Cloudland 
Cyn.  State  Park,  8.V.1952,  O.  Peck  (CNC).  Idaho:  Kootenai 
Co.,  1  male,  Carlin  Bay,  Lk.  Coeur  d'Alene,  16  km  N  of  Har- 
rison, 640  m,  14/20.  VII.  1981,  W.  J.  Turner  (AC),  1  male,  217 
23. VIII.  1977  (WSU),  1  male,  24/28. VII.  1977  (WSU);  Latah 
Co.,  42  males,  Lost  Creek,  19.3  km  ENE  of  Potlatch,  823  m, 
5. VIII.  1979,  W.  J.  Turner  (WSU),  6  males  (AC),  1  male, 
1. VII.  1980  (WSU),  1  male,  7. VII.  1980  (WSU),  1  male  (AC), 
7  males,  23.  VII.  1980  (WSU),  1  male  (AC),  2  males,  Styrchnine 
Creek,  24.1  km  ENE  Potlatch,  884  m,  1/3.VH.1980  (WSU), 
1  male,  7/9. VII.  1980  (WSU),  1  male,  Physocarpus,  1. VII.  1980 
(WSU),  4  males,  1 1 .3  km  NNE  of  Moscow,  823  m,  24.  VII.  1980, 
W.  J.  Turner  (WSU),  5  males,  Big  Meadow  Rec.  Area,  1 1.3 
km  N  of  Troy,  914  m,  31. VII.  1979,  W.  J.  Turner  (WSU),  3 
males  (AC),  1  male,  L.  Sand  Creek  nr.  Bonami  Creek,  25.7 
km  E  of  Potlatch,  884  m,  9.  VIII.  1979,  W.  J.  Turner  (WSU). 
Iowa:  Boone  Co.,  1  male,  Ledges  State  Park,  23.VI.1961,  J.  L. 
Laffoon  (UICM).  Michigan:  Midland  Co.,  1  male,  5. VII.  1951, 
R.  R.  Dreisbach  (USNM).  New  Jersey:  Monmouth  Co.,  1  male, 
Long  Branch,  1  l.VL,  C.  W.  Johnson  (MCZC).  Oregon:  Baker 
Co.,  3  males,  Up.  Goose  Creek,  54.7  km  SE  of  Union,  1,268 
m,  13/1 9. VII.  1975,  E.  J.  Davis  (WSU),  1  male  (AC),  1  male, 
Low.  Goose  Creek,  57.9  km  SE  of  Union,  1,219  m,  13/. 
19.VII.1975,  E.  J.  Davis  (WSU),  1  male,  Velvet  Creek,  22.1 
km  SE  of  Union,  1,439  m,  13/1 9.  VII.  1975,  E.  J.  Davis  (AC); 
Grant  Co.,  1  male,  40  km  N  Mt.  Vernon,  24.VII.1974,  P.  H. 
Arnaud  (CAS);  Union  Co.,  1  male,  Low.  Lick  Creek,  41.8  km 
SE  of  Union,  1,305  m,  2 1/23. VII.  1977,  E.  J.  Davis  (WSU). 
Virginia:  Montgomery  Co.,  1  male,  Blacksburg,  640  m, 
28. VI.  1962,  J.  G.  Chillcott  (AC);  Roanoke  Co.,  1  male,  Mt. 
Roanoke,  1 8. V.I 965,  J.  G.  Chillcott  (AC).  Washington:  Asotin 
Co.,  1  male,  6.4  km  S  of  Anatone,  1,097  m,  12. VIII.  1980,  W. 
Turner  (WSU),  1  male,  Fields'  Spring  State  Park,  1,067-1,219 
m,  30. VI.  1975,  W.  J.  Turner  (WSU);  Chelan  Co.,  1  male, 
Lucerne,  29. VIII.  19 19,  A.  L.  Melander  (USNM).  CANADA: 
Manitoba:  1  male,  Ninette,  "Betula  glandulosa,  Populus  bal- 
samifera  associate,"  15. VII.  1958,  J.  G.  Chillcott  (CNC).  On- 
tario: Ottawa,  1  male,  26.  VI.  1 958,  J.G.  Chillcott  (AC),  1  male, 
9.VI.1962,  J.  R.  Vockeroth  (CNC),  1  male,  Maynooth, 
22.VI.1953,  J.  F.  McAlpine  (CNC).  Quebec:  1  male,  Old  Chel- 
sea, Summit  King  Mtn.,  351  m,  25. VI.  1962,  J.  R.  Vockeroth 
(CNC),  1  male,  9. VIII.  1961  (CNC).  Saskatchewan:  1  male, 
Uranium  City,  59°34'N  108°36'W,  18. VI.  1962,  J.  G.  Chillcott 
(CNC),  1  male,  Wallwort,  52°33'N  104°03'W,  20. VII.  1 942,  J. 
D.  Ritchie  (UK). 

REMARKS.— Development  of  the  ventral  setae 
and  tubercle  of  the  posterior  femur  shows  con- 
siderable variation.  Specimens  from  the  Atlantic 
Coast  have  long  setae,  a  distinct  nipple-like  tu- 
bercle, and  genitalia  confirming  substantially  to 
Figures  25  and  27.  Specimens  collected  between 
Ontario  and  California  have  short  setae,  and  more 
often  a  gentle  prominence  rather  than  a  tubercle, 
with  genitalia  resembling  Figures  24  and  26.  Cal- 


ROGERS:  AMERICAN  DRAPET1S 


189 


28 


29 


FIGURES  28-30.    Drapetis  latipennis.  Figure  28.  Posterior  view  of  male  hind  leg,  Douglas  Co.,  Kansas.  Figures  29  (ventrolateral 
view)  and  30  (dorsolateral  view).  Male  genitalia,  Orange  Co.,  California. 


ifornia  material  often  shows  well-developed  tu- 
bercles and  genitalia  similar  to  Atlantic  Coast 
specimens.  Abdominal  tergites  VI  and  VII  are 
mostly  polished  in  specimens  from  Lost  Creek, 


Idaho.  Specimens  from  Velvet  Creek,  Oregon, 
and  Coeur  d'Alene  Lake,  Idaho,  lack  all  pro- 
pleural  tomentum.  Sternite  V  is  undivided  in 
specimens  from  Big  Meadow  Creek,  Idaho. 


190 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  7 


31 


32 


mm 


FIGURES  3 1-33.    Drapetis  Solaris  n.  sp.,  paratype  male.  Figure  3 1 .  Anterior  view  of  femur  of  center  leg.  Figures  32  (ventrolateral 
view)  and  33  (dorsolateral  view).  Genitalia. 


Individuals  from  the  Pacific  Coast  have  so  far 
been  collected  only  in  mountainous  regions, 
where  wing  infumation  occurs  only  sporadically 
and  is  coincident  with  generalized  melanism. 

Drapetis  latipennis  Melander 

(Figs.  9,  28-30) 

Drapetis  latipennis  Melander,  1902:209. 

DIAGNOSIS.— Three  strong  pairs  of  verticals; 
gena  obscured  by  eye  in  profile;  propleural  to- 
mentum  profuse,  extending  to  anteroventral  lat- 
eral corner  of  humerus;  posterior  femur  bearing 


five  to  seven  long,  sturdy  anteroventral  bristles 
along  distal  quarter;  posterior  tibia  bearing  a 
group  of  long,  down  turned  setae  near  distal  third 
of  extensor  surface;  genitalia  as  in  Figures  29,  30. 
TYPES.  —  Lectotype  (here  designated):  male 
(USNM),  USA:  labelled,  "Lawrence  Kans." 
Paralectotypes,  same  locality:  1  male  (USNM), 
3  females  (AMNH  730).  A  male  from  Milwau- 
kee, Wisconsin  (AMNH  729)  included  in  the 
original  type  series  is  Drapetis  infumata. 

OTHER  MATERIAL  EXAMINED.— USA:  California:  Orange  Co., 
2  males,  Irvine  Ranch  nr.  Bonita  Creek.,  "on  Platanus," 
24. V.I 963,  E.  I.  Schlinger  (UCR).  Illinois:  Champaign  Co.,  1 


ROGERS:  AMERICAN  DRAPETIS 


191 


0.1  mm 


FIGURES  34-36.    Drapetis  torulosa  n.  sp.,  paratype  male.  Figure  34.  Anterior  view  of  center  leg.  Tomentose  area  indicated 
by  stippling.  Figures  35  (ventrolateral  view)  and  36  (dorsolateral  view).  Genitalia. 


male,  Urbana,  29.VI.1915  (IHNS).  Kansas:  Douglas  Co.,  1 
male,  Atherton  (USNM).  New  York:  Ulster  Co.,  1  male,  Cher- 
rytown,  8/1 8.VIII.  1971,  P.  and  B.  Wygodzinsky  (AMNH). 

REMARKS.— Flattened  spines  of  abdominal 
tergites  IV  and  V  are  long,  thick  and  erect.  The 
peculiar  long  setae  on  the  posterior  tibia  are  pres- 
ent in  both  sexes,  but  are  shorter  in  females; 
Drapetis  assimilis  Meigen  has  less  developed  long 
setae  in  the  same  location. 

Drapetis  Solaris  n.  sp. 

(Figures  4,  11,31-33) 

DIAGNOSIS.— Center  femur  bearing  ventrally 
directed  setae  across  center  of  anterior  surface, 
Figure  31;  genitalia  as  in  Figures  32,  33. 


DESCRIPTION.— Male.  Length  1.6  mm.  Head 
black;  three  pairs  of  verticals,  outer  two  short; 
frons  width  at  ventral  end  one-sixth  length  of 
third  antennal  segment;  face  linear;  gena  entirely 
visible  in  profile;  third  antennal  segment  shaped 
as  in  Figure  4,  arista  seven  times  its  length.  Tho- 
rax black,  except  posterior  edge  of  mesopleuron 
and  dorsal  edge  of  sternopleuron  dark  brown; 
humerus  bearing  an  inconspicuous  bristle;  pro- 
pleural  tomentum  limited  to  ventral  margin;  hal- 
ter light  brown.  Wing  1.6  mm  long,  neuration  as 
in  Figure  1 1 .  Legs  and  coxae  chiefly  dark  brown, 
slightly  lighter  at  knees,  on  tarsal  articles  and 
anterior  tibia,  and  in  ventral  half  of  anterior  coxa; 
center  femur  bearing  a  single  row  of  ventral  spines 
along  entire  femoral  length,  spines  longer  proxi- 


192 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  7 


mally,  anterior  surface  of  femur  bearing  ventrally 
directed  setae  and  an  unusually  sturdy  antero- 
ventral  distal  bristle;  center  tibia  very  slightly 
bowed;  a  row  of  very  long  ventral  setae  along 
entire  length  of  posterior  femur.  Abdomen  lat- 
erally bearing  long,  flattened  spines  on  tergites 
IV  and  V. 

TYPES. -Holotype:  male  (CAS  13455),  MEX- 
ICO: labelled,  "MEXICO:  Mexico:  Teotihuacan 
Pyramid  to  the  Sun.  27.XII.1970,  P.  H.  and  M. 
Arnaud/Coll.  at  flowers  Cassia  tomentosa  L.f. 
Fam.  Leguminosae  Det.  D.  Breedlove."  Para- 
types:  3  males,  same  data  (CAS). 

REMARKS.— The  face  of  one  paratype  is  con- 
stricted ventrally,  and  another  has  no  discernible 
outer  vertical.  The  species  is  named  in  reference 
to  its  type  locality. 

Drapetis  torulosa  n.  sp. 

(Figs.  5,  8,  34-36) 

DIAGNOSIS.— Three  pairs  of  strong  verticals; 
anterodorsal  surface  of  center  femur  with  a  to- 
mentose  callosity  erupting  at  proximal  quarter 
and  gradually  subsiding  distally,  Figure  34,  pos- 
teroventral  surface  bearing  one  row  of  bristles, 
bristles  lengthening  distally;  center  tibia  lacking 
flexor  spines;  genitalia  as  in  Figures  35,  36. 

DESCRIPTION.  — Male.  Length  1.4  mm.  Head 
black;  three  pairs  of  verticals;  frons  width  at  ven- 
tral end  one-eighth  length  of  third  antennal  seg- 
ment; face  strongly  constricted  at  ventral  end; 
gena  entirely  visible  in  profile;  third  antennal 
segment  shaped  as  in  Figure  5,  arista  five  times 
its  length.  Thorax  dark  brown;  propleuron  com- 
pletely polished,  three  weak  pale  setae  along  ver- 
tical suture;  halter  dark  yellow.  Wing  1.6  mm 
long,  neuration  as  in  Figure  8.  Legs  and  coxae 
yellow,  except  tarsi  yellowish  brown;  center  fe- 
mur with  one  very  strong,  nearly  erect  antero- 
ventral  distal  bristle  near  knee,  and  strong  yellow 
posteroventral  bristles  in  a  single  row,  bristles 
lengthening  distally  along  proximal  three-quar- 
ters of  femur;  anterodorsal  surface  of  center  fe- 
mur thickly  tomentose,  with  a  long  swelling  be- 
ginning at  proximal  quarter  and  distally  subsiding 
across  remaining  length  of  femur;  center  tibia 
lacking  flexor  spines;  posterior  femur  lacking 
ventral  setae,  bearing  a  rounded  basal  lump  on 
posteroventral  surface  just  distad  of  trochanter. 
Abdomen  laterally  bearing  tiny  flattened  spines 
on  tergites  IV  and  V,  closely  appressed  to  body. 


TYPES. -Holotype:  male  (CAS  13459),  MEX- 
ICO: labelled,  "MEXICO:  Baja  California:  Agua 
Caliente  (San  Carlos),  18.5  km  east  of  Maneade- 
ro,  6. VII.  1973,  Paul  H.  Arnaud,  Jr."  Paratypes, 
same  data,  1  male  (AC),  2  males  (PHA). 

REMARKS.— Abdominal  sternite  V  of  a  macer- 
ated paratype  is  divided  through  the  center.  This 
species  is  named  in  reference  to  the  peculiar 
modification  of  the  center  femur. 

ACKNOWLEDGMENTS 

Assistance  of  the  following  persons  is  grate- 
fully recognized:  for  loans  of  specimens:  P.  H. 
Arnaud,  Jr.,  G.  W.  Byers,  S.  I.  Frommer,  L.  V. 
Knutson,  the  late  J.  L.  Laffoon,  A.  F.  Newton, 
J.  A.  Powell,  R.  O.  Schuster,  H.  J.  Teskey,  F.  C. 
Thompson,  W.  J.  Turner,  the  late  P.  W.  Wygod- 
zinsky;  for  loan  of  equipment:  H.  V.  Daly;  for 
help  in  manuscript  preparation:  C.  C.  Magowan, 
M.  B.  McAdow,  and  E.  I.  Schlinger;  for  funding: 
B.  A.  Czuba  and  the  late  A.  P.  Philips. 

LITERATURE  CITED 

BIGOT,  J.  F.  M.  1857.  Essai  d'une  classification  generate  et 
synoptique  de  1'ordre  des  Insectes  Dipt6res.  Ann.  Soc.  Ento- 
mol.Fr.  26:551-564. 

CHVALA,  M.  1975.  The  Tachydromiinae  (Dipt.  Empididae) 
of  Fennoscandia  and  Denmark,  Vol.  3.  Scandinavian  Sci- 
ence Press,  Klampenborg.  336  pp. 

COLLIN,  J.  E.  1961.  Empididae.  British  flies,  Vol.  VI.  Cam- 
bridge University  Press,  Cambridge.  782  pp. 

GRIFFITHS,  G.  C.  D.  1972.  The  phylogenetic  classification  of 
Diptera  Cyclorrhapha,  with  special  reference  to  the  structure 
of  the  male  postabdomen.  Dr.  W.  Junk,  The  Hague.  340  pp. 

KOVALEV,  V.  G.  1 972.  Diptera  of  the  genera  Drapetis  Mg. 
and  Crossopalpus  Bigot  (Empididae)  of  the  European  part 
of  the  USSR.  Entomol.  Obozr.  61:173-196. 

MEIGEN,  J.  W.  1822.  Systematische  Beschreibung  der  be- 
kannten  europaischen  zweifliigeligen  Insekten,  Vol.  III. 
Hamm.  416  pp. 

MELANDER,  A.  L.  1 902.  A  monograph  of  the  North  American 
Empididae.  Part  I.  Trans.  Am.  Entomol.  Soc.  28:195-367, 
pis.  5-9. 

.    1918.   The  dipterous  genus  Drapetis  Meigen  (family 

Empididae).  Ann.  Entomol.  Soc.  Am.  11:183-221. 

-.    1928.  Diptera.  Fam.  Empididae.  Genera  Insectorum 


185:1-434. 
.   1965.  Family  Empididae  (Empidae,  Hybotidae).  Pp. 

446-48 1  in  A  catalog  of  the  Diptera  of  America  north  of 

Mexico.  A.  Stone,  C.  W.  Sabrosky,  W.  W.  Wirth,  R.  H. 

Foote,  and  J.  R.  Coulson,  eds.  U.S.D.A.,  Agric.  Res.  Serv., 

Agric.  Handb.  276. 
ROGERS,  E.   1983.  The  Neotropical  species  of  Drapetis  Meigen 

(Diptera:  Empididae).  Syst.  Entomol.  8:431-452. 


PROCEEDINGS 

OF  THE 

CALIFORNIA  ACADEMY  OF  SCIENCES 


Vol.  46,  No.  8,  pp.  193-207,  2  figs.  December  20,  1989 


ECOLOGICAL,  EVOLUTIONARY,  AND  CONSERVATION 

IMPLICATIONS  OF  FEEDING  BIOLOGY  IN  OLD  WORLD 

CAT  SNAKES,  GENUS  BOIGA  (COLUBRIDAE) 

By 

Harry  W.  Greene* 

Museum  of  Vertebrate  Zoology, 
University  of  California,  Berkeley,  California  94720 

*  Research  Associate,  Department  of  Herpetology, 

Fellow,  California  Academy  of  Sciences, 
Golden  Gate  Park,  San  Francisco,  California  94118 


ABSTRACT:  Cat  snakes  forage  actively  and  sometimes  consume  multiple  prey  items  at  a  site.  Some  species 
use  an  unusual  style  of  constriction  and  venom  to  immobilize  prey,  which  is  typically  swallowed  head-first. 
Approximately  300  prey  items  from  21  species  demonstrate  intra-  and  interindividual,  ontogenetic,  and  geo- 
graphic dietary  variation.  Most  species  of  Boiga  are  small  to  moderately  long  snakes  that  eat  mainly  lizards, 
or  huge  snakes  that  eat  lizards  as  juveniles  and  mainly  mammals  and/or  birds  as  adults.  Probably  predation 
on  birds  by  introduced  B.  irregularis  on  Guam  is  a  retained  primitive  feeding  pattern,  whereas  that  species' 
diet  of  rodents  on  New  Guinea  and  euryphagy  by  B.  dendrophila  throughout  its  range  are  correlated  with 
local  ecological  factors.  With  few  exceptions  (e.g.,  a  mouse-deer  in  B.  dendrophila),  prey/predator  mass  ratios 
are  not  large  compared  to  other  snakes.  Morphological  and  ecological  considerations  suggest  that,  like  B. 
irregularis,  several  other  species  of  cat  snakes  could  have  catastrophic  impact  as  introduced  predators  on 
island  bird  and  bat  faunas. 

Received  February  16, 1989.  Accepted  July  6,  1989. 


evolutionary,  and  conservation  implications  of 

The  Old  World  genus  Boiga  includes  approx-  that  variation. 

imately  25  species  of  moderate  to  large,  rear-  The  feeding  biology  of  cat  snakes  is  of  special 
fanged  colubrids  (Leviton  1970;  Kroon  1973;  interest  for  at  least  three  reasons: 
Rasmussen  1979),  often  called  cat  or  tree  snakes  (1)  Boiga  is  morphologically  and  ecologically 
because  of  their  nocturnal,  arboreal  habits  and  diverse,  and  thus  offers  interesting  material  for 
vertical  pupils.  This  report  surveys  published  in-  broader  evolutionary  considerations.  For  ex- 
formation  on  feeding  in  cat  snakes,  and  analyzes  ample,  although  rear-fanged  snakes  play  prom- 
additional  data  based  on  the  examination  of  mu-  inent  roles  in  discussions  of  venom  evolution 
seum  specimens.  It  addresses  the  nature  of  di-  (e.g.,  Smith  and  Bellairs  1947;  Kardong  1979, 
etary  and  behavioral  variation  within  and  among  1 980;  Savitzky  1 980),  an  understanding  of  that 
2 1  species  of  Boiga,  and  discusses  ecological,  problem  is  hampered  by  a  lack  of  information 

[193] 


194 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  8 


on  the  feeding  biology  and  phylogenetic  rela- 
tionships of  informative  taxa  (Cadle  1983). 

(2)  Cat  snakes  might  be  important  arboreal 
predators  on  vertebrates  in  many  Old  World 
tropical  ecosystems,  and  three  to  five  species  of 
Boiga  sometimes  occur  sympatrically  (e.g.,  Tay- 
lor 1922a;  Lim  1955;  Lloyd  et  al.  1968;  Kroon 
1973;  Inger  and  Colwell  1977).  Autecological 
analyses  provide  data  for  subsequent  commu- 
nity-level research,  information  that  could  not 
be  obtained  for  snakes  at  a  single  site  during  a 
short-term  study  (Greene  1988a). 

(3)  Predation  by  Boiga  irregularis  was  a  key 
factor  in  the  recent,  precipitous  decline  of  bird 
populations  on  the  Pacific  island  of  Guam,  where 
that  species  was  inadvertently  introduced  in  the 
1940s  (Savidge  1984,  1987,  1988;  Conry  1988; 
Fritts  1988).  Island  populations  of  fruit  bats,  al- 
ready widely  threatened  by  direct  human  per- 
secution, might  be  at  risk  as  well  (Wiles  1988). 
Data  on  the  natural  diets  of  cat  snakes  could 
prove  useful  in  managing  this  problem  and  pre- 
venting additional  catastrophes. 

METHODS 

Stomach  contents  were  examined  through 
ventral  slits  in  specimens  in  the  Academy  of  Nat- 
ural Sciences  of  Philadelphia  (ANSP);  American 
Museum  of  Natural  History,  New  York  (AMNH); 
California  Academy  of  Sciences,  San  Francisco 
(CAS);  Field  Museum  of  Natural  History,  Chi- 
cago (FMNH);  Museum  of  Comparative  Zool- 
ogy, Harvard  University,  Cambridge  (MCZ); 
Museum  of  Southwestern  Biology,  University  of 
New  Mexico,  Albuquerque  (MSB);  Museum  of 
Vertebrate  Zoology,  University  of  California, 
Berkeley  (MVZ);  and  National  Museum  of  Nat- 
ural History,  Washington  (USNM).  I  avoided 
especially  soft,  brittle,  or  otherwise  fragile  spec- 
imens; all  others  for  each  species  were  examined 
to  preclude  a  bias  for  large  prey  caused  by  open- 
ing only  snakes  with  externally  palpable  food 
items.  My  results  also  incorporate  data  on  10 
New  Guinea  specimens  of  Boiga  irregularis  that 
contained  prey  (R.  Shine,  in  litt.)  in  the  Austra- 
lian Museum  (AM)  and  Queensland  Museum 
(QM),  and  specific  published  information  on 
feeding.  I  took  care  to  account  for  redundancy 
among  literature  records  and  museum  specimens 
(e.g.,  Schmidt  1927  and  Pope  1935;  Smith  1916 
and  1930),  and  sometimes  obtained  information 
not  mentioned  previously. 


Direction  of  ingestion,  identity,  and  linear  di- 
mensions of  prey  items  were  recorded  whenever 
possible.  Intact  prey  items  (or  a  reference  spec- 
imen of  comparable  size)  and  the  predators  were 
weighed  after  blotting  and  draining  them  briefly 
on  paper  towels.  Abbreviations  refer  to  snout- 
vent  length  (SVL),  total  length  (TL),  maximum 
total  length  (MTL),  and  prey/predator  mass  ratio 
(MR).  All  bird  or  lizard  eggs  in  a  stomach  were 
counted  as  one  item,  because  they  might  repre- 
sent a  single  location  event  and  because  some- 
times the  exact  number  of  items  was  impossible 
to  determine. 

SPECIES  ACCOUNTS 
Boiga  angulata  (Peters  1861) 

Philippine  Islands,  MTL  1.4  m  (Leviton  1970). 
Four  snakes  contained  five  agamid  lizards  ( 1  Ca- 
lotessp.,3Dracosp.,x=  1.3/snake;  Griffin  1910; 
CAS  154175,  CAS-SU  24179;  FMNH  61620). 
MRs  for  two  Draco  sp.  in  snakes  with  SVLs  of 
646  and  1,079  mm  were  0.39  and  0.12,  respec- 
tively. Two  Draco  sp.  and  an  unidentified  agamid 
were  swallowed  head-first. 

Boiga  blandingi  (Hallowell  1857), 
Blanding's  Tree  Snake 

Africa,  MTL  2.7  m  (Pitman  1974;  Goodman 
1985).  Fifteen  specimens  contained  at  least  30 
prey  items  (x  =  2.0/snake):  a  1.8  m  TL  snake 
contained  1  agamid  lizard  (Agama  agamd)  and 
ca.  8  bats  (Eptesicus  tenuipennis,  Vespertilioni- 
dae;  Tadarida  sp.,  Molossidae);  5  each  contained 
1  bat;  1  contained  bats  and  1  lizard;  1  contained 
a  chameleonid  lizard;  1  contained  4  birds;  1  con- 
tained a  sunbird  (Nectarinafamosa);  4  each  con- 
tained 1  bird,  2  swallowed  head-first;  and  1  con- 
tained 3  bats,  swallowed  tail-first,  MRs  =  0.03/ 
each  (Laurent  1956;  Woodward  1960;  Barry 
1961;  Cozens  1961;  Dunger  1961;  Menzies  1961; 
Pitman  1962;  Hedges  1983;  FMNH  179319-20; 
MCZ  54007).  Three  adults  in  one  tree  each  con- 
tained young  weaver  birds,  Plesiositagra  cucul- 
latus,  and  others  contained  fledgling  birds  and 
immature  rats  (Jones  1961). 

Boiga  ceylonensis  (Giinther  1858), 
Sri  Lankan  Cat  Snake 

India  and  Sri  Lanka,  MTL  1 .6  m  (Smith  1 943). 
Forty-two  prey  items  included  8  frogs  (1  Rana 


GREENE:  OLD  WORLD  CAT  SNAKES 


195 


limnocharis,  Ranidae;  5  Philautus  sp.,  Rhaco- 
phoridae),  21  lizards  (1  Gekkonidae;  2  Calotes 
sp.,  3  C.  versicolor,  1  Charasia  dor  sails,  3  Salea 
horsfieldi,  Agamidae),  3  birds  (1  fledgling),  and 
2  mice  (Wall  1919  [including  snakes  referred  to 
as  B.  nuchalis],  1923, 1924;Hutton  1949;  Taylor 
1 953;  FMNH  1 3 1 380, 1 67006).  Two  lizards  were 
swallowed  head-first. 

Boiga  cyanea  (Dumeril  and  Bibron  1854), 
Green  Cat  Snake 

India  and  Southeast  Asia,  MTL  2.3  m  (Smith 
1943).  Seven  snakes  contained  1  snake  (a  female 
B.  cyanea,  1,350  mm  TL,  in  a  1,260  mm  TL 
male)  and  6  birds  (1  sparrow  [Passer  sp.,  Pas- 
seridae],  1  probable  wagtail  [Motacilla  sp.,  Mo- 
tacillidae,  in  a  snake  of  821  mm  SVL],  1  week- 
old  domestic  chick  [Gallus  gallus,  Phasianidae]; 
Smith  and  Kloss  1915;  Saint  Girons  1972;  Frith 
1977;  MCZ  58260).  One  bird  had  been  swal- 
lowed head-first. 

Boiga  cynodon  (Boie  1827), 
Dog-toothed  Cat  Snake 

Southeast  Asia  and  Indo- Australian  Archipel- 
ago; MTL  2.8  m  (Smith  1 943;  Saint  Girons  1 972; 
Kroon  1973;  Tweedie  1983;  FMNH  131809). 
Nineteen  snakes  contained  20  items  (x  =  1.05/ 
snake),  including  3  lizards  (Draco  sp.,  D.  volans, 
Agamidae;  Gekko  gecko,  Gekkonidae),  1 2  birds, 
4  or  more  bird  eggs,  and  1  rodent  (Flower  1899; 
Auffenberg  1980;  ANSP  26421;  CAS  8533, 
125173;  FMNH  53454-6,  71628,  131809, 
131811,  150883,  158652-3,  168006,  183749). 
One  lizard,  1  rodent,  and  5  birds  were  swallowed 
head-first.  A  small  snake  (SVL  =  360  mm)  had 
eaten  a  Draco  sp.,  and  a  juvenile  (SVL  =  640 
mm)  contained  a  relatively  large  bird  (MR  = 
0.44).  An  intact  bird's  egg,  maximum  diameter 
51  mm,  was  swallowed  by  a  snake  with  head 
length  of  52.2  mm. 

Boiga  dendrophila  (Boie  1827), 
Mangrove  Snake 

Southeast  Asia  and  nearby  archipelagos,  MTL 
2.8  m  (Burger  1975;  Tweedie  1983).  Fifty-four 
prey  items  in  52  snakes  (Jc  =  1 .04/snake)  included 
1  frog,  15  lizards  (9  scincids  [2  Lamprolepis 
smaragdinum];  2  lacertids,  Tachydromus  sexlin- 
eatus;  3  agamids  [2  Calotes  sp.,  1  C.  cristatelus]; 


1  varanid,  Varanus  salvator);  1  snakes  (1  B.  cyn- 
odon, 2  Chrysopelea  ornata,  1  Psammodynastes 
pulverulentus,  Colubridae;  1  Calloselasma  rho- 
dostoma,  Viperidae);  17  mammals  (2  bats  [1 
Eonycteris  spelea,  Pteropodidae],  1  mouse-deer 
[Tragulus  javanicus,  Tragulidae],  14  rodents 
[probably  mostly  Muridae,  including  Rattus  sp.]); 
and  13  birds  (Flower  1899;  Smith  1916,  1930; 
Taylor  1922b;  Brongersma  1934;  Harrison  1962; 
Leviton  1970;  Lim  1973;  Minton  and  Dunson 
1978;  Smith  1985;  ANSP  264 17-8;  CAS  12362- 
3,  15313;  CAS-SU  28525-6;  FMNH  53460, 
53462,  53466,  68912,  71630,  96609,  120393, 
131808,  150879,  158679,  178602,  230061). 

Juvenile  Boiga  dendrophila  eat  mainly  ecto- 
therms  (lizards  and  rarely  snakes),  whereas  adults 
eat  mainly  birds  and/or  mammals  (5/6  times  vs. 
25/34  items,  respectively,  using  Brongersma's 
[1934]  sample  and  museum  records;  P  =  0.01, 
Fisher  Exact  Test).  Sample  sizes  are  insufficient 
for  statistical  comparisons,  but  I  detect  no  evi- 
dence for  major  geographic  variation  in  the  diet 
of  adults:  frequencies  of  lizards,  snakes,  birds, 
and  mammals,  respectively,  are  1,  1,  5,  and  3 
for  Borneo;  5,  2,  2,  and  5  for  the  Philippine  Is- 
lands (all  of  the  mammals  were  in  Palawan  man- 
grove snakes);  2,  1,3,  and  0  for  Sulawesi;  and  1, 
3,  0,  and  2  for  mainland  Asia.  The  largest  snake 
containing  a  lizard  and  the  smallest  containing 
an  endotherm  had  SVLs  of  1 , 1 1 2  mm  and  840 
mm,  respectively.  Four  lizards,  3  snakes,  3  birds, 
4  rodents,  and  1  adult  mouse-deer  were  swal- 
lowed head-first,  the  latter  by  a  1.7  m  TL  snake. 
MRs  for  2  lizards  and  a  bird  were  0.10-0.18  (x 
=  0.14). 

Boiga  dightoni  (Boulenger  1894) 

India,  MTL  1.3  m  (Smith  1943).  One  adult 
(SVL  =  915  mm)  contained  an  agamid  lizard, 
Calotes  versicolor,  swallowed  head-first,  MR  = 
0.24  (FMNH  217699  [Inger  et  al.  1984]). 

Boiga  drapiezii  (Boie  1827), 
White-spotted  Cat  Snake 

Southeast  Asia,  MTL  2.1  m  (Tweedie  1954, 
1983).  Eight  prey  items  included  5  lizards  (1 
Gekkonidae;  1  Draco  sp.,  1  Gonocephalus  sp., 
Agamidae),  1  rhacophorid  frog,  and  2  birds  (Lim 
1956;  FMNH  131804,  138594,  196805-6).  Both 
agamids  and  the  frog  were  swallowed  head-first. 
MRs  for  the  gecko  and  Draco  were  both  0.25. 


196 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  8 


Boiga  forsteni  (Dumeril  and  Bibron  1854) 

Sri  Lanka  and  India,  MTL  2.3  m  (de  Silva 
1 980).  Stomachs  contained  birds  and  a  bat  (Wall 
1921). 


Boiga  gokool  (Gray  1834), 
Eastern  Gamma  Cat  Snake 

Asia  and  India,  MTL  1.2m  (Smith  1 943).  Two 
snakes  contained  a  mouse  and  an  agamid  lizard, 
the  latter  swallowed  head-first  (Wall  1910;  MCZ 
58261). 


Boiga  irregularis  (Merrem  1790), 
Brown  Tree  Snake 

New  Guinea,  northern  Australia,  and  nearby 
archipelagos;  MTL  2.3  m  (Fritts  1988).  Fifty- 
nine  snakes  contained  65  items  (1-3  per  stom- 
ach, x  =  1.10),  including  2  frogs  (1  Rana  sp., 
Ranidae),  19  lizards  (1  agamid,  Gonocephalus 
modestus;  10  scincids  [2  Emoia  cyanura,  3  Lam- 
prolepis  smaragdinum],  8  gekkonids  [1  Gehyra 
sp.,  2  G.  mutilata,  1  Gekko  vittatus]),  13  birds 
(1  passerine,  1  owlet-nightjar  [Aegotheles  ben- 
nettii,  Aegothelidae],  2  starlings  [Aplonis  sp., 
Sturnidae],  1  quail  [Coturnix  chinensis,  Phasian- 
idae]),  and  30  mammals  (1  marsupial  bandicoot, 
Peramelidae;  1  bat;  1  shrew,  Croddura  sp.;  17 
rodents  (5  juveniles,  1  adult  Rattus  exulans,  1 
R.  rattus,  1  Mus  musculus,  all  murids;  Rooij  1917; 
Kopstein  1926;  Hediger  1933;  Pendleton  1947; 
McDowell  1984;  AM  4553, 11825, 12443, 12542, 
14858,  86923,  87463,  87466;  AMNH  40233, 
42339,  42371,  59083,  62031-2,  65513,  73962, 
85727-8,  85731,  95553,  95607,  101080,  101083; 
CAS  49910,  94027,  103388,  113600,  121224; 
FMNH  14076,  41980;  QM  4381,  13.1280; 
USNM  6197,  61918,  61920,  159971,  195594, 
213431,  215938,  215942,  215947,  216007, 
216009,  237136,  237634).  Stomachs  with  mul- 
tiple prey  included  1  with  2  E.  cyanura,  1  with 
2  G.  mutilata,  1  with  2  young  Aplonis  sp.,  and  1 
with  3  nestling  rodents  (apparently  a  single 
species).  Three  "bird"  records  consisted  of  1  egg, 
several  eggs,  and  1  bird  plus  2-5  eggs. 

Forty-eight  measurable  individuals  containing 
prey  had  SVLs  of  292-1,710  mm.  A  285  mm 
SVL  specimen  had  an  obvious  umbilical  scar, 
and  six  females  containing  enlarged  ova  or  shelled 
oviductal  eggs  had  SVLs  of  840-1,305  mm.  On 


those  bases,  I  divided  snakes  containing  prey  into 
"juveniles"  (equal  to  or  less  than  750  mm  SVL) 
and  "adults"  (greater  than  750  mm  SVL).  Ju- 
veniles take  significantly  more  ectothermic  prey 
(snake,  frogs,  lizards)  vs.  endothermic  prey  (birds, 
mammals),  whereas  adults  show  the  opposite  bias 
(14:3  vs.  6:32,  respectively;  P  <  0.01,  x2  test). 
The  largest  snake  containing  a  lizard  was  1 ,000 
mm  SVL,  the  smallest  containing  an  endotherm 
was  646  mm  SVL,  and  the  shift  from  emphasis 
on  one  prey  type  to  the  other  occurs  at  ca.  800- 
900  mm  SVL  (Fig.  1).  Larger  snakes  ate  larger 
prey  than  did  smaller  snakes  (Fig.  2).  Prey/pred- 
ator mass  ratios  ranged  from  0.004-0.241  (x  = 
0.106,  n  =  13),  and  the  largest  item  was  a  32  g 
starling  (Aplonis  sp.)  in  a  152  g  Boiga  irregularis 
from  the  Bismark  Archipelago. 

The  65  prey  items  are  from  throughout  the 
species'  range  except  for  Australia,  where  the  diet 
of  this  species  is  under  study  by  R.  Shine.  There 
is  significant  geographic  variation  in  diet  (P  < 
0.01,  x2  test)  if  38  records  for  endothermic  prey 
are  divided  into  those  from  mainland  New 
Guinea  vs.  those  from  smaller  islands  (all  others 
except  Sulawesi,  including  the  Solomon,  Bis- 
marck, and  Molucca  groups).  Snakes  from  New 
Guinea  ate  mammals  more  frequently  than  they 
ate  birds  (19:3),  in  agreement  with  summaries  of 
unspecified  numbers  of  prey  (Room  1 974;  Parker 
1982)  and  general  comments  about  diet  on  the 
tags  of  AMNH  74506  and  82334.  The  single  diet 
record  from  Sulawesi,  also  a  "large  island,"  was 
of  a  mammal.  Snakes  from  "smaller  islands" 
consumed  birds  and  their  eggs  more  frequently 
than  they  ate  mammals  (10:6). 

Direction  of  ingestion  was  determined  for  36 
prey  items.  Twenty-eight  were  swallowed  head- 
first, including  2  frogs,  8  lizards,  1 1  birds,  1  bat, 
1  snake,  and  5  rodents.  Seven  were  swallowed 
tail-first,  including  2  lizards,  1  bird,  and  4  ro- 
dents. One  lizard  was  swallowed  bent  doubled. 

Savidge  (1 988)  analyzed  494  prey  items  in  353 
Boiga  irregularis  from  an  introduced  population 
on  Guam.  As  in  natural  populations,  those  snakes 
ate  primarily  lizards  as  juveniles  and  birds  and/ 
or  mammals  as  adults.  Her  mean  of  1.4  items/ 
snake  is  higher  than  I  observed,  perhaps  because 
she  counted  eggs  as  individual  items  in  a  stom- 
ach. Wiles  (1988)  reported  three  juvenile  fruit 
bats  in  the  stomach  of  a  2.5  m  TL  snake  on 
Guam,  and  indirect  evidence  that  these  snakes 
were  raiding  bat  roosts. 


GREENE:  OLD  WORLD  CAT  SNAKES 


197 


Boiga  jaspidea  (Dumeril  and  Bibron  1854), 
Jasper  Cat  Snake 

Southeast  Asia,  MTL  1.5  m  (Taylor  1965; 
Tweedie  1983).  Stomachs  contained  1  agamid 
lizard  (Calotes  sp.),  1  bird,  and  1  tree-mouse 
(Chiropodomys gliroides,  Muridae)  in  a  1,213  mm 
TL  snake  (Lim  1956,  1967). 

Boiga  kraepelini  Stejneger  (1902) 

East  Asia  and  nearby  islands,  MTL  1.6  m 
(Kuntz  1963).  Four  snakes  contained  2  agamid 
lizards  (swallowed  head-first),  2  birds,  and  2  bird 
eggs,  ready  to  hatch  (Pope  1935;  CAS  18891-2; 
FMNH  24986). 

Boiga  multifasciata  (Blyth  1860), 
Himalayan  Cat  Snake 

Himalayan  region,  MTL  1.06  m  (Smith  1943). 
One  contained  an  agamid  lizard,  Japalura  var- 
iegata(Wa\\  1909). 


Boiga  multomaculata  (Boie  1827), 
Large-spotted  Cat  Snake 

East  Asia,  Java,  Sumatra,  and  Borneo;  MTL 
1.2  m  (Smith  1943).  Stomachs  contained  6  birds 
(1  Motadlla  sp.,  Motacillidae)  and  8  agamid  liz- 
ards (Calotes  sp.,  C.  versicolor);  three  lizards  (1 
with  MR  =  0.58)  and  a  bird  were  swallowed 
head-first  (Wall  1926a,  b;  Schmidt  1927;  Pope 
1935;  FMNH  26451,  105704,  178666;  MSB 
40769). 

Boiga  nigriceps  (Gunther  1863), 
Dark-headed  Cat  Snake 

South  Asia  and  nearby  archipelagos;  MTL  1.7 
m  (Rooij  1917).  One  snake  contained  a  bird, 
swallowed  head-first  (FMNH  128150). 

Boiga  ocellata  Kroon  (1973) 

South  Asia;  MTL  1.7  m  (Kroon  1973).  Two 
adults  contained  a  bird  and  a  bird  egg  (FMNH 
191998-9). 

Boiga  ochracea  (Gunther  1868), 
Tawny  Cat  Snake 

Asia,  MTL  1.3m  (Smith  1 943).  Ten  prey  items 
included  8  agamid  lizards  (2  Japalura  variegata, 


3  Calotes  sp.),  1  fledgling  bird,  and  1  mass  of 
bird  eggs;  three  of  the  agamids  were  swallowed 
head-first  (Wall  1909,  1926a,  as  B.  hexagonatus; 
CAS  12365,  95252). 

Boiga  pulverulenta  (Fischer  1856), 
Powdered  Tree  Snake 

Africa,  MTL  1.3  m  (Pitman  1974).  Ten  snakes 
contained  13  prey  items  (x  =  1.3/snake):  3  mice 
(2  juvenile  Praomys  sp.,  Muridae,  in  1  snake),  2 
agamid  lizards,  and  8  birds  (Werner  1909;  Pit- 
man 1974;  FMNH  4017,  59006-7;  LACM 
38688,  46328;  MCZ  29354,  29357;  USNM 
223775).  Two  lizards  and  6  birds  were  swallowed 
head-first;  2  snakes  containing  lizards  were  330 
and  850  mm  SVL,  whereas  5  containing  birds 
were  ca.  700-780  mm  SVL.  MRs  were  0.12  for 
each  lizard  and  0. 1 5  for  each  of  4  birds  in  2 
snakes  (x  —  1 .4).  One  record  of  a  bat  in  this 
species  (Woodward  1960)  is  based  on  a  misi- 
dentified  B.  blandingi  (see  above). 

Boiga  schultzei  Taylor  (1923) 

Philippine  Islands;  MTL  1.4  m  (Taylor  1923). 
A  juvenile,  SVL  310  mm,  had  swallowed  2  gek- 
konid  lizards,  Gehyra  mutilata,  head  first  (CAS 
62153). 

Boiga  trigonata  (Schneider  in  Bechstein  1 802), 
Gamma  Cat  Snake 

Southern  Asia,  India,  Sri  Lanka;  MTL  1.2  m 
(Smith  1 943). ". . .  the  commonest  contents  were 
birds,  chickens  and  eggs"  (Lindberg  1932).  Four- 
teen snakes  contained  1 5  prey  (x  =  1 .07  items/ 
snake),  including  1 1  lizards  (Acanthodactylus  sp., 
Lacertidae;  at  least  4  Calotes  sp.,  2  C.  versicolor, 
Agamidae;  1  scincid;  1  set  of  3  lizard  eggs)  and 
3  birds  (1  Saxicoloides  fulica,  Muscicapidae);  two 
lizards  and  1  bird  were  swallowed  head-first  (Wall 
1907a,  1919,  1921;  Minton  1966;  Sharma  and 
Vazirani  1977;  CAS  13686;  FMNH  131958, 
166998,  171761).  Ants  and  grasshoppers  in  the 
stomach  of  a  small  Indian  specimen  (Sunder- 
singh  1960)  were  probably  remnants  from  the 
stomach  contents  of  a  prey  lizard. 

DISCUSSION 

INDIVIDUAL  AND  GEOGRAPHIC  DIETARY 
VARIATION.  — Intraspecific  variation  in  snake 
diets  can  include  intra-  and  interindividual,  on- 


198 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  8 


togenetic,  seasonal,  and  geographic  components 
(Arnold  1977;  Greene  1984;  Seib  1985;  Mushin- 
sky  1987;  Macias  Garcia  and  Drummond  1988; 
Savidge  1988).  At  one  extreme,  Thamnophis  ele- 
gans  (Colubridae)  shows  extensive  geographic 
variation,  and  litter  mates  have  genetic  predis- 
positions for  alternative  prey  types  (Arnold  1977); 
conversely,  the  diet  of  Micrurus  fulvius  (Elapi- 
dae)  varies  only  slightly  at  all  sizes  and  through- 
out the  year  over  a  large  geographic  range  (Greene 
1984). 

In  studies  of  museum  specimens,  evidence  for 
intraindividual  dietary  variation  comes  from 
multiple  prey  types  in  single  stomachs,  as  in  a 
Boiga  ceylonensis  that  contained  a  bird  and  two 
mice,  and  two  B.  blandingi  that  each  contained 
bats  and  a  lizard.  All  such  records  for  B.  angu- 
lata,  B.  cynodon,  B.  dendrophila,  B.  irregularis, 
B.  kraepelini,  B.  ochracea,  B.  pulverulenta,  B. 
schultzei,  and  B.  trigonata  apparently  are  mul- 
tiples of  one  prey  species,  as  are  several  records 
for  B.  blandingi.  Of  353  B.  irregularis  on  Guam 
that  contained  prey,  67  had  more  than  one  item 
of  the  same  "prey  type"  (e.g.,  lizards  and/or  liz- 
ard eggs,  birds  and/or  bird  eggs);  14  had  lizards 
and  birds  or  mammals,  as  expected  for  small 
adults;  and  only  five  (<6%  of  stomachs  with 
multiple  items)  contained  birds  and/or  their  eggs 
and  mammals  (Savidge  1988).  These  data  sug- 
gest that  individual  cat  snakes  rarely  eat  birds 
and  mammals,  at  least  within  a  short  time  pe- 
riod. Some  species,  however,  often  consume  more 
than  one  of  the  same  prey  type  (and  even  same 
prey  species),  and  perhaps  this  is  especially  true 
for  B.  blandingi.  Surely  the  clumped  distribution 
of  roosting  bats  and  nestling  birds  facilitates  mul- 
tiple ingestion. 

There  is  significant  geographic  dietary  varia- 
tion in  Boiga  irregularis  within  its  natural  range, 
and  microgeographic  variation  in  predation  on 
rodents  vs.  birds  in  urban  vs.  forest  habitats  by 
introduced  snakes  on  Guam  (data  from  Savidge 
[1988,  table  1],  x2  test,  P  <  0.01).  The  data  for 
other  species  of  Boiga  are  inadequate  to  distin- 
guish geographic  from  individual  and  ontoge- 
netic  variation. 

Perhaps  diet  is  plastic  at  some  localities,  or 
throughout  all  or  part  of  the  range  of  some  species 
of  Boiga.  If  so,  interspecific  and  geographic  pat- 
terns could  result  from  individual  responses  to 
local  prey  characteristics  and/or  availability, 
rather  than  from  dietary  evolution.  The  infre- 


quent occurrence  of  multiple  prey  types  in  in- 
dividual stomachs  of  several  species  of  Boiga  is 
consistent  with  stable  prey  preferences,  whereas 
there  evidently  has  been  a  rapid  local  response 
by  B.  irregularis  to  shifting  prey  densities  on 
Guam  (Savidge  1988,  and  see  above;  Green  well 
et  al.  [1984]  described  a  possible  example  in- 
volving Thamnophis  sirtalis  and  birds  on  an  is- 
land, and  Chiszar  [1989]  discussed  the  behav- 
ioral basis  for  prey  switching  in  B.  irregularis). 
Genetic  differentiation  and  phenotypic  plasticity 
are  not  mutually  exclusive,  in  that  response  to 
local  conditions  could  override  genealogically 
constrained  prey  preferences,  and  a  thorough  as- 
sessment of  these  possibilities  for  cat  snakes  will 
require  studies  of  the  ontogeny  and  control  of 
feeding  behavior  in  several  populations. 

PREY/PREDATOR  SIZE  RELATIONSHIPS  AND 
ONTOGENETIC  VARIATION.— As  in  other  snakes 
(Greene  1984;  Seib  1984,  1985),  larger  individ- 
uals of  Boiga  eat  larger  prey,  and  also  continue 
to  eat  relatively  small  prey  (Fig.  2).  Most  MRs 
for  eight  species  of  Boiga  are  not  especially  large 
compared  with  those  of  other  snakes  (0.004-0.58, 
x  =  0.16,  n  =  32;  cf.  Voris  and  Moffett  1981; 
Greene  1983,  1984,  1989;  Seib  1984,  1985;Jayne 
et  al.  1988).  However,  a  1.7  m  B.  dendrophila 
would  weigh  no  more  than  1  kg  (pers.  obs.)  and 
lesser  mouse-deer  weigh  0.7-2.0  kg  (Lekagul  and 
McNeely  1977),  so  the  latter  is  a  relatively  large 
item.  A  bandicoot  in  B.  irregularis  and  prey 
snakes  longer  than  the  B.  cyanea  and  B.  dendro- 
phila that  contained  them  (Smith  and  Kloss  1915; 
Smith  1916)  likewise  were  surely  relatively  heavy; 
a  captive  B.  irregularis  consumed  a  domestic  ro- 
dent with  MR  =  0.60  (Chiszar  1989).  Records 
for  B.  irregularis  and  B.  pulverulenta  demon- 
strate that  a  cat  snake  can  ingest  at  least  a  third 
of  its  mass  if  it  eats  more  than  one  item  in  a  nest 
or  roost. 

Data  from  its  natural  range  and  the  introduced 
Guam  population  agree  that  Boiga  irregularis 
changes  from  eating  mainly  lizards  to  eating 
mainly  birds  and/or  mammals  at  a  SVL  of  ca. 
80-100  cm.  Similar  shifts  evidently  occur  in  B. 
dendrophila  and  B.  pulverulenta,  and  seem  likely 
for  other  large  species  of  Boiga  (Fig.  1).  Perhaps 
that  pattern  reflects  simply  an  ability  of  larger 
snakes  to  swallow  larger  prey,  since  slopes  and 
intercepts  for  endothermic  vs.  ectothermic  prey 
are  similar  (analysis  of  covariance  for  data  in  Fig. 
2,  P  >  0.05).  However,  large  snakes  eat  endo- 


GREENE:  OLD  WORLD  CAT  SNAKES 


199 


V) 

I«H 

o 

CD 

DC 


CD 

12H 


•    Endothermic  prey 
U    Ectothermic  prey 


Boiga  irregularis 

N  =  47 


1 


o 
o 

0) 
DC 


I 1 1 r— 

20    30    40    50    60    70    80    90    100110120130140150160170180190200210 


B   8- 


SVL  (cm) 


6- 


4H 


CD 
.O 


•    Endothermic  prey 
ED    Ectothermic  prey 


(17  species) 

N  =  69 


1 1.. 1. 1 


i 1 1 1 1 1 1 1— r 

20    30    40    50    60    70    80    90   100110120130140150160170180190200210 


SVL  (cm) 


FIGURE  1 .    Records  for  ectothermic  and  endothermic  prey  with  respect  to  predator  snout-vent  length  in  cat  snakes,  genus 
Boiga.  For  this  comparison,  multiple  items  of  one  prey  type  in  a  single  stomach  were  counted  once. 


therms  that  are  smaller  than  the  ectotherms  tak- 
en by  smaller  snakes  (Fig.  2),  implying  that  larger 
snakes  fail  to  encounter  small  ectotherms  or  have 
undergone  a  real  change  in  prey  preference. 

FEEDING  BEHAVIOR.— Most  Boiga  prey  mainly 
on  diurnal,  arboreal  lizards  (especially  agamids 
of  the  genus  Calotes)  and  diurnal  birds.  Studies 
on  free-living  and  captive  B.  irregularis  suggest 
that  these  nocturnal  hunters  must  search  for  rath- 
er than  ambush  inactive  prey  (Fritts  et  al.  1987; 
Savidge  1988;  Chiszar  1989;  Chiszar  etal.  1985, 
1 988).  Jones  (in  Pitman  1958)  saw  three  B.  blan- 


dingi  "pushing  their  heads  into  one  nest  after  the 
other  and  taking  the  young  [weaver  birds].  ..." 
However,  a  B.  dendrophila  constricting  an  Eo- 
nycteris  spelea  in  a  tree  after  dark  probably  am- 
bushed the  bat  while  it  was  feeding  on  or  ap- 
proaching fruit,  as  that  species  typically  roosts 
in  caves  (Harrison  1962). 

Some  Boiga  cynodon,  B.  dendrophila,  and  B. 
trigonata  constrict  prey  with  the  tail  as  well  as 
hold  and  presumably  envenomate  an  animal  with 
their  teeth  (Wall  1907b,  1921;  Barach  1952; 
Murphy  1977;  however,  Lim  et  al.  [1960]  stated 


200 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  8 


2.0- 


O. 

D) 


0.5- 


0 


A 
A 


0.0 

0.0  0.5  1.0  1.5  2.0          2.5  3.0 

Log  Predator  Mass 

FIGURE  2.  Relationship  between  log  predator  mass  and  log 
prey  mass  for  Boiga  irregularis  (triangles,  n  =  8)  and  eight 
other  species  of  Boiga  (circles,  n  =  13).  Hollow  symbols  in- 
dicate ectothermic  prey;  solid  symbols  indicate  endothermic 
prey.  For  this  comparison,  multiple  items  of  one  prey  type  in 
a  single  stomach  were  counted  once. 


that  B.  cynodon  does  not  constrict  prey).  Boiga 
multomaculata  constricts  lizards,  but  the  precise 
method  is  unknown  (Campden-Main  1 970).  Boi- 
ga blandingi  and  B.  ceylonensis  do  not  constrict 
prey,  whereas  B.  forsteni  and  B.  irregularis  vary 
in  this  regard  (Groves  1973;  Murphy  1977;  de 
Silva  1980;  Parker  1982;  Shine  and  Schwaner 
1985;  Fritts  1988).  Young  (1983)  observed  that 
B.  blandingi  controlled  prey  with  loops  of  the 
body,  but  did  not  constrict;  that  B.  cyanea  over- 
powered live  mice  with  constriction  "in  a  man- 
ner similar  to  that  of  a  rat  snake  or  a  king  snake"; 
that  B.  cynodon  did  not  constrict,  but  controlled 
live  prey  with  loose  coils  of  the  anterior  body; 
and  that  B.  dendrophila  was  "an  efficient  and 
powerful  constrictor . . .  upon  warm  blooded  prey, 
but  it  rarely  employed  this  behavior  ...  on 
snakes."  Captive  B.  irregularis  swallow  small  prey 
immediately  and  apply  constricting  coils  to  rel- 
atively heavy  prey  (Chiszar  1989). 

Additional  inter-  and  intraspecific  compari- 
sons will  be  required  to  elucidate  the  variation 
in  and  evolution  of  prey-immobilization  behav- 
ior in  Boiga.  Constriction  with  the  tail  is  rare 
among  snakes  (Murphy  1977;  Greene  1977)  and 
absent  in  the  outgroup  to  Boiga  (see  below),  sug- 
gesting three  possibilities:  (1)  It  is  a  derived  state 
for  Boiga,  either  as  yet  unobserved  or  secondarily 
lost  in  most  species;  (2)  some  other  derived  at- 
tribute of  Boiga  (e.g.,  arboreality,  long  tails,  mul- 
tiple constriction  of  nestlings)  favors  this  type  of 
constriction  and  it  arose  repeatedly  in  the  genus; 
or  (3)  constriction  with  the  tail  is  a  derived  at- 


tribute for  a  monophyletic  subset  of  Boiga,  a 
behavioral  synapomorphy  for  those  species. 

The  venom  of  B.  blandingi  is  neurotoxic  and 
can  kill  a  large  rodent  or  a  small  monkey  in  less 
than  an  hour  (Wakeman  1966;  Groves  1973; 
Levinson  et  al.  1976),  and  at  least  some  other 
species  have  venom  that  is  slightly  or  moderately 
toxic  to  humans  (e.g.,  Deraniyagala  1955;  Whit- 
aker  1970;  Burger  1975;  Minton  and  Dunson 
1978;  Cox  1988;  Fritts  1988).  A  captive  adult  B. 
dendrophila  (TL  ca.  1.8  m)  had  difficulty  sub- 
duing and  swallowing  adult  laboratory  rats  (Ba- 
rach  1952),  but  those  animals  can  be  powerful 
opponents  for  captive  snakes  (pers.  obs.). 

Head-first  prey  ingestion  is  the  norm  in  nature 
for  Boiga  irregularis  (28/36  items,  P  <  0.01,  x2 
test),  as  it  is  for  many  other  snakes  (e.g.,  Greene 
1976,  1983,  1984,  1989;  Ananjeva  and  Orlov 
1982;  Voris  and  Voris  1983;  Seib  1984,  1985). 
Of  6 1  cases  for  which  direction  of  ingestion  was 
determined  for  1 6  other  species  of  Boiga,  58  were 
swallowed  head-first;  3  bats  were  eaten  tail-first. 

EVOLUTIONARY  IMPLICATIONS.— An  evolu- 
tionary consideration  of  feeding  necessitates  the 
categorization  of  dietary  variation  into  alterna- 
tive states  and  the  identification  of  appropriate 
outgroups,  so  that  the  distribution  of  states  among 
populations,  taxa,  and  outgroups  can  be  used  to 
specify  transformations  during  the  history  of  that 
clade  (see  Arnold  1981;  Kluge  1987;  Losos  and 
Greene  1988).  A  phylogenetic  consideration  of 
feeding  in  Boiga  and  its  relatives,  although  based 
on  incomplete  data,  has  implications  that  justify 
the  following  preliminary  analysis. 

The  precise  relationships  of  Boiga  to  other  col- 
ubrids  are  unknown;  Rasmussen  (1979,  1985) 
placed  it  with  the  African  genera  Crotaphopeltis, 
Dipsadoboa,  and  Telescopus,  so  these  are  used 
here  as  a  collective  outgroup.  Species  of  Crota- 
phopeltis are  terrestrial  and  feed  on  toads.  Species 
of  Dipsadoboa  are  arboreal  and  feed  on  frogs  and 
lizards.  The  Eurasian  species  of  Telescopus  evi- 
dently are  terrestrial,  whereas  the  African  T.  dhara 
and  T.  semiannulatus  are  arboreal;  all  eat  lizards 
and  occasionally  birds  and  bats.  Species  in  all 
three  genera  are  nocturnal  and  reach  MTLs  of 
ca.  1  m  (Corkhill  1932;  Pitman  1974;  Broadley 
1983). 

Most  species  of  Boiga  are  small  to  moderately 
long  snakes  that  eat  lizards  throughout  their  lives, 
or  large  snakes  that  eat  lizards  as  juveniles  and 
endotherms  as  adults.  Division  of  the  second  of 


GREENE:  OLD  WORLD  CAT  SNAKES 


201 


those  qualitative  groups  into  four  subgroups 
yields  a  total  of  five  dietary  character  states: 

Diet-1  consists  primarily  of  lizards  throughout 
life,  with  an  occasional  bird  or  mammal.  Boiga 
ceylonensis  possesses  this  diet,  as  do  perhaps  B. 
angulata,  B.  dightoni,  B.  gokool,  and  B.  ocracea. 
These  species  have  MTLs  of  1.2-1.6  m. 

Diet-2  consists  of  lizards  in  juveniles,  mainly 
of  birds  in  adults.  Boiga  cynodon,  some  popu- 
lations of  B.  irregularis,  and  B.  multomaculata 
exhibit  this  diet,  as  do  perhaps  B.  drapiezii,  B. 
forsteni,  B.  kraepeleni,  B.  nigriceps,  B.  ocellata, 
and  B.  trigonata.  These  species  have  MTLs  of 
1.2-2.8  m. 

Diet-3  consists  of  lizards  in  juveniles,  mainly 
of  rodents  in  adults.  Only  Boiga  irregularis  (MTL 
=  2.3  m)  on  New  Guinea  exhibits  this  state. 

Diet-4  consists  of  lizards  in  juveniles,  mainly 
of  mammals  and  birds  in  adults.  Boiga  blandingi 
and  perhaps  B.  cyanea,  B.  jaspidea,  and  B.  pul- 
verulentus  exhibit  this  state,  and  have  MTLs  of 
1.3-2.8  m.  However,  sample  sizes  are  quite  small 
except  for  the  first  species,  and  in  no  case  can 
the  possibility  of  geographic  variation  for  diet-2 
and  diet-3  be  excluded.  Also,  bats  are  probably 
functionally  equivalent  to  birds  in  shape  and  ar- 
boreal roosting  behavior  (Cundall  and  Greene, 
unpubl.;  many  other  snakes  that  eat  bats  also  eat 
birds,  Schaetti  1 984),  so  that  B.  blandingi  per- 
haps should  be  assigned  diet-2. 

Diet- 5  consists  of  lizards  in  juveniles,  mainly 
of  snakes,  birds,  and  mammals  in  adults.  Stom- 
ach contents  and  observations  on  captives  (Young 
1983)  suggest  that  Boiga  dendrophila  is  more  eu- 
ryphagic  than  other  species  in  the  genus. 

Ingroup  and  outgroup  comparisons  indicate 
that  Boiga  is  primitively  nocturnal  and  perhaps 
derivatively  arboreal.  Diet-1  and  relatively  small 
MTL  are  found  in  a  few  species  and  in  the  out- 
group,  but  lacking  knowledge  of  intrageneric  re- 
lationships, we  cannot  determine  if  their  pres- 
ence in  Boiga  is  primitive  or  a  reversal.  Diet-2 
and  large  size  are  widespread  in  species  of  Boiga 
and  absent  in  the  outgroup  (with  the  possible 
exception  of  diet-2  in  some  Telescopus),  but  we 
cannot  be  sure  if  those  characteristics  are  derived 
for  the  genus  or  within  Boiga,  perhaps  more  than 
once. 

Diet-2  is  probably  primitive  for  adult  Boiga 
irregularis,  based  on  widespread  occurrence  in 
the  genus,  suggesting  that  diet-3  is  derived  within 
that  species,  but  the  reasons  for  intraspecific  geo- 


graphic variation  are  not  known.  Birds  are  rarely 
eaten  by  B.  irregularis  in  New  Guinea,  although 
the  island  has  a  rich  avifauna  (Beehler  et  al.  1 986). 
Several  species  of  New  Guinea  birds  exhibit  nest- 
ing attributes  that  are  possibly  antipredator  spe- 
cializations (J.  M.  Diamond,  University  of  Cal- 
ifornia, Los  Angeles,  in  litt.),  but  perhaps  so  do 
birds  on  other  islands  and  the  mainland  that  are 
commonly  eaten  by  other  species  of  Boiga.  A 
niche  shift  in  response  to  competitive  pressure 
seems  unlikely,  because  no  other  New  Guinea 
snakes  are  known  to  feed  heavily  on  birds 
(McDowell  1969, 1972, 1975, 1984;  Shine  1980a; 
Parker  1982;  Malnate  and  Underwood  1988). 
Perhaps  New  Guinea  arboreal  rodents  are  so 
much  more  diverse  and  numerous  than  on  near- 
by archipelagos  (Laurie  and  Hill  1954;  Menzies 
and  Dennis  1979)  that  they  provide  "better"  food 
per  unit  effort  or  are  simply  encountered  more 
often  than  birds. 

Based  on  ingroup  and  outgroup  comparisons, 
diet-4  and  diet-5  are  probably  derived  within 
Boiga.  Euryphagy  evidently  reaches  a  maximum 
for  the  genus  in  B.  dendrophila  and  is  correlated 
with  a  derived  preference  for  riparian  habitats 
(Taylor  1965;  Frith  1977;  Minton  and  Dunson 
1 978;  Dring  1 979;  Cranbrook  and  Furtado  1 988; 
this  proclivity  is  not  reported  for  other  species 
of  Boiga  or  the  outgroup  genera).  Some  theoret- 
ical foraging  models  predict  that  habitat  spe- 
cialization results  in  broader  diets  and  use  of 
suboptimal  prey  (e.g.,  MacArthur  and  Pianka 
1966;  Krebs  et  al.  1983).  The  relevance  of  those 
models  for  cat  snakes  could  be  explored  with 
captive  animals,  in  part,  by  testing  the  presump- 
tion that  alternative  prey  types  vary  in  profitabil- 
ity (see  Godley  1980). 

My  findings  on  feeding  biology  in  Boiga  un- 
derscore some  cautionary  lessons  with  regard  to 
the  evolution  of  rear-fanged  snakes: 

(1)  Ontogenetic  and/or  geographic  variation 
can  be  sufficiently  large  that  a  proportional  rep- 
resentation of  overall  diet  for  a  species  will  ob- 
scure functional,  biogeographic,  and  phyloge- 
netic  correlates:  the  overall  sample  for  Boiga 
irregularis  from  throughout  its  natural  range  con- 
sists of  approximately  32%  lizards,  21%  birds, 
and  44%  mammals,  but  most  individuals  feed 
entirely  on  lizards  as  juveniles  and  on  either 
mainly  rodents  or  mainly  birds  as  adults. 

(2)  The  preceding  analysis  assumes  that  diets 
really  are  attributes  of  populations  and  taxa,  and 


202 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  8 


that  they  have  evolved  conservatively  (i.e.,  with 
minimal  homoplasy).  High  rates  of  reversals  and 
convergence  can  only  be  detected  by  detailed 
phylogenetic  analysis,  so  there  is  as  yet  no  basis 
for  postulating  this  alternative  for  Boiga.  The 
other  possibility,  that  individual  dietary  plastic- 
ity accounts  fully  for  geographic  and  systematic 
patterns,  is  discussed  above  (see  INDIVIDUAL  AND 
GEOGRAPHIC  DIETARY  VARIATION). 

(3)  There  is  no  evidence  that  the  venom  ap- 
paratus of  Boiga  is  functionally  associated  with 
feeding  on  prey  that  are  especially  heavy  or  oth- 
erwise costly,  although  rare  ingestion  of  large  prey 
remains  a  possibility  (see  B.  dendrophila,  above, 
and  Greene  1986a). 

(4)  Rear  fangs  arose  at  the  level  of  Boiga  or, 
more  likely,  a  more  inclusive  group  (Rasmussen 
1 979, 1 985).  Whatever  the  ecological  role  of  ven- 
om injection  (if  any)  in  extant  cat  snakes,  its 
possible  adaptive  origin  (if  any)  in  the  clade  con- 
taining Boiga  cannot  be  identified  until  addi- 
tional information  on  suprageneric  relationships 
and  natural  history  are  available  (see  Cadle  1983; 
Greene  1986a;  Schaefer  and  Lauder  1986).  To 
the  extent  those  limitations  apply  to  other  clades, 
conclusions  about  the  adaptive  significance  of 
rear-fanged  dentition  are  premature. 

ECOMORPHOLOGICAL  SUITES  AND  CONVERGENT 

ANALOGUES.— Qualitative  ecomorphological  an- 
alogs have  been  identified  among  certain  geo- 
graphically and  phylogenetically  distant  snakes 
(e.g.,  Shine  1980a,  b,  c;  Henderson  and  Binder 
1980),  and  suggestive  parallels  can  be  drawn  be- 
tween species  of  Boiga  and  certain  other  taxa. 
Some  resemblance  to  Afro-Asian  species  of  Te- 
lescopus  might  be  either  convergent  or  homol- 
ogous (Rasmussen  1979,  1985);  the  two  genera 
are  largely  allopatric.  Rear-fanged  colubrids  of 
the  Malagasy  genus  Lycodryas  (Rax worthy  1988) 
and  neotropical  genus  Imantodes  (especially  the 
I.  cenchoa  species  group,  Myers  1982)  are  ar- 
boreal, nocturnal,  and  resemble  small  species  of 
Boiga  in  general  body  form  (e.g.,  large  head  and 
eyes,  slender  neck)  and  diet.  The  New  World, 
rear-fanged  genus  Trimorphodon  is  said  to  be 
related  to  Boiga  (Dowling  et  al.  1983)  or  to  sev- 
eral genera  of  New  World  colubrines  (Minton 
and  Salanitro  1972;  Cadle  1988;  R.  Lawson,  CAS, 
pers.  comm.).  Both  species  of  Trimorphodon  are 
arboreal,  nocturnal,  and  feed  largely  on  sleeping, 
diurnal  iguanid  lizards;  only  certain  populations 
of  T.  biscutatus  occasionally  include  birds  and 


bats  in  the  diet  (Greene,  unpubl.  data).  A  non- 
venomous,  diurnal  neotropical  colubrid,  Pseustes 
poecilonotus,  resembles  B.  dendrophila  in  size, 
diet,  and  defensive  behavior;  however,  even  ju- 
venile P.  poecilonotus  eat  small  birds  (Greene 
1979,  and  unpubl.  data).  I  detect  no  particular 
convergent  resemblance,  beyond  arboreality,  ir- 
rascibility  (Greene  1988b),  and  nocturnality,  be- 
tween Boiga  and  Australian  elapids  of  the  genus 
Hoplocephalus  (Shine  1983). 

CAT  SNAKES  AS  INTRODUCED  PREDATORS.— 
This  study  provides  potentially  useful  insights 
with  regard  to  the  threat  that  introduced  cat 
snakes  pose  for  island  prey  faunas: 

(1)  Heavy  predation  on  birds  by  Boiga  irre- 
gularis  on  Guam  probably  reflects  the  primitive 
diet  for  this  species  (see  above),  rather  than  a 
feeding  shift  in  response  to  special  local  condi- 
tions. It  is  not  surprising  that  this  snake  makes 
heavy  inroads  on  island  bird  populations,  es- 
pecially if  acceptable  alternative  prey  (e.g.,  ro- 
dents, see  above)  are  unavailable  or  rare,  and/or 
if  the  native  avifauna  has  evolved  behavioral  and 
demographic  attributes  in  the  absence  of  pred- 
ators (see  Jackson  [1974]  for  a  bird  with  behav- 
ioral defenses  against  nest  predation  by  snakes). 
The  effects  of  an  introduced  snake  on  native  is- 
land prey  populations  might  be  all  the  more  rapid 
in  the  absence  of  natural  predators  on  snakes 
(Greene  1988a). 

(2)  Neonate  Boiga  irregularis  are  evidently  too 
small  to  eat  endothermic  prey,  and  starvation  of 
juveniles  is  probably  a  major  source  of  natural 
mortality  in  some  snake  populations  (e.g.,  Fitch 
1 949).  Potential  control  measures  for  introduced 
Boiga  might  include  manipulating  the  availabil- 
ity or  characteristics  (e.g.,  artificial  toxicity)  of 
lizard  prey. 

(3)  Boiga  irregularis  is  a  moderately  large  snake 
with  a  venom  delivery  system  and  constricting 
behavior  that  could  permit  it  to  subdue  a  wide 
range  of  prey  sizes  and  types.  Like  several  other 
species  of  Boiga,  it  possesses  a  relatively  wide 
skull  and  long  dentaries  (Marx  and  Rabb  1972), 
derived  characteristics  that  are  likely  associated 
with  increased  gape  (Rasmussen  1979;  Arnold 
1983),  thereby  facilitating  a  diet  of  birds,  eggs, 
and  bats  (Cundall  and  Greene,  unpubl.  data). 
Although  not  so  highly  specialized  as  to  be  un- 
able to  forage  on  the  ground,  evidently  all  species 
of  Boiga  climb  well  and  are  nocturnal. 

Each  of  the  above  attributes  of  cat  snakes  might 


GREENE:  OLD  WORLD  CAT  SNAKES 


203 


enhance  their  ability  to  feed  on  sleeping  diurnal 
birds  and  their  eggs,  or  on  roosting  and  foraging 
bats.  For  those  reasons  and/or  because  of  known 
predation  on  birds  and  bats,  all  moderate  to  large 
Boiga  are  potentially  catastrophic  invaders  of 
isolated  island  faunas.  Species  for  special  concern 
include  B.  blandingi,  B.  cyanea,  B.  cynodon,  B. 
dendrophila,  B.  drapeizii,  B.  forsteni,  B.  irregu- 
laris,  B.  jaspidea,  B.  kraepelini,  B.  nigriceps,  B. 
ocellata,  B.  pulverulenta,  B.  schnitzel,  and  B.  tri- 
gonata.  There  is  a  clear  potential  for  additional 
introductions:  recently  individual  B.  irregularis 
have  reached  several  Pacific  islands  (Fritts  1 988), 
at  least  one  B.  dendrophila  landed  on  Taiwan  in 
a  shipment  of  lumber  (A.  H.  T.  Yu,  pers.  comm.), 
and  several  species  of  large  Boiga  are  commonly 
available  in  the  pet  trade  (pers.  obs.). 

EPILOGUE 

I  have  never  seen  a  Boiga  in  the  field,  and  this 
research  was  made  possible  by  the  efforts  of  nu- 
merous collectors,  curators,  authors  of  anec- 
dotes, and  sponsors  of  natural  history  publica- 
tions. Few  of  those  individuals  intended  their 
specimens  for  studies  on  the  evolution  of  feeding, 
and  even  fewer  imagined  that  their  labors  would 
have  implications  for  an  urgent  conservation 
problem.  Nevertheless,  starting  from  scratch  to 
learn  something  comparable  about  the  broader 
context  for  avian  predation  by  introduced  cat 
snakes  on  Guam  would  have  required  many  years 
of  work.  Therein  surely  lie  some  lessons  (see  also 
Greene  1986b;  Greene  and  Losos  1988). 

ACKNOWLEDGMENTS 

I  thank  P.  Alberch,  J.  E.  Cadle,  R.  C.  Drewes, 
R.  F.  Inger,  A.  E.  Leviton,  H.  Marx,  R.  W. 
McDiarmid,  C.  W.  Myers,  H.  L.  Snell,  and  J.  W. 
Wright  for  access  to  specimens;  B.  Beehler,  W. 

C.  Brown,  R.  I.  Crombie,  J.  M.  Diamond,  T.  H. 
Fritts,  M.  LeCroy,  W.  Z.  Lidicker,  C.  Luke,  E. 

D.  Pierson,  R.  Shine,  A.  H.  T.  Yu,  R.  Zusi,  and 
R.  G.  Zweifel  for  identifying  prey  items,  supply- 
ing data,  or  other  help;  A.  E.  Leviton  for  patiently 
encouraging  my  interests  in  Asiatic  snakes;  J.  E. 
Cadle  for  critical  comments  on  the  manuscript; 
and  the  U.S.  Fish  and  Wildlife  Service,  National 
Science  Foundation  (BSR  83-00346),  and  Field 
Museum  of  Natural  History  (Karl  P.  Schmidt 
Fund)  for  financial  support. 


LITERATURE  CITED 

ANANJEVA,  N.  B.  AND  N.  L.  ORLOV.  1982.  Feeding  behaviour 
of  snakes.  Vertebrata  Hungarica  21:25-31. 

ARNOLD,  E.  N.  1981.  Estimating  phytogenies  at  low  taxo- 
nomic  levels.  Z.  Zool.  Syst.  Evolutionforsch.  19:1-35. 

ARNOLD,  S.  J.  1977.  Polymorphism  and  geographic  variation 
in  the  feeding  behavior  of  the  garter  snake  Thamnophis  ele- 
gans.  Science  197:676-678. 

.  1983.  Morphology,  performance  and  fitness.  Amer. 

Zool.  23:347-361. 

AUFFENBERG,  W.  1980.  The  herpetofauna  of  Komodo,  with 
notes  on  adjacent  areas.  Bull.  Florida  State  Mus.  25:39-156. 

BARACH,  J.  P.  1952.  Some  observations  on  the  mangrove 
snake  in  captivity.  Herpetologica  8:102-103. 

BARRY,  D.  H.  1961.  Correspondence:  bat-eating  snakes.  Ni- 
ger. Field  26:42-43. 

BECHSTEIN,  J.  M.  1 802.  Herrn  De  la  Cepede's  Naturgeschich- 
te  der  Amphibien  oder  der  eyerlegenden  vierfussigen  Thiere 
und  der  Schlangen.  Eine  Fortsetzung  von  Buffon's  Natur- 
geschichte  aus  dem  Franzosischen  ubersetzt  und  mil  An- 
merkungen  und  Zusatsen  versehen,  Vol.  4.  Industrie  Comp- 
toir,  Weimar. 

BEEHLER,  B.  M.,  T.  K.  PRATT,  AND  D.  A.  ZIMMERMAN.  1986. 
Birds  of  New  Guinea.  Princeton  Univ.  Press,  Princeton,  New 
Jersey. 

BLYTH,  E.  1860.  Report  of  curator,  zoological  department. 
J.  Asiatic  Soc.  Bengal  29:87-1 15. 

BOIE,  F.  1827.  Bemerkungen  tiber  der  Familien  und  Gattun- 
gen  der  Ophidier.  Isis  von  Oken  20:506-566. 

BOULENGER,  G.  A.  1894.  Description  of  a  new  snake  found 
in  Travancore  by  Mr.  S.  Dighton,  Pirmaad.  J.  Bombay  Nat. 
Hist.  Soc.  8:528. 

BROADLEY,  D.  G.  1983.  FitzSimons' snakes  of  southern  Af- 
rica. Delta  Books,  Johannesburg,  South  Africa. 

BRONGERSMA,  L.  D.  1 934.  Contributions  to  Indo-Australian 
herpetology.  Zool.  Meded.  Rijksmus.  Nat.  Hist.  17:161-251. 

BURGER,  W.  L.  1975.  A  case  of  mild  envenomation  by  the 
mangrove  snake,  Boiga  dendrophila.  Snake  6:99-100. 

CADLE,  J.  E.  1983.  Problems  and  approaches  in  the  inter- 
pretation of  the  evolutionary  history  of  venomous  snakes. 
Mem.  Inst.  Butantan  46:255-274. 

.  1988.  Phylogenetic  relationships  among  advanced 

snakes:  a  molecular  perspective.  Univ.  California  Publ.  Zool. 
119:1-77. 

CAMPDEN-MAIN,  S.  M.  1970.  A  field  guide  to  the  snakes  of 
South  Vietnam.  Smithsonian  Inst.,  Washington,  D.C. 

CHISZAR,  D.  A.  1989.  Behavior  of  the  brown  tree  snake  (Boi- 
ga irregularis):  a  study  in  applied  comparative  psychology. 
In  press  in  Contemporary  issues  in  comparative  psychology. 
D.  A.  Dewsbury,  ed.  Sinauer  Associates,  Sunderland,  Mas- 
sachusetts. 

CHISZAR,  D.,  D.  CARRILLO,  P.  RAND,  AND  J.  CHISZAR.  1985. 
Nocturnal  activity  in  captive  brown  tree  snakes  (Boiga  ir- 
regularis). Bull.  Maryland  Herpetol.  Soc.  21:115-118. 

CHISZAR,  D.,  K.  KANDLER,  R.  LEE,  AND  H.  M.  SMITH.  1988. 
Stimulus  control  of  predatory  attack  in  the  brown  tree  snake 
(Boiga  irregularis).  2.  Use  of  chemical  cues  during  foraging. 
Amphibia-Reptilia  9:77-88. 

CONRY,  P.  J.  1 988.  High  nest  predation  by  brown  tree  snakes 
on  Guam.  Condor  90:478-482. 

CORKHILL,  N.  L.  1932.  The  snakes  of  Iraq.  J.  Bombay  Nat. 
Hist.  Soc.  35:552-572. 


204 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  8 


Cox,  M.  J.  1988.  Serious  effects  from  the  bite  of  the  red  cat- 
eye  snake,  Boiga  nigriceps.  Bull.  Chicago  Herp.  Soc.  23:162. 

COZENS,  B.  1 96 1 .  Correspondence:  bat-eating  snakes.  Niger. 
Field  26:44. 

CRANBROOK,  E.  AND  J.  I.  FURTADO.  1988.  Freshwaters.  Pp. 
225-250  in  Key  environments:  Malaysia.  E.  Cranbrook,  ed. 
Pergamon  Press,  Oxford. 

DERANIYAGALA,  P.  E.  P.  1955.  A  colored  atlas  of  some  ver- 
tebrates from  Ceylon,  Vol.  3.  Serpentoid  Reptilia.  Govt. 
Press,  Colombo. 

DE  SILVA,  P.  H.  D.  H.  1980.  Snake  fauna  of  Sri  Lanka,  with 
special  reference  to  skull,  dentition  and  venom  in  snakes. 
Dept.  Govt.  Printing,  Colombo,  Sri  Lanka. 

DOWLING,  H.  G.,  R.  HIGHTON,  G.  C.  MAHA,  AND  L.  R.  MAXSON. 
1983.  Biochemical  evaluation  of  colubrid  snake  phylogeny. 
J.Zool.  201:309-329. 

DRING,  J.  C.  M.  1979.  Amphibians  and  reptiles  from  north- 
ern Trengganu,  Malaysia,  with  descriptions  of  two  new  geck- 
os: CnemaspisanA  Cyrtodactylus.  Bull.  Brit.  Mus.  (Nat.  Hist.) 
34:181-241. 

DUMERIL,  A.  M.  C.  AND  G.  BiBRON.  1854.  Erpetologie  gene- 
rale  ou  histoire  naturelle  complete  des  reptiles.  Paris. 

DUNGER,  G.  T.  1961.  Correspondence:  bird-eating  snakes. 
Niger.  Field  26:43. 

FISCHER,  J.  G.  1856.  Neue  Schlangen  des  Hamburgischen 
Naturhistorischen  Museums.  Abh.  Naturwiss.  Ver.  Ham- 
burg 3:81. 

FITCH,  H.  S.  1949.  Study  of  snake  populations  in  central 
California.  Amer.  Midi.  Nat.  41:1-150. 

FLOWER,  S.  S.  1 899.  Notes  on  a  collection  of  reptiles  made 
in  the  Malay  Peninsula  and  Siam  from  November  1896  to 
September  1898,  with  a  list  of  species  recorded  from  those 
countries.  Proc.  Zool.  Soc.  London  1899:600-696. 

FRITH,  C.  B.  1977.  A  survey  of  the  snakes  of  Phulet  Island 
and  the  adjacent  mainland  areas  of  peninsular  Thailand. 
Nat.  Hist.  Bull.  Siam  Soc.  26:263-316. 

FRITTS,  T.  H.  1988.  The  brown  tree  snake,  Boiga  irregularis, 
a  threat  to  Pacific  islands.  U.S.  Fish  Wild.  Serv.  Biol.  Rept. 
88:1-36. 

FRITTS,  T.  H.,  N.  J.  SCOTT,  JR.,  AND  J.  A.  SAVIDGE.  1987. 
Activity  of  the  arboreal  brown  tree  snake  (Boiga  irregularis) 
on  Guam  determined  by  electrical  outages.  Snake  19:5 1-58. 

GODLEY,  J.  S.  1980.  Foraging  ecology  of  the  striped  swamp 
snake,  Regina  alleni,  in  southern  Florida.  Ecol.  Monogr.  50: 
411-436. 

GOODMAN,  J.  D.  1985.  Two  record  size  Blanding's  tree  snakes 
from  Uganda.  East  Afr.  Nat.  Hist.  Soc.  Bull.  1985:56-57. 

GRAY,  J.  E.  1834.  Illustrations  of  Indian  zoology  chiefly  se- 
lected from  the  collection  of  Maj.-Gen.  Hardwicke,  Vol.  2. 
London. 

GREENE,  H.  W.  1976.  Scale  overlap,  a  directional  sign  stim- 
ulus for  prey  ingestion  by  ophiophagous  snakes.  Z.  Tierpsy- 
chol.  41:113-120. 

.    1977.  Phylogeny,  convergence,  and  snake  behavior. 

Ph.D.  Dissertation,  Univ.  Tennessee,  Knoxville. 

-.    1979.   Behavioral  convergence  in  the  defensive  dis- 


.    1 986a.  Diet  and  arboreality  in  the  emerald  monitor, 

Varanus  prasinus,  with  comments  on  the  study  of  adapta- 
tion. Fieldiana,  Zool.  (new  ser.)  31:1-12. 

-.   1 986b.  Natural  history  and  evolutionary  biology.  Pp. 


plays  of  snakes.  Experientia  35:747-748. 

-.    1983.   Dietary  correlates  of  the  origin  and  radiation 


99-108  in  Predator-prey  relationships:  perspectives  and  ap- 
proaches from  the  study  of  lower  vertebrates.  M.  E.  Feder 
and  G.  V.  Lauder,  eds.  Univ.  Chicago  Press,  Chicago. 
.    1988a.    Species  richness  in  tropical  predators.  Pp. 


259-280  in  Tropical  rainforests:  diversity  and  conservation. 

F.  Almeda  and  C.  M.  Pringle,  eds.  Cal.  Acad.  Sci.,  San  Fran- 
cisco. 
.    1988b.  Antipredator  mechanisms  in  reptiles.  Pp.  1- 

152  in  Biology  of  the  Reptilia,  Vol.  16.  Ecology  B.  Defense 

and  life  history.  C.  Gans  and  R.  B.  Huey,  eds.  Alan  R.  Liss, 

Inc.,  New  York. 
.   1989.  Defensive  behavior  and  feeding  biology  of  the 

Asian  mock  viper,  Psammodynastes  pulverulentus  (Colu- 

bridae),  a  specialized  predator  on  scincid  lizards.  Chin.  Herp. 

Res.  2:21-32. 
GREENE,  H.  W.  AND  J.  B.  Losos.    1988.   Systematics,  natural 

history,  and  conservation.  BioScience  38:458-462. 
GREENWELL,  M.  G.,  M.  HALL,  AND  O.  J.  SEXTON.    1984.  Phe- 

notypic  basis  for  a  feeding  change  in  an  insular  population 

of  garter  snakes.  Devi.  Psychobiol.  17:457-463. 
GRIFFIN,  L.  E.   1910.  A  list  of  snakes  from  the  island  of  Polillo, 

P.  I.,  with  descriptions  of  a  new  genus  and  two  new  species. 

Philippine  J.  Sci.  5:211-215. 
GROVES,  F.    1973.    Reproduction  and  venom  in  Blanding's 

tree  snake.  Int.  Zoo  Yrbk.  13:106-108. 
GUNTHER,  A.    1858.    Catalogue  of  colubrine  snakes  in  the 

collection  of  the  British  Museum.  London. 
.    1863.  Third  account  of  new  species  of  snakes  in  the 

collection  of  the  British  Museum.  Ann.  Mag.  Nat.  Hist.,  Ser. 

3,  12:348-365. 
.    1868.  Sixth  account  of  new  species  of  snakes  in  the 


of  snakes.  Amer.  Zool.  23:431-441. 

.   1984.  Feeding  behavior  and  diet  of  the  eastern  coral 

snake,  Micrurus  fulvius.  Spec.  Publ.  Mus.  Nat.  Hist.  Univ. 
Kansas  10:147-162. 


collection  of  the  British  Museum.  Ann.  Mag.  Nat.  Hist.,  Ser. 
4,  1:1-8. 

HALLOWELL,  E.  1857.  Notice  of  a  collection  of  reptiles  from 
the  Gaboon  Country,  West  Africa,  recently  presented  to  the 
Academy  of  Natural  Sciences  of  Philadelphia,  by  Dr.  Henry 
A.  Ford.  Proc.  Acad.  Nat.  Sci.  Philadelphia  1857:48-73. 

HARRISON,  T.  1962.  Cat  snake  kills  fruit  bat.  Malay.  Nat.  J. 
16:153-154. 

HEDGES,  N.  G.  1983.  Reptiles  and  amphibians  of  East  Africa. 
Kenya  Literature  Bureau,  Nairobi. 

HEDIGER,  H.  1933.  Uber  die  von  Herrn  Dr.  A.  Buehler  auf 
der  Admiralitats-Gruppe  und  einigen  benachbarten  Inseln 
gesammelten  Reptilien  und  Amphibien.  Verh.  Naturforsch. 
Ges.  Basil  44: 1-25. 

HENDERSON,  R.  W.  AND  M.  H.  BINDER.  1 980.  The  ecology 
and  behavior  of  vine  snakes  (Ahaetulla,  Oxybelis,  Thelotor- 
nis,  Uromacef):  a  review.  Contr.  Biol.  Geol.  Milwaukee  Publ. 
Mus.  37:1-38. 

HUTTON,  A.  F.  1949.  Notes  on  the  snakes  and  mammals  of 
the  High  Wavy  Mountains,  Madura  District,  S.  India.  J. 
Bombay  Nat.  Hist.  Soc.  48:454-460. 

INGER,  R.  F.  AND  R.  K.  COLWELL.  1977.  Organization  of 
contiguous  communities  of  amphibians  and  reptiles  in  Thai- 
land. Ecol.  Monogr.  47:229-253. 

INGER,  R.  F.,  H.  B.  SHAFFER,  M.  KOSHY,  AND  R.  BAKDE.  1 984. 
A  report  on  a  collection  of  amphibians  and  reptiles  from  the 
Ponmudi,  kerala,  South  India.  J.  Bombay  Nat.  Hist.  Soc. 
81:551-570. 


GREENE:  OLD  WORLD  CAT  SNAKES 


205 


JACKSON,  J.  A.  1974.  Gray  rat  snakes  versus  red-cockaded 
woodpeckers:  predator-prey  adaptations.  Auk  91:342-347. 

JAYNE,  B.  C.,  H.  K.  VORIS,  AND  K.  B.  HEANG.  1988.  Diet, 
feeding  behavior,  growth,  and  numbers  of  a  population  of 
Cerberus  rhynchops  (Serpentes:  Homalopsinae)  in  Malaysia. 
Fieldiana,  Zool.  (new  ser.)  50:1-15. 

JONES,  T.  1961.  Notes  on  bat  eating  snakes.  Niger.  Field  26: 
69-70. 

KARDONG,  K.  V.  1979.  "Protovipers"  and  the  evolution  of 
snake  fangs.  Evolution  33:433-443. 

.    1980.    Evolutionary  patterns  in  advanced  snakes. 

Amer.  Zool.  20:269-282. 

KLUGE,  A.  G.  1987.  Cladistic  relationships  in  the  Gekko- 
noidea  (Squamata,  Sauria).  Misc.  Publ.  Mus.  Zool.  Univ. 
Michigan  173:1-54. 

KOPSTEIN,  F.  1926.  Reptilien  von  den  Molukken  und  den 
benachbarten  Inseln.  Zool.  Mededel.  9:71-1 12. 

KREBS,  J.  R.,  D.  W.  STEPHENS,  AND  W.  J.  SUTHERLAND.  1983. 
Perspectives  in  optimal  foraging.  Pp.  165-216  in  Perspec- 
tives in  ornithology:  essays  presented  for  the  centennial  of 
the  American  Ornithologists'  Union.  A.  H.  Brush  and  G.  A. 
Clark,  Jr.,  eds.  Cambridge  Univ.  Press,  Cambridge. 

KROON,  C.  1973.  A  new  colubrid  snake  (Boiga)  from  south- 
eastern Asia.  Copeia  1973:580-586. 

KUNTZ,R.  E.  1963.  Snakes  of  Taiwan.  U.S.  Naval  Res.  Unit 
No.  2,  Taipei. 

LAURENT,  R.  F.  1956.  Contribution  a  Pherpetologie  de  la 
region  des  Grands  Lacs  de  I'Afrique  centrale.  Ann.  Mus. 
Roy.  Congo  Belg.  Ser.  8,  48:1-390. 

LAURIE,  E.  M.  O.  AND  J.  E.  HILL.  1954.  List  of  land  mammals 
of  New  Guinea,  Celebes  and  adjacent  islands,  1758-1952. 
Brit.  Mus.  (Nat.  Hist.),  London. 

LEKAGUL,  B.  AND  J.  A.  MCNEELY.  1977.  Mammals  of  Thai- 
land. Kurusapha  Ladprao  Press,  Bangkok. 

LEVINSON,  S.  R.,  M.  H.  EVANS,  AND  F.  GROVES.  1976.  A 
neurotoxic  component  of  the  venom  from  Blanding's  tree 
snake  (Boiga  blandingi).  Toxicon  14:307-312. 

LEVITON,  A.  E.  1970.  Contributions  to  a  review  of  Philippine 
snakes,  XI.  The  snakes  of  the  genus  Boiga.  Philippine  J.  Sci. 
97:291-314. 

LIM,  B.  L.  1955.  Snakes  collected  near  Kuala  Lampur.  Malay. 
Nat.  J.  9:122-125. 

.    1956.    The  natural  food  of  some  Malayan  snakes. 

Malay.  Nat.  J.  10:139-143. 

-.    1967.  Snakes  collected  in  Gunong  Brinchang,  Cam- 


eron Highlands,  Pahang.  Malay.  Nat.  J.  20:121-127. 
.    1973.  Unusual  predators  of  the  mouse-deer.  Malay. 


Nat.  J.  26:170. 

LIM,  B.  L.,  J.  P.  J.  WATERS,  AND  A.  R.  BEN  OMAR.  1960. 
Observations  on  some  captive  snakes.  Malay.  Nat.  J.  14: 
181-187. 

LINDBERG,  K.  1932.  Snakes  on  the  Barsi  Light  Railway  (Dec- 
can).  J.  Bombay  Nat.  Hist.  Soc.  35:690-697. 

LLOYD,  M.,  R.  F.  INGER,  AND  F.  W.  KING.  1968.  On  the 
diversity  of  reptile  and  amphibian  species  in  a  Bornean  rain 
forest.  Amer.  Nat.  102:497-515. 

Losos,  J.  B.  AND  H.  W.  GREENE.  1988.  Ecological  and  evo- 
lutionary implications  of  diet  in  monitor  lizards.  Biol.  J. 
Linn.  Soc.  35:379^07. 

MACARTHUR,  R.  H.  AND  E.  R.  PiANKA.  1 966.  On  the  optimal 
use  of  a  patchy  environment.  Amer.  Nat.  100:603-609. 

MACIAS  GARCIA,  C.  AND  H.  DRUMMOND.  1988.  Seasonal  and 
ontogenetic  variation  in  the  diet  of  the  Mexican  garter  snake, 


Thamnophis  eques,  in  Lake  Tecomulco,  Hidalgo.  J.  Her- 

petol.  22:129-134. 
MALNATE,  E.  V.  AND  G.  UNDERWOOD.    1988.    Australasian 

natricine  snakes  of  the  genus  Tropidonophis.  Proc.  Acad. 

Nat.  Sci.  Philadelphia  140:59-201. 
MARX,  H.  AND  G.  B.  RABB.    1972.   Phyletic  analysis  of  fifty 

characters  of  advanced  snakes.  Fieldiana,  Zool.  63:1-321. 
MCDOWELL,  S.  B.   1 969.  Toxicocalamus,  a  New  Guinea  genus 

of  snakes  of  the  family  Elapidae.  J.  Zool.  159:443-511. 
.    1972.   The  species  of  Stegonotus  (Serpentes,  Colu- 

bridae)  in  Papua  New  Guinea.  Zool.  Meded.  47:6-26. 
-.    1975.  A  catalogue  of  the  snakes  of  New  Guinea  and 


the  Solomons,  with  special  reference  to  those  in  the  Bernice 
P.  Bishop  Museum.  Part  II.  Anilioidea  and  Pythoninae.  J. 
Herp.  9:1-79. 
.   1984.  Results  of  the  Archbold  expeditions.  No.  1 12. 


The  snakes  of  the  Huon  Peninsula,  Papua  New  Guinea. 

Amer.  Mus.  Novitates  2775:1-28. 

MENZIES,  J.  I.    1961.  Correspondence:  bat-eating  snakes.  Ni- 
ger. Field  26:44. 
MENZIES,  J.  I.  AND  E.  DENNIS.    1 979.  Handbook  of  New  Guinea 

rodents.  Wau  Ecol.  Inst.  Handbk.  (7). 
MERREM,  B.   1 790.  Bietrage  zur  Naturgeschichte.  Priv.  Publ., 

Duisburg  and  Lemgo. 
MINTON,  S.  A.    1966.   A  contribution  to  the  herpetology  of 

West  Pakistan.  Bull.  Amer.  Mus.  Nat.  Hist.  134:27-184. 
MINTON,  S.  A.  AND  W.  A.  DUNSON.    1978.   Observations  on 

the  Palawan  mangrove  snake,  Boiga  dendrophila  multidncta 

(Reptilia,  Serpentes,  Colubridae).  J.  Herp.  12:107-108. 
MINTON,  S.  A.  AND  S.  K.  SALANITRO.    1972.   Serological  re- 
lationships among  some  colubrid  snakes.  Copeia  1 972:246- 

252. 
MURPHY,  J.  B.    1977.    An  unusual  method  of  immobilizing 

prey  by  the  dog-tooth  cat  snake,  Boiga  cynodon.  Copeia 

1977:182-184. 
MUSHINSKY,  H.  R.    1987.    Foraging  ecology.  Pp.  302-334  in 

Snakes:  ecology  and  evolutionary  biology.  R.  A.  Seigel,  J.  T. 

Collins,  and  S.  S.  Novak,  eds.  Macmillan  Publ.  Co.,  New 

York. 
MYERS,  C.  W.    1 982.   Blunt-headed  vine  snakes  (Imantodes) 

in  Panama,  including  a  new  species  and  other  revisionary 

notes.  Amer.  Mus.  Novitates  2738:1-50. 
PARKER,  F.    1 982.    The  snakes  of  Western  Province.  Div. 

Wildlife,  Papua  New  Guinea. 
PENDLETON,  R.  C.   1947.  A  snake  "den"  tree  on  Guadalcanal 

Island.  Herpetologica  3:189-190. 
PETERS,  W.    1861.    Eine  zweite  Ubersicht  der  von  Hrn.  F. 

Jagor  auf  Malacca,  Java,  Borneo  und  den  Philippinen  gesam- 

melten  unde  dem  Kgl.  Zoologischen  Museum  uberstandten 

Schlangen.  Monatsb.  Akad.  Wiss.  Berlin  1861:683-691. 
PITMAN,  C.  R.  S.   1958.  Snake  and  lizard  predators  on  birds. 

Bull.  Brit.  Ornithol.  Club  78:82-86. 
.   1 962.  More  snake  and  lizard  predators  of  birds.  Bull. 

Brit.  Ornithol.  Club  82:45-55. 
.    1974.   A  guide  to  the  snakes  of  Uganda.  Wheldon 

and  Welsey,  Ltd.,  Codicote. 
POPE,  C.  H.    1935.   The  reptiles  of  China.  Amer.  Mus.  Nat. 

Hist.,  New  York. 
RASMUSSEN,  J.  B.    1979.    An  intergeneric  analysis  of  some 

boigine  snakes— Bogert's  ( 1 940)  Group  XIII  and  XIV  (Boigi- 

nae,  Serpentes).  Dansk.  Naturhist.  For.  Vidensk.  Meddel. 

141:97-155. 
.    1985.   A  re-evaluation  of  the  systematics  of  the  Af- 


206 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  8 


rican  rear-fanged  snakes  of  Bogert's  Groups  XIII-XVI,  in- 
cluding a  discussion  of  some  evolutionary  trends  within 
Caenophidia.  Pp.  531-548  in  Proceedings  of  the  Interna- 
tional Symposium  on  African  vertebrates:  systematics,  phy- 
logeny  and  evolutionary  ecology.  K.-L.  Schuchmann,  ed. 
Zool.  Forsch.  Mus.  Alexander  Koenig,  Bonn. 

RAXWORTHY,  C.  J.  1988.  Reptiles,  rainforest  and  conserva- 
tion in  Madagascar.  Biol.  Conserv.  43:181-211. 

Rootr,  N.  1917.  The  reptiles  of  the  Indo-Australian  Archi- 
pelago. E.  J.  Brill  Publ.,  Leiden. 

ROOM,  P.  M.  1974.  Lizards  and  snakes  from  the  Northern 
District  of  Papua,  New  Guinea.  Brit.  J.  Herpetol.  5:38-46. 

SAINT  GIRONS,  H.  1972.  Les  serpents  du  Cambodge.  Mem. 
Mus.  Nat.  Hist.  Natur.,  nouv.  ser.,  74:1-170. 

SAVIDOE,  J.  A.  1 984.  Guam:  paradise  lost  for  wildlife.  Biol. 
Conserv.  30:305-317. 

.  1987.  Extinction  of  an  island  forest  avifauna  by  an 

introduced  snake.  Ecology  68:660-668. 

.   1988.  Food  habits  of  Boiga  irregularis,  an  introduced 


predator  on  Guam.  J.  Herp.  22:275-282. 

SAVITZKY,  A.  H.  1980.  The  role  of  venom  delivery  strategies 
in  snake  evolution.  Evolution  34:1 194-1204. 

SCHAEFER,  S.  A.  AND  G.  V.  LAUDER.  1 986.  Historical  trans- 
formation of  functional  design:  evolutionary  morphology  of 
feeding  mechanisms  in  loricarioid  catfishes.  Syst.  Zool.  35: 
489-508. 

SCHAETTI,  B.  1 984.  Fledermause  als  Nahrung  von  Schlangen. 
Bonn.  Zool.  Beitr.  35:335-342. 

SCHMIDT,  K.  P.  1927.  The  reptiles  of  Hainan.  Bull.  Amer. 
Mus.  Nat.  Hist.  54:395^65. 

SEIB,  R.  L.  1 984.  Prey  use  in  three  syntopic  neotropical  racers. 
J.  Herpetol.  18:412-420. 

.  1985.  Euryphagy  in  a  tropical  snake,  Coniophanes 

fissidens.  Biotropica  17:57-64. 

SHARMA,  R.  C.  AND  T.  G.  VAZIRANI.  1977.  Food  and  feeding 
habits  of  some  reptiles  of  Rajasthan.  Rec.  Zool.  Survey  India 
73:77-93. 

SHINE,  R.  1 980a.  Ecology  of  the  Australian  death  adder  Acan- 
thophis  antarcticus  (Elapidae):  evidence  for  convergence  with 
the  Viperidae.  Herpetologica  36:281-289. 

.  1 980b.  Reproduction,  feeding  and  growth  in  the  Aus- 
tralian burrowing  snake  Vermicella  annulata.  J.  Herpetol. 
14:71-77. 

.  1 980c.  Ecology  of  eastern  Australian  whipsnakes  of 

the  genus  Demansia.  J.  Herpetol.  14:381-389. 

.  1983.  Arboreality  in  snakes:  ecology  of  the  Australian 

elapid  genus  Hoplocephalus.  Copeia  1983:198-205. 

SHINE,  R.  AND  T.  SCHWANER.  1 985.  Prey  constriction  be  ven- 
omous snakes:  a  review,  and  new  data  on  Australian  species. 
Copeia  1985:1067-1071. 

SMITH,  B.  E.  1985.  A  systematic  survey  of  the  squamate 
reptiles  of  eastern  Mindanao,  Philippine  Islands,  with  notes 
on  ecomorphology.  M.S.  Thesis,  Louisiana  State  Univ.  78 
pp. 

SMITH,  M.  A.  1916.  On  a  collection  of  reptiles  and  batra- 
chians  from  peninsular  Siam.  J.  Nat.  Hist.  Soc.  Siam  2:  MS- 
HI. 

.  1930.  The  Reptilia  and  Amphibia  of  the  Malay  Pen- 
insula. Bull.  Raffles  Mus.  3:1-149. 

.  1943.  The  fauna  of  British  India,  Ceylon,  and  Burma, 

including  the  whole  of  the  Indo-Chinese  sub-region,  Reptilia 
and  Amphibia,  Vol.  III.  Serpentes.  Taylor  and  Francis,  Lon- 
don. 


SMITH,  M.  AND  A.  o'A.  BELLAIRS.  1947.  The  head  glands  of 
snakes,  with  remarks  on  the  evolution  of  the  parotid  glands 
and  teeth  of  the  Opisthoglypha.  Zool.  J.  Linn.  Soc.  41:351- 
370. 

SMITH,  M.  A.  AND  C.  B.  KLOSS.  1915.  On  reptiles  and  ba- 
trachians  from  the  coast  and  islands  of  south-east  Siam.  J. 
Nat.  Hist.  Soc.  Siam  1:237-249. 

STEJNEGER,  L.  1 902.  A  new  opisthoglyp  snake  from  Formosa. 
Proc.  Biol.  Soc.  Washington  15:1 15-1 17. 

SUNDERSINGH,  P.  P.  H.  1 960.  Reptilian  fauna  of  Pilani  with 
a  note  on  the  food  contents.  Part  I— Ophidia.  Proc.  Rajas- 
than Acad.  Sci.  7:52-59. 

TAYLOR,  E.  H.  1922a.  Herpetological  fauna  of  Mount  Maki- 
ling.  Philippine  Agricul.  11:127-139. 

.  1922b.  The  snakes  of  the  Philippine  Islands.  Phil- 
ippine Bur.  Sci.  Publ.  16:1-312. 

.  1923.  Additions  to  the  herpetological  fauna  of  the 

Philippine  Islands,  III.  Philippine  J.  Sci.  22:515-557. 

.  1953.  Report  on  a  collection  of  Ceylonese  serpents. 

Univ.  Kansas  Sci.  Bull.  35:1615-1624. 

.  1 965.  The  serpents  of  Thailand  and  adjacent  waters. 

Univ.  Kansas  Sci.  Bull.  45:609-1096. 

TWEEDIE,  M.  W.  F.  1954.  Notes  on  Malayan  reptiles  No.  3. 
Bull.  Raffles  Mus.  25:107-1 17. 

.  1983.  The  snakes  of  Malaya,  3rd  ed.  Singapore  Nat. 

Printers,  Singapore. 

VORIS,  H.  K.  AND  M.  W.  MOFFETT.  1981.  Size  and  proportion 
relationship  between  the  beaked  sea  snake  and  its  prey.  Bio- 
tropica 13:15-19. 

VORIS,  H.  K.  AND  H.  H.  VORIS.  1983.  Feeding  strategies  in 
marine  snakes:  an  analysis  of  evolutionary,  morphological, 
behavioral  and  ecological  relationships.  Amer.  Zool.  23:41 1- 
425. 

WAKEMAN,  B.  N.  1966.  Uganda's  poisonous  snakes:  further 
observations  on  feeding  habits.  Uganda  J.  30:101-103. 

WALL,  F.  1907a.  Notes  on  snakes  collected  in  Fyzabad.  J. 
Bombay  Nat.  Hist.  Soc.  18:101-129. 

.  1907b.  A  popular  treatise  on  the  common  Indian 

snakes.  Part  7.  Echis  carinatus,  Boiga  trigonatus.  J.  Bombay 
Nat.  Hist.  Soc.  18:525-554. 

.  1909.  Notes  on  snakes  from  the  neighbourhood  of 

Darjeeling.  J.  Bombay  Nat.  Hist.  Soc.  19:337-357. 

.  1910.  Notes  on  snakes  collected  in  upper  Assam. 

Part  II.  J.  Bombay  Nat.  Hist.  Soc.  19:825-845. 

.    1919.    Notes  on  a  collection  of  snakes  made  in  the 


Nilgiri  Hills  and  the  adjacent  Wynaad.  J.  Bombay  Nat.  Hist. 
Soc.  26:552-584. 

1921.  Ophidia  taprobanica,  or  the  snakes  of  Ceylon. 


Govt.  Printer,  Colombo. 
.    1923.   Notes  on  snakes  collected  on  Annasigalla  Es- 


tate from  August,  1920,  to  December,  1921.  Spolia  Zeylan- 

ica  12:252-270. 
.    1 924.   Notes  on  Ceylon  snakes  collected  by  Mr.  W. 

W.  A.  Phillips.  Spolia  Zeylanica  13:71-88. 
.    1926a.  Notes  on  snakes  collected  in  Burma  in  1924. 


J.  Bombay  Nat.  Hist.  Soc.  30:805-821. 
.    1926b.   Snakes  collected  in  Burma  in  1925.  J.  Bom- 


bay Nat.  Hist.  Soc.  31:558-566. 

WERNER,  F.  1 909.  Uber  neue  oder  seltene  Reptilien  des  Na- 
turhistorischen  Museums  in  Hamburg.  Mitt.  Naturhist.  Mus. 
Hamburg  26:205-247. 


GREENE:  OLD  WORLD  CAT  SNAKES  207 

WHITAKER,  R.    1970.   Slight  reaction  from  bites  of  the  rear-  WOODWARD,  S.  F.   1960.  A  bat-eating  snake.  Niger.  Field  25: 

fanged  snakes  Boiga  ceylonensis  (Gunther)  and  Dryophis  na-  1 72-1 74. 

5M/M5  (Lacepede).  J.  Bombay  Nat.  Hist.  Soc.  67:113.  YOUNG,  R.  A.   1983.  Feeding  behavior  of  captive  Boiga.  Bull. 

WILES,  G.  J.    1988.   The  status  of  fruit  bats  on  Guam.  Pac.  Chicago  Herp.  Soc.  18:89-91. 

Sci.  41:148-157. 


CALIFORNIA  ACADEMY  OF  SCIENCES 

Golden  Gate  Park 
San  Francisco,  California  94 1 1 8 


PROCEEDINGS 

OF  THE 

CALIFORNIA  ACADEMY  OF  SCIENCES 


Vol.  46,  No.  9,  pp.  209-220,  4  figs.,  1  table. 


December  20,  1989 


NEW  SPECIES  AND  TAXONOMIC  NOTES  ON  MEXICAN  AND 
CENTRAL  AMERICAN  MELASTOMATACEAE 

By 

Frank  Almeda 

Department  of  Botany,  California  Academy  of  Sciences,  Golden  Gate  Park, 
San  Francisco,  California  94118-4599 


ABSTRACT:  Descriptions,  diagnostic  illustrations,  and  discussions  are  presented  for  four  new  species  of 
Miconia:  M.  grayumii  from  Costa  Rica  and  Nicaragua;  M.  ibarrae  from  Mexico;  M.  iteophylla  and  M.peltata 
from  Panama.  A  reappraisal  of  the  floral  morphology  ofLeandra  consimilis  necessitates  its  transfer  to  Miconia 
and  the  adoption  of  a  new  name,  M.  ligulata.  A  chromosome  number  of  n  =  11  is  reported  for  Pilocosta 
oerstedii  subsp.  campanensis  for  the  first  time.  To  reflect  the  uniqueness  and  relationships  of  subsp.  campanensis 
it  is  elevated  to  specific  status  based  on  a  reconsideration  of  morphological,  chromosomal,  and  ecogeographic 
attributes. 

Received  June  9, 1989.  Accepted  August  22, 1989. 


INTRODUCTION 

Recent  field  work  in  Central  America  has  gen- 
erated a  rich  reservoir  of  specimens  upon  which 
to  base  a  treatment  of  the  Melastomataceae  for 
Flora  Mesoamericana.  In  the  course  of  studying 
this  material  I  continue  to  detect  many  new  taxa, 
significant  range  extensions,  nomenclatural 
problems  in  need  of  adjustment,  and  other  in- 
formation that  enhances  our  understanding  of 
previously  described  taxa.  This  paper  records 
some  of  these  discoveries  and  observations  to 
make  the  names  available  for  use  in  other  re- 
gional checklists  and  floristic  treatments  of  the 
family. 

Miconia  grayumii  Almeda,  sp.  nov. 

(Fig.  1) 

TYPE. -COSTA  RICA.  Heredia:  Finca  La  Selva,  Field  Sta- 
tion of  the  Organization  for  Tropical  Studies  on  the  Rio  Puerto 
Viejo  E  of  its  junction  with  the  Rio  Sarapiqui,  elev.  ca.  100 


m,  22  Oct.  1982,  McDowell  576  (holotype:  CAS!;  isotype: 
DUKE). 

Frutex  vel  arbuscula  1-5  m.  Ramuli  obscure  quadrangulati 
demum  teretes  sicut  petioli  foliorum  subtus  venae  primariae 
inflorescentiaque  modice  granuloso-furfuracei.  Petioli  (2-)6- 
21  mm  longi;  lamina  5.5-12.7  x  1.2-4.3  cm  anguste  elliptica 
apice  anguste  gradatimque  acuminate  basi  late  acuta  vel  ob- 
tusa,  3-plinervata,  membranacea  et  undulato-denticulata  vel 
integra.  Inflorescentia  2.5-5  cm  longa  multiflora;  flores  5-meri, 
pedicellis  (ad  anthesim)  2-2.5  mm  longis,  bracteolis  0.5  x  0.25 
mm  persistentibus.  Hypanthium  (ad  torum)  1  mm  longum 
sparsiuscule  caduceque  granuloso-f urfuraceum;  lobis  interior- 
ibus  0.25  x  0.5  late  deltoideis,  dentibus  exterioribus  0.25-0.5 
mm  eminentibus.  Petala  1.5-2  x  0.5  mm  obovato-oblonga 
intus  minute  papillosa  extus  glabra.  Stamina  isomorphica  gla- 
bra;  filamenta  1-1.5  mm  longa;  antherarum  thecae  0.5-0.75 
x  0.25  mm  anguste  oblongae,  poro  paulo  dorsaliter  inclinato; 
connectivum  nee  prolongatum  nee  appendiculatum.  Stylus  2.5- 
3  mm  glaber;  ovarium  5-loculare  et  %  inferum  glabro. 

Shrub  or  small  trees  1-5  m  tall.  Older  cauline 
internodes  subquadrate  to  terete,  sparsely  to 
moderately  ferrugineous  scurfy-pulverulent  to 
glabrate.  Distal  branchlets,  petioles,  elevated  pri- 


[209] 


210 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  9 


FIGURE  1.  Miconia  grayumii  Almeda.  A,  habit,  x  ca.  %;  B,  representative  leaf  (lower  surface),  x  1;  C,  flower  (at  anthesis) 
with  all  petals  and  stamens  removed,  x  1 5;  D,  petal  (adaxial  surface),  x  ca.  22;  E,  stamens,  dorsal  view  (left)  and  lateral  view 
(right),  x30;  F,  mature  berry,  x  ca.  5;  G,  seeds,  x  ca.  15.  (A  from  Folsom  8914;  B  from  Smith  152;  C-F  from  the  holotype; 
G  from  Hammel  12011.) 


mary  leaf  veins  below,  and  inflorescences  mod- 
erately scurfy-pulverulent.  Leaves  of  a  pair 
somewhat  unequal  in  size;  petioles  (2-)6-2 1  mm 
long;  blades  membranaceous,  5.5-12.7  cm  long 
and  1.2-4.3  cm  wide,  narrowly  elliptic,  apex  at- 
tenuate to  gradually  acuminate,  base  broadly 


acute  to  obtuse  and  typically  inaequilateral  but 
not  decurrent  on  the  petiole,  margin  undulate- 
denticulate  to  entire,  3-plinerved,  the  innermost 
pair  of  elevated  primaries  diverging  from  the  me- 
dian nerve  in  opposite  or  irregularly  alternate 
fashion  at  successive  points  above  the  blade  base, 


ALMEDA:  MEXICAN  AND  CENTRAL  AMERICAN  MELASTOMATACEAE 


211 


the  transverse  secondaries  prominulous  and 
spaced  3-5  mm  apart  at  the  widest  portion  of 
the  blade,  glabrous  above  at  maturity,  sparingly 
scurfy-pulverulent  and/or  obscurely  glandular- 
pulverulent  to  glabrate  below.  Inflorescence  a 
paniculiform  dichasium  2.5-5  cm  long  divari- 
cately branched  from  the  base,  terminal  but 
sometimes  becoming  pseudolateral  because  of 
axillary  bud  elongation;  bracts  of  rachis  nodes 
persistent,  lance-triangular,  0.5-3  mm  long,  0.25 
mm  wide,  glabrous  above,  sparsely  scurfy-pul- 
verulent to  glabrate  below;  bracteoles  sessile  and 
persistent,  paired  and  fused  into  a  short  incon- 
spicuous nodal  collar  or  elevated  interpetiolar 
ridge,  lance-triangular  to  subulate,  0.5  mm  long, 
0.25  mm  wide,  glabrous  above  and  sparingly 
scurfy-pulverulent  to  glabrous  below,  margin  en- 
tire. Pedicels  2-2.5  mm  long,  sparingly  and  de- 
ciduously  scurfy-pulverulent  to  glabrate.  Hypan- 
thium  (at  anthesis)  globose,  1  mm  long  to  the 
torus,  sparsely  and  deciduously  scurfy-pulveru- 
lent. Calyx  lobes  5,  rounded-deltoid,  hyaline  and 
glabrous,  0.25  mm  long,  ca.  0.5  mm  wide  basally 
between  sinuses,  the  margins  entire;  exterior  ca- 
lyx teeth  5,  subulate,  0.25-0.5  mm  long,  exceed- 
ing and  obscuring  the  lobes,  essentially  glabrous 
throughout;  torus  glabrous  adaxially.  Petals  5, 
obscurely  papillose  adaxially,  otherwise  gla- 
brous, white,  obovate-oblong,  rounded  apically, 
1.5-2  mm  long,  0.5  mm  wide  distally.  Stamens 
10,  isomorphic;  filaments  glabrous,  complanate, 
1-1.5  mm  long;  anthers  0.5-0.75  mm  long,  0.25 
mm  wide,  yellow,  oblong,  rounded  to  truncate 
at  the  apex  with  a  retuse  to  somewhat  dorsally 
inclined  terminal  pore;  connective  thickened 
dorsally  but  not  prolonged  or  appendaged  at  the 
filament  insertion.  Ovary  (at  anthesis)  ca.  %  in- 
ferior, 5-celled,  globose,  glabrous  apically.  Style 
erect  and  conspicuously  curved  distally,  gla- 
brous, 2.5-3  mm  long;  stigma  punctiform  to 
truncate.  Berry  blue-black  or  purple  at  maturity, 
globose,  3-5  mm  long  and  3-5  mm  in  diameter. 
Seeds  angulate-pyramidate,  0.5-1  mm  long,  beige, 
smooth  with  obscure  but  distinct  verruculose  an- 
gles on  the  convex  face. 

PHENOLOGY.— Flowering  and  fruiting  occur 
sporadically  from  February  through  November. 

DISTRIBUTION.— Rain  forests  and  stream  banks 
in  the  Caribbean  lowlands  of  Costa  Rica  and 
Nicaragua  at  40-180  m. 

ADDITIONAL  SPECIMENS  EXAMINED.— COSTA  RICA.  Here- 
dia:  Finca  La  Selva,  OTS  Field  Station  on  Rio  Puerto  Viejo, 


trail  at  perimeter  of  successional  plots,  13  Feb.  1981,  Folsom 
8914  (CAS);  Finca  La  Selva,  OTS  Field  Station,  West  Bound- 
ary, 800-1,200  m  line,  19  Feb.  1981,  Folsom  9041  (CAS);  W 
of  San  Jose  on  Rio  Sarapiqui,  across  Rio  Sardinal  at  edge  of 
LomasSardinal,  4  Feb.  1983,  Garwoodetal.  1070  (EM);  Finca 
La  Selva,  Loop  Trail,  near  SW  Trail,  15  Jul.  1979,  Grayum 
1834  (CAS);  Finca  La  Selva,  East  Boundary  Trail,  2,300  m 
line,  19  Jul.  1979,  Grayum  &  Sperry  1854  (CAS);  Finca  La 
Selva,  forested  knoll  E  of  Boundary  Trail  on  way  to  Q.  Ar- 
boleda  swamp,  14  Mar.  1980,  Hammel8076  (CAS);  Finca  La 
Selva,  bank  along  Rio  Puerto  Viejo,  550  S  x  1,500  E  m  grid, 
23  Jun.  1980,  Hammel  9103  (CAS);  Finca  La  Selva,  near  Loop 
Trail,  100  m  S,  28  Aug.  1980,  Hammel  9603  (CAS);  Finca  La 
Selva,  on  ridge  in  plot  HI,  4  May  1982,  Hammel  12011  (CAS): 
Finca  La  Selva,  West  Boundary,  1,200  m  S,  28  May  1982, 
Hammel  12576  (CAS);  Finca  La  Selva,  El  Sura  Trail,  1,200 
m  line,  25  Sep.  1982,  McDowell  199  (CAS);  Finca  La  Selva, 
around  1,850  m  E,  1,950  m  S,  1  Sep.  1981,  Smith  152  (CAS). 
Limon:  Cerro  Coronel,  E  of  Rio  Zapote,  ca.  1.5  km  from  Rio 
Colorado,  10°40'N,  83°40'W,  13  Mar.  1987,  Stevens  etal.  24795 
(CAS).  NICARAGUA.  Zelaya:  Nueva  Guinea,  Colonia  Yo- 
laina,  13  Aug.  1982,  Araquistain  3069  (CAS);  Cano  Costa  Ri- 
quita,  1.8  km  SW  of  Colonia  Naciones  Unidas,  11°43'N, 
84°18'W,  6-7  Nov.  1977,  Stevens  4971  (CAS). 

The  first  collections  ofAficonia  grayumii,  made 
just  over  a  decade  ago,  were  initially  referred  to 
M.  brenesii  Standley  because  of  similarities  in 
foliar  morphology,  indument  type,  and  the  di- 
varicately branched  dichasia.  Critical  study  of 
the  improved  material  now  available  necessi- 
tates the  recognition  of  two  species  based  on  con- 
sistent morphological  and  ecological  differences. 
All  of  the  perceived  differences  between  M.  gra- 
yumii and  M.  brenesii  involve  details  of  the  flow- 
ers and  fruits.  The  latter  has  broadly  deltoid  calyx 
teeth  that  do  not  greatly  exceed  or  obscure  the 
calyx  lobes,  petals  that  are  smooth  on  both  sur- 
faces, and  anthers  that  are  cuneate  and  broadest 
distally  with  flaring  dorsally  inclined  pores.  This 
broad  pore  and  its  thick  connective  that  is  pro- 
longed for  a  short  distance  (0.25  mm)  below  the 
thecae  give  the  anthers  of  M  brenesii  an  unusual 
dorsally  arcuate  conformation  when  dry.  Some 
of  the  most  readily  observed  differences  between 
M.  grayumii  and  M.  brenesii  involve  stylar  and 
seed  characters.  In  M.  brenesii  the  straight  style 
is  terminated  by  a  capitellate  stigma  and  its  seeds, 
although  pyramidate  like  M.  grayumii,  have 
completely  smooth  polished  angles.  There  are 
also  pronounced  ecogeographic  differences  be- 
tween these  two  species.  In  Costa  Rica,  M.  bre- 
nesii is  largely  confined  to  a  more  westerly  dis- 
tribution at  higher  elevations  (900-1,600  m) 
whereas  M.  grayumii  occurs  in  the  eastern  low- 
lands (40-180  m)  that  extend  northward  to  Nic- 
aragua. 


212 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  9 


Despite  the  many  vegetative  similarities  be- 
tween M.  brenesii  and  M.  grayumii,  the  latter 
appears  to  be  closest  to  M.  ligulata  Almeda  (also 
discussed  below).  Petal,  stamen,  and  seed  mor- 
phology are  identical  in  these  two  species.  How- 
ever, in  M.  ligulata  the  indument  consists  of  pi- 
noid  hairs,  the  leaves  are  larger  (12-28  x  4-10 
cm)  and  decurrent  on  the  petiole,  the  torus  is 
puberulent  adaxially,  the  petals  are  glabrous 
throughout,  and  the  style  is  straight. 

This  species  is  named  for  Michael  Grayum 
who  first  collected  this  species  at  the  La  Selva 
Field  Station  in  1979.  His  subsequent  field  work 
in  many  other  areas  of  Costa  Rica  has  added 
numerous  interesting  plants  to  the  flora  of  the 
region. 

Miconia  ibarrae  Almeda,  sp.  nov. 

(Fig.  2) 

TYPE.— MEXICO.  Veracruz:  road  to  Lazaro  Cardenas,  4  km 
W  of  Estacion  de  Biologia  Tropical  Los  Tuxtlas,  95°04'  and 
95°09'W,  18°36'N,  municipio  San  Andres  Tuxtla,  elev.  400  m, 
3  Jun.  1986,  Sinica  738  (holotype:  CAS!;  isotype:  MEXU). 

Arbor  parva  4-9  m.  Ramuli  quadrangulati  demum  teretes 
sicut  foliorum  subtus  venae  primarieae  inflorescentia  hypan- 
thiaque  modice  vel  sparse  stellulato-furfuracei  demum  glabra- 
ti.  Petioli  0.8-2.1  cm  longi;  lamina  5-11.5  x  2.5^t.l  cm  el- 
liptica  vel  elliptico-lanceolata  apice  acuminata  basi  acuta  vel 
obtusa,  3(-5)-plinervata,  membranacea  et  undulato-denticu- 
lata.  Panicula  5-10  cm  longa  multiflora;  flores  5-meri,  pedi- 
cellis  (ad  anthesim)  0.5  mm  longis,  bracteolis  0.25-0.5  x  0.25 
mm  persistentibus.  Hypanthium  (ad  torum)  2-2.5  mm  Ion- 
gum;  calyx  0.5  mm  longus  truncatus  vel  vix  undulatus,  den- 
tibus  exterioribus  ca.  0.25  mm  eminentibus.  Petala  2.5-3  x 
1 .5  mm  obovato-oblonga,  extus  et  intus  dense  papillosa.  Stam- 
ina isomorphica  glabra;  filamenta  2  mm  longa;  antherarum 
thecae  2-2.5  x  0.5  mm  oblongae,  poro  ventraliter  inclinato; 
connectivum  nee  prolongatum  nee  appendiculatum.  Stylus  3.5- 
5  mm  glaber;  ovarium  5-6-loculare  omnino  inferum  apice 
glabro. 

Trees  4-9  m  tall.  Uppermost  branchlets,  veg- 
etative buds,  and  inflorescences  sparsely  to  mod- 
erately stellate-furfuraceous,  the  distal  branchlet 
nodes  bearing  prominent  interpetiolar  lines  or 
ridges.  Leaves  of  a  pair  slightly  unequal  in  size; 
petioles  0.8-2.1  cm  long;  blades  membrana- 
ceous,  5-1 1.5  cm  long  and  2.5-4.1  cm  wide,  el- 
liptic to  elliptic-lanceolate,  apex  acuminate,  base 
acute  to  obtuse,  margin  undulate-denticulate,  3(- 
5)-plinerved,  the  innermost  pair  of  elevated  pri- 
maries diverging  from  the  median  nerve  in  op- 
posite or  alternate  fashion  above  the  blade  base, 
above  glabrous  at  maturity,  below  essentially  gla- 
brous or  sparingly  and  deciduously  stellate-fur- 
furaceous on  the  elevated  primary  nerves.  Inflo- 


rescence a  terminal  multiflowered  panicle  5-10 
cm  long;  bracts  of  the  rachis  nodes  persistent, 
narrowly  lance-triangular,  1-1.5  mm  long,  0.25 
mm  wide  at  the  base,  sparingly  stellulate-furfura- 
ceous  to  glabrate;  bracteoles  sessile  and  persis- 
tent, subulate,  0.25-0.5  mm  long,  0.25  mm  wide, 
sparingly  stellulate-furfuraceous,  margin  entire. 
Pedicels  0.5  mm  long  but  lengthening  to  1  mm 
in  fruit,  sparingly  stellate.  Hypanthia  (at  anthe- 
sis)  campanulate,  2-2.5  mm  long  to  the  torus, 
sparingly  and  deciduously  stellulate-furfura- 
ceous to  glabrous.  Calyx  tube  ca.  0.5  mm  long, 
truncate  or  obscurely  undulate  with  5  inconspic- 
ous  and  early  deciduous  triangular  calyx  teeth 
ca.  0.25  mm  long  that  barely  project  beyond  ca- 
lyx rim;  torus  glabrous  adaxially.  Petals  5,  con- 
spicuously papillose  on  both  surfaces,  white,  ob- 
ovate-oblong,  rounded  to  irregularly  undulate 
apically,  2.5-3  mm  long,  1.5  mm  wide  distally. 
Stamens  10,  isomorphic;  filaments  glabrous, 
complanate,  2  mm  long;  anthers  2-2.5  mm  long, 
ca.  0.5  mm  wide,  erect  to  slightly  incurved  api- 
cally, yellow,  linear-oblong,  laterally  compressed 
and  deeply  channeled  ventrally  between  thecae, 
rounded  at  the  apex  with  a  ventrally  inclined 
terminal  pore;  connective  thickened  dorsally  but 
not  prolonged  or  appendaged  at  the  filament  in- 
sertion. Ovary  (at  anthesis)  completely  inferior, 
5-6 -celled,  globose  to  depressed-conic,  the  apex 
fluted  and  glabrous  but  becoming  rounded  in 
fruit.  Style  somewhat  declinate  and  incurved  dis- 
tally, glabrous,  3.5-5  mm  long;  stigma  puncti- 
form  to  truncate.  Berry  globose,  greenish-yellow 
when  plump  and  fleshy,  4-5  mm  long  and  5-5.5 
mm  in  diameter.  Seeds  irregularly  pyramidate, 
1.5  mm  long,  beige,  smooth  and  vernicose  with 
a  rounded  or  bluntly  angulate  convex  face. 

PHENOLOGY.— Flowering  and  fruiting  speci- 
mens have  been  collected  from  April  through 
July  and  June  through  August,  respectively. 

DISTRIBUTION.— Rain  forests  of  the  munici- 
pios  of  Catemaco  and  San  Andres  Tuxtla  in  Ve- 
racruz, Mexico  at  100-500  m. 

ADDITIONAL  SPECIMENS  EXAMINED.  — MEXICO.  Veracruz: 
Municipio  Catemaco,  6  km  SW  of  Sontecomapan,  1 2  May 
1 972,  Beaman  5953  (F,  MEXU);  Estacion  de  Biologia  Tropical 
Los  Tuxtlas,  Cerro  Vigia,  21  May  1981,  Gentry  &  Lott  32208 
(CAS,  MEXU,  MO,  US);  Municipio  San  Andres  Tuxtla,  La- 
guna  Escondida,  2  km  NE  of  Estacion  de  Biologia  Tropical 
Los  Tuxtlas,  22  Jul.  1983,  Ibarra  754  (CAS);  Estacion  de  Bio- 
logia Tropical  Los  Tuxtlas,  95°04'  and  95°09'W,  18°34'  and 
18°36'N,  Lot  67,  24  Jun.  1984,  Ibarra  &  Cedilla  1792  (CAS, 
TEX),  Estacion  de  Biologia  Tropical  Los  Tuxtlas,  95°04'  and 
95°09'W,  1S°34'  and  18°36'N,  Lot  67,  21  Aug.  1984,  Ibarra 


ALMEDA:  MEXICAN  AND  CENTRAL  AMERICAN  MELASTOMATACEAE 


213 


FIGURE  2.  Miconia  ibarrae  Almeda.  A,  habit,  x  %;  B,  representative  leaf  (lower  surface),  x  %;  C,  representative  flower  with 
petals  and  stamens  removed  (left)  and  floral  bud  (right),  x  ca.  7;  D,  petal  (adaxial  surface),  x  9;  E,  stamens,  lateral  view  (left) 
and  ventral  view  (right),  x  8;  F,  mature  berry,  x  6;  G,  seeds,  x  ca.  6.  (A-E  from  the  holotype;  F  from  Ibarra  754;  G  from  Ibarra 
&  Sinica  1931.) 


&  Sinica  1931  (CAS);  Laguna  Escondida,  2.5  km  NE  of  Es- 
tacion  de  Biologia  Tropical  Los  Tuxtlas,  95°04'  and  95°09'W, 
18°34'  and  18°36'N,  23  Apr.  1985,  Ibarra  at  al.  2407  (CAS). 

This  restricted  species  is  defined  by  its  sparse 
cover  of  white  stellate-furfuraceous  hairs  on  the 
lower  leaf  surfaces,  prominent  interpetiolar  lines 


or  ridges  on  distal  branchlets,  undulate-dentic- 
ulate leaves,  truncate  to  obscurely  undulate  ca- 
lyx, papillose  petals,  and  irregularly  pyramidate 
seeds  that  are  smooth  and  vernicose  on  the 
rounded  or  bluntly  angulate  convex  face.  In  leaf 
shape,  calyx  development,  stamen  morphology, 


214 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  9 


and  petal  details,  Miconia  ibarrae  is  most  similar 
to  M.  fulvostellata  L.  O.  Williams,  a  more  wide- 
spread species  that  also  occurs  in  the  Los  Tuxtlas 
region  (Ibarra  and  Sinica  1987).  In  M.  fulvostel- 
lata, the  branchlets,  petioles,  lower  leaf  surfaces, 
inflorescences,  and  hypanthia  (at  anthesis)  are 
densely  covered  with  a  ferrugineous  stellate  in- 
dument  that  conceals  the  surfaces  of  these  struc- 
tures. A  superficial  comparison  of  these  two 
species  may  give  the  impression  that  M.  fulvo- 
stellata is  a  densely  pubescent  extreme  of  M. 
ibarrae.  However,  M.  fulvostellata  also  differs 
from  the  new  species  by  a  suite  of  other  diag- 
nostic characters:  its  style  is  straight,  its  stigma 
is  capitate,  and  its  pyramidate  seeds  have  pus- 
ticulate  angles  on  the  convex  face  and  a  raphe 
on  the  concave  side  invested  with  a  soft  hyaline 
tissue  that  collapses  irregularly  when  dry. 

This  species  is  named  for  Guillermo  Ibarra  M., 
coauthor  of  a  useful  checklist  of  the  plants  of  the 
Tropical  Biological  Station  at  Los  Tuxtlas,  Ve- 
racruz. It  was  through  his  generosity  that  a  fine 
set  of  flowering  and  fruiting  material  was  made 
available  for  this  study.  In  the  published  check- 
list, M.  ibarrae  is  listed  as  Miconia  sp.  (Ibarra 
and  Sinica  1987). 

Miconia  iteophylla  Almeda,  sp.  nov. 

(Fig.  3) 

TYPE.— PANAMA.  Code:  along  Rio  San  Juan  below  its 
junction  with  Rio  Tife,  elev.  1,200  ft  (366  m),  11  Jun.  1978, 
Hammel  3393  (holotype:  CAS!;  isotype:  MO). 

Frutex  0.5-1.5  m.  Ramuli  primum  obscure  quadrangulati 
demum  teretes  sicut  laminarum  subtus  venae  primariae  pe- 
tiolique  modice  granuloso-furfuracei.  Petioli  2-1 1  mm  longi; 
lamina  4.5-9  x  0.6-1.7  cm  anguste  elliptica  apice  acuminata 
basi  acuta  vel  acuminata,  3-plinervata,  membranacea  et  in- 
tegra.  Panicula  2-7  cm  longa  multiflora;  flores  5-meri,  pedi- 
cellis  (ad  anthesim)  1-2  mm  longis,  bracteolis  0.5  x  0.25  mm 
persistentibus.  Hypanthium  (ad  torum)  1-1.5  mm  longum  sparse 
caduceque  granuloso-furfuraceum;  calycis  tubus  0.25  mm  lon- 
gus,  lobis  subtruncatus  vel  vix  undulatus,  dentibus  exterioribus 
subulatis  0.25  mm  eminentibus.  Petala  glabra  2-2.5  x  0.75 
mm  oblongo-elliptica  apice  rotundato.  Stamina  isomorphica 
glabra;  filamenta  1-1.5  mm  longa;  antherarum  thecae  1.25- 
1.5  x  0.25  mm  angustae  oblongae,  poro  paulo  dorsaliter  in- 
clinato;  connectivum  nee  prolongatum  nee  appendiculatum. 
Stylus  3.5  mm  glaber;  ovarium  5-loculare  et  %  inferum  apicem 
versus  sparse  granuloso-puberulum. 

Shrubs  0.5-1.5  m  tall.  Uppermost  branchlets 
subquadrate  to  terete,  moderately  covered  with 
a  brown  scurfy-pulverulent  indument,  the  distal 
branchlet  nodes  bearing  interpetiolar  lines  or 
ridges.  Leaves  of  a  pair  somewhat  unequal  in 


size;  petioles  2-1 1  mm  long;  blades  membrana- 
ceous,  4.5-9  cm  long  and  0.6-1.7  cm  wide,  nar- 
rowly elliptic,  apex  acuminate,  base  acute  to  acu- 
minate, margin  entire,  3-plinerved,  the  inner  pair 
of  elevated  primaries  diverging  from  the  median 
nerve  in  opposite  or  subopposite  fashion,  the 
transverse  secondaries  prominulous  and  spaced 
l-2(-3)  mm  apart  at  the  widest  portion  of  the 
blade,  sparingly  and  deciduously  scurfy-pulver- 
ulent above  when  young,  deciduously  scurfy-pul- 
verulent on  the  elevated  primaries  and  higher 
order  venation  below.  Inflorescence  a  terminal 
multiflowered  panicle  2-7  cm  long  typically 
branching  0.6-3  cm  above  the  node  initiating  the 
inflorescence,  the  rachis  beset  with  brown  scurfy 
hairs;  bracts  of  the  rachis  nodes  sessile  and  per- 
sistent, 1.5-3.5  mm  long,  0.25-0.5  mm  wide, 
glabrous  above,  sparsely  and  deciduously  scurfy- 
pulverulent  below;  bracteoles  sessile  and  persis- 
tent, paired  and  fused  laterally  into  a  short  nodal 
collar  or  elevated  ridge,  linear-oblong  to  narrow- 
ly and  bluntly  triangular,  0.5  mm  long,  ca.  0.25 
mm  wide,  essentially  glabrous.  Pedicels  1-2  mm 
long,  moderately  to  densely  scurfy-pulverulent. 
Hypanthia  (at  anthesis)  campanulate,  1-1.5  mm 
long  to  the  torus,  deciduously  scurfy-pulverulent. 
Calyx  tube  0.25  mm  long,  subtruncate,  hyaline 
and  entire,  the  lobes  barely  discernible  as  5  de- 
pressed triangular  undulations;  exterior  calyx 
teeth  5,  subulate,  0.25  mm  long,  exceeding  and 
obscuring  the  lobes,  glabrous  throughout;  torus 
fimbrillate-puberulent  adaxially.  Petals  5,  gla- 
brous, white,  elliptic-oblong,  rounded  apically, 
2-2.5  mm  long,  0.75  mm  wide.  Stamens  10,  iso- 
morphic;  filaments  glabrous,  complanate,  1-1.5 
mm  long;  anthers  1.25-1.5  mm  long,  0.25  mm 
wide,  yellow,  linear-oblong,  truncate,  rounded  or 
emarginate  at  the  apex  with  a  somewhat  dorsally 
inclined  terminal  pore;  connective  thickened 
dorsally  but  not  prolonged  or  appendaged  at  the 
filament  insertion.  Ovary  (at  anthesis)  %  inferior, 
5 -celled,  globose,  apex  fluted  and  sparingly  glan- 
dular puberulent  but  becoming  rounded  and  gla- 
brous in  fruit.  Style  straight,  glabrous,  3.5  mm 
long;  stigma  truncate  to  capitellate.  Berry  glo- 
bose, red  turning  purple-black  when  mature,  2- 
4  mm  long  and  3-4  mm  in  diameter.  Seeds  ir- 
regularly angulate-pyramidate,  0.5  mm  long,  rus- 
ty brown,  smooth,  and  vernicose  with  polished 
angles  on  the  convex  face. 

PHENOLOGY.— Flowering  specimens  have  been 
collected  in  June  and  July;  fruiting  collections 


ALMEDA:  MEXICAN  AND  CENTRAL  AMERICAN  MELASTOMATACEAE 


215 


B 


FIGURE  3.  Miconia  iteophylla  Almeda.  A,  habit,  x  ca.  1;  B,  representative  leaf  (lower  surface),  x  ca.  1;  C,  flower  (at  anthesis) 
with  all  petals  and  stamens  removed,  x  10;  D,  petal  (adaxial  surface),  x  19;  E,  stamens,  dorsal  view  (left)  and  lateral  view  (right), 
x  ca.  14;  F,  mature  berry,  x  ca.  7;  G,  seeds,  x20.  (A-E  from  Hammel  3330;  F  from  the  holotype;  G  from  Antonio  3669.) 


have  been  made  in  February,  June,  and  Decem- 
ber. 

DISTRIBUTION. —Rain  forests  on  the  Caribbean 
slope  of  Panama  in  Bocas  del  Toro  and  Code 
provinces  where  it  typically  grows  near  moving 
water  at  200-450  m. 


ADDITIONAL  SPECIMENS  EXAMINED.— PANAMA.  Bocas  del 
Toro:  upper  Rio  San  Pedro,  13  Jul.  1979,  Gordon  59Db  (MO). 
Code:  trail  from  Cano  Susio  to  Cerro  Tife  on  the  Atlantic 
slope,  3  Feb.  1980,  Antonio  3669  (CAS);  Atlantic  slope  NW 
of  El  Cope,  along  Rio  San  Juan  near  fork  with  Rio  Tife,  9  Jun. 
1978,  Hammel  3330  (CAS);  Caribbean  side  of  divide  at  El 
Cope,  3  Feb.  1983,  Hamilton  &  Davidse  2628  (CAS);  between 


216 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  9 


Cano  Susio  and  waterfall  at  base  of  Cerro  Tife,  13  Dec.  1980, 
Sytsma  et  al.  2541  (CAS). 

Miconia  iteophylla  is  distinguished  by  its  brown 
scurfy-pulverulent  indument,  narrow  elliptic 
leaves,  linear-oblong  petals,  and  pyramidate  seeds 
with  smooth  polished  angles  on  the  convex  face. 
It  is  related  to  a  small  group  of  species  that  in- 
cludes M.  ligulata  and  M.  peltata,  both  of  which 
are  also  treated  herein.  These  three  species  share 
a  scurfy-pulverulent  indument,  linear-oblong 
petals,  a  deciduously  glandular  puberulent  ovary 
apex,  and  a  torus  that  is  puberulent  adaxially. 
The  brown  scurfy  indument  is  superficially  sim- 
ilar in  all  of  these  taxa.  In  M.  iteophylla,  however, 
it  consists  of  branlike  scales  unlike  the  pinoid 
hairs  of  M  ligulata  and  M.  peltata.  The  seeds  of 
M.  iteophylla  and  M.  peltata  are  identical  in  shape 
and  sculptural  details  which  suggests  that  these 
two  species  may  be  more  closely  related  to  one 
another  than  either  is  to  M.  ligulata.  In  general 
aspect  M.  iteophylla  resembles  M.  ligulata,  which 
is  readily  differentiated  by  its  larger  (12-2  8  x  4- 
10  cm)  5-plinerved  leaf  blades  that  are  decurrent 
on  the  petiole,  and  pyramidate  seeds  with  ver- 
ruculose  angles  on  the  convex  face. 

The  epithet  for  this  species  is  derived  from  the 
Greek  words  itea,  willow,  and  phyllon,  leaf,  in 
reference  to  the  narrow  willowlike  leaf  blades. 
This  species  is  referred  to  locally  as  "ugua"  (fide 
Gordon  59Db,  MO)  by  the  Guaymi,  the  largest 
Indian  group  in  Panama  (Gordon  1982). 

Miconia  ligulata  Almeda,  nom.  nov. 

Leandra  consimilis  Gleason,  Ann.  Missouri  Bot.  Gard.  45:268. 
1958. 

TYPE. -PANAMA.  Las  Minas,  Allen  2702  (holotype:  NY). 
Non  Miconia  consimilis  Pilger,  Verh.  Bot.  Ver.  Brand.  47: 1 70. 
1 905.  This  latter  entity  is  a  taxonomic  synonym  ofGraffenrieda 
gracilis  (Triana)  L.  O.  Williams  (Williams  1 963;  Wurdack  1 970). 

In  the  protologue,  Gleason  described  the  petals 
of  this  species  as  oblong-lanceolate.  This  is  un- 
derstandable in  view  of  the  fact  that  the  distal 
petal  margins  tend  to  become  involute  and  su- 
perficially appear  acute  when  pressed  and  dried. 
Traditionally  berry-fruited  species  in  the  Mico- 
nieae  with  terminal  inflorescences  and  acute  pet- 
als have  been  placed  in  Leandra.  The  petals  of 
Leandra  consimilis  are  actually  linear-oblong  and 
rounded  to  obtuse  apically  like  several  described 
species  in  Miconia.  To  rectify  its  taxonomic 
placement,  Leandra  consimilis  is  here  trans- 


ferred to  Miconia.  This  requires  the  adoption  of 
a  new  name  because  the  epithet  consimilis  is  al- 
ready pre-empted  in  Miconia.  When  Gleason  de- 
scribed this  species  it  was  known  from  only  the 
Panamanian  holotype.  Wurdack  (1978)  reported 
it  from  Costa  Rica  south  to  Venezuela  and  I  have 
recently  recorded  that  its  range  extends  north- 
ward to  Nicaragua  (Almeda,  in  press).  An  ex- 
panded description  is  provided  here  to  reflect  the 
range  of  variation  exhibited  by  the  species 
throughout  its  known  range. 

Shrubs  or  small  trees  (1.5-)4-9  m  tall.  Up- 
permost branches  rounded-quadrate  to  quadri- 
sulcate,  the  young  branchlets,  petioles,  primary 
leaf  veins  below,  and  inflorescences  moderately 
to  densely  covered  with  a  brown  scurfy  indument 
of  pinoid  hairs.  Leaves  of  a  pair  equal  to  some- 
what unequal  in  size;  petioles  0.7-2.5  cm  long; 
blades  membranaceous,  1 2-28  cm  long  and  4- 
1 0  cm  wide,  elliptic,  apex  attenuate  to  long-acu- 
minate, base  gradually  tapering  and  decurrent  on 
the  petiole,  margin  entire  to  undulate,  5-pli- 
nerved, the  innermost  pair  of  elevated  primaries 
diverging  from  the  median  nerve  in  opposite  or 
alternate  fashion,  the  transverse  secondaries 
prominulous  and  spaced  3-7  mm  apart  at  the 
widest  portion  of  the  blade,  sparsely  scurfy-pul- 
verulent to  glabrate  above,  moderately  and  de- 
ciduously scurfy-pulverulent  on  the  secondary 
and  higher  order  veins  below.  Inflorescence  a 
terminal  multiflowered  panicle  4-1 1.5  cm  long 
typically  branching  (0.5-)1. 5-3.5  cm  above  the 
node  initiating  the  inflorescence;  bracts  of  the 
rachis  nodes  persistent,  1-5  mm  long,  0.25-1 
mm  wide,  glabrous  above,  moderately  covered 
with  scurfy-pulverulent  or  short-stalked  pinoid 
hairs  below;  bracteoles  sessile  and  persistent, 
paired  and  fused  laterally  to  form  a  nodal  ridge 
or  shallow  collar,  narrowly  triangular,  0.5  mm 
long,  0.25-0.5  mm  wide,  glabrous  above  and 
sparingly  scurfy-pulverulent  to  glabrate  below. 
Pedicels  0.5-1  mm  long  with  an  indument  like 
that  of  the  bracteoles.  Hypanthia  (at  anthesis) 
campanulate,  1-1.5  mm  long  to  the  torus,  de- 
ciduously scurfy-pulverulent  to  stellulate-furfu- 
raceous.  Calyx  tube  0.25  mm  long,  subtruncate, 
hyaline,  and  entire  or  undulate  with  5  rounded- 
triangular  lobes  0.25  mm  long;  exterior  calyx  teeth 
5,  subulate,  0.25  mm  long,  equaling  or  exceeding 
and  obscuring  the  lobes,  glabrous  throughout; 
torus  sparsely  fimbrillate-puberulent  or  glandu- 
lar puberulent  adaxially.  Petals  5,  glabrous,  white, 


ALMEDA:  MEXICAN  AND  CENTRAL  AMERICAN  MELASTOMATACEAE 


217 


linear-oblong,  obtuse  to  rounded  apically,  2-3.5 
mm  long,  0.75-1  mm  wide.  Stamens  10,  iso- 
morphic;  filaments  glabrous,  complanate,  1  mm 
long;  anthers  0.75-1 .25  mm  long,  0.25  mm  wide, 
yellow,  linear-oblong,  truncate  to  slightly  emar- 
ginate  at  the  apex  with  a  somewhat  dorsally  in- 
clined terminal  pore;  connective  thickened  dor- 
sally but  not  prolonged  or  appendaged  at  the 
filament  insertion.  Ovary  (at  anthesis)  %  to  % 
inferior,  5 -celled,  globose,  apex  lobed  or  fluted 
and  sparsely  glandular  puberulent  but  becoming 
rounded  and  glabrous  in  fruit.  Style  straight,  gla- 
brous, 3-4  mm  long;  stigma  truncate  but  not 
conspicuously  dilated.  Berry  globose,  pink  but 
turning  blue-purple  when  mature,  2-4  mm  long 
and  3-4  mm  in  diameter.  Seeds  angular-pyrami- 
date,  ca.  0.5  mm  long,  brown,  smooth  with  ver- 
ruculose  angles  on  the  convex  face. 

PHENOLOGY.  — Flowering  sporadically  from 
January  through  October,  fruiting  specimens  have 
been  collected  from  August  through  January. 

DISTRIBUTION.— Widespread  in  wet  evergreen 
forests  on  the  Caribbean  slope  of  Central  Amer- 
ica from  eastern  Nicaragua  (Zelaya)  south  and 
east  through  Costa  Rica  and  Panama  to  north- 
western Venezuela  (Zulia)  from  sea  level  to  1 , 1 00 
m. 


REPRESENTATIVE  SPECIMENS  EXAMINED.— NICARAGUA. 
Zelaya:  SW  flank  of  Cerro  Hormiguero,  1 8  Apr.  1979,  Grijalva 
440  (CAS).  COSTA  RICA.  Alajuela:  between  San  Lorenzo  and 
Los  Angeles  de  San  Ramon,  above  Rio  San  Lorenzo,  10°14'N, 
84°32'W,  20  Sep.  1978,  Burger  &  Antonio  11176  (CAS).  Here- 
dia:  road  between  Puerto  Viejo  and  La  Virgen,  near  Chilamate, 
30  May  1 982,  Hammel  12670  (CAS);  Finca  La  Selva,  the  OTS 
Field  Station  on  the  Rio  Puerto  Viejo  just  E  of  its  junction 
with  the  Rio  Sarapiqui,  1  Aug.  1980,  Hammel  9403  (CAS). 
Limon:  2  airline  km  SSE  of  Islas  Buena  Vista  in  the  Rio  Col- 
orado, 83°40'W,  10°40'N,  13-14Sep.  1 986,  Davidse  &  Herrera 
31064  (CAS);  Hacienda  Tapezco-Hda.  La  Suerte,  29  air  km 
W  of  Tortuguero,  10°30'N,  83°47'W,  22  Aug.  1979,  Davidson 
&  Donahue  8665  (CAS);  Cerro  Coronel,  E  of  Rio  Zapote,  along 
and  above  new  road  within  1  km  of  Rio  Colorado,  10°40'N, 
83°40'W,  13-14  Sep.  1986,  Stevens  &  Montiel  24356  (CAS). 
PANAMA.  Code:  La  Mesa  above  El  Valle  de  Anton,  N  of 
Cerro  Gaital,  8°37'N,  80°06'W,  26  Jul.  1984,  de  Nevers  et  al. 
3524  (CAS);  foothills  and  summit  of  Cerro  Caracoral,  near  La 
Mesa  and  N  of  El  Valle  de  Anton,  10  Sep.  1981,  Knapp  1092 
(CAS);  Atlantic  slope  of  the  Continental  Divide  near  sawmill 
above  El  Cope,  8°40'N,  80°36'W,  1 3  Feb  1982,  Knapp  &  Dres- 
sier 3  506  (CAS);  Rio  Guanche,  9°30'N,  79°39'W,  15  Oct.  1980, 
Sytsma  1666  (CAS).  Comarca  de  San  Bias:  El  Llano-Carti 
Road,  Continental  Divide,  9°19'N,  78°55'W,  25  Aug.  1984,  de 
Nevers  3760  (CAS).  Darien:  Cerro  Pirre  valley  between  Pirre 
and  next  most  southerly  peak,  10-20  Jul.  1977,  Folsom  4429 
(CAS).  Panama:  top  of  Cerro  Campana,  8  Sep.  1982,  D'Arcy 
15092  (CAS);  Pipeline  road,  N  of  Gamboa,  9°10'N,  79°45'W, 


4  Aug.  1984,  de  Nevers  &  Hews  3635  (CAS).  VENEZUELA. 
Zulia:  Cano  Helena,  Sierra  Perija,  Delascio  &  Benkowski  3197 
(US.) 

Miconia  ligulata  belongs  to  the  group  of  three 
species  discussed  above  under  M.  iteophylla. 
Among  its  close  allies  it  is  unique  in  having  wide 
leaf  blades  (4-10  cm)  that  are  gradually  tapered 
at  the  base  and  decurrent  on  the  petiole.  The 
verruculose  angles  on  the  convex  seed  face  con- 
stitute another  distinctive  but  less  conspicuous 
feature  of  M.  ligulata.  The  new  name  for  this 
species  is  derived  from  ligula,  Latin  for  little 
tongue  or  strap,  in  reference  to  the  elongate  strap- 
like  petals. 

Miconia  peltata  Almeda,  sp.  nov. 
(Fig.  4) 

TYPE.— PANAMA.  Panama:  near  Cerro  Jefe,  along  road  to- 
wards Alto  Pacora,  forested  slopes  ca.  850  m,  ca.  9°15'N, 
79°30'W,  27  Dec.  1985,  McPherson  7882  (holotype:  CAS!; 
isotype:  MO). 

Frutex  vel  arbor  parva  3-6  m.  Ramuli  primum  paulo  com- 
pressi  demum  teretes  sicut  petioli  foliorum  subtus  venae  pri- 
mariae  inflorescentiaque  pilis  pinoideis  ca.  0.25  mm  longis 
plus  minusve  caducis  dense  induti.  Petioli  1.3-5  cm  longi; 
lamina  peltata  6. 7-14.5  x  3.3-8.3cmovatavelelliptico-ovata 
apice  acuminata  basi  rotundata  vel  subcordata,  5-7-nervata, 
subcoriacea  et  undulato-denticulata  vel  subintegra.  Inflores- 
centia  3-4  cm  longa  multiflora,  ramis  oppositis  divaricatis; 
flores  5-meri,  pedicellis  (ad  anthesim)  1  mm  longum,  pilis 
plerumque  breviuscule  stipitato-stellatis  vel  stellulato-pinoide- 
is  praeditum.  Calycis  tubus  0.25  mm  longus,  lobis  ca.  0.25 
mm  altis  late  triangularibus,  dentibus  exterioribus  triangulari- 
bus  0.25  mm  eminentibus.  Petala  glabra  2.5  x  0.5  mm  ob- 
longa-elliptica  apice  rotundato  vel  rotundato-obtuso.  Stamina 
isomorphica  glabra;  filamenta  1  mm  longa;  antherarum  thecae 
0.75-1  x  0.25  mm  anguste  oblongae,  poro  paulo  dorsaliter 
inclinato;  connectivum  nee  prolongatum  nee  appendiculatum. 
Stylus  1.5  mm  glaber;  ovarium  5-loculare  et  %  inferum  apice 
sparsiuscule  glanduloso-puberulo. 

Shrubs  or  small  trees  3-6  m  tall.  Uppermost 
branchlets  compressed  and  two-sided  but  be- 
coming rounded  with  age.  Distal  branchlets,  pet- 
ioles, primary  leaf  veins  below,  and  inflores- 
cences densely  and  somewhat  deciduously 
covered  with  a  rusty  brown  indument  of  pinoid 
hairs.  Leaves  of  a  pair  somewhat  unequal  in  size, 
peltate  with  the  petiole  attached  5-8  mm  from 
the  basal  edge,  petioles  1.3-5  cm  long;  blades 
subcoriaceous,  6.7-14.5  cm  long  and  3.3-8.3  cm 
wide,  ovate  to  elliptic-ovate,  apex  acuminate,  base 
broadly  rounded  to  subcordate,  margin  incon- 
spicuously undulate-denticulate  to  subentire,  5- 
7-nerved  abaxially,  the  transverse  secondaries 
prominulous  and  2-3  mm  apart  at  the  widest 


218 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  9 


FIGURE  4.  Miconia  peltata  Almeda.  A,  habit,  x  ca.  %;  B,  representative  leaf  (lower  surface),  x  ca.  %;  C,  petal  (adaxial 
surface),  x  1 2;  D,  stamens,  dorsal  view  (left)  and  lateral  view  (right),  x  ca.  1 7;  E,  mature  berry  and  bracteoles,  x  6;  F,  enlargement 
of  pinoid  hairs  on  fruiting  pedicel,  x  ca.  60;  G,  seeds,  x  16.  (A,  B,  E,  F  from  Tyson  3440;  C,  D,  G  from  the  holotype.) 


portion  of  the  blade,  glabrous  above,  the  sec- 
ondaries and  higher  order  veins  below  sparingly 
beset  with  spreading  pinoid  hairs  and  an  incon- 
spicuous appressed  glandular  puberulence.  Inflo- 
rescence a  multiflowered  paniculiform  dicha- 
sium  3-4  cm  long  divaricately  branched  from 
the  base;  bracts  of  the  rachis  nodes  sessile  and 


persistent,  1.5-4  mm  long,  0.5  mm  wide,  sparse- 
ly beset  with  pinoid  hairs;  bracteoles  sessile  and 
persistent,  subulate,  0.25-0.5  mm  long,  mostly 
less  than  0.25  mm  wide,  essentially  glabrous 
throughout.  Pedicels  1  mm  long.  Hypanthia  (at 
anthesis)  campanulate  to  subglobose,  1.5  mm 
long  to  the  torus,  moderately  to  sparsely  covered 


ALMEDA:  MEXICAN  AND  CENTRAL  AMERICAN  MELASTOMATACEAE 


219 


TABLE  1 .    Comparison  of  chromosome  numbers,  pollen  size,  and  ecogeographic  ranges  in  Pilocosta. 


Haploid 
number 

Pollen  diameter 

Ecogeographic  range 

P.  campanensis 
P.  erythrophylla 
P.  nana 

P.  oerstedii 

n  =  11 
n  =  7,  14 
n  =  33 

n=  18 

16.5-1  8.8  urn 
14.1-16.5  /tm 
25.5-30.6  /tm 

17.6-2  1.2  Mm 

Central  Panama  at  300-1,000  m 
Central  Costa  Rica  at  1,480-2,000  m 
Costa  Rica  through  central  Panama  to  Colombia  (Magdalena) 
and  Ecuador  (Pichincha)  at  450-1,800  m 
Central  Costa  Rica  to  western  Panama  at  900-2,000  m 

with  stipitate-stellate  or  short  pinoid  hairs.  Calyx 
tube  0.25  mm  long,  subtruncate,  hyaline  and  en- 
tire, with  5  triangular  lobes  ca.  0.25  mm  long; 
exterior  calyx  teeth  5,  bluntly  triangular,  0.25 
mm  long,  glabrous  throughout;  torus  inconspic- 
uously nmbrillate  puberulent  adaxially.  Petals  5, 
glabrous,  reddish-pink,  oblong-elliptic,  rounded 
to  obtuse  apically,  2.5  mm  long,  0.5  mm  wide. 
Stamens  10,  isomorphic;  filaments  glabrous, 
complanate,  1  mm  long;  anthers  0.75-1  mm  long, 
0.25  mm  wide,  pale  yellow,  oblong,  rounded  to 
subtruncate  at  the  apex  with  a  somewhat  dorsally 
inclined  terminal  pore;  connective  thickened 
dorsally  but  not  prolonged  or  appendaged  at  the 
filament  insertion.  Ovary  (at  anthesis)  %  inferior, 
5-celled,  globose,  apex  fluted  and  sparingly  glan- 
dular puberulent  but  rounded  and  glabrous  in 
fruit.  Style  straight,  glabrous,  1.5  mm  long;  stig- 
ma truncate.  Berry  globose,  purple  when  mature, 
2-2.5  mm  long  and  3-4  mm  in  diameter.  Seeds 
irregularly  angulate-pyramidate,  0.5  mm  long, 
brown,  smooth  with  polished  angles  on  the  con- 
vex face. 

PHENOLOGY.— Flowering  and  fruiting  speci- 
mens have  been  collected  in  October,  December, 
and  February. 

DISTRIBUTION.— Endemic  to  Cerro  Jefe  in  cen- 
tral Panama  at  850-1,000  m. 

ADDITIONAL  SPECIMENS  EXAMINED.— PANAMA.  Panama: 
Cerro  Jefe,  7  Oct.  1986,  Aranda  &  Valdespino  185  (CAS,  MO, 
PMA);  E  slope  of  Cerro  Jefe,  8  Feb.  1966,  Tyson  3440  (MO- 
2  sheets). 

Miconia  peltata  also  belongs  to  the  group  of 
three  species  discussed  above  under  M.  iteo- 
phylla.  Because  of  its  ovate  to  elliptic-ovate  pel- 
tate leaves  it  is  recognizable  even  in  sterile  con- 
dition. In  addition  to  its  unusual  leaves,  M.  peltata 
differs  from  closely  allied  species  in  having  red- 
dish-pink petals  and  an  inflorescence  that  is  di- 
varicately branched  from  the  base.  The  epithet 
peltata  is  derived  from  the  Greek  word,  pelte, 


small  shield.  This  refers  to  the  attachment  of  the 
petioles  to  the  lower  foliar  surface  away  from  the 
margin  of  the  blade. 

Pilocosta  campanensis  (Almeda  &  Whiffin) 
Almeda,  stat.  nov. 

Pilocosta  oerstedii  (Triana)  Almeda  &  Whiffin  subsp.  campa- 
nensis Almeda  &  Whiffin,  Syst.  Hot.  5(3):306.  1980  [1981]. 

TYPE.— PANAMA.  Panama:  Cerro  Campana,  10  Dec.  1967, 
Lewis  et  al.  3069  (holotype:  MO!;  isotypes:  COL,  DUKE!,  F!, 
K!,  NY!,  UC!). 

When  Trevor  Whiffin  and  I  described  this  as 
a  subspecies  of  Pilocosta  oerstedii  we  based  our 
decision  on  similarities  in  habit  and  androecial 
details  (Almeda  and  Whiffin  1981).  We  also  en- 
tertained the  possibility  of  describing  it  as  a  sub- 
species of  P.  nana  because  of  similarities  in  the 
kind  and  distribution  of  hypanthial  pubescence. 
At  the  time  it  was  the  only  taxon  in  the  genus 
for  which  we  had  no  chromosome  information. 
Based  on  information  then  available  we  recog- 
nized two  subspecies  under  P.  oerstedii  because 
of  their  non-overlapping  geographic  distribution 
and  elevational  displacement.  A  recent  study  of 
chromosome  numbers  in  neotropical  Melasto- 
mataceae  (Almeda  and  Chuang,  in  prep.)  reveals 
that  this  taxon  has  a  haploid  number  of  n  =  11 
(count  based  on  field  collected  buds  of  Almeda 
et  al.  5880,  CAS).  Based  on  morphological,  cy- 
tological,  and  ecogeographic  data,  it  is  clear  that 
Pilocosta  oerstedii  subsp.  campanensis  has  all  the 
attributes  that  characterize  species  in  this  genus 
(Table  1).  I  therefore  elevate  it  to  species  status. 
An  alternative  disposition  would  be  to  make  it 
a  subspecies  of  P.  nana.  This,  however,  would 
make  no  sense  in  view  of  their  pronounced  mor- 
phological and  cytological  differences  and  the  fact 
that  I  have  found  individuals  of  these  two  taxa 
growing  next  to  one  another  near  La  Mesa  in 
Code  Province,  Panama. 


220 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  9 


A  chromosome  number  of  n  =  1 1  in  P.  cam- 
panensis  reconfirms  our  previous  assertion  re- 
garding the  complex  cytological  evolution  within 
this  small  natural  genus.  It  also  sheds  new  light 
on  the  relationship  of  P.  nana  and  provides  some 
clues  on  the  evolutionary  polarity  of  character 
states  judged  to  be  of  taxonomic  value  in  Pilo- 
costa.  With  n  =  33,  P.  nana  appears  to  be  a 
hexaploid  derivative  of  P.  campanensis.  The 
polyploid  gain  in  P.  nana  has  evidently  been 
accompanied  by  the  derivation  of  an  herbaceous 
habit  and  subisomorphic  androecium  from  the 
woody  habit  and  dimorphic  androecium  of  P. 
campanensis. 

ACKNOWLEDGMENTS 

Fieldwork  for  this  study  was  supported,  in  part, 
by  U.S.  National  Science  Foundation  Grant  BSR 
8614880  (Flora  Mesoamericana)  and  the  In- 
House  Research  Fund  of  the  California  Academy 
of  Sciences.  For  special  assistance  during  the 
course  of  this  study  I  thank  Gerrit  Davidse,  Barry 
E.  Hammel,  Guillermo  Ibarra  M.,  and  Gordon 
McPherson.  I  am  also  grateful  to  the  Missouri 
Botanical  Garden  and  the  Smithsonian  Tropical 
Research  Institute  for  logistical  support  in  the 
field;  Carolyn  Rendu  for  the  illustrations;  Tsan- 


lang  and  Fei-mei  Chuang  for  assistance  with 
chromosome  counts;  and  the  curators  and  staffs 
of  the  following  herbaria  (acronyms  fide  Holm- 
gren et  al.  1981)  for  providing  special  loans,  gifts, 
and/or  assistance  during  study  visits:  BM,  BR, 
CR,  DUKE,  F,  K,  MA,  MEXU,  MO,  NY,  P, 
PMA,  US. 

LITERATURE  CITED 

ALMEDA,  F.  In  press.  Melastomataceae.  In  Flora  de  Nica- 
ragua. W.  D.  Stevens,  ed.  Missouri  Botanical  Garden,  St. 
Louis. 

ALMEDA,  F.  AND  T.  WHIFFIN.  1981.  Pilocosta,  a  new  genus 
of  tropical  American  Melastomataceae.  Syst.  Bot.  5(3):294- 
311. 

GORDON,  B.  L.  1982.  A  Panama  forest  and  shore.  The  Box- 
wood Press,  Pacific  Grove,  California. 

HOLMGREN  P.  K.,  W.  KEUKEN,  AND  E.  K.  SCHOFIELD.  1981. 
Index  Herbariorum.  Part  1, 7th  ed.  The  herbaria  of  the  world. 
RegnumVeg.  106:l^t52. 

IBARRA  M.,  G.  AND  S.  SINICA  C.  1987.  Listados  floristicos 
de  Mexico.  VII.  Estacion  de  Biologia  Tropical  Los  Tuxtlas, 
Veracruz:  1-51.  Universidad  Nacional  Autonoma  de  Mex- 
ico, Mexico. 

WILLIAMS,  L.  O.  1 963.  Tropical  American  plants  V.  Fieldiana 
Bot.  29(10):545-586. 

WURDACK.J.  J.  1970.  Certamen  Melastomataceis  XV.  Phy- 
tologia  20:369-389. 

.  1978.  Suplemento  a  las  Melastomaceas  de  Venezue- 
la. Acta  Bot.  Venez.  13:125-172. 


CALIFORNIA  ACADEMY  OF  SCIENCES 

Golden  Gate  Park 
San  Francisco,  California  94 1 1 8 


PROCEEDINGS 

OF  THE 

CALIFORNIA  ACADEMY  OF  SCIENCES 


Vol.  46,  No.  10,  pp.  221-242,  14  figs.  December  20,  1989 


REVIEW  OF  THE  EELPOUT  GENUS 

PACHYCARA  ZUGMAYER,  1911  (TELEOSTEI:  ZOARCIDAE), 
WITH  DESCRIPTIONS  OF  SIX  NEW  SPECIES 

By 
M.  Eric  Anderson 

Department  of  Ichthyology,  California  Academy  of  Sciences,  Golden  Gate  Park, 
San  Francisco,  California  94118 


ABSTRACT:  The  eelpout  genus  Pachycara  Zugmayer,  1911  is  revised  to  include  12  species,  6  of  which  are 
reported  here  as  new:  P.  shcherbachevi  from  the  Indian  Ocean;  P.  sulkai  and  P.  crossacanthum  from  the 
Atlantic;  and  P.  mesoporum,  P.  pammelas,  and  P.  rimae  from  the  Pacific.  Pachycara  is  a  primitive  lycodine 
eelpout  genus  characterized  by  its  large  body  size,  small  head  pores,  lack  of  mental  crests,  and  low  para  sphenoid 
wing.  It  is  most  closely  related  to  Lycodes  Reinhardt,  1832,  and  Thermarces  Rosenblatt  and  Cohen,  1986. 
The  species  of  Pachycara  are  characterized  by  their  lateral  line  patterns,  variably  present  or  absent  pelvic 
fins,  counts  of  the  axial  skeleton,  gill  raker  morphology,  head  pore  patterns,  and  branchiostegal  ray  number. 
The  genus  is  distributed  in  the  eastern  Pacific,  Antarctic,  Atlantic  and  Indian  oceans  from  the  upper  continental 
slope  to  the  abyss  (200-4,780  m). 

Received  February  21, 1989.  Accepted  July  1,  1989. 

,  Considerable  taxonomic  confusion  exists  in  the 

literature  on  eelpouts,  especially  for  the  rare  deep- 

The  eelpout  genus  Pachycara  was  erected  by  sea  species.  As  was  found  earlier  (Anderson 
Zugmayer  (191  la,  b)  for  a  species  known  from  1982a),  comparison  of  recently  collected  speci- 
a  single  specimen  taken  in  the  abyssal  North  At-  mens  of  Pachycara  with  older,  published  mate- 
lantic.  Anderson  (1988a)  considered  the  Antarc-  rial  was  difficult  because  of  a  lack  of,  or  poor 
tic  genus  Austrolycichthys  Regan,  1 9 1 3  a  junior  state  of,  types  and  the  use  of  non-diagnostic  char- 
synonym  of  Pachycara,  broadening  the  diagnosis  acters  by  previous  authors.  As  a  contribution  to 
chiefly  by  rejecting  the  presence  or  absence  of  the  clarification  of  the  systematics  of  Pachycara, 
pelvic  fins  as  a  generic  character  in  eelpouts  (see  I  provide  descriptions  and/or  diagnoses  for  all 
also  Anderson  1988b).  Anderson  and  Peden  species,  list  synonymies,  provide  a  key,  distri- 
(1988)  reported  the  genus  from  the  North  Pacific  butional  maps  and  illustrations  to  facilitate  iden- 
for  the  first  time,  further  broadening  its  limits,  tification,  and  compare  Pachycara  to  its  putative 
and  described  two  new  species.  They  cited  my  sister  genera. 

unpublished  dissertation  (Anderson  1984)  as  a  Prior  to  the  postwar  rejuvenation  of  deep-sea 
source  of  information  on  the  genus.  This  paper  exploration,  specimens  here  referred  to  Pachy- 
is  an  update  of  that  study,  revised  to  include  1 2  cam  were  few  in  collections,  and,  in  fact,  con- 
species,  six  of  which  are  new.  sisted  only  of  the  types  of  P.  bulbiceps  (Garman, 

[221] 


222 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  10 


1 899)  (=P.  obesa  Zugmayer,  1 9 1 1 ),  P.  suspectum 
(Garman,  1 899),  P.  crassiceps  (Roule,  1916),  and 
about  60  specimens  of  P.  brachycephalum  (Pap- 
penheim,  1912).  The  more  important  recent  col- 
lections of  deep-sea  fishes  that  resulted  in  the 
capture  of  Pachycara  specimens  are  briefly  dis- 
cussed below. 

Specimens  from  the  Pacific  Ocean  were  first 
taken  during  the  round-the-world  expedition  of 
the  Danish  ship  GALATHEA  which  collected 
the  rare  P.  suspectum  in  the  Gulf  of  Panama  in 
1952.  This  expedition  had  in  the  previous  year 
collected  the  holotype  of  P.  shcherbachevi  in  the 
Bay  of  Bengal.  The  bulk  of  the  North  Pacific 
species  were  taken  by  free  vehicle  set  line  traps 
by  personnel  of  the  Scripps  Institution  of  Ocean- 
ography off  Mexico  and  southern  California,  by 
bottom  trawls  operated  from  ships  of  Oregon 
State  University,  and  by  Canadian  experimental 
sablefish  (Anoplopoma  fimbrid)  traps  off  the 
Queen  Charlotte  Islands  (summarized  in  An- 
derson and  Peden  1988).  Specimens  from  the 
southeastern  Pacific  have  come  from  operations 
of  DSV  ALVIN  in  the  Galapagos  Rift  Zone  (Co- 
hen and  Haedrich  1983),  Chilean  fisheries  re- 
connaissance, supplied  to  the  author  by  Ismael 
Kong,  University  of  Antofagasta,  and  deep-sea 
bottom  trawls  by  the  ship  ANTON  BRUUN 
during  the  National  Science  Foundation's  South- 
eastern Pacific  Biological  Oceanographic  Pro- 
gram (SEPBOP).  Explorations  in  the  Atlantic 
Ocean  began  with  the  capture  of  the  type  species, 
P.  obesa  (=P.  bulbiceps)  by  the  expeditions  of 
Prince  Albert  I  of  Monaco  in  the  Bay  of  Biscay 
(Zugmayer  191  la,  b).  Subsequently,  other  east- 
ern Atlantic  material  has  come  from  expeditions 
of  the  Institute  of  Oceanographic  Sciences,  U.K. 
(Merrett  and  Marshall  1981),  the  BENGUELA 
VII  expedition  of  the  Institute  de  Investigaciones 
Pesqueras  de  Barcelona  (now  Institute  de  Cien- 
cias  del  Mar),  fish  trapping  cruises  of  the  Office 
de  la  Recherche  Scientifique  et  Technique  Outre- 
Mer,  Dakar,  Senegal,  exploratory  fisheries  cruis- 
es of  the  USSR  (Golovan'  1978),  and  similar 
cruises  by  the  South  African  ship  AFRICANA 
around  the  Cape  of  Good  Hope.  Specimens  of 
Pachycara  from  the  northwestern  Atlantic  are 
relatively  few  and  were  taken  by  bottom  trawls 
in  the  Middle  Atlantic  Bight  (Musick  1979;  Hae- 
drich et  al.  1980),  the  Caribbean  Sea  (Anderson 
et  al.  1986),  by  bottom  traps  around  Puerto  Rico, 
supplied  to  the  author  by  Dannie  A.  Hensley, 


University  of  Puerto  Rico,  and  from  operations 
of  DSV  ALVIN  in  the  Gulf  of  Mexico  (Paull  et 
al.  1984). 

METHODS  AND  MATERIALS 

Measurements  were  made  with  dial  calipers  or 
ocular  micrometer  to  the  nearest  0.1  mm.  Os- 
teological  observations  were  made  on  cleared  and 
stained  specimens  (Dingerkus  and  Uhler  1977) 
and  drawings  made  with  the  aid  of  a  camera 
lucida.  Definitions  of  characters  and  measure- 
ments follow  those  of  Anderson  (1982a,  1984). 
Institutional  abbreviations  are  as  listed  in  Lev- 
iton  et  al.  (1985),  or  as  emended  by  Leviton  and 
Gibbs(1988). 

This  review  is  based  on  249  specimens,  76- 
597  mm  standard  length  (SL),  housed  in  21  mu- 
seum collections  in  12  countries  (listed  by  ab- 
breviation in  each  account;  measurements  of 
specimens  in  SL).  All  type  specimens  were  ex- 
amined except  the  holotype  of  Lycenchelys  cras- 
siceps Roule,  1916,  thought  to  be  lost.  A  lecto- 
type  of  Lycodes  brachycephalus  Pappenheim, 
1912,  is  selected.  Full  accounts  and  illustrations 
are  provided  for  all  species  except  those  recently 
reviewed  (P.  brachycephalum  [see  Anderson 
1988a],  P.  bulbiceps,  P.  gymninium,  P.  lepinium, 
and  P.  suspectum  [see  Anderson  and  Peden 
1988]),  but  expanded  diagnoses  for  these  species 
are  provided  which  include  other,  important 
characters,  for  ease  of  reference  under  one  cover. 
The  osteological  section  of  the  description  of  the 
genus  is  intentionally  brief,  and  the  reader  should 
consult  Anderson  (1982a)  for  a  more  thorough 
account  of  a  generalized  zoarcid  (which  includes 
Pachycara}. 

Species  accounts  are  presented  alphabetically 
by  ocean  basin  in  the  following  order:  Atlantic, 
Indian,  Pacific,  and  Antarctic. 

Abbreviations  for  bone  names  used  in  the  text 
figures  are  as  follows: 

act— actinosts 

ang— anguloarticular 

boc — basioccipital 

br  r— branchiostegal  ray 

cerhy— ceratohyal 

cl— cleithrum 

cor— coracoid 

den— dentary 

ect— ectopterygoid 

epihy— epihyal 


ANDERSON:  REVIEW  OF  EELPOUT  GENUS  PACHYCARA 


223 


epo— epioccipital  (epiotic) 

exo— exoccipital 

fr— frontal 

hyom — hyomandibula 

hyphy — hypohyals 

inop — interopercle 

inthy— interhyal 

lat  eth— lateral  ethmoid 

mes — mesethmoid 

met— metapterygoid 

op— opercle 

pal— palatine 

par— parietal 

pas— parasphenoid 

pel — postcleithrum 

pel— pelvic  (basipterygium) 

pmx — premaxilla 

pop— preopercle 

ptem — posttemporal 

pto— pterotic 

pts — pterosphenoid 

quad— quadrate 

scap— scapula 

sop— subopercle 

sph— sphenotic 

supcl — supracleithrum 

urhy— urohyal 

Pachycara  Zugmayer,  1911 

Pachycara  Zugmayer,  191  la:  12  (original  description;  expand- 
ed in  Zugmayer,  1 9 1 1  b:  1 34-1 36,  pi.  VI,  fig.  6.  Type  species: 
Pachycara  obesa  Zugmayer,  1911  (=Maynea  bulbiceps  Gar- 
man,  1 899),  by  monotypy. 

Austrolycichthys  Regan,  1 9 1 3:244-245,  fig.  2  (type  species:  Ly- 
codes  brachycephalus  Pappenheim,  1 9 1 2,  by  subsequent  des- 
ignation of  Jordan,  1920:550). 

Pachycarichthys  Whitley,  1931:334  (improper  emendation). 

Pachychara  (lapsus  calami).  Fowler,  1936:1056,  fig.  437;  An- 
driashev,  1986:170,  text  fig. 

DIAGNOSIS.  — Body  robust,  tail  short;  skin  firm; 
mental  crests  (cartilaginous  ridges  on  dentary) 
absent;  parasphenoid  wing  not  extending  above 
mid-height  of  trigeminofacialis  foramen;  pala- 
topterygoid  (palatal)  arch  well  developed,  with 
ectopterygoid  and  mesopterygoid  broadly  artic- 
ulating with  quadrate;  cephalic  lateralis  pores  few, 
small,  rounded;  precaudal  vertebrae  21-32;  sub- 
orbital  bones  6-8;  pelvic  fins  present  or  absent, 
their  presence  individually  variable  in  two 
species;  scales,  body  lateral  line,  vomerine  and 
palatine  teeth,  pseudobranch  (except  in  one 
species),  and  pyloric  caeca  present. 


DESCRIPTION.— Head  large,  ovoid,  snout  gen- 
erally bluntly  rounded,  never  dorsoventrally  de- 
pressed in  adults;  young  juveniles  with  some- 
what depressed  heads,  but  snout  not  as  long  and 
tapering  as  lycenchelyine  (Anderson  1984)  gen- 
era. Largest  males  with  wider  head  than  in  fe- 
males owing  to  greater  development  of  adductor 
mandibulae  (in  those  species  with  sample  size 
large  enough  to  test  statistically).  Body  elongate, 
robust,  subcircular  in  cross-section,  its  greatest 
depth  midway  between  pectoral  base  and  anus; 
body  height  at  anal  fin  origin  7.7-16.9%  SL  in 
specimens  over  1 60  mm  SL.  Tail  high,  increas- 
ingly laterally  compressed  posteriad.  Skin  firm, 
not  gelatinous,  covering  all  fins.  Scales  minute, 
cycloid,  imbedded,  covering  body  usually  in- 
cluding abdomen  (except  in  smallest  specimens 
and  P.  rimae),  bases  of  vertical  fins,  pectoral  base 
and  pectoral  axil.  Complete  mediolateral  and 
ventral  (Andriashev  1954,  fig.  142)  lateral  line(s) 
developed  as  single  row  of  superficial  neuromasts 
("free  lateralis  organs";  Springer  and  Freihofer 
1976),  their  origins  variable  and  usually  species 
specific.  Eye  large  and  ovoid  in  young,  often  rel- 
atively small  and  rounded  in  adults,  often  en- 
tering dorsal  profile  of  head.  Nostril  single,  tu- 
bular, at  snout  tip,  nasal  tube  not  reaching  upper 
lip  except  in  some  young.  Pectoral  fin  large,  of 
14-19  rays,  middle  rays  longest  in  large  speci- 
mens, lowermost  rays  thickened,  slightly  exsert- 
ed  at  tips.  Pectoral  base  low  on  body,  its  origin 
just  below  midline,  insertion  on  abdomen.  Gill 
slit  extending  ventrally  to  lower  margin  of  pec- 
toral base  or  below.  Ovary  single.  Pseudobranch 
filaments  2-7  (absent  in  P.  rimae).  Two  nublike 
pyloric  caeca,  usually  better  defined  in  young. 
Stomach  and  intestines  pale.  Peritoneum  black. 
Reproductive  mode  unknown,  probably  ovipa- 
rous. 

Mouth  terminal  or  subterminal,  lips  fleshy  but 
not  thickened.  Upper  lip  free  across  snout,  lower 
lip  adnate  at  dentary  symphysis.  Oral  valve  (pal- 
atal membrane)  moderately  developed,  usually 
reaching  anterior  margin  of  vomer,  but  often 
constricted  at  sides  except  in  very  small  speci- 
mens. 

Cephalic  lateralis  pores  small,  rounded,  rela- 
tively few.  Six  (usually;  eight  in  P.  crassiceps) 
suborbital  bones  from  which  arise  5-7  pores.  One 
to  four  postorbital  pores  present,  their  number, 
like  those  of  suborbitals,  individually  variable 
within  a  species.  Two  nasal  (anterior  supraor- 


224 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  10 


hyom 


pmx 


pop 


quad 


den 


ang 


FIGURE  1.    Jaws,  suspensorium,  and  opercular  bones  ofPachycara  bulbiceps.  BMNH  1981.6.16:15,  332  mm  SL;  right  lateral 
view. 


bital)  pores  present  at  snout  tip  arising  from  fo- 
ramina in  anterior  and  posterior  ends  of  nasal 
bone.  Interorbital  pore(s)  absent.  Occipital  pores 
absent  except  in  P.  mesoporum  and  in  two  of  six 
specimens  of  P.  crossacanthum.  Preopercular  and 
mandibular  canals  joined;  preoperculomandibu- 
lar  pores  and  bone  foramina  system  in  plesio- 
morphic  state  for  family  (and  no  variation  not- 
ed), with  four  pores  emanating  from  dentary,  one 
from  anguloarticular,  and  three  from  preopercle 
(Fig.  1).  Head  speckled  with  small,  white  super- 
ficial neuromasts,  particularly  on  nape  and 
cheeks;  no  pattern  to  these  discernible. 

Neurocranium  well  ossified,  somewhat  box- 
like  (Fig.  2).  Lateral  ethmoid  broad,  not  strongly 
decurved,  but  deeply  sculptured  at  surface  for 
attachment  of  palatine  ligaments.  Anterior,  cup- 
shaped  section  of  mesethmoid  broad,  deep.  Para- 
sphenoid  wing  low,  not  reaching  mid-height  of 
trigeminofacialis  foramen,  broadly  articulating 
with  frontal  and  pterosphenoid.  Pterosphenoid 
almost  completely  forming  anterior  border  of  tri- 
geminofacialis foramen.  Frontals  unfused,  with 
deep,  open  channel  anteriorly  passing  supraor- 
bital  lateralis  canal.  Sphenotic  and  parietal  sep- 
arated by  frontal  and  pterotic.  Lateral  protuber- 
ance for  attachment  of  levator  arcus  palatini 
muscle  a  weak  ridge.  Parietals  separated  from 
midline  by  supraoccipital,  with  no  channel 
through  posterior  margin  bracing  supratemporal 
commissure  (which  is  incomplete  or  absent  in 
all  but  two  species).  Supraoccipital  large,  anterior 
ramus  extending  well  under  frontals,  mesial  crest 
low;  supraoccipital  narrowly  contacting  exoccip- 


ital  posteriorly.  Prootic  and  sphenotic  articulat- 
ing dorsally  forming  hyomandibular  condyle. 
Hyomandibular  foramen  well  separated  from 
condyle  (Fig.  2A).  Prootic  also  articulating  dor- 


epo 


lot  eth 


exo 


epo 


exo 


FIGURE  2.    Neurocranium  ofPachycara  gymninium,  BCPM 
980-100,  281  mm  SL:  (A)  left  lateral  view;  (B)  dorsal  view. 


ANDERSON:  REVIEW  OF  EELPOUT  GENUS  PACHYCARA 


225 


inthy 


epihy 


plem 


hyphy 


urhy 


FIGURE  3.  Right  hyoid  bar  ofPachycara  gymninium,  BCPM 
980-100,  281  mm  SL;  lateral  view. 

sally  with  pterotic  and  intercalar,  which  form  a 
ridge  extending  onto  exoccipital.  Single  pair  of 
small,  barrel-like  lateral  extrascapulars  present. 
Foramina  passing  ninth  (glossopharyngeal)  and 
tenth  (vagus)  cranial  nerves  located  on  low,  lat- 
eral ridge  well  away  from  margins  of  exoccipital. 

Teeth  present  on  jaws  (absent  on  maxilla), 
vomer,  palatine  and  pharyngeal  bones.  Palato- 
pterygoid  (palatal)  arch  well  developed,  ecto- 
pterygoid  and  mesopterygoid  broadly  overlapping 
anterior  and  dorsal  margins  of  quadrate  (Fig.  1). 
Hyomandibular  posterior  ramus  not  elongated. 
Opercle  (posteriorly)  and  subopercle  poorly  os- 
sified. Metapterygoid  relatively  large  and  thick. 

Most  or  all  of  ceratohyal-epihyal  juncture 
smooth,  space  between  bones  cartilage-filled; 
bone  weakly  interdigitating  dorsally  in  large 
specimens  of  one  observed  species  (Fig.  3).  Hy- 
pohyals  separated  by  wide,  cartilage-filled  space. 
Interhyal  ventral  surface  with  concave  fossa;  dor- 
sal tip  not  greatly  sculptured.  Six  branchiostegal 
rays  present  (except  in  P.  rimae),  with  four  ar- 
ticulating with  ceratohyal  and  two  with  epihyal. 
Urohyal  large  and  broad,  with  well  developed 
fossae  dorsally  for  attachment  of  sternohyoideus 
muscle. 

Ceratobranchial  five  (lower  pharyngeals)  den- 
tigerous,  unfused.  Three  pairs  of  ossified  infra- 
pharyngobranchials  (upper  pharyngeals),  asso- 
ciated with  gill  arches  2-4.  Three  ossified 
basibranchials;  fourth  an  irregular  cartilaginous 
pad.  Uncinate  processes  on  epibranchials  three 
and  four  only.  Gill  rakers  present  on  epibran- 
chials, ceratobranchials,  and  usually  one  or  two 
on  hypobranchials,  numbering  10-21  on  first 
arch. 

Posttemporal  ventral  ramus  weak  or  absent 
(Fig.  4).  Supracleithrum  with  thin,  poorly  ossi- 
fied posterior  lamina;  median  ridge  well  devel- 
oped. Scapular  foramen  enclosed  by  bone;  scap- 


FIGURE  4.  Left  pectoral  girdle  ofPachycara  bulbiceps,  CAS 
6 1 220,  445  mm  SL;  lateral  view.  Pelvic  bones  greatly  reduced 
in  size  in  this  species. 

ula  with  well  developed  posterior  strut.  Four 
actinosts  present  bearing  all  pectoral  rays  but 
dorsalmost  2-3  (borne  by  scapular  strut).  Pelvic 
fins  absent  (P.  bulbiceps),  present  (10  species),  or 
variably  present  and  absent  (P.  mesoporum  and 
P.  sulaki),  with  three  (rarely  two)  soft  rays.  Sin- 
gle, needlelike  postcleithrum  present.  Cleithrum 
with  deep  dorsal  and  ventral  grooves  for  muscle 
attachments. 

Vertebrae  symmetrical.  Epipleural  ribs  on  ver- 
tebrae 1-21  or  22.  Pleural  ribs  on  third  to  ulti- 
mate precaudal  vertebrae.  Dorsal  fin  origin  as- 
sociated with  vertebrae  2-8,  with  no  free  dorsal 
pterygiophores.  First,  or  first  and  second,  dorsal 
element  a  "flexible  spine"  (unsegmented  and  un- 
branched,  but  bilaterally  divided  at  base);  oc- 
casionally branched  at  tip,  but  this  may  be  a 
preservation  artifact.  Last  dorsal  ray  associated 
with  second  through  fifth  (usually  fourth)  preural 
vertebrae.  Anal  fin  origin  associated  with  ante- 
penultimate to  ultimate  precaudal  vertebrae,  with 
1-9  anal  pterygiophores  (all  bearing  soft  rays) 


226 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  10 


inserted  anterior  to  haemal  spine  of  first  caudal 
vertebrae.  Last  anal  ray  associated  with  second 
(rarely  third)  preural  vertebrae.  Caudal  fin  with 
1-3  epural  and  3-6  upper  and  lower  hypural  rays, 
totaling  8-12  rays. 

COMPARATIVE  REMARKS.— Recently,  Arnulf  et 
al.  (1987)  elaborated  on  the  osteology  and  rela- 
tionships of  Thermarces  Rosenblatt  and  Cohen, 
1986,  thought  to  be  derived  from  Pachycara  by 
the  latter  authors.  Thermarces  was  diagnosed  by 
Rosenblatt  and  Cohen,  in  part,  on  the  basis  of 
four  characters  shown  by  Arnulf  et  al.  (1987)  to 
have  different  states  than  those  originally  de- 
scribed (those  of  original  description  given  first): 
(1)  one  or  two  suborbital  ossifications  vs.  six;  (2) 
pectoral  radials  unossified  vs.  ossified;  (3)  two 
pairs  of  infrapharyngobranchials  vs.  three;  and 
(4)  postcleithrum  absent  vs.  present.  The  speci- 
men of  Thermarces  cerberus  (SIO  82-46)  studied 
by  Rosenblatt  and  Cohen  was  poorly  ossified  and 
I  have  observed,  on  subsequent  restaining,  a 
postcleithrum,  six  suborbital  bones,  four  pecto- 
ral radials,  and  a  cartilaginous  third  pharyngeal 
tooth  plate.  These  structures,  although  bony, 
picked  up  very  little  or  no  alizarin-red  stain, 
however.  The  skeletons  of  many  deep-sea  zoar- 
cids  are  relatively  unossified  (Anderson  and 
Hubbs  1981)  and  endochondral  bone  may  not 
fully  ossify  in  individuals  of  some  species  living 
in  possibly  environmentally  stressful  habitats. 
Stresses  to  normal  calcium  metabolism  in  the 
hydrothermal  vent  habitat  of  Thermarces  may 
include  high  dissolved  sulphide  and  metal  ion 
levels  (J.  M.  Edmonds,  Massachusetts  Institute 
of  Technology,  pers.  comm.),  making  alizarin 
staining  difficult  (Anderson  1 984).  This  seems  to 
have  been  the  case  with  the  specimen  cleared 
and  stained  by  Rosenblatt  and  Cohen  (1986)  for 
illustration. 

On  the  basis  of  the  detailed  osteology  pre- 
sented by  Arnulf  et  al.  (1987)  and  my  studies, 
Thermarces  does  appear  to  be  derived  from,  and 
very  close  to,  Pachycara,  although  Arnulf  et  al. 
suggested  Thermarces  might  be  placed  in  a  new 
subfamily  someday.  However,  Thermarces  is 
separable  from  Pachycara  by  only  four  charac- 
ters: ( 1 )  pelvic  bone  absent;  (2)  scales  absent;  (3) 
lateral  line  absent;  (4)  flesh  gelatinous.  Although 
Arnulf  et  al.  presented  a  detailed  osteological  de- 
scription of  Thermarces,  their  discussion  on  re- 
lationships and  characters  is  flawed.  For  exam- 
ple, their  table  1  is  a  very  limited  assemblage  of 


essentially  non-diagnostic  characters  of  higher 
groups,  some  of  which  are  redundant.  Their 
subfamily  "Brotulinae"  refers  to  the  Parabro- 
tulidae  (Nielsen  1973,  1986;  Anderson  1986). 
Subfamily  Neozoarcinae  was  placed  in  Stichaei- 
dae  by  Anderson  (1984).  Melanostigma  was 
shown  not  to  possess  a  basisphenoid  bone  by 
Anderson  and  Hubbs  (1981).  The  attribution  and 
interpretation  of  many  characters  of  zoarcids  by 
Arnulf  et  al.  are  mostly  incorrect  and  a  detailed 
discussion  of  these  and  an  osteological  analysis 
will  be  presented  elsewhere. 


KEY  TO  SPECIES  OF  PACHYCARA 

1A.  Branchiostegal  rays  six;  pseudobranch 
present 2 

IB.   Branchiostegal  rays  four  or  five;  pseu- 
dobranch absent ..  Pachycara  rimae  n.  sp. 
Galapagos  Rift  Zone 

2A.  Occipital  (supratemporal)  pores  absent 
(rarely  present  in  P.  crossacanthum;  pre- 
caudal  vertebrae  26-28 3 

2B.   Single,  mesial  occipital  pore  present; 

precaudal  vertebrae  24-26 

Pachycara  mesoporum  n.  sp. 

Peru  and  Chile 

3A.  Lateral  line  with  mediolateral  and  ven- 
tral branches*;  precaudal  vertebrae  23- 
32;  dorsal  fin  origin  associated  with  ver- 
tebrae 3-8;  anal  fin  origin  associated  with 
vertebrae  23-3 1 4 

3B.   Lateral  line  of  ventral  branch  only;  pre- 
caudal vertebrae  21-23;  dorsal  fin  origin 
associated  with  vertebrae  2-3;  anal  fin 
origin  associated  with  vertebrae  20-22 
........  Pachycara  suspectum  (Garman,  1899) 

Eastern  Tropical  Pacific 

4A.  Dorsal  fin  origin  associated  with  verte- 
brae 3-7;  precaudal  vertebrae  23-30; 
head  length  1 1.8-19.9%  SL 5 

4B.  Dorsal  fin  origin  associated  with  verte- 
bra 8;  precaudal  vertebrae  32;  head 

length  1 1.4%  SL 

Pachycara  shcherbachevi  n.  sp. 

Indian  Ocean 

5A.  Gill  rakers  on  lower  limb  (ceratobran- 
chial)  of  first  arch  blunt,  triangular 6 


*  Lateral  line  condition  unknown  in  P.  shcherbachevi  n.  sp. 


ANDERSON:  REVIEW  OF  EELPOUT  GENUS  PACHYCARA 


227 


80 

60 

40 

20 

0 

20 

40 
60 


FIGURE  5.    Distribution  of  selected  Pachycara  species.  Circles:  P.  bulbiceps;  squares:  P.  crassiceps;  hexagons:  P.  crossacanthum; 
triangles:  P.  sulaki;  star:  P.  shcherbachevi. 


5B.   Gill  rakers  on  lower  limb  of  first  arch 

bi-  or  tricuspid 

Pachycara  crossacanthum  n.  sp. 

Eastern  Tropical  Atlantic 
6A.  Origin  of  ventral  branch  of  lateral  line 

just  posterior  to  last  postorbital  pore  ..    7 
6B.   Origin  of  ventral  branch  of  lateral  line 
on  or  behind  vertical  through  posterior 

third  of  pectoral  fin 

Pachycara  sulaki  n.  sp. 

Western  Tropical  Atlantic 

7A.  Pelvic  fins  present 8 

7B.   Pelvic  fins  absent 

Pachycara  bulbiceps  (Garman,  1899) 

NE  Pacific  and  N  Atlantic 
8A.  Suborbital  pores  along  ventral  ramus 

(beneath  eye)  six 9 

8B.   Suborbital  pores  beneath  eye  five 

Pachycara  brachycephalum 

(Pappenheim,  1912) 
Antarctica 

9A.  Origin  of  mediolateral  branch  of  lateral 
line  on  or  behind  vertical  through  pec- 
toral axil 1 0 

9B.  Origin  of  mediolateral  branch  of  lateral 
line  on  nape,  just  posterodorsal  to  last 

postorbital  pore 

_ Pachycara  pammelas  n.  sp. 

Chile 


1 OA.  Precaudal  vertebrae  27-3 1  origin  of  me- 
diolateral branch  of  lateral  line  anterior 
to  vertical  through  posterior  third  of 

pectoral  fin 1 1 

10B.  Precaudal  vertebrae  23-26;  origin  of 
mediolateral  lateral  line  at  or  posterior 
to  vertical  through  posterior  margin  of 

pectoral  fin Pachycara  lepinium 

Anderson  and  Peden,  1988 
Northeastern  Pacific 

11  A.  Total  vertebrae  102-109;  caudal  verte- 
brae 73-80;  D  96-103;  A  77-84;  pec- 
toral fin  length  63.0-70.8%  HL 

Pachycara  gymninium 

Anderson  and  Peden,  1988 
Northeastern  Pacific 

1  IB.  Total  vertebrae  1 1 1-118;  caudal  verte- 
brae 84-88;  D  105-1 12;  A  86-92;  pec- 
toral fin  length  81.6-97.0%  HL 

Pachycara  crassiceps  (Roule,  1916) 

Eastern  Atlantic 

Pachycara  bulbiceps  (Garman,  1899) 
(Figs.  1,  4,  5) 

Maynea  bulbiceps  Garman,  1899: 140- 141,  pi.  E,  fig.  1  (original 
description.  Type  locality:  Gulf  of  Panama).  McAllister  and 
Rees,  1964:106,  107.  Pearcy  et  al.,  1982:387,  399,  400. 

Pachycara  obesa  Zugmayer,  191  la:12;  191  lb:134-136,  pi.  VI, 
fig.  6  (type  locality:  Bay  of  Biscay).  Markle  and  Sedberry, 


228 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  10 


1978:22-25,  fig.  1.  Merrett  and  Marshall,  1981:240.  An- 

driashev,  1 973:547.  Haedrich  and  Merrett,  1988:1335, 1338 

(as  obesum). 
-Paehychara  obesa  Zugmayer  (lapsus  calami).  Fowler,  1936: 

1056,  fig.  437.  Andriashev,  1986:1 149,  text  fig. 
Pachycara  bulbiceps  (Garman).  Anderson  and  Peden,  1988: 

84-88,  figs.  1-3. 

MATERIAL  EXAMINED.— See  Anderson  and  Peden  (1988:84- 
85). 

DIAGNOSIS.— Pelvic  fins  absent;  mediolateral 
lateral  line  originating  posterior  to  pectoral  fin 
margin;  scales  absent  on  nape;  vertebrae  25-31 
+  82-91  =  112-119;  D  104-114;  A  86-97;  P, 
16-19;  dorsal  fin  origin  associated  with  vertebrae 
4-7;  gill  rakers  0-4  +  11-15=1 1-19;  branchios- 
tegal  rays  six;  suborbital  pores  6-7;  postorbital 
pores  1-3;  head  length  1 1.8-15.0%  SL;  pectoral 
fin  length  8.7-1 2.9%  SL. 

DISTRIBUTION.— Both  sides  of  the  North  At- 
lantic at  depths  of  2,400-4,780  m  and  in  the 
northeastern  Pacific  from  the  Queen  Charlotte 
Islands,  Canada,  to  the  Gulf  of  Panama  at  depths 
of  2,60 1-4,000  m  (Fig.  5). 

Pachycara  crassiceps  (Roule,  1916) 

(Figs.  5,  6) 

Lycenchelys  crassiceps  Roule,  1916:16  (original  description. 
Type  locality:  Bay  of  Biscay;  based  on  two  syntypes  appar- 
ently lost). 

Lycodes  macrops  (non  Gunther  1880).  Vaillant,  1888:306- 
309. 

Lycodes  mucosus  (non  Richardson  1855).  Vaillant,  1888:31 1- 
312. 

Lycenchelys  crassiceps.  Roule,  1919:64-66,  pi.  V,  fig.  1.  Mer- 
rett and  Marshall,  1981:240. 

Lycodes  sp.  Golovan',  1974:289. 

Lycodes  atlanticus  (non  Jensen  1 902).  Golovan',  1978:226. 

Lycodes  crassiceps.  Andriashev,  1986:1 137,  text  fig. 

MATERIAL  EXAMINED. -BMNH  1 98 1 .6.23:3  (448  mm);  Por- 
cupine Sea  Bight,  SW  of  Ireland  (51°14.7'N,  13°16.3'W); 
CHALLENGER  sta.  50509;  trawl,  1,490-1,523  m;  3  June  1979. 
ZIN  45347  (415  mm);  off  Cabo  Barbas,  Western  (Spanish) 
Sahara  (22°20'N,  1 7°30'W);  ZVEZDA  KRYMA  sta.  260;  trawl, 
1,540-1,620  m;  16  Feb.  1973.  BMNH  1981.6.23:1,  2  (432- 
518  mm);  off  Cap  Blanc,  Mauritania  (20°09.1'N,  18°08.8'W); 
DISCOVERY  sta.  9 1 33-7;  trawl,  2, 1 30-2, 1 9 1  m;  26  Nov.  1 976. 
CAS  55587  (337  mm);  off  Senegal  (13°12.5'N,  17°43.4'W);  L. 
AMARO;  trap,  1,300  m;  10  Feb.  1983.  ZIN  45348  (509  mm); 
off  Cape  Fria,  Namibia  (18°26'S,  14°06'E);  EVRIKA;  trawl, 
1,100-1,125  m;  22  Dec.  1975.  HPB  168/1985  (513  mm);  off 
Namibia  (23°05'S,  12°41'E);  BENGUELA  VII  sta.  2;  trawl, 
652  m;  9  Aug.  1984.  PEM  13784  (395  mm);  off  Cape  of  Good 
Hope,  South  Africa  (33°40.0'S,  1 7°25.8'E);  AFRICANA  cruise 
060,  sta.  A6999-01-04B;  trawl,  923  m;  5  Mar.  1988.  SAM 
uncat.  (396  mm)  off  Cape  of  Good  Hope  (34°54.9'S,  1 8°1 2. 1  'E); 
AFRICANA  cruise  060,  sta.  A7038-16-04B;  trawl,  917  m;  14 
Mar.  1988.  SAM  31601  (189  mm);  off  Cape  of  Good  Hope 
(33°37.5'S,  17°24.5'E);  AFRICANA  cruise  060,  sta.  A7002- 


06-03B;  trawl,  668  m;  8  Mar.  1988.  RUSI  28210  (380  mm); 
off  Cape  of  Good  Hope  (34°25.6'S,  17°41.2'E);  AFRICANA 
sta.  A4310;  trawl,  760  m;  7  July  1986. 

COUNTS  AND  MEASUREMENTS.— Vertebrae  27- 
30  +  84-90  =  111-118;  D  105-112;  A  86-94; 
C  10-12;  P,  17-19;  P2  3;  vomerine  teeth  6-8; 
palatine  teeth  8-17;  gill  rakers  3  +  12-13;  bran- 
chiostegal  rays  six;  pseudobranch  filaments  4-5. 
Following  measurements  in  percent  SL:  head 
length  12.6-14.5;  head  width  8.0-10.3;  pectoral 
fin  length  10.7-13.3;  predorsal  length  17.3-21.6; 
preanal  length  38. 1-43.7;  body  height  10.7-13.6; 
gill  slit  length  5.5-7.4;  caudal  fin  length  2.3-4.6. 
Following  measurements  in  percent  HL:  head 
width  59.4-70.7;  upper  jaw  length  38.7-45.4; 
pectoral  fin  length  81.6-97.0;  snout  length  17.4- 
23.4;  eye  diameter  17.3-20.8;  gill  slit  length  40.4- 
50.7;  interorbital  width  8.5-11.4;  interpupillary 
width  25.4-33.3;  pelvic  fin  length  9.9-18.6  Pec- 
toral base/length  ratio  32.8-44.3. 

DIAGNOSIS.  — Pelvic  fins  present;  mediolateral 
lateral  line  originating  above  pectoral  fin  just 
posterior  to  pectoral  base;  scales  present  on  nape; 
vertebrae  27-30  +  84-90  =111-118;  gill  rakers 
15-21;  pectoral  fin  length  81.6-97.0%  HL. 

DESCRIPTION.— Head  ovoid,  dorsal  profile  of 
snout  more  steeply  sloping  in  large  adults  than 
juveniles.  Scales  extending  anteriorly  onto  nape 
to  about  two  eye  diameters  behind  posterior 
margin  of  eye  in  large  specimens;  scales  present 
on  abdomen  to  isthmus,  pectoral  base  and  axil, 
and  extending  onto  pectoral  fin  up  to  about  half 
its  length;  scales  present  on  unpaired  fins  almost 
to  their  margin.  Eye  circular,  entering  dorsal  pro- 
file of  head  in  large  specimens.  Gill  slit  extending 
to  ventral  margin  of  pectoral  base.  Lobe  at  dorsal 
margin  of  gill  slit  not  formed  in  some  large  spec- 
imens; weak  and  rounded  in  juveniles.  Pectoral 
fin  origin  at  or  slightly  below  body  midline,  in- 
sertion on  abdomen;  posterior  margin  of  fin 
wedge-shaped  (cf.  Lycodes),  pectoral  rays  4-8 
longest;  ventralmost  three  or  four  rays  thickened, 
tips  slightly  exserted. 

Mouth  subterminal,  slightly  oblique,  upper  jaw 
extending  posteriorly  to  middle  of  eye.  No  epi- 
dermal prickles  evident  on  snout  or  lips.  Oral 
valve  reaching  anterior  edge  of  vomer  and  co- 
alesced with  lateral  margin  of  plate  opposite 
vomer.  Jaw  teeth  small,  conical;  dentary  with  4- 
5  irregular  rows  anteriorly,  blending  into  single 
posterior  row;  premaxilla  with  3-4  anterior  rows, 
blending  into  single  posterior  row.  Vomerine  teeth 
in  irregular  patch,  longer  and  sharper  in  juveniles 


ANDERSON:  REVIEW  OF  EELPOUT  GENUS  PACHYCARA 


229 


FIGURE  6. 
pelvic  fins. 


Pachycara  crassiceps  (Roule),  BMNH  1 98 1 .6.23: 1-2,  518m  SL,  from  off  Cap  Blanc,  Mauritania.  Arrow  indicates 


than  adults.  Palatine  teeth  not  retrorse,  in  double 
row  at  least  anteriorly,  in  all. 

Cephalic  lateralis  system  with  two  or  three 
postorbital  pores  (pores  one  and  four,  arising  from 
frontal  and  lateral  extrascapular,  or  pores  one, 
three,  and  four).  Two  pairs  of  anterior  supraor- 
bital  (nasal)  pores,  one  set  mesial  to  nasal  tube, 
the  other  posteriorly.  Six  or  seven  suborbital 
pores,  six  arising  from  ventral  ramus  of  bone 
chain  and  one  (absent  in  three  specimens;  absent 
on  one  side  only  in  one  specimen)  from  ascend- 
ing ramus  behind  eye  just  ventral  to  first  post- 
orbital  pore.  Eight  preoperculomandibular  pores, 
four  arising  from  dentary,  one  from  anguloartic- 
ular,  and  three  from  preopercle.  Interorbital  and 
occipital  pores  absent.  Body  lateral  line  system 
with  mediolateral  branch  originating  above  pec- 


toral fin  about  one  eye  diameter  posterior  to  pec- 
toral base;  ventral  branch  originating  just  pos- 
terior to  fourth  postorbital  pore;  both  branches 
complete  to  tail  tip.  Free  superficial  neuromasts 
(about  8-10)  forming  dorsal  "lateral  line"  orig- 
inating above  pectoral  base  on  horizontal  through 
middle  of  eye  and  extending  posteriorly  to  ver- 
tical through  posterior  margin  of  pectoral  fin  or 
slightly  anteriorly. 

Dorsal  fin  origin  associated  with  vertebrae  five 
or  six.  Anal  fin  origin  associated  with  ultimate 
precaudal  vertebrae  (numbers  25-30),  with  3-5 
anal  pterygiophores  inserted  anterior  to  haemal 
spine  of  first  caudal  vertebra.  Last  dorsal  ray 
associated  with  fourth  preural  vertebra,  last  anal 
ray  associated  with  second  preural  vertebra.  Cau- 
dal fin  with  two  epural,  4-5  upper  and  4-5  lower 


230 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  10 


FIGURE 


•^^^^^^^•^•••^^^••^^ 
7.    Pachycara  crossacanthum  n.  sp.,  CAS  62408,  paratype,  227  mm  SL,  from  off  Senegal. 


hypural  rays.  Gill  rakers,  small,  dorsalmost  on 
lower  limb  (ceratobranchial)  triangular,  sharply 
pointed,  ventralmost  blunt;  ventralmost  four 
rakers  in  ZIN  45347  with  tricuspid  tips.  Bran- 
chiostegal  rays  six.  Pseudobranch  filaments  long, 
simple. 

Color  uniformly  black  (adults)  or  dark  choc- 
olate brown  (juveniles),  in  recently  preserved 
material.  Eye  blue.  Abdomen  with  bluish  tinge 
in  juveniles,  black  in  adults.  Lining  of  orobran- 
chial  cavity  black,  lateral  regions  of  palate  pale. 

DISTRIBUTION.— Known  in  eastern  Atlantic 
from  SW  of  Ireland  to  off  South  Africa  on  upper 
and  lower  continental  slopes  at  depths  of  652- 
2,191  m(Fig.  5). 

REMARKS.— This  species  was  originally  de- 
scribed from  two  syntypes,  neither  of  which  were 
found.  However,  a  neotype  is  not  selected  here 
as  the  syntypes  have  not  been  sufficiently  searched 
for,  and  diagnostic  features  (counts  of  the  axial 
skeleton  from  x-radiography)  published  in  Roule 
(1919)  permit  the  distinction  of  P.  crassiceps  from 
its  close  congener,  P.  crossacanthum  (below). 
Pectoral  fin  ray  counts  in  Roule  (1919)  are,  how- 
ever, undoubtedly  erroneous. 

Pachycara  crossacanthum  n.  sp. 

(Figs.  5,  7,  8) 

HOLOTYPE.— CAS  55586  (female,  370  mm);  off  Senegal 
(12°58.0'N,  17°41.8'W);  L.  AMARO  sta.  Casamance;  fish  trap, 
900  m;  B.  Seret,  8  Feb.  1983. 

PARATYPES.-MNHN  1988-1170  (3;  227-378  mm);  col- 
lected with  holotype.  CAS  62408  (female,  227  mm);  offSenegal 
(15'47.8'N,  17°07.1'W);  L.  AMARO,  sta.  6;  fish  trap,  900  m; 
D.  Gaertner,  29  Mar.  1984.  MNHN  1988-1169  (female,  353 
mm);  offPointe  Banda,  Gabon  (04°00'S,  10°08'E);  NIZERY 
sta.  14-76;  fish  trap,  1,050  m;  A.  Crosnier,  20  May  1976. 

COUNTS  AND  MEASUREMENTS.— Vertebrae  26- 
28  +  75-79  =  101-107;  D  97-103;  A  77-83;  C 
10-12  P,  17-18;  P2  3;  vomerine  teeth  5-14;  pal- 
atine teeth  6-16;  gill  rakers  3  +  12-13;  branchi- 
ostegal  rays  six;  pseudobranch  filaments  4-5. 
Following  measurements  in  percent  SL:  head 
length  14.8-16.9;  head  width  8.6-9.5;  pectoral 


fin  length  9.2-10.5;  predorsal  length  17.6-19.7; 
preanal  length  41.3-44.2;  body  height  9.6-1 1.8; 
gill  slit  length  5.8-6.7;  caudal  fin  length  3.4-5.0. 
Following  measurements  in  percent  HL:  head 
width  52.7-64.5;  upper  jaw  length  35.5-45.6; 
pectoral  fin  length  54.6-66.0;  snout  length  19.1- 
23.2;  eye  diameter  1 5.2-20.2;  gill  slit  length  34.9- 
41.8;  interorbital  width  6.9-10.3;  interpupillary 
width  21.1-25.3;  pelvic  fin  length  12.2-18.0. 
Pectoral  base/length  ratio  38.9-50.9. 

DIAGNOSIS.— Pelvic  fins  present;  dorsal  gill 
rakers  on  first  arch  bi-  or  tricuspid  at  tips  (adults 
and  subadults);  scales  absent  on  nape  and  ab- 
domen; postorbital  pores  four;  vertebrae  26-28 
+  75-79  =  101-107;  origin  of  mediolateral  lat- 
eral line  near  dorsal  margin  of  pectoral  base;  oc- 
cipital pores  present  or  absent. 

DESCRIPTION.— Head  ovoid,  snout  not  bluntly 
rounded  (as  some  species),  but  steep  anteriorly. 
Scales  extending  anteriorly  on  body  in  wedge- 
shaped  pattern  to  vertical  from  half  pectoral  fin 


mm 


FIGURE  8.  Outer  surface  of  right  first  gill  arch  of  holotype 
of  Pachycara  crossacanthum  n.  sp.,  showing  furcated  gill  rak- 
ers; gill  lamellae  not  shown. 


ANDERSON:  REVIEW  OF  EELPOUT  GENUS  PACHYCARA 


231 


FIGURE  9.    Pachycara  sulaki  n.  sp.,  USNM  29281 1,  holotype,  189  mm  SL,  from  off  Puerto  Rico. 


length  posterior  to  pectoral  margin  (usually  in 
smallest  specimens)  to  vertical  through  middle 
of  pectoral;  scales  absent  on  abdomen,  nape,  pec- 
toral axil  and  base.  Eye  ovoid,  entering  dorsal 
profile  of  head.  Gill  slit  short,  not  reaching  ven- 
tral margin  of  pectoral  base.  Lobe  at  dorsal  mar- 
gin of  gill  slit  rounded  and  weakly  developed  in 
three  specimens  (including  holotype),  squared- 
off  and  well  denned  in  others.  Pectoral  fin  origin 
well  below  body  midline,  insertion  on  abdomen; 
posterior  margin  of  fin  evenly  rounded,  middle 
rays  longest;  ventralmost  four  or  five  rays  thick- 
ened, tips  slightly  exserted. 

Mouth  subterminal,  oblique,  upper  jaw  ex- 
tending posteriorly  to  middle  of  eye,  or  slightly 
in  advance.  No  epidermal  prickles  evident  on 
snout  or  lips.  Oral  valve  weak,  not  reaching  an- 
terior edge  of  vomer,  and  not  coalesced  with  lat- 
eral margin  of  palate  opposite  vomer.  Jaw  teeth 
in  large  specimens  long,  sharp,  retrorse  ante- 
riorly; dentary  with  4-5  irregular  rows  anteriorly, 
blending  into  single  posterior  row;  premaxilla 
with  two  (males)  or  three  (females)  anterior  rows, 
blending  into  single  posterior  row.  Vomerine  teeth 
in  irregular  patch;  palatine  teeth  not  retrorse,  in 
double  row  in  large  females,  in  single  row  in  all 
others. 

Cephalic  lateralis  system  with  four  postorbital 
pores  arising  from  frontal  (pore  one),  pterotic 
(pore  two),  between  pterotic  and  lateral  extra- 
scapular  (pore  three)  and  just  posterior  to  lateral 
extrascapular  (pore  four).  Two  pairs  of  anterior 
supraorbital  (nasal)  pores,  one  set  mesial  to  nasal 
tube,  the  other  posteriorly.  Seven  suborbital 
pores,  six  arising  from  ventral  ramus  of  bone 
chain  and  one  from  ascending  ramus  behind  eye 
just  ventral  to  first  postorbital  pore.  Eight  pre- 
operculomandibular  pores,  four  arising  from 
dentary,  one  from  anguloarticular  and  three  from 
preopercle.  Interorbital  and  occipital  pores  ab- 
sent, except  in  CAS  62408  which  has  three  mi- 
nute occipital  pores  and  MNHN  1988-11 70,  227 
mm  SL,  which  has  a  single,  right  lateral  occipital 
pore.  Body  lateral  line  with  mediolateral  branch 


originating  in  pectoral  axil  near  dorsal  margin  of 
pectoral  base;  ventral  branch  originating  just 
posterior  to  fourth  postorbital  pore;  both  branch- 
es complete  to  tail  tip. 

Dorsal  fin  origin  associated  with  vertebra  four. 
Anal  fin  origin  associated  with  ultimate  precau- 
dal  vertebra  (numbers  26-28),  with  three  anal 
pterygiophores  inserted  anterior  to  haemal  spine 
of  first  caudal  vertebra.  Last  dorsal  ray  associated 
with  fourth  preural  vertebra,  last  anal  ray  asso- 
ciated with  second  preural  vertebra.  Caudal  fin 
with  1-2  epural,  4-5  upper  and  4-6  lower  hypural 
rays. 

Gill  rakers  on  upper  limb  (epibranchial)  of  first 
arch  long,  sharply  pointed  in  young,  bi-  or  tri- 
furcate  in  large  specimens;  rakers  on  lower  limb 
(ceratobranchial)  tubular,  those  of  large  speci- 
mens with  bi-or  tricuspid  tips  (Fig.  8);  lowermost 
1-4  rakers  of  some  specimens  simple,  pointed. 
Branchiostegal  rays  six.  Pseudobranch  filaments 
4-5,  tips  furcate  in  large  specimens. 

Color  uniformly  dark  brown,  margins  of  dor- 
sal, anal  and  pectoral  fins  black;  eye  and  abdo- 
men bluish.  Peritoneum  and  inner  palate  black. 
Lips  and  lining  of  lateral  sides  of  palate  white. 

DISTRIBUTION.— Known  off  western  tropical 
Africa  from  Senegal  to  Gabon  on  the  upper  con- 
tinental slope  at  depths  of  900-1,050  m  (Fig.  5). 

ETYMOLOGY. —From  the  Greek  updaaos  (fringe) 
and  'ciKCivQa  (thorn)  in  reference  to  the  species' 
furcate  gill  rakers. 

Pachycara  sulaki  n.  sp. 

(Figs.  5,  9) 

Zoarcid  fish.  Paull  et  al.,  1984:965,  966. 
Pachycara  sp.  Anderson  et  al.,  1986:800. 

HOLOTYPE.— USNM  2928 1 1  (male,  1 89  mm);  Mona  Passage 
off  Puerto  Rico  (1 8°28.7'N,  67°20.6'W);  PEZMAR,  Deep  Trap- 
ping Project,  sta.  36;  medium  fish  trap,  2,000  m;  D.  A.  Hensley, 
12-13  Feb.  1986. 

PARATYPES.— USNM  233627  (female,  161  mm);  Venezuela 
Basin,  Caribbean  Sea  (13°31.3-25.7'N,  64°45.1-40.4'W);  USNS 
BARTLETT  sta.  E092;  trawl,  3,510-3,469  m;  M.  D.  Richard- 
son and  party,  28  Nov.  1981.  SIO  87-33  (female,  133  mm); 


232 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  10 


Florida  Escarpment,  Gulf  of  Mexico  (26°05'N,  84°54'W);  DSV 
ALVIN  dive  1771;  Floe  Sucker,  3,303  m;  C.  Paull  and  party, 
7  Nov.  1986. 

COUNTS  AND  MEASUREMENTS.— Vertebrae  27- 
30  +  80-85  =  107-1 13;  D  99-107;  A  81-89;  C 
9-1 1;  P,  16-17;  P2  0  or  3;  vomerine  teeth  4-8; 
palatine  teeth  4-7;  gill  rakers  2-3  +  9-13  =  11- 
16;  branchiostegal  rays  six;  pseudobranch  fila- 
ments 3-4.  Following  measurements  in  percent 
SL:  head  length  14.8-17.7;  head  width  7.3-9.0; 
pectoral  fin  length  9.8-10.6;  predorsal  length 
20.8-23.5;  preanal  length  39.0-41 .7;  body  height 
8.3-9.5;  gill  slit  length  4.8-6.5;  caudal  fin  length 
2. 1-3.8.  Following  measurements  in  percent  HL: 
head  width  47.8-56.1;  upper  jaw  length  36.4- 
50.0;  pectoral  fin  length  55.5-68.8;  snout  length 
1 6.4-20. 1 ;  eye  diameter  1 7.8-23.9;  gill  slit  length 
28.0-42.1;  interorbital  width  6.4-10.5;  interpu- 
pillary  width  23.8-32.8;  pelvic  fin  length  5.7- 
13.9  (in  two).  Pectoral  base/length  ratio  38.9- 
40.0. 

DIAGNOSIS.— Pelvic  fins  present  or  absent;  or- 
igin of  mediolateral  lateral  line  just  posterior  to 
last  (fourth)  postorbital  pore;  origin  of  ventral 
lateral  line  on  or  behind  vertical  through  pos- 
terior third  of  pectoral  fin;  suborbital  pores  5-6; 
dorsal  fin  origin  associated  with  vertebrae  6-7; 
vertebrae  27-30  +  80-85  =  107-1 1 1;  predorsal 
length  20.8-23.5%  SL. 

DESCRIPTION.— Head  ovoid,  long,  dorsal  pro- 
file evenly  tapering;  snout  bluntly  rounded.  Scales 
extending  anteriorly  on  body  in  wedge-shaped 
pattern  to  vertical  through  middle  of  pectoral  fin 
(all  three  specimens  juveniles;  adults  probably 
more  densely  scaled);  scales  absent  in  these  ju- 
veniles on  dorsal  fin  anteriorly  (present  on  pos- 
terior third),  abdomen,  head,  nape,  pectoral  base 
and  axil.  Eye  circular,  entering  dorsal  profile  of 
head.  Gill  slit  reaching  ventral  margin  of  pectoral 
base  or  just  above  it.  Lobe  at  dorsal  margin  of 
gill  slit  weak,  rounded.  Pectoral  fin  origin  just 
below  body  midline,  insertion  on  abdomen;  pos- 
terior margin  of  fin  evenly  rounded,  middle  rays 
longest;  ventralmost  five  rays  thickened,  tips 
slightly  exserted.  Pelvic  fins  present  except  in 
SIO  87-33;  of  three  soft  rays  each  in  other  two 
specimens. 

Mouth  subterminal,  oblique,  upper  jaw  ex- 
tending posteriorly  to  middle  of  eye  in  females 
(paratypes),  or  its  rear  margin  in  male  holotype. 
No  epidermal  prickles  evident  on  snout  or  lips. 
Oral  valve  well  developed  in  these  small  speci- 


mens, overlapping  anterior  edge  of  vomer  and 
not  coalesced  with  lateral  margins  of  plate  op- 
posite vomer.  Jaw  teeth  relatively  long  and  sharp 
in  male,  smaller  and  conical  in  females;  outer 
jaw  teeth  retrorse;  both  jaws  (all  specimens)  with 
two  irregular  rows  of  teeth  blending  into  single, 
posterior  row.  Vomerine  teeth  in  irregular  patch, 
enlarged  in  male;  palatine  teeth  not  retrorse,  in 
single  row. 

Cephalic  lateralis  system  with  two  postorbital 
pores  arising  from  frontal  (pore  one)  and  lateral 
extrascapulars  (pore  four).  Two  pairs  of  anterior 
supraorbital  (nasal)  pores,  one  set  mesial  to  nasal 
tube,  the  other  posteriorly.  Five  (USNM  233627) 
or  six  suborbital  pores,  all  arising  from  ventral 
ramus  of  bone  chain.  Eight  preoperculomandib- 
ular  pores,  four  arising  from  dentary,  one  from 
anguloarticular  and  three  from  preopercle.  In- 
terorbital and  occipital  pores  absent.  Superficial 
neuromasts  prominent  on  head  an  anterior  por- 
tion of  body,  with  series  on  preopercle,  behind 
eye  running  along  occiput  in  converging  row, 
around  rictus  of  mouth,  and  snout  between  nos- 
trils. Body  lateral  line  with  mediolateral  branch 
originating  just  behind  posteriormost  (number 
four)  postorbital  pore;  this  branch  undulating 
across  body  somewhat,  then  straightening  above 
anus;  ventral  branch  originating  on  or  slightly 
behind  (holotype)  posterior  third  of  pectoral  fin; 
both  branches  complete  to  tail  tip. 

Dorsal  fin  origin  associated  with  vertebrae  6- 
7.  Anal  fin  origin  associated  with  penultimate  or 
ultimate  precaudal  vertebrae  (numbers  26-30), 
with  3-5  anal  pterygiophores  inserted  anterior  to 
haemal  spine  of  first  caudal  vertebra.  Last  dorsal 
ray  associated  with  fourth  preural  vertebra,  last 
anal  ray  associated  with  second  preural  vertebra. 
Caudal  fin  with  two  epural,  3-5  upper  hypural 
and  four  lower  hypural  rays. 

Gill  rakers  on  upper  limb  (epibranchial)  and 
dorsalmost  rakers  on  lower  limb  (ceratobran- 
chial)  narrow  and  sharply  pointed;  ventralmost 
ceratobranchial  rakers  blunt,  triangular.  Bran- 
chiostegal rays  six.  Pseudobranch  filaments  long, 
numbering  3-4. 

Live  color  notes  or  photographs  of  all  three 
specimens  available,  each  one  quite  different. 
Holotype  uniformly  dark  chocolate  brown,  as 
most  other  congeners,  unpaired  fins  black,  eye 
blue.  Paratype  USNM  233627  from  the  Carib- 
bean Sea  with  lustrous  dark  blue  body,  typical 
of  many  deep-sea  fishes  such  as  halosaurs  and 


ANDERSON:  REVIEW  OF  EELPOUT  GENUS  PACHYCARA 


233 


alepocephalids  (Markle  1980,  fig.  1),  head  and 
pectoral  fins  black.  Paratype  SIO  87-33  relatively 
blanched,  body  uniformly  sandy  gray,  occiput, 
nape  and  pectoral  axil  occupied  by  a  wide,  white 
band,  eye  and  abdomen  blue,  unpaired  fins  trans- 
lucent. 

DISTRIBUTION.— Known  only  from  three  spec- 
imens taken  in  abyssal  waters  of  the  tropical 
northwestern  Atlantic  at  depths  of  2,000-3,510 
m  (Fig.  5). 

ETYMOLOGY.— Named  after  Kenneth  J.  Sulak, 
intrepid  explorer  of  the  great  murky  depths  of 
the  Atlantic  Ocean,  to  honor  his  many  contri- 
butions to  knowledge  of  the  deep-sea  fishes  of 
this  and  adjacent  regions. 

Pachycara  shcherbachevi  n.  sp. 

(Figs.  5,  10) 

HOLOTYPE.-ZMUC  P-761147  (female,  236  mm  SL);  Bay 
of  Bengal  (15°54.0'N,  90°17.0'E);  GALATHEA  sta.  314;  trawl 
(HOT/4800),  2,600  m;  1930  hr;  3  May  1951. 

COUNTS  AND  MEASUREMENTS.— Vertebrae  32 
+  90  =  122;  D  112;  A  94;  C  12;  P,  17;  P2  3; 
vomerine  teeth  two;  palatine  teeth  4/4;  gill  rakers 
2+14;  branchiostegal  rays  six;  pseudobranch 
filaments  three.  Following  measurements  in  per- 
cent SL:  head  length  1 1.4;  head  width  7.7;  pec- 
toral fin  length  1 0.0;  predorsal  length  16.6;  preanal 
length  not  measurable;  body  height  7.7;  caudal 
fin  length  3.0.  Following  measurements  in  per- 
cent HL:  head  width  67.7;  upper  jaw  length  39.0; 
pectoral  fin  length  87.7;  snout  length  16.0;  eye 
diameter  23.8;  interorbital  width  13.8;  interpu- 
pillary  width  28.6;  pelvic  fin  length  8.9.  Pectoral 
base/length  ratio  36.0. 

DIAGNOSIS.— Pelvic  fins  present;  vertebrae  32 
+  90  =  122;  dorsal  fin  origin  associated  with 
vertebra  eight;  head  length  1 1 .4%  SL. 

DESCRIPTION.  — Specimen  trawl  damaged— 
completely  skinned,  body  torn  before  anus;  scales, 
lateral  line,  and  most  head  pores  not  evident. 
Head  ovoid,  small;  its  length  shorter  in  relation 
to  SL  than  any  Pachycara  specimen,  irregardless 
of  size.  Eye  large,  ovoid,  entering  dorsal  profile 
of  head  when  viewed  laterally.  Gill  opening  dam- 
aged, tissue  covering  branchiostegal  membranes 
torn,  but  adhering  on  left  side  in  places,  and,  with 
skin  along  pectoral  base,  gill  opening  appearing 
to  extend  ventrally  to  just  above  lower  pectoral 
base.  Dorsal  margin  of  gill  slit  with  short,  ante- 
riorly directed  emargination.  Pectoral  fin  origin 


well  below  body  midline,  insertion  on  abdomen; 
posterior  margin  of  fin  (from  reconstruction  uti- 
lizing both  fins)  evenly  rounded,  middle  rays 
longest;  ventralmost  four  rays  thickened. 

Mouth  subterminal,  upper  jaw  (premaxilla  and 
maxilla  with  some  flesh)  posteriorly  relatively 
high,  somewhat  rectangular.  Oral  valve  reaching 
anterior  edge  of  vomer  and  coalesced  with  lateral 
margin  (left)  of  palate  opposite  vomer.  Jaw  teeth 
small,  conical,  retrorse  anteriorly;  dentary  with 
14  teeth  in  two  irregular  rows  anteriorly,  blend- 
ing into  single,  posterior  row;  premaxillae  with 
1 3  (right)  or  1 1  (left)  teeth,  also  in  two  irregular 
rows  anteriorly,  blending  into  single,  posterior 
row.  Two  small  vomerine  teeth.  Four  palatine 
teeth  on  both  bones,  in  short  row. 

Cephalic  lateralis  pores  evident  from  connec- 
tive tissue  tubules  and  impressions  in  exposed 
subdermal  lipid  layer,  not  all  pores  remaining. 
Presumably  two  pairs  of  anterior  supraorbital 
pores  mesial  to  nasal  tube.  Six  suborbital  pores, 
all  arising  from  ventral  ramus  of  bone  chain  un- 
der eye.  Eight  preoperculomandibular  pores,  four 
arising  from  dentary,  one  from  anguloarticular, 
and  three  from  preopercle.  Interorbital  and  oc- 
cipital pores  presumably  absent.  Postorbital  pores 
one  (behind  eye)  and  four  (anterodorsal  upper 
end  of  gill  slit)  observed,  others  (pore  two  and 
three)  may  or  may  not  be  present  in  the  species. 

Dorsal  fin  origin  associated  with  vertebra  eight; 
posteriormost  origin  in  genus,  except  for  P.  ri- 
mae,  n.  sp.,  which  also  has  dorsal  origin  with 
eighth  vertebra.  Anal  fin  origin  associated  with 
penultimate  precaudal  vertebra  (number  31),  with 
four  anal  pterygiophores  inserted  anterior  to  hae- 
mal spine  of  first  caudal  vertebra.  Last  dorsal  ray 
associated  with  fourth  preural  vertebra;  last  anal 
ray  associated  with  second  preural  vertebra.  Cau- 
dal fin  with  two  epural,  five  upper  hypural,  and 
five  lower  hypural  rays. 

Gill  rakers  small,  sharply  pointed  and  pyrami- 
dal dorsally;  ventralmost  five  much  smaller  than 
dorsalmost,  but  distinct.  Branchiostegal  rays  six. 
Pseudobranch  filaments  three,  very  small. 

Color  unknown,  but  bits  of  black  skin  adhere 
to  head,  body,  and  fins.  Perhaps  all  black  in  life. 
Peritoneum  black.  Lining  of  orobranchial  cham- 
ber dark  brown  in  alcohol,  probably  black  in  life 
also. 

DISTRIBUTION.— Known  from  a  single  speci- 
men from  the  abyssal  Bay  of  Bengal,  northwest 
of  the  Andaman  Islands,  in  2,600  m  (Fig.  5). 


234 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  10 


FIGURE  10. 
trawl). 


Pachycara  shcherbachevi  n.  sp.,  ZMUC  P-761 147,  holotype,  236  mm  SL,  Bay  of  Bengal  (specimen  damaged  by 


ETYMOLOGY.— Named  after  Yuri  Nikolaevich 
Shcherbachev,  P.  P.  Shirshov  Institute  of  Ocean- 
ology,  Academy  of  Sciences,  USSR,  friend  and 
colleague,  in  honor  of  his  pioneering  contribu- 
tions to  knowledge  of  the  deep-sea  fishes  of  the 
Indian  Ocean. 


Pachycara  gymninium  Anderson  and  Peden,  1988 

(Fig.  11) 

Lycodes  sp.  Hubbs  et  al.,  1979:14. 
Lycenchelys  "E."  Pearcy  et  al.,  1982:387. 
Pachycara  gymninium  Anderson  and  Peden,   1988:88-91, 
fig.  5. 


ANDERSON:  REVIEW  OF  EELPOUT  GENUS  PACHYCARA 


235 


MATERIAL  EXAMINED.  — See  Anderson  and  Peden  (1988:88- 
89). 

DIAGNOSIS.  —  Pelvic  fins  present,  their  length 
4.8-11.3%  HL;  mediolateral  lateral  line  origi- 
nating in  pectoral  axil  just  posterior  to  vertical 
through  pectoral  base;  scales  absent  on  nape,  or, 
if  present,  not  extending  anterior  to  line  con- 
necting anterodorsal  edges  of  gill  slits;  vertebrae 
28-31  +  73-80  =  102-109;  D  96-103;  A  77- 
84;  P,  14-18;  dorsal  fin  origin  associated  with 
vertebrae  3-5;  anal  fin  origin  associated  with  ver- 
tebrae 27-31;  gill  rakers  3-5  +  10-14  =  13-19; 
branchiostegal  rays  six;  suborbital  pores  6-7; 
postorbital  pores  2-3;  head  length  12.0-15.6% 
SL;  pectoral  fin  length  63.0-70.8%  HL. 

DISTRIBUTION.— Eastern  North  Pacific  off  the 
Queen  Charlotte  Islands,  Canada,  south  to  the 
Gulf  of  California  at  depths  of  1,829-3,225  m 
(Fig.  11). 

Pachycara  lepinium  Anderson  and  Peden,  1988 

(Fig  11) 

Lycodes  sp.  Hubbs  et  al.,  1979:14. 

Lycenchelys  "D."  Pearcy  et  al.,  1982:387. 

Pachycara  lepinium  Anderson  and  Peden,  1988:91-92,  fig.  6. 

MATERIAL  EXAMINED.— See  Anderson  and  Peden  (1988:91). 

DIAGNOSIS.— Pelvic  fins  present,  their  length 
11.5-17.3%  HL;  mediolateral  lateral  line  origi- 
nating posterior  to  pectoral  fin  margin;  scales 
present  on  nape;  vertebrae  23-26  +  80-94  = 
105-120;  D  99-113;  A  85-98;  P,  15-18;  dorsal 
fin  origin  associated  with  vertebrae  4-5;  anal  fin 
origin  associated  with  vertebrae  23-26;  gill  rak- 
ers 2-5  +  11-15  =  12-19;  branchiostegal  rays 
six;  suborbital  pores  six;  postorbital  pores  2-3; 
head  length  13.8-15.9%  SL;  pectoral  fin  length 
72.5-89.9%  HL. 

DISTRIBUTION.— Eastern  North  Pacific  off  the 
Queen  Charlotte  Islands,  Canada,  south  to  off 
Cabo  Colnett,  Baja  California  Norte,  Mexico,  at 
depths  of  1,728-2,907  m  (Fig.  1 1). 

Pachycara  suspectum  (Garman,  1899) 

(Fig.  11) 

Phucocoetes suspectus  Garman,  1 899: 1 37,  pi.  XXX,  figs.  3,  3a. 
Pachycara  suspectum  (Garman).  Anderson  and  Peden,  1988: 
88,  fig.  4. 

MATERIAL  EXAMINED.— See  Anderson  and  Peden  (1988:88). 

DIAGNOSIS.— Pelvic  fins  present;  lateral  line  of 
ventral  branch  only;  scales  absent  on  nape;  ver- 
tebrae 21-23  +  84-87  =  105-110;  D  100-106; 


100U 


80 


60C 


40C 


FIGURE  1 1 .  Distribution  of  eastern  Pacific  species  of  Pa- 
chycara. Closed  circles:  P.  gymninium;  squares:  P.  lepinium; 
hexagon:  P.  rimae,  triangles:  P.  mesoporum;  stars:  P.  suspec- 
tum; open  circles:  P.  pammelas. 

A  85-89;  P,  16;  dorsal  fin  origin  associated  with 
vertebrae  2-3;  anal  fin  origin  associated  with  ver- 
tebrae 20-22;  gill  rakers  1-2+14-15  =  16;  bran- 
chiostegal rays  six;  suborbital  pores  seven;  post- 
orbital  pores  three;  head  length  15.6-17.7%  SL; 
pectoral  fin  length  57.4-58.6%  HL. 

DISTRIBUTION.— Eastern  North  Pacific  in  the 
Gulf  of  California  south  to  the  Gulf  of  Panama 
at  depths  of  915-1,280  m  (Fig.  1 1). 

Pachycara  rimae  n.  sp. 

(Figs.  11,  12) 

Zoarcidae  (partim).  Cohen  and  Haedrich,  1983:376. 
Pachycara  (?).  Cohen  et  al.,  1985:229. 

HoLOTYPE.-LACM  44699-1  (male,  403  mm  SL);  vie.  Ga- 
lapagos Rift  hydrothermal  vents  (00°47.8'N,  86°09.9'W);  DSV 
ALVIN;  baited  trap,  2,500  m;  J.  Corliss,  16  Mar.  1977. 


236 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  10 


FIGURE  12.    Pachycara  rimae  n.  sp.,  LACM  44699-1,  holotype,  403  mm  SL,  Galapagos  Rift  Zone.  Illustration  of  right  side 
shown  in  left  lateral  view  (left  side  damaged). 


COUNTS  AND  MEASUREMENTS.— Vertebrae  26 
+  67  =  93;  D  86;  A  70;  C  9;  P,  1 5;  P2  2;  vomerine 
teeth  three;  palatine  teeth  8-9;  gill  rakers  3  +  7; 
branchiostegal  rays  four  or  five;  pseudobranch 
filaments  absent.  Following  measurements  in 
percent  SL:  head  length  13.7;  head  width  10.1; 
pectoral  fin  length  9.3;  predorsal  length  21.6; 
preanal  length  44.2;  body  height  14.1;  gill  slit 
length  6.0;  caudal  fin  length  1 .7.  Following  mea- 
surements in  percent  HL:  head  width  74.2;  upper 
jaw  length  54.3;  pectoral  fin  length  68.1;  snout 
length  18.9;  eye  diameter  13.0;  gill  slit  length 
43.7;  interorbital  width  12.7;  interpupillary  width 
42.1;  pelvic  fin  length  4.2.  Pectoral  base/length 
ratio  42.7. 

DIAGNOSIS.— Pelvic  fins  nublike,  of  two  soft 
rays;  vertebrae  26  +  67  =  93;  D  86;  A  70;  lateral 
line  of  mediolateral  branch  only;  dorsal  fin  origin 
associated  with  vertebra  eight;  pseudobranch  ab- 
sent; branchiostegal  rays  four  or  five;  suborbital 
pores  five;  gill  rakers  10. 

DESCRIPTION.— Head  deep,  rounded,  some- 
what shorter  than  similarly  sized  congeners.  Snout 
steeply  sloping  anteriorly.  Left  side  of  body  and 
tail  damaged,  but  skin  pliable  around  head  and 
dorsum  owing  to  subdermal  lipid  layer.  Scales 
extending  anteriorly  on  body  in  wedge-shape 
pattern  to  vertical  just  anterior  to  posterior  mar- 
gin of  pectoral  fin;  scales  absent  on  head,  nape, 
dorsum,  pectoral  base,  abdomen,  and  unpaired 
fins.  Eye  small,  circular,  entering  dorsal  profile 
of  head  when  viewed  laterally.  Gill  slit  short, 
extending  ventrally  to  opposite  pectoral  ray  1 2. 
Lobe  at  dorsal  margin  of  gill  slit  weakly  devel- 
oped, gill  slit  not  continued  anteriorly.  Pectoral 
fin  origin  well  below  body  midline;  left  pectoral 
fin  badly  damaged  in  capture,  right  fin  deformed, 
apparently  injured  in  life,  with  tips  of  dorsalmost 
four  rays  missing  (Fig.  1 2);  six  ventralmost  pec- 
toral rays  thickened,  slightly  exserted  at  tips. 

Mouth  terminal,  oblique,  upper  jaw  extending 


posteriorly  to  middle  of  eye  or  beyond  (snout 
region  twisted  to  one  side).  No  epidermal  prick- 
les evident  on  snout  or  lips.  Oral  valve  just  reach- 
ing anterior  margin  of  vomer  and  coalesced  with 
lateral  margins  of  palate  opposite  vomer.  Teeth 
small,  conical,  sharp.  Jaw  teeth  in  double  row 
anteriorly,  blending  into  single,  posterior  row. 
Vomerine  teeth  three,  in  an  arc;  palatine  teeth 
in  single  row. 

Cephalic  lateralis  system  with  two  postorbital 
pores  arising  from  frontal  (pore  one)  and  lateral 
extrascapular  (pore  four).  Two  pairs  of  anterior 
supraorbital  (nasal)  pores,  one  set  anteromesial 
to  nasal  tube,  the  other  posteromesially.  Five 
suborbital  pores,  all  arising  from  ventral  ramus 
of  suborbital  bone  chain.  Eight  preoperculoman- 
dibular  pores,  four  arising  from  dentary,  one  from 
anguloarticular,  and  three  from  preopercle.  In- 
terorbital and  occipital  (supratemporal)  pores 
absent.  Body  lateral  line  mediolateral,  originat- 
ing posterior  to  rear  margin  of  pectoral  fin,  com- 
plete to  tail  tip;  no  ventral  branch. 

Dorsal  fin  origin  associated  with  vertebra  eight. 
Anal  fin  origin  associated  with  ultimate  precau- 
dal  vertebra  (number  26),  with  two  anal  pteryg- 
iophores  inserted  anterior  to  haemal  spine  of  first 
caudal  vertebra.  Last  dorsal  ray  associated  with 
fourth  preural  vertebra,  last  anal  ray  associated 
with  second  preural  vertebra.  Caudal  fin  with 
two  epural,  four  upper  hypural,  and  four  lower 
hypural  rays. 

Gill  rakers  very  short,  triangular,  lowermost 
rakers  on  first  arch  mere  nubs;  raker  denticles 
absent.  Branchiostegal  rays  four  on  right  side  (two 
articulating  with  ceratohyal,  two  with  epihyal) 
and  five  on  left  side  (three  articulating  with  cer- 
atohyal, two  with  epihyal).  Pseudobranch  absent. 

Color  uniformly  light  brown  (Cohen  et  al.  1985: 
229);  eye  and  abdomen  bluish.  Peritoneum  and 
lining  of  orobranchial  chamber  dark  brown  in 
alcohol,  probably  black  in  life.  Head  and  right 


ANDERSON:  REVIEW  OF  EELPOUT  GENUS  PACHYCARA 


237 


FIGURE  13.    Pachycara  mesoporum  n.  sp.,  CAS  62406,  holotype,  485  mm  SL,  from  off  Antofagasta,  Chile. 


pectoral  fin  with  scattered  whitish  blotches  and 
streaks. 

DISTRIBUTION.— Known  only  from  the  holo- 
type taken  in  the  abyssal  southeastern  Pacific  in 
2,500  m. 

ETYMOLOGY.— From  the  Latin  "rima"  (fis- 
sure). A  genitive  noun  in  apposition,  alluding  to 
the  species'  capture  at  the  Galapagos  Rift  Zone 
(Cohen  and  Haedrich  1983;  Cohen  et  al.  1985). 

Pachycara  mesoporum  n.  sp. 

(Figs.  11,  13) 

HOLOTYPE.— CAS  62406  (male,  485  mm);  off  Antofagasta, 
Chile;  Chilean  fisheries  vessel;  trawl,  depth  uncertain  (ca.  300 
?  m);  I.  Kong  Urbina,  18  Apr.  1986. 

PARATYPES.-LACM  437 1 7- 1  (5;  1 70-2 1 7  mm);  offTrujillo, 
Peru  (08°26'S,  80°36.5'W);  ANTON  BRUUN  sta.  650E;  free 
vehicle  set  line  trap,  1,830  m;  L.  Knapp  and  party,  8-9  June 
1966.  ZIN  48394  (male,  189  mm);  off  Laguna  Grande,  Peru 
( 1 4°44'S,  76°1 2' W);  DMITRY  MENDELEEV  cr.  20,  sta.  1 654; 
Galathea  trawl,  1,495-1,430  m;  N.  V.  Parin,  2335-0120  hr, 
1 9-20  Mar.  1978.  MNHNC  6653  (female,  3 1 8  mm);  off  Arica, 
Chile  (18°26'S,  70°37'W);  TIBERIADES;  trawl,  760  m;  I.  Kong 
Urbina,  17  Feb.  1981. 

COUNTS  AND  MEASUREMENTS.— Vertebrae  24- 
26  +  78-87  =  103-1 13;  D  96-108;  A  80-90;  C 
9-10;  P,  15-18;  P2  absent  or  three;  vomerine 
teeth  3-12;  palatine  teeth  2-11;  gill  rakers  2-3 
+  1 1-16  =  13-18;  branchiostegal  rays  six;  pseu- 
dobranch  filaments  3-4.  Following  measure- 
ments in  percent  SL:  head  length  12.6-15.9;  head 
width  6.1-8.2;  pectoral  fin  length  8.8-12.1;  pre- 
dorsal  length  15.1-18.1;  preanal  length  33.6-38.1; 
body  height  7.8-9.5;  gill  slit  length  4.1-6.9;  cau- 
dal fin  length  2.1-3.2.  Following  measurements 
in  percent  HL:  head  width  44.6-52.5;  upper  jaw 
length  34.7-51.7;  pectoral  fin  length  57.1-81.1; 
snout  length  16.0-21.2;  eye  diameter  17.2-22.9; 
gill  slit  length  28.1-44.3;  interorbital  width  5.9- 
9.0;  interpupillary  width  24.0-29.6;  pelvic  fin 
length  8.0-8.8  (in  two  of  eight).  Pectoral  base/ 
length  ratio  32.8^6.3. 

DIAGNOSIS.— Pelvic  fins  present  or  absent;  sin- 
gle, mesial  occipital  pore  present:  postorbital 


pores  3-4;  mediolateral  lateral  line  originating 
slightly  anterior  to  vertical  through  posterior 
margin  of  pectoral  fin;  ventral  lateral  line  origi- 
nating just  posterior  to  last  postorbital  pore;  ver- 
tebrae 24-26  +  78-87  =  103-113. 

DESCRIPTION.— Following  description  based  on 
one  adult  male  (holotype),  one  adult  female 
(MNHNC  6653),  three  juvenile  males,  and  three 
juvenile  females.  Head  ovoid  in  juveniles  and 
adult  female,  more  triangular  in  adult  male.  Male 
with  dorsoventrally  depressed  head  and  straight- 
er  snout  than  others;  adult  female  and  juveniles 
with  bluntly  rounded  snouts.  Scales  extending 
anteriorly  to  vertical  through  dorsal  fin  origin, 
or  slightly  anterior  to  it;  scales  absent  on  nape 
(rare  or  absent  anterior  to  line  connecting  upper 
edge  of  gill  slit  and  dorsal  fin  origin).  Scales  pres- 
ent on  unpaired  fins  (almost  to  their  margin  pos- 
teriorly), abdomen  to  isthmus,  in  pectoral  axil 
and  on  pectoral  base  except  adult  female  and  one 
juvenile  male.  Eye  ovoid,  more  rounded  in  adult 
male,  entering  dorsal  profile  of  head  except  in 
adult  female;  eye  diameter  17.2%  HL  in  adult 
male,  20.2-22.9%  HL  in  others.  Gill  slit  short, 
not  reaching  ventral  margin  of  pectoral  base  ex- 
cept in  adult  male,  in  which  it  reaches  slightly 
beyond  lowest  pectoral  ray;  gill  slit  length  42.2% 
HL  in  male,  28.1-34.0%  HL  in  juveniles,  and 
44.3%  HL  in  adult  female  owing  to  its  anterior 
extension  (below).  Lobe  at  dorsal  margin  of  gill 
slit  small,  squared-off  in  holotype  and  some  ju- 
veniles; dorsal  margin  not  extending  anteriorly 
in  some  juveniles;  operculum  broadly  rounded 
and  extending  anteriorly  almost  one  eye  diam- 
eter in  adult  female.  Pectoral  fin  origin  on  body 
midline  in  smallest  juveniles,  well  below  midline 
in  adults,  insertion  on  abdomen;  posterior  mar- 
gin of  fin  usually  evenly  rounded,  middle  rays 
longest,  however,  female  with  left  fin  distinctly 
wedge-shaped  and  right  fin  injured,  with  tips  of 
upper  and  middle  rays  missing  (fin  deeply 
notched).  Ventralmost  four  or  five  pectoral  rays 
thickened,  tips  slightly  exserted. 


238 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  10 


Mouth  subterminal,  oblique,  snout  and  upper 
jaw  more  produced  in  adult  male  than  others; 
upper  jaw  extending  posteriorly  to  rear  margin 
of  eye  in  adult  male  (upper  jaw  length  5 1 .7%  HL), 
or  to  middle  of  eye  in  others  (37. 1-42. 1%  HL  in 
adult  female  and  juveniles).  No  epidermal  pric- 
kles evident  on  snout  or  lips.  Oral  valve  weak, 
not  reaching  anterior  edge  of  vomer  in  adults, 
but  not  coalesced  with  lateral  margin  of  palate 
opposite  vomer  in  any  specimen.  Jaw  teeth  in 
adult  male  long,  caniniform,  slightly  retrorse  (Fig. 
1 3);  dentary  with  three  irregular  rows  anteriorly, 
blending  into  single,  posterior  row;  premaxilla 
with  two  irregular  anterior  rows,  blending  into 
single,  posterior  row;  outer  row  on  premaxilla 
about  five  times  as  large  as  inner  row.  Jaw  teeth 
in  adult  female  and  juveniles  small,  blunt,  con- 
ical; dentary  with  four  irregular  rows  anteriorly 
(juveniles  with  2-3  rows),  blending  into  single, 
posterior  row;  premaxilla  with  three  irregular  an- 
terior rows  (juveniles  with  two),  blending  into 
single,  posterior  row.  Vomerine  teeth  in  irregular 
patch  or  gradual  arc  (holotype).  Vomerine  and 
palatine  teeth  retrorse  and  caniniform  in  adult 
male,  small,  blunt  and  conical  in  others;  palatine 
teeth  in  irregular,  double  row  in  adult  female,  in 
single  row  in  others. 

Cephalic  lateralis  system  with  three  postor- 
bital  pores  arising  from  frontal  (pore  one),  be- 
tween pterotic  and  lateral  extrascapular  (pore 
three),  and  just  posterior  to  lateral  extrascapular 
(pore  four)  except  in  adult  female  in  which  pore 
two  is  present  on  both  sides,  arising  from  pterot- 
ic. Two  pairs  of  anterior  supraorbital  (nasal) 
pores,  one  set  mesial  to  nasal  tube,  the  other 
posteriorly  (posteromesially  in  adult  male).  Usu- 
ally seven  suborbital  pores,  with  six  arising  from 
ventral  ramus  of  bone  chain  and  one  from  as- 
cending ramus  behind  eye.  Exceptions  to  this  are: 
(1)  LACM  43717-1,  192  mm,  which  has  five 
pores  arising  from  ventral  ramus  and  one  from 
ascending  ramus  on  right  side,  and  five  from 
ventral  ramus  and  two  from  ascending  ramus  on 
left  side;  and  (2)  LACM  437 1 7- 1 , 1 72  mm,  which 
has  suborbital  pores  two  and  four  oddly  doubled 
(pore  at  these  positions  are  situated  one  atop  the 
other)  and  no  third  pore,  giving  seven  pores  aris- 
ing from  ventral  ramus  and  one  from  ascending 
ramus.  Eight  preoperculomanibular  pores,  four 
arising  from  dentary,  one  from  anguloarticular 
and  three  from  preopercle.  Interorbital  pores  ab- 
sent. Occipital  (supratemporal)  canal  with  single, 


mesial  pore  in  all.  Body  lateral  line  with  medio- 
lateral  branch  originating  on  vertical  about  one 
eye  diameter  anterior  to  posterior  margin  of  pec- 
toral fin;  mediolateral  branch  not  visible  in  great- 
ly faded  juveniles  in  LACM  43717-1;  ventral 
branch  originating  just  posterior  to  fourth  post- 
orbital  pore,  gradually  descending  across  abdo- 
men to  above  anal  fin  base;  both  branches  com- 
plete to  tail  tip. 

Dorsal  fin  origin  associated  with  vertebrae  4- 
5.  Anal  fin  origin  associated  with  ultimate  pre- 
caudal  vertebra  (numbers  24-26),  with  3-4  anal 
pterygiophores  inserted  anterior  to  haemal  spine 
of  first  caudal  vertebra.  Last  dorsal  ray  associated 
with  second,  fourth,  or  fifth  preural  vertebrae, 
last  anal  ray  associated  with  second  preural  ver- 
tebra. Caudal  fin  with  1-2  epural,  four  upper 
hypural,  and  4-5  lower  hypural  rays.  Pelvic  fins 
absent  except  in  MNHNC  6653  and  ZIN  48394 
which  have  small,  fleshy  fins  of  three  rays  each. 

Gill  rakers  on  upper  limb  (epibranchial)  of  first 
arch  moderately  long,  sharply  pointed;  ventral- 
most  rakers  on  lower  limb  (ceratobranchial)  of 
adult  female  and  juveniles  small,  but  sharply 
pointed,  those  of  adult  male  blunt,  with  squared- 
off  tips.  Branchiostegal  rays  six.  Pseudobranch 
filaments  long,  numbering  3—4. 

Recently  preserved  specimens  (collected  1978- 
1986)  uniformly  dark  brown  in  alcohol,  probably 
black  in  life.  Eye  blue.  Peritoneum,  lining  of  gill 
cavity  and  inner  palate  black.  Outer  palate  and 
lips  pale. 

DISTRIBUTION.— Known  from  eight  specimens 
taken  off  western  South  America  from  the  upper 
slope  to  the  edge  of  the  Peru-Chile  Trench,  re- 
liably known  from  depths  of  760-1,830  m  (Fig. 
11).  Capture  depth  of  holotype  unknown;  from 
fisheries  vessel. 

ETYMOLOGY.— From  the  Greek  n'taos  (middle) 
and  iropos  (hole,  pore)  in  reference  to  the  unique 
condition  of  the  species'  occipital  canal  with  its 
single,  mesial  pore. 

REMARKS.— This  is  the  second  species  of 
Pachycara  known  in  which  pelvic  fins  are  present 
or  absent  (the  other  is  P.  sulaki)  and  the  fourth 
zoarcid  reported  with  this  condition  (others  are 
Crossostomus  chilensis  [see  Gosztonyi  1 977]  and 
Lycenchelys  monstrosa  [see  Anderson  1982b]). 
The  two  specimens  with  pelvic  fins  (ZIN  48394 
and  MNHNC  6653)  also  exhibit  slightly  elevated 
vertebral  counts  (111  and  113  vs.  103-106  for 
others),  higher  gill  raker  counts  (16  and  18  vs. 


ANDERSON:  REVIEW  OF  EELPOUT  GENUS  PACHYCARA 


239 


FIGURE  14.    Pachycara  pammelas  n.  sp.,  CAS  62407,  holotype,  336  mm  SL,  from  off  Punta  Lobos,  Chile. 


13-15  for  others),  and  higher  pectoral  ray  counts 
(16  and  18  vs.  15-16  for  others).  It  was  tempting 
to  consider  the  presence  of  two  closely  related 
species  with  mesial  occipital  pores  in  the  sample 
of  eight  specimens.  However,  this  unique  con- 
dition, plus  the  identical  lateral  lines,  squama- 
tion,  and  state  of  dentition,  makes  it  apparent 
that  the  noted  anatomical  differences  cannot  be 
considered  above  the  range  of  individual  vari- 
ation found  in  either  other  Pachycara  species 
represented  by  larger  sample  sizes,  or  other  zoar- 
cids. 


Pachycara  pammelas  n.  sp. 

(Figs.  11,  14) 

HOLOTYPE.— CAS  62407  (female,  336  mm);  off  Punta  Lobos, 
Chile  (20°50'S,  70°27'W);  Chilean  fisheries  vessel;  trawl,  610 
m;  I.  Kong  Urbina,  21  Feb.  1981. 

PARATYPES.-SIO  72-183  (2;  96-192  mm);  off  Arica,  Chile 
(18°44.7^t0.2'S,  70-40.7-35. 1'W);  South  Tow  Exped.,  1,097- 
l,152m;R.  L.  Wisner  and  party,  1 152-1 330  hr,  7  May  1972. 
LACM  43732-1  (2;  152-217  mm);  off  Coquimbo,  Chile 
(30°06.0'S,  71°42.6'W);  ANTON  BRUUN  sta.  708;  free  ve- 
hicle set  line  trap,  1 ,000  m;  L.  W.  Knapp  and  party,  1 4  Aug. 
1966.  LACM  43726-1  (2;  226-231  mm);  off  Valpariso,  Chile 
(33°39.0'S,  72°09.5'W);  ANTON  BRUUN  sta.  LWK  66-41; 
trawl,  1,170-1,480  m;  L.  W.  Knapp  and  party,  10  Aug.  1966. 

COUNTS  AND  MEASUREMENTS.— Vertebrae  23- 
26  +  81-88  =  105-1 14;  D  99-109;  A  83-92;  C 
10-1 1;  P,  15-18;  P2  3;  vomerine  teeth  3-9;  pal- 
atine teeth  2-14;  gill  rakers  1-3  +  12-17  =  14- 
20;  branchiostegal  rays  six;  pseudobranch  fila- 
ments 2-5.  Following  measurements  in  percent 
SL:  head  length  12.8-15.6;  head  width  5.7-9.0; 
pectoral  fin  length  7.1-9.2  predorsal  length  1 5.0- 
18.4;  preanal  length  34.0-37.0;  body  height  6.3- 
10.5;  gill  slit  length  4.0-6.4;  caudal  fin  length  1 .6- 
2.7.  Following  measurements  in  percent  HL:  head 
width  39.9-67.4;  upper  jaw  length  34.1-52.7; 
pectoral  fin  length  50.8-64.2;  snout  length  15.6- 
22.2;  eye  diameter  19.5-30.4;  gill  slit  length  25.7- 
50.2;  interorbital  width  6.7-8.9;  interpupillary 
width  23.7-33.2;  pelvic  fin  length  7.4-13.3.  Pec- 
toral base/length  ratio  33.0-51.4. 


DIAGNOSIS.— Pelvic  fins  present;  origin  of  me- 
diolateral  lateral  line  immediately  posterodorsal 
last  postorbital  pore;  origin  of  ventral  lateral  line 
immediately  posterior  to  last  postorbital  pore; 
vertebrae  23-26  +  81-88  =  105-1 14;  no  scales 
on  nape. 

DESCRIPTION.— Following  description  based  on 
one  adult  female  (holotype),  two  juvenile  males, 
and  four  juvenile  females.  Head  ovoid,  snout 
bluntly  rounded  in  females  giving  head  more 
rounded  appearance  than  males.  Males  (1 52  and 
1 92  mm)  with  head  elongate,  dorsoventrally  de- 
pressed, snout  long,  gently  tapering.  Scales  ex- 
tending anteriorly  on  body  to  nape  where  present 
in  patch  extending  in  front  of  dorsal  fin  origin  in 
holotype  (absent  there  in  smaller  paratypes); 
scales  present  on  unpaired  fins  (except  96  mm 
juvenile),  abdomen,  pectoral  axil  and  base  (latter 
in  holotype  only).  Eye  ovoid,  entering  dorsal  pro- 
file of  head.  Gill  slit  short,  extending  just  to  lower 
margin  of  pectoral  base  or  slightly  above.  Lobe 
at  dorsal  margin  of  gill  slit  weak,  gill  slit  not 
extending  anteriorly  in  largest  male;  lobe  round- 
ed or  squared-off.  Pectoral  fin  origin  well  below 
body  midline  in  larger  specimens  (closer  to  mid- 
line  in  smallest  two),  insertion  on  abdomen;  pos- 
terior margin  of  fin  evenly  rounded,  middle  rays 
longest;  ventralmost  four  or  five  rays  thickened, 
tips  very  slightly  exserted. 

Mouth  subterminal,  oblique,  snout  more  pro- 
duced in  males  than  females;  upper  jaw  extend- 
ing posteriorly  to  middle  of  eye  in  females,  to 
just  beyond  its  posterior  margin  in  largest  male. 
No  epidermal  prickles  evident  on  snout  or  lips. 
Oral  valve  well  developed,  overhanging  anterior 
margin  of  vomer  in  holotype,  coalesced  with  sides 
of  palate  posterior  to  vomer  in  all.  Jaw  teeth 
small,  blunt,  conical;  dentary  with  four  irregular 
rows  anteriorly  in  holotype,  blending  into  single, 
posterior  row  consisting  of  2-4  teeth;  dentary 
teeth  in  two  to  four  (largest  two  paratypes,  LACM 
43726- 1)  irregular  anterior  rows  in  others,  blend- 
ing into  single,  posterior  row.  Premaxilla  with 
two  anterior  rows  of  teeth,  blending  into  single, 


240 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  10 


posterior  row  except  in  smallest  specimen  (SIO 
72-183,  96  mm)  which  has  single  row  of  nine 
teeth.  Vomerine  teeth  in  patch  or  arc  (small  spec- 
imens); palatine  teeth  blunt,  not  retrorse,  in  dou- 
ble row  in  holotype,  single  row  in  others. 

Cephalic  lateralis  system  with  three  postor- 
bital  pores  arising  from  frontal  (pore  one),  be- 
tween pterotic  and  lateral  extrascapular  (pore 
three),  and  posterior  to  lateral  extrascapular  (pore 
four).  Largest  male  (SIO  72-183,  192  mm)  with 
pores  one  and  four  only.  Two  pairs  of  anterior 
supraorbital  (nasal)  pores,  one  set  mesial  to  nasal 
tube,  the  other  posteriorly.  Seven  suborbital 
pores,  six  arising  from  ventral  ramus  of  bone 
chain  and  one  from  ascending  ramus  behind  eye 
just  ventral  to  first  postorbital  pore;  eight  sub- 
orbital  pores  on  right  side  only  in  LACM  43732- 1 , 
152  mm).  Eight  preoperculomandibular  pores, 
four  arising  from  dentary,  one  from  anguloartic- 
ular,  and  three  from  preopercle.  Interorbital  and 
occipital  pores  absent.  Body  lateral  line  difficult 
to  detect  in  faded  paratypes,  clear  in  holotype; 
mediolateral  branch  originating  about  half  an  eye 
diameter  or  less  posterodorsal  to  last  (number 
four)  postorbital  pore;  ventral  branch  originating 
immediately  posterior  to  postorbital  pore  four; 
both  branches  complete  to  tail  tip. 

Dorsal  fin  origin  associated  with  vertebra  4- 
5.  Anal  fin  origin  associated  with  ultimate  or 
penultimate  precaudal  vertebrae  (numbers  23- 
26),  with  2-4  anal  pterygiophores  inserted  an- 
terior to  haemal  spine  of  first  caudal  vertebra. 
Last  dorsal  ray  associated  with  fourth  preural 
vertebra,  last  anal  ray  associated  with  second 
preural  vertebra.  Caudal  fin  with  two  epural,  4- 
5  upper  hypural  and  four  lower  hypural  rays. 

Gill  rakers  on  upper  limb  (epibranchial)  of  first 
arch  simple,  sharply  pointed;  rakers  on  lower 
limb  (ceratobranchial)  laterally  compressed,  with 
squared-off  tips;  six  lower  limb  rakers  in  holo- 
type with  shallowly  bicuspid  tips.  Branchiostegal 
rays  six.  Pseudobranch  filaments  2-5,  minute  in 
smallest  specimen. 

Holotype  uniformly  black  (paratypes  faded), 
eye  blue.  Peritoneum,  gill  cavity  and  inner  palate 
black.  Outer  palate  and  inner  surface  of  lips  pale. 
Adult  female  with  ripe  ova  measuring  3.9-4.2 
mm  in  diameter  (not  counted). 

DISTRIBUTION.  — Known  from  seven  speci- 
mens captured  along  the  edge  of  the  Peru-Chile 
Trench  off  Chile  at  depths  of  610-1,480  m  (Fig. 
11). 


ETYMOLOGY.— From  the  Greek  iran  (an  inten- 
sifier  meaning  "all")  and  ntXas  (black),  a  noun 
in  apposition,  alluding  to  the  coloration  of  the 
species. 

Pachycara  brachycephalum  (Pappenheim,  1912) 

Lycodes  brachycephalus  Pappenheim,  1912:179-180,  pi.  X, 

fig.  3. 
Austrolycichthys  brachycephalus  (Pappenheim).  Regan,  1913: 

244-245,  fig.  2;  Waite,  1 9 1 6: 1 5,  pi.  I,  text  fig.  2, 2A;  Norman, 

1938:81,  83,  84. 
Pachycara  brachycephalus  (Pappenheim).  Anderson,  1988a: 

74-77,  figs.  16-19. 

MATERIAL  EXAMINED.— See  Anderson  (1988a:74-75). 

LECTOTYPE  (herein  designated).— ZMB  18929  (male,  169 
mm);  Wilhelm  Land,  Antarctica;  GAUSS  winter  station;  fish 
traps,  385  m;  Deutsche  Sudpolar  Exped.,  1902. 

DIAGNOSIS.— Pelvic  fins  present;  mediolateral 
lateral  line  originating  just  posterior  to  pectoral 
fin  margin;  ventral  lateral  line  originating  above 
dorsal  edge  of  gill  slit;  scales  absent  on  nape; 
vertebrae  23-27  +  71-86  =  95-1 13;  D  88-107; 
A  72-90;  P  16-18;  dorsal  fin  origin  associated 
with  vertebrae  3-7;  gill  rakers  0-3  +  9-13  =  9- 
16;  branchiostegal  rays  six;  suborbital  pores  5  + 
1  (5  +  0  in  one);  postorbital  pores  1-2;  head 
length  14.0-19.9%  SL;  pectoral  fin  length  10.0- 
12.7%SL. 

DISTRIBUTION.— Coastal  Antarctica  in  inner 
basins  and  on  outer  slope  at  depths  of  200- 
1,810m. 


DISCUSSION 

So  few  specimens  of  Pachycara  are  known  for 
most  species  that  it  is  difficult  to  consider  the 
evolutionary  history  of  the  genus.  Phylogenet- 
ically,  Pachycara  seems  to  have  arisen  as  a  deep- 
slope  sister  taxon  or  precursor  of  Lycodes,  dif- 
ferentiated by  its  reduced  parasphenoid  wing 
(apomorphy),  retention  of  the  oral  valve  and  ce- 
phalic pores,  and  lack  of  developed  mental  crests 
(plesiomorphies;  see  Anderson  1984).  Ther- 
marces  is  derived  with  respect  to  Pachycara  in 
four  reductive  characters:  ( 1 )  absence  of  pelvic 
bones;  (2)  absence  of  scales;  (3)  absence  of  lateral 
lines;  and  (4)  gelatinous  flesh. 

The  present  distributions  of  Pachycara  species 
indicate  two  main  historical  biogeographic  tracks: 
(1)  North  Atlantic-eastern  Pacific;  and  (2)  west- 
ern South  America-Antarctica  (P.  brachycepha- 


ANDERSON:  REVIEW  OF  EELPOUT  GENUS  PACHYCARA 


241 


/wm).  With  only  one  specimen  known,  the  origin 
and  distribution  of  P.  shcherbachevi  (Indian 
Ocean)  remains  a  mystery,  but  it  appears  closer 
to  P.  bulbiceps  than  the  two  West  African  species 
in  its  high  vertebral  count,  small  head,  and  re- 
trograde dorsal  fin  origin.  Recently,  I  reviewed 
the  mid-Miocene  tectonic  history  of  the  Amer- 
ican tropics  as  it  relates  to  the  distribution  of 
deep-sea  bottom  fishes  (Anderson  1988a),  and 
Pachycara  in  particular  (Anderson  and  Peden 
1988).  I  concluded  that  the  closing  of  the  Pan- 
amanian seaway  to  deep-demersal  fishes  about 
7-10  million  years  ago  did  not  sunder  Atlantic 
and  Pacific  stocks  of  P.  bulbiceps  (if  there  were 
any),  and  that  the  present  two  populations  pos- 
sibly exchange  genes  through  as  yet  unsampled 
South  Pacific  and  South  Atlantic  populations 
(Anderson  and  Peden  1988).  This  type  of  distri- 
bution (Fig.  5)  invites  speculation  on  the  origin 
and  relationships  of  P.  sulaki,  presently  known 
only  from  the  western  tropical  Atlantic.  This 
species  expresses  states  of  three  characters  that 
link  it  with  southeastern  Pacific  forms  (P.  me- 
soporum  and  P.  rimae)  that  may  have  phyloge- 
netic  utility:  (1)  possession  of  only  five  suborbital 
pores  along  the  ventral  ramus  in  some  individ- 
uals; (2)  dorsal  fin  origin  retrograde,  associated 
with  vertebrae  six  to  seven  (P.  sulaki)  or  eight 
(P.  rimae);  and  (3)  absence  of  pelvic  fins  in  some 
individuals  (P.  sulaki  and  P.  mesoporuni).  How- 
ever, a  phylogenetic  analysis  of  Pachycara  is  pre- 
cluded by  uncertainty  about  the  extent  of  distri- 
bution and  the  variability  of  many  character 
states,  particularly  of  the  axial  skeleton,  denti- 
tion, cephalic  pores,  and  the  occasional  lack  of 
pelvic  fins.  This,  in  turn,  produces  a  cautionary 
appraisal  of  the  specific  status  of  the  present  sam- 
ples of  P.  bulbiceps  (Pacific  and  Atlantic  popu- 
lations) and  P.  mesoporum  (range  of  variation) 
that  only  future  sampling  may  clarify. 


ACKNOWLEDGMENTS 

For  specimen  donations  or  loans,  photo- 
graphs, and  data  compilation  I  am  grateful  for 
the  help  of  C.  Allue,  A.  P.  Andriashev,  M.-L. 
Bauchot,  D.  M.  Cohen,  A.  Cornejo  C.,  J.  F.  Gras- 
sle,  K.  E.  Hartel,  D.  A.  Hensley,  P.  A.  Hulley, 
S.  Jewett,  S.  R.  Johnson,  I.  Kong  U.,  C.  Karrer, 
R.  J.  Lavenberg,  D.  F.  Markle,  R.  Melendez,  N. 
R.  Merrett,  A.  V.  Neyelov,  J.  G.  Nielsen,  W.G. 


Pearcy,  A.  E.  Peden,  B.  Ranchod,  M.  D.  Rich- 
ardson, C.  R.  Robins,  R.  H.  Rosenblatt,  J.  A. 
Seigel,  B.  Seret,  C.  R.  Smith,  D.  L.  Stein,  H.  J. 
Walker,  and  R.  R.  Wilson.  W.  I.  Follett  provided 
financial  assistance  for  part  of  this  study.  K.  Klitz 
rendered  the  specimen  illustrations  (Figs.  7,  9, 
12-14). 


LITERATURE  CITED 

ANDERSON,  M.  E.  1982a.  Revision  of  the  fish  genera  Gym- 
nelus  Reinhardt  and  Gymnelopsis  Soldatov  (Zoarcidae),  with 
two  new  species  and  comparative  osteology  of  Gymnelus 
viridis.  Natl.  Mus.  Nat.  Sci.,  Publ.  Zool.  (17):  1-76. 

.    1982b.  Biological  results  of  the  University  of  Miami 

Deep-Sea  Expeditions.  1 36.  A  new  eelpout  (Teleostei:  Zoar- 
cidae) from  the  eastern  tropical  Pacific  Ocean.  Bull.  Mar. 
Sci.  32(1):207-212. 

-.    1 984.   On  the  anatomy  and  phylogeny  of  the  Zoar- 


cidae (Teleostei:  Perciformes).  Ph.D.  Dissertation,  College 
of  William  and  Mary,  Williamsburg,  Virginia.  254  pp. 
-.   1 986.  Family  no.  95:  Parabrotulidae.  P.  343  in  Smiths' 


sea  fishes.  M.  M.  Smith  and  P.  C.  Heemstra,  eds.  MacMillan 
South  Africa,  Ltd. 
.    1988a.    Studies  on  the  Zoarcidae  of  the  southern 


hemisphere.  I.  The  Antarctic  and  subantarctic  regions.  Ant. 

Res.  Ser.  47:59-113. 
.   1 988b.  A  new  genus  of  California  eelpout  (Teleostei: 

Zoarcidae)  based  on  Maynea  californica  Starks  and  Mann, 

1911.  Proc.  Calif.  Acad.  Sci  45(5):89-96. 
ANDERSON,  M.  E.,  R.  E.  CRABTREE,  H.  J.  CARTER,  K.  J.  SULAK, 

AND  M.  D.  RICHARDSON.    1986.    Distribution  of  demersal 

fishes  of  the  Caribbean  Sea  found  below  2,000  meters.  Bull. 

Mar.  Sci.  37(3):794-807. 
ANDERSON,  M.  E.  AND  C.  L.  HUBBS.    1981.  Redescription  and 

osteology  of  the  northeastern  Pacific  fish  Derepodichthys 

alepidotus  (Zoarcidae).  Copeia  1 98 1(2):34 1-352. 
ANDERSON,  M.  E.  AND  A.  E.  PEDEN.    1988.  The  eelpout  genus 

Pachycara  (Teleostei:  Zoarcicdae)  in  the  northeastern  Pacific 

Ocean,  with  descriptions  of  two  new  species.  Proc.  Calif. 

Acad.  Sci.  46(3):83-94. 
ANDRIASHEV,  A.  P.    1954.   Fishes  of  the  northern  seas  of  the 

USSR.  Zool.  Inst,  Acad.  Sci.,  Contr.  Fauna  USSR  53:1- 

566. 
.    1973.   Zoarcidae.  In  Check-list  of  the  fishes  of  the 

north-eastern  Atlantic  and  of  the  Meditrranean.  J.-C.  Hu- 

reau  and  T.  Monod,  eds.  Unesco,  Paris.  1:540-547. 

-.    1986.  Zoarcidae  (including  Lycodidae).  In  Fishes  of 


the  northeastern  Atlantic  and  the  Mediterranean.  P.  J.  P. 

Whitehead,  M.-L.  Bauchot,  J.-C.  Hureau,  J.  Nielsen,  and  E. 

Tortonese,  eds.  Unesco,  Paris.  3:1 130-1 150. 
ARNULF,  I.,  F.  J.  MEUNIER,  AND  P.  GEISTDOERFER.    1987.  Os- 

teologiede  Thermarces  cerberus  Rosenblatt  and  Cohen,  1986, 

Zoarcidae  des  sources  hydrothermales  du  Pacifique  Est,  sui- 

vie  d'une  discussion  sur  sa  classification.  Cybium  1 1  (2):  141- 

158. 
COHEN,  D.  M.  AND  R.  L.  HAEDRICH.    1983.  The  fish  fauna  of 

the  Galapagos  thermal  vent  region.  Deep-Sea  Res.  30(4A): 

371-379. 
COHEN,  D.  M.,  R.  H.  ROSENBLATT,  AND  R.  L.  HAEDRICH.   1985. 


242 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  10 


Identity  of  thermal  vent  fishes  in  the  eastern  Pacific:  an 
interim  report.  Biol.  Soc.  Wash.,  Bull.  6:229-230. 

DINGERKUS,  G.  AND  L.  UHLER.  1977.  Enzyme  clearing  of 
alcian  blue  stained  whole  small  vertebrates  for  demonstra- 
tion of  cartilage.  Stain  Tech.  52(4):229-232. 

FOWLER,  H.  W.  1 936.  The  marine  fishes  of  west  Africa.  Bull. 
Amer.  Mus.  Nat.  Hist.  70(2):607-1493. 

CARMAN,  S.  1 899.  Reports  of  an  exploration  off  the  west 
coasts  of  Mexico,  Central  and  South  America,  and  off  the 
Galapagos  Islands,  in  charge  of  Alexander  Agassiz,  by  the 
ALBATROSS  during  1891.  XXVI.  The  fishes.  Mems.  Mus. 
Comp.  Zool.  24:1-431. 

GOLOVAN',  G.  G.  1 974.  Preliminary  data  on  the  composition 
and  distribution  of  the  bathyal  ichthyofauna  (in  the  Cap 
Blanc  area).  Oceanol.,  Akad.  Nauk  SSSR  14(2):288-290. 

.  1978.  Composition  and  distribution  of  the  ichthy- 
ofauna of  the  continental  slope  of  northwestern  Africa.  Tru- 
dy Inst.  Oceanol.  1 1 1:195-258. 

GoszTONYi,  A.  E.  1977.  Results  of  the  research  cruises  of 
FRV  "Walther  Herwig"  to  South  America.  XLVIII.  Revi- 
sion of  the  South  American  Zoarcidae  (Osteichthyes,  Blen- 
nioidei),  with  the  description  of  three  new  genera  and  five 
new  species.  Arch.  FischWiss.  27(3):  19 1-249. 

GUNTHER,  A.  1 880.  Report  on  the  shore  fishes  procured  dur- 
ing the  voyage  of  H.M.S.  "Challenger"  in  the  years  1873- 
1876.  Rep.  Sci.  Res.  Challenger  1(6):  1-82. 

HAEDRICH,  R.  L.  AND  N.  R.  MERRETT.  1 988.  Summary  atlas 
of  deep  living  demersal  fishes  in  the  North  Atlantic  Basin. 
J.Nat.  Hist.  22:1325-1362. 

HAEDRICH,  R.  L.,  G.  T.  ROWE,  AND  P.  T.  POLLONI.  1980.  The 
megabenthic  fauna  in  the  deep  sea  south  of  New  England, 
USA.  Mar.  Biol.  57:165-179. 

HUBBS,  C.  L.,  W.  I.  FOLLETT,  AND  L.  J.  DEMPSTER.  1979.  List 
of  the  fishes  of  California.  Occ.  Pap.  Calif.  Acad.  Sci.  (133): 
1-51. 

JENSEN,  A.  S.  1902.  Ichthyologiske  studier.  HI.  Om  nogle  ny 
arter  af  slaegten  Lycodes.  Vidensk.  Meddel.  Naturh.  Foren. 
Kbhvn.  1901:205-214. 

JORDAN,  D.  S.  1920.  The  genera  of  fishes,  part  IV.  Stanf. 
Univ.  Publ.,  Univ.  Ser.  (43):41 1-576. 

LEVTTON,  A.  E.  AND  R.  H.  GIBBS,  JR.  1988.  Standards  in 
herpetology  and  ichthyology.  Standard  symbolic  codes  for 
institution  resource  collections  in  herpetology  and  ichthy- 
ology. Supplement  No.  1 :  additions  and  corrections.  Copeia 
1988(l):280-282. 

LEVITON,  A.  E.,  R.  H.  GIBBS,  JR.,  E.  HEAL,  AND  C.  E.  DAWSON. 
1985.  Standards  in  herpetology  and  ichthyology:  part  I. 
Standard  symbolic  codes  for  institutional  resource  collec- 
tions in  herpetology  and  ichthyology.  Copeia  1985(3):802- 
832. 

MARKLE,  D.  F.  1 980.  A  new  species  and  a  review  of  the  deep- 
sea  fish  genus  Asquamiceps  (Salmoniformes,  Alepocephali- 
dae).  Bull.  Mar.  Sci.  30(l):45-53. 

MARKLE,  D.  F.  AND  G.  R.  SEDBERRY.  1978.  A  second  spec- 
imen of  the  deep-sea  fish,  Pachycara  obesa,  with  a  discussion 
of  its  classification  and  a  checklist  of  other  Zoarcidae  off 
Virginia.  Copeia  1978(l):22-25. 

MCALLISTER,  D.  E.  AND  E.  I.  S.  REES.  1964.  A  revision  of 
the  eelpout  genus  Melanostigma,  with  a  new  genus  and  with 
comments  on  Maynea.  Natl.  Mus.  Canada,  Bull.  199:85- 
109. 


MERRETT,  N.  R.  AND  N.  B.  MARSHALL.  1981.  Observations 
on  the  ecology  of  deep-sea  bottom  living  fishes  collected  off 
northwest  Africa  (08°-27°N).  Prog.  Oceanogr.  9:185-244. 

MUSICK,  J.  A.  1979.  Community  structure  of  fishes  on  the 
continental  slope  and  rise  off  the  middle  Atlantic  coast  of 
the  United  States.  Va.  Inst.  Mar.  Sci.,  Spec.  Sci.  Rept.  (96): 
1-53. 

NIELSEN,  J.  G.  1973.  Parabrotulidae.  In  Check-list  of  the 
fishes  of  the  north-eastern  Atlantic  and  of  the  Mediterra- 
nean. J.-C.  Hureau  and  T.  Monod,  eds.  Unesco,  Paris.  1: 
548-549. 

.  1986.  Parabrotulidae.  In  Fishes  of  the  north-eastern 

Atlantic  and  the  Mediterranean.  P.  J.  P.  Whitehead,  M.-L. 
Bauchot,  J.-C.  Hureau,  J.  Nielsen,  and  E.  Tortonese,  eds. 
Unesco,  Paris.  3:1151-1152. 

NORMAN,  J.  R.  1938.  Coast  fishes,  pt.  HI.  The  Antarctic 
region.  Discovery  Repts.  18:1-104. 

PAPPENHEIM,  P.  1912.  Die  Fische  der  Deutschen  Siidpolar- 
Expedition  1901-1903. 1.  Die  Fische  der  Antarktis  und  Su- 
bantarktis.  Deutsche  Sudpolar  Exped.  13:160-182. 

PAULL,  C.  K.,  B.  HECKER,  R.  COMMEAU,  R.  P.  FREE- 
MAN-LYNDE,  C.  NEUMANN,  W.  P.  CORSO,  S.  GOLUBIE,  J.  E. 
HOOK,  E.  SIKES,  AND  J.  CURRAY.  1984.  Biological  com- 
munities at  the  Florida  Escarpment  resemble  hydrothermal 
vent  taxa.  Science  226(4677):965-967. 

PEARCY,  W.  G.,  D.  L.  STEIN,  AND  R.  S.  CARNEY.  1982.  The 
deep-sea  benthic  fish  fauna  of  the  northeastern  Pacific  Ocean 
on  Cascadia  and  Tufts  Abyssal  Plains  and  adjoining  conti- 
nental slopes.  Biol.  Oceanogr.  l(4):375-428. 

REGAN,  C.  T.  1913.  The  Antarctic  fishes  of  the  Scottish  Na- 
tional Antarctic  Expedition.  Trans.  Royal  Soc.  Edinburgh 
49(2):229-292. 

RICHARDSON,  J.  1855.  Account  of  the  fish.  In  Last  of  the 
Arctic  voyages.  E.  Belcher,  ed.  2:1-30. 

ROSENBLATT,  R.  H.  AND  D.  M.  COHEN.  1986.  Fishes  living 
in  deepsea  thermal  vents  in  the  tropical  eastern  Pacific,  with 
descriptions  of  a  new  genus  and  two  new  species  of  eelpouts 
(Zoarcidae).  Trans.  San  Diego  Soc.  Nat.  Hist.  21(4):71-79. 

ROULE,  L.  1916.  Notice  preliminaire  sur  quelques  especes 
nouvelles  ou  rares  des  poissons  provenant  de  croisieres  de 
S.A.S.  le  Prince  de  Monaco.  Bull.  Inst.  Oceanogr.  Monaco 
(320):  1-32. 

.  1919.  Poissons  provenant  des  campagnes  du  yacht 

"Princesse  Alice"  (1891-1913)  et  du  yacht  "Hirondelle  II" 
(1914).  Result.  Camp.  Sci.  Prince  Albert  I  52:1-191. 

SPRINGER,  V.  G.  AND  W.  C.  FRIEHOFER.  1976.  Study  of  the 
monotypic  fish  family  Pholidichthyidae  (Perciformes). 
Smithson.  Contr.  Zool.  (2 16):  1^*3. 

VAILLANT,  L.  1888.  Expeditions  scientifiques  du  Travailleur 
etdu  Talisman  pendeant  les  annees  1880, 1881, 1882,  1883. 
Poissons.  G.  Masson,  Paris.  406  pp. 

WATTE,  E.  R.  1916.  Fishes.  Sci.  Rept.,  ser.  C,  Australasian 
Ant.  Exped.  1911-1914.  3(l):l-92. 

WHTTLEY,  G.  P.  1931.  New  names  for  Australian  fishes.  Austr. 
Zool.  6(4):3 10-334. 

ZUGMAYER,  E.  191  la.  Diagnoses  des  poissons  nouveaux 
provenant  des  campagnes  du  yacht  "Princesse  Alice"  ( 1 90 1- 
1910).  Bol.  Inst.  Oceanogr.,  Monaco  (193):1-14. 

.  1 9 1 1  b.  Poissons  provenant  des  campagnes  du  yacht 

"Princesse  Alice."  Result.  Camp.  Sci.  Prince  Albert  I  35:1- 
174. 


CALIFORNIA  ACADEMY  OF  SCIENCES 

Golden  Gate  Park 
San  Francisco,  California  94118 


PROCEEDINGS 

OF  THE 

CALIFORNIA  ACADEMY  OF  SCIENCES 


Vol.  46,  No.  11,  pp.  243-277,  13  figs.,  3  tables. 


December  20, 1989 


A  PHYLOGENETIC  REVISION  OF  THE  PHALLOSTETHID  FISHES 
(ATHERINOMORPHA,  PHALLOSTETHIDAE) 

By 
Lynne  R.  Parent! 

California  Academy  of  Sciences,  Golden  Gate  Park, 
San  Francisco,  California  94118 


ABSTRACT:  The  Phallostethidae  (including  Neostethidae)  comprises  a  group  of  small  to  minute  atherino- 
morph  fishes  distinguished  from  all  other  teleosts  by  the  presence  in  males  of  a  complex,  bilaterally  asymmetric, 
subcephalic  copulatory  organ,  the  priapium.  I  recognize  19  species  in  four  genera:  Phallostethus  Regan,  1913, 
Neostethus  Regan,  1916,  Gulaphallus  Herre,  1925,  and  Phenacostethus  Myers,  1928.  This  includes  one  new 
species,  Neostethus  robertsi,  from  Luzon.  Phallostethids  are  widely  distributed  throughout  coastal  brackish 
waters  from  Thailand,  Singapore,  Malaysia,  and  the  Philippines,  as  well  as  inland  freshwater  streams  of 
Luzon  and  Borneo. 

The  three  objectives  of  this  review  are  to:  (1)  describe  the  anatomy  and  variation  of  the  priapium  and 
propose  synonymy  and  probable  homology  of  its  many  and  varied  osteological  and  myological  modifications; 
(2)  present  taxonomic  synonymies  and  diagnoses  of  the  19  species  in  four  genera;  and  (3)  hypothesize  phy- 
logenetic  relationships  among  the  species  by  grouping  into  higher  categories  based  primarily  on  the  presence 
of  shared  derived  characters  of  the  priapium. 

Received  November  18,  1988.  Accepted  September  2,  1989. 


TABLE  OF  CONTENTS 

Abstract 

Introduction 

Materials  and  Methods 

Relationships  of  Phallostethids  to  other 

Atherinomorphs 

Synonymy  and  Homology  of  Priapial 

Structures 

Phylogenetic  Analysis 

Key  to  Phallostethidae 

Classification  of  Phallostethidae 

Systematic  Accounts 

Family  Phallostethidae 

Subfamily  Phallostethinae 

Tribe  Phallostethini 

Genus  Phallostethus  Regan,  1913 


Phallostethus  dunckeri  Regan, 

1913 267 

243  Genus  Phenacostethus  Myers,  1928     267 

244  Phenacostethus  smithi 

245  Myers,  1928 268 

Phenacostethus  posthon 

247  Roberts,  1971   268 

Phenacostethus  trewavasae 

248  Parenti,  1986 268 

255  Tribe  Neostethini 269 

264  Genus  Neostethus  Regan,  1916  ...     269 

265  Neostethus  lankesteri 

266  Regan,  1916 269 

266  Neostethus  bicornis 

267  Regan,  1916 269 

267  Neostethus  amaricola 

267  (Villadolid  and  Manacop,  1935)  270 


[243] 


244 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 1 


Neostethus  borneensis 

Herre,  1939  270 

Neostethus  villadolidi 

Herre,  1942 271 

Neostethus  zamboangae 

Herre,  1942  271 

Neostethus  palawanensis 

(Myers,  1935) 271 

Neostethus  thessa  (\urich,  1937)    272 
Neostethus  ctenophorus 

(Aurich,  1937) 272 

Neostethus  robertsi,  new  species  272 

Subfamily  Gulaphallinae 273 

Tribe  Gulaphallini  273 

Genus  Gulaphallus  Herre,  1925  ...    273 
Gulaphallus  eximius 

Herre,  1925  274 

Gulaphallus  mirabilis 

Herre,  1925  274 

Gulaphallus  falcifer 

Manacop,  1936 274 

Gulaphallus  bikolanus 

(Herre,  1926) 275 

Gulaphallus  panayensis 

(Herre,  1942) 275 

Conclusions 275 

Acknowledgments 276 

Literature  Cited 276 

INTRODUCTION 

Regan  (1913)  named  the  first  known  phallo- 
stethid  species  Phallostethus  dunckeri  in  honor 
of  Georg  Duncker  (1904),  who  first  reported  this 
unusual  group  of  fishes  from  Johore  on  the  Malay 
Peninsula.  During  the  following  three  decades, 
attention  focused  on  description  of  new  species 
and  genera  (e.g.,  Regan  1916;  Herre  1925,  1926; 
Myers  1928;  Villadolid  and  Manacop  1935; 
Manacop  1936),  anatomy  of  the  complex  cop- 
ulatory  organ,  the  priapium  (e.g.,  Bailey  1936; 
Aurich  1937;  TeWinkel  1939;  Woltereck  1942a, 
b),  and  consideration  of  the  relationships  of  phal- 
lostethids  to  other  fishes  (e.g.,  Smith  1 927;  Myers 
1935;  Hubbs  1944). 

Presence  of  a  priapium  led  to  speculation  that 
phallostethids  must  be  live  bearers  (e.g.,  Regan 
1913).  However,  Smith  (1927)  observed  female 
phallostethids  in  Thailand  depositing  eggs.  Vil- 
ladolid and  Manacop  (1935)  confirmed  that  fe- 
males of  Gulaphallus  mirabilis  lay  fertilized  eggs 
following  copulation.  Structure  of  ovaries  and 
testes,  presence  of  sperm  bundles  (spermato- 


zeugmata),  and  mode  of  internal  fertilization  are 
currently  under  investigation  (Grier  and  Parenti, 
in  prep.). 

Early  descriptions  of  phallostethids  followed 
the  convention  of  naming  morphologically  dis- 
tinctive populations  not  just  as  new  species  but 
as  new  genera  as  well.  In  the  last  comprehensive 
review  of  phallostethids,  Herre  (1942)  recog- 
nized 18  species  classified  in  10  genera,  one  of 
which  had  two  subgenera  (Table  1). 

Interest  in  phallostethid  systematics  was  re- 
newed with  description  of  a  new  species,  Phena- 
costethus  posthon  Roberts,  197 la  and  a  detailed 
osteological  study  of  Neostethus  bicornis  Regan 
(formerly  in  Ceratostethus)  (Roberts  1 97 1  b).  De- 
spite extensive  work  by  Roberts,  numerous  ques- 
tions remained  concerning  the  relationships  of 
phallostethids  to  other  atherinomorphs,  limits 
of,  and  relationships  among,  genera,  and  total 
number  of  recognizable  species.  It  was  with  the 
goal  of  answering  these  questions  that  I  began 
my  studies  of  phallostethids. 

A  preliminary  review  resulted  in  detailed  os- 
teology of  Phallostethus  dunckeri  and  the  pro- 
posal that  phallostethids  are  the  sister  group  of 
the  Indo-Pacific  atherinoid,  Dentatherina  mer- 
ceri  Patten  and  Ivantsoff,  1983  (Parenti  1984).  I 
described  a  new  species,  Phenacostethus  trewa- 
vasae  Parenti,  1 986a  and  brought  the  total  num- 
ber of  then  recognized  species  to  20.  Homology 
of  priapial  structures,  always  a  challenge  to  stu- 
dents of  phallostethids,  was  considered  by  ex- 
amining development  of  the  pelvic  fins  of  one 
species,  Gulaphallus  falcifer  Manacop  (formerly 
in  Manacopus),  in  which  there  are  bilaterally 
asymmetric  adult  females  as  well  as  males  (Pa- 
renti 1986b).  I  (Parenti  1986c)  reviewed  the  as- 
sociation between  bones  and  teeth  in  teleost  fish- 
es because  in  some  phallostethid  species  there  is 
a  perichondrally  ossified  lower  jaw  bone,  the 
paradentary,  that  may  possess  teeth. 

Much  of  the  information  compiled  on  phal- 
lostethid fishes  has  been  inaccessible  to,  or  ig- 
nored by,  systematists  because  of  the  array  of 
peculiar  names  applied  to  parts  of  the  complex 
priapium  and  because  of  the  small  size  of  phal- 
lostethids. Further,  no  previous  classifications 
summarize  cladistic  relationships  adequately; 
rather,  they  emphasize  gross  phenetic  differences 
among  species. 

Phallostethid  fishes  are  distributed  throughout 
coastal  mainland  and  insular  Southeast  Asia  from 
Thailand  to  the  Philippines  (Fig.  1),  entirely  west 


PARENTI:  PHALLOSTETHID  FISHES 


245 


of  Wallace's  Line,  one  of  several  hypothetical 
boundaries  separating  the  Asian  from  the  Aus- 
tralian biota  (see  Whitmore  1987).  Cladistic  re- 
lationships of  phallostethids  can  be  used  to  pre- 
sent a  working  hypothesis  of  the  distributional 
history  of  one  segment  of  the  Indo-Australian 
biota  (see  methodology  in  Humphries  and  Pa- 
renti  1986).  Results  from  this  study  will  be  com- 
bined with  those  of  other  analyses  to  prepare  a 
historical  biogeographic  analysis  of  the  Indo- 
Australian  atherinomorph  fishes  (Parenti,  in 
prep.). 

This  review  has  three  objectives:  (1)  to  de- 
scribe the  anatomy  and  variation  of  the  priapium 
and  propose  synonymy  and  probable  homology 
of  its  numerous  osteological  and  myological 
modifications;  (2)  to  present  taxonomic  synon- 
ymies and  diagnoses  of  the  herein  recognized  1 9 
species,  with  one  described  as  new,  classified  in 
four  genera;  and  (3)  to  hypothesize  relationships 
among  the  species,  grouping  them  into  higher 
categories,  primarily  on  synapomorphies  of  the 
priapium. 

MATERIALS  AND  METHODS 

I  observed  osteological  characters  on  speci- 
mens counterstained  with  alcian  blue  and  aliza- 
rin red  S  (Dingerkus  and  Uhler  1977)  or  stained 
solely  with  alizarin.  I  examined  specimens  with 
a  Zeiss  SV8  dissecting  microscope  and  recorded 
data  using  a  drawing  tube  and  photomicrography 
attachments.  Representatives  of  1 8  of  the  1 9  cur- 
rently recognized  phallostethid  species  were  ex- 
amined as  part  of  my  recent  studies.  No  speci- 
mens of  Neostethus  ctenophorus  (Aurich,  1937) 
were  available;  characters  are  from  the  original 
description. 

Meristic  data  (except  for  scale  counts)  were 
tabulated  from  cleared  and  stained  specimens. 
In  recording  meristic  data  (Table  2),  the  terminal 
half-centrum  is  counted  as  one  vertebra,  and  all 
fin  spines  and  rays  are  counted,  including,  for 
example,  the  last  two  anal  rays  which  articulate 
with  just  one  pterygiophore. 

Institutional  abbreviations  follow  the  Stan- 
dard Symbolic  Codes  for  Institutional  Research 
Collections  in  Herpetology  and  Ichthyology 
(Leviton  et  al.  1985).  The  following  abbrevia- 
tions are  defined:  Dist.,  district;  Is.,  island;  Prov., 
province;  R.,  river;  and  SL,  standard  length.  Cat- 
alog numbers  for  phallostethid  material  exam- 
ined are  given  in  the  text. 


TABLE  1.    Comparison  of  Herre's  (1942)  classification  of 
Phallostethids  with  that  presented  herein. 


Herre(1942) 


Present  paper 


Phallostethus  dunckeri 
Phenacostethus  smithi 


Mirophallus  bikolanus 
Solenophallus  thessa 
Gulaphallus  eximius 
Gulaphallus  mirabilis 
Ceratostethus  bicornis 
Manacopus  falcifer 
Neostethus  (Neostethus) 

amaricola 
Neostethus  (Neostethus) 

lankesteri 
Neostethus  (Neostethus) 

siamensis 
Neostethus  (Neostethus) 

villadolidi 
Neostethus  (Sandakanus) 

borneensis 
Neostethus  (Sandakanus) 

coronensis 
Neostethus  (Sandakanus) 

panayensis 
Neostethus  (Sandakanus) 

zamboangae 

Ctenophallus  ctenophorus 
Plectrostethus 
palawanensis 


Phallostethus  dunckeri 
Phenacostethus  smithi 
Phenacostethus  posthon 
Phenacostethus  trewavasae 
Gulaphallus  bikolanus 
Neostethus  thessa 
Gulaphallus  eximius 
Gulaphallus  mirabilis 
Neostethus  bicornis 
Gulaphallus  falcifer 
Neostethus  amaricola 

Neostethus  lankesteri 

synonym  of  Neostethus 

lankesteri 
Neostethus  villadolidi 

Neostethus  borneensis 

synonym  of  Neostethus 

borneensis 
Gulaphallus  panayensis 

Neostethus  zamboangae 

Neostethus  robertsi 
Neostethus  ctenophorus 
Neostethus  palawanensis 


The  theory  of  phylogenetic  reconstruction  fol- 
lowed is  that  commonly  referred  to  as  phyloge- 
netic systematics  or  cladistics,  as  outlined  by 
Hennig  (1966)  and  modified  by  numerous  au- 
thors (see  review  by  Wiley  1981).  Genera  and 
other  higher  taxa  are  defined  as  monophyletic  by 
the  sharing  of  homologous  derived  characters, 
termed  synapomorphies.  Species  are  defined  as 
the  smallest,  recognizable,  morphologically  dis- 
tinctive groups  of  males  and  females  which  may 
share  homologous  derived  characters,  termed 
autapomorphies. 

For  priapial  characters,  a  state  that  is  unique 
to  a  particular  phallostethid  species,  or  group  of 
species,  is  treated  as  derived.  This  is  a  special 
case  of  the  method  of  outgroup  comparison  (see 
Maddison  et  al.  1 984)  which  is  used  to  determine 
polarity  of  non-priapial  characters.  The  follow- 
ing atherinomorph  specimens  were  examined  for 
outgroup  comparison:  Bedotia  sp.,  aquarium 
material,  CAS  44367  (1  of  3  counterstained); 
Dentatherina  merceri,  Moluccas,  USNM  230374 


246 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 1 


INDIAN 
OCEAN 


10 


FIGURE  1.  Distribution  of  phallostethids.  Solid  circles,  type  localities  of  the  19  currently  recognized  phallostethid  species:  1. 
Phallostethus  dunckeri  Regan,  Muar  R.;  2.  Phenacostethus  smithi  Myers,  Bangkok;  3.  Phenacostethus posthon  Roberts,  a.  Khlong 
Kla  Sohm,  b.  Khlong  Langu;  4.  Phenacostethus  trewavasae  Parenti,  Baram  R.;  5.  Neostethus  lankesteri  Regan,  a.  Muar  R.,  b. 
Singapore,  c.  Chantabun  R.;  6.  Neostethus  bicornis  Regan,  Kuala  Langat;  7.  Neostethus  palawanensis  (Myers),  Ulugan  Bay;  8. 
Neostethus  thessa  (Aurich),  Lake  Mainit;  9.  Neostethus  borneensis  Herre,  a.  Sandakan  Bay,  b.  Coron;  10.  Neostethus  villadolidi 
Herre,  a.  Misamis,  b.  Dumaguete;  1 1 .  Neostethus  zamboangae  Herre,  Zamboanga;  1 2.  Neostethus  amaricola  (Villadolid  and 
Manacop),  Pasay;  13.  Neostethus  robertsi,  new  species,  Calasiao  R.;  14.  Neostethus  ctenophorus  (Aurich),  Laguna  de  Bay;  15. 
Gulaphallus eximius  Herre,  Nueva  Vizcaya;  1 6.  Gulaphallus  mirabilis  Herre,  Ibo  R.;  1 7.  Gulaphallusfalcifer  Manacop,  Pampanga 
and  Nueva  Ecija;  18.  Gulaphallus  bikolanus  (Herre),  lakes  Bato  and  Lanigay;  and  19.  Gulaphallus  panayensis  (Herre),  Panay. 
Solid  triangle,  unsubstantiated  report  of  unidentified  phallostethid  from  off  northwestern  Sumatra  by  Aurich  (1937). 


(2  of  42  counterstained);  Hypoatherina  ovalaua, 
Vanuatu,  Malekula  Is.,  CAS  62564  (2  of  1 0  coun- 
terstained); Melanotaenia  affinis,  northern  New 
Guinea,  CAS  40573  (4  of  3 1  counterstained),  and 
Pseudomugil  signifer,  New  South  Wales,  Austra- 
lia, CAS  62565  (6  of  62  counterstained). 

The  pattern  of  reduction  or  loss  of  a  character, 
which  is  common  in  atherinomorphs,  is  useful 
as  a  source  of  information  for  phylogenetic  re- 
construction when  correlated  with  other,  more 


complex  characters  (Parenti  1 986b).  Ontogenetic 
data  are  used  to  infer  homology  among  character 
states  (Patterson  1 982)  and  as  an  additional  source 
of  information  for  character  polarity  (Nelson 
1978).  Ontogenetic  transformation  and  pattern 
of  reduction  or  loss  are  the  most  reliable  sources 
of  information  on  polarity  of  states  of  the  pria- 
pium.  Outgroup  comparison  is  limited  because 
a  priapium  is  not  found,  even  in  a  simplified 
condition,  outside  the  Phallostethidae. 


PARENTI:  PHALLOSTETHID  FISHES 


247 


TABLE  2.    Meristic  characters  of  phallostethids.  Dl  =  first  dorsal  rays,  D2  =  second  dorsal  rays,  A  =  anal  rays,  PI  =  pectoral 
rays,  BR  =  branchiostegal  rays,  V  =  vertebrae,  SLS  =  scales  in  lateral  series. 


Species 

Dl 

D2 

A 

PI 

BR 

V 

SLS 

N.  lankesteri 

2 

6-7 

16-17 

11-12 

5 

34-35 

31-32 

N.  bicornis 

1 

5-6 

15-16 

10-11 

5-6 

36-37 

31-34 

N.  palawanensis 

1 

5-6 

15-16 

11 

5 

34-35 

30-32 

N.  thessa 

0 

8-10 

17-22 

12-13 

5 

36-37 

32-34 

N.  borneensis 

1-2 

5-6 

13-16 

10-12 

5 

31-32 

25-27 

N.  robertsi 

2 

6 

17-18 

10-11 

5 

36 

30 

N.  villadolidi 

2 

5-6 

14-16 

9 

5-6 

35-36 

30-32 

N.  zamboangae 

2 

5-6 

15-17 

9 

5-6 

35-36 

29-32 

N.  amaricola 

1-2 

5-6 

16 

10-11 

5-6 

35-36 

32 

N.  ctenophorus* 

2 

5-7 

14-17 

10-11 

— 

— 

31-33 

G.  eximius 

1-2 

7-8 

15-18 

11 

5 

36-37 

52-58 

G.  mirabilis 

1-2 

7 

17-18 

10-12 

5 

36-37 

36-38 

G.falcifer 

1-2 

8 

17 

10 

5 

35 

32 

G.  bikolanus 

0 

6 

14-17 

11-13 

5 

35-36 

32 

G.  panayensis 

1-2 

6-7 

13-15 

11 

5 

31-33 

28-30 

P.  dunckeri 

0 

8-10 

26-28 

9-10 

4 

40 

b 

P.  trewavasae 

1 

6 

14-15 

9-10 

5 

34 

b 

P.  position 

1 

5-6 

14-15 

9-10 

4 

34-35 

b 

P.  smilhi 

1 

6-7 

14-15 

9-10 

4-5 

33-35 

b 

"From  Aurich(  1937). 

b  Because  Phallostethini  scales  are  small  and  deciduous,  it  is  difficult  to  record  an  accurate  count. 


RELATIONSHIPS  OF  PHALLOSTETHIDS  TO 
OTHER  ATHERINOMORPHS 

None  has  questioned  monophyly  of  the  phal- 
lostethids, i.e.,  all  phallostethid  males  have  a 
priapium,  and  none  has  argued  that  the  priapium 
is  not  a  uniquely  evolved  structure.  I  present  no 
argument  here  to  contradict  phallostethid  mono- 
phyly. However,  I  acknowledge  that  information 
from  some  future  study  could  lead  us  to  conclude 
that  not  all  priapia  are  homologous. 

Perhaps  the  only  currently  controversial  aspect 
of  phallostethid  systematics  is  their  relationship 
to  other  acanthopterygian  fishes.  The  following 
four  issues,  with  references  for  arguments  pro 
and  con,  summarize  the  debate: 

1.  Atherinomorphs  are  monophyletic,  al- 
though their  relationship  to  other  acanthomorph 
teleosts  remains  obscure  (Rosen  1964;  Rosen  and 
Parenti  1981;  Collette  et  al.  1984;  Parenti  1981, 
1984;  Grier  and  Collette  1987). 

Since  Rosen  (1964)  formally  recognized  the 
series  Atherinomorpha,  which  comprises  the 
atherinoids,  cyprinodontiforms,  and  beloni- 
forms,  the  group  has  been  accepted  by  ichthy- 
ologists, although  with  some  reservation  (e.g., 
Roberts  1971b).  However,  it  is  generally  agreed 
that  systematic  differences  between  the  Ather- 


inidae  and  Cyprinodontidae  are  not  as  great  as 
had  been  thought  by  Hubbs  (1944).  Traditional 
groups  such  as  the  Percesoces— which  contains 
the  mullets  (Mugiloidea),  polynemids  (Polyne- 
moidea),  and  phallostethids  (Phallostethoidea), 
according  to  the  classification  of  Myers  (1937)— 
are  no  longer  recognized.  However,  possible  ath- 
erinomorph  relatives  may  be  among  constituents 
of  these  taxa,  such  as  the  mugilids.  Confirmation 
of  the  sister  group  of  the  Atherinomorpha,  as 
well  as  other  acanthomorph  taxa,  is  the  subject 
of  ongoing  studies  by  numerous  systematic  ich- 
thyologists, including  myself,  and  will  not  be  dis- 
cussed further  here. 

2.  Phallostethids  are  atherinomorphs  (Rosen 
and  Parenti  1981;  Parenti  1984;  White  etal.  1984; 
Ivantsoff  et  al.  1987),  or  may  be  related  to  poly- 
nemids (Bailey  1936)  or  to  gobioids  (Springer 
1983). 

Determining  whether  phallostethids  are  ath- 
erinomorphs requires  a  well-corroborated  defi- 
nition of  the  series  Atherinomorpha;  phallo- 
stethids either  fit  the  definition  or  they  do  not. 
Rosen  and  Parenti  (1981)  listed  defining  char- 
acters of  the  Atherinomorpha,  which  includes  a 
derived  ethmoid  region  and  hyobranchial  ap- 
paratus. Recent  workers  have  corroborated  the 


248 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 1 


proposal  that  the  Atherinomorpha  is  monophy- 
letic  (e.g.,  Grier  and  Collette  1987),  and  that 
phallostethids  are  atherinomorphs  (Parent!  1 984; 
Ivantsoffetal.  1987). 

Bailey  (1936)  hypothesized  a  close  relationship 
between  phallostethids  and  polynemids  based  on 
the  shared  close  association  of  the  pelvic  and 
pectoral  girdles.  Springer  (1983)  cited  shared  re- 
ductions between  phallostethids  and  gobioids  to 
suggest,  in  a  footnote,  their  possible  close  rela- 
tionship. To  endorse  the  hypothesis  that  phal- 
lostethids are  atherinomorphs  requires  that  the 
characters  shared  by  phallostethids  and  poly- 
nemids or  gobioids  be  treated  as  convergent. 
However,  Springer's  (1983)  proposal  expands  the 
list  of  possible  atherinomorph  close  relatives  to 
polynemids,  mullets,  and  gobioids. 

3.  Atherinoids,  including  the  silversides  (or 
hardyheads),  rainbowfishes,  and  phallostethids, 
are  not  monophyletic  (Rosen  and  Parenti  1981; 
Parenti  1984;  Ivantsoffetal.  1987)oratherinoids 
are  monophyletic  (White  et  al.  1984). 

Rosen  and  Parenti  (1981)  and  Parenti  (1984) 
proposed  that  atherinoids  do  not  constitute  a 
monophyletic  group  because  they  exhibit  trans- 
formation series  for  certain  characters,  such  as 
transformation  of  the  first  dorsal  fin  from  ple- 
siomorphic  atherinoids  (e.g.,  Bedotia  and  Me- 
lanotaenid)  in  which  it  is  well  developed,  to  phal- 
lostethids in  which  it  is  reduced  or  absent.  This 
view  was  supported  by  Ivantsoffet  al.  (1987). 
White  et  al.  (1984)  proposed  two  developmental 
characters  to  support  their  view  that  atherinoids 
are  monophyletic.  This  issue  will  be  clarified  when 
there  is  a  well-corroborated  proposal  of  relation- 
ships among  the  major  groups  of  atherinomorph 
fishes,  a  study  under  progress. 

4.  Phallostethids  are  the  sister  group  of  Den- 
tatherina  Patten  and  Ivantsoff,  1983  (Parenti 
1984),  or  the  close  relationship  of  phallostethids 
to  Dentatherina  is  not  supported  (Ivantsoffet  al. 
1987). 

Parenti  (1984)  proposed  a  close  relationship 
between  Dentatherina  merceri  and  phallosteth- 
ids based  on,  among  other  characters,  presence 
of  a  paradentary  bone.  At  issue  is  whether  the 
paradentary  elements— ossifications  of  the  labial 
ligament— are  homologous  in  phallostethids  and 
Dentatherina.  The  phallostethid  paradentary 
bone  is  composed  of  acellular  bone  surrounding 
hyaline  cartilage  (Parenti  1986c;  Ivantsoffet  al. 
1987).  The  Dentatherina  paradentary  bone  is 


composed  of  hyaline  cartilage  with  a  calcified 
core  (Ivantsoffet  al.  1987). 

Ivantsoffet  al.  (1987)  cited  the  histological  dif- 
ferences between  the  phallostethid  and  Dentath- 
erina paradentary  elements  to  conclude  that  the 
bones  are  not  homologs  and,  therefore,  cannot 
be  used  to  propose  a  close  relationship  between 
two  taxa.  They  also  rejected  homology  based  on 
the  conclusion  that  calcification  or  ossification 
in  a  labial  ligament  could  result  as  a  response  to 
stress.  I  concur  that  if  the  bones  are  not  homol- 
ogous they  cannot  be  used  to  propose  a  close 
relationship  between  two  taxa.  However,  I  reject 
the  functional  argument  against  homology.  The 
labial  ligament  could  become  calcified  or  ossified 
in  any  taxon  in  which  it  occurs;  within  atherino- 
morphs, it  is  reported  ossified  only  in  phallos- 
tethids and  Dentatherina.  In  my  opinion,  addi- 
tional characters  are  needed  to  accept  or  refute 
the  proposal  of  a  phallostelhid-Dentatherina  sis- 
ter group. 

Phallostethids  need  no  additional  support  to 
be  accepted  as  a  monophyletic  group;  a  seem- 
ingly endless  number  of  derived  characters  could 
be  listed  to  define  them  (see  Synonymy  and  Ho- 
mology of  Priapial  Structures).  Further  research 
should  focus  on  the  question:  to  which  group  of 
Old  World  atherinoids  are  phallostethids  most 
closely  related?  Hubbs  (1944:70)  summarized  the 
importance  of  this  type  of  question  in  system- 
atics:  "Whether  or  not  a  suborder  or  an  order  be 
recognized  for  the  Phallostethidae,  on  the  basis 
of  the  bizarre  copulatory  organ  (the  priapium) 
and  of  other  structures,  is,  like  most  questions 
of  ranking,  of  no  great  inherent  significance. 
Whether  the  relationship  of  the  family  is  with 
Cyprinodontidae  or  with  the  Atherinidae,  or  with 
neither,  is  a  problem  of  greater  meaning." 

SYNONYMY  AND  HOMOLOGY  OF 
PRIAPIAL  STRUCTURES 

Since  Regan  (1913)  first  described  the  complex 
anatomy  of  the  phallostethid  priapium,  students 
of  this  unusual  group  of  fishes  have  struggled  to 
attain  a  stable,  appropriate  set  of  terms  for  the 
numerous  priapial  structures  (Regan  1916;  Bai- 
ley 1936;  Aurich  1937).  Nearly  every  paper  dis- 
cussing priapial  anatomy  contains  errors  of  iden- 
tification or  proposed  homology.  A  common 
mistake  is  to  refer  to  the  same  structure  by  more 
than  one  name  in  a  single  paper. 


PARENTI:  PHALLOSTETHID  FISHES 


249 


Proliferation  of  names  for  priapial  structures 
was  encouraged  by  Regan  (1916),  who  believed 
that  the  priapium  was  a  new  morphological 
structure  with  no  known  homolog  among  teleost 
fishes.  For  example,  Regan  gave  the  adjectival 
name  "anterior  infrasulcar"  to  a  bone  which  I 
consider  the  homolog  of  one  of  the  pelvic  bones. 

No  doubt,  a  stable  nomenclature  has  not  been 
attained  because  homology  of  the  priapial  struc- 
tures has  not  been  well  understood,  and  because 
structure  of  the  priapium  may  vary  greatly  from 
one  phallostethid  genus  to  another.  Bailey  (1 936) 
concluded  that  priapia  of  Phenacostethus  smithi 
and  Gulaphallus  mirabilis,  and  hence  of  all  phal- 
lostethids,  correspond  bone-for-bone.  Further- 
more, he  believed  that  these  structures  could  be 
identified  in  polynemids,  close  relatives  of  phal- 
lostethids  according  to  Bailey.  However,  he  did 
not  examine  a  species  of  Neostethus,  which  has 
bony  elements  not  found  in  either  Phenacoste- 
thus or  Gulaphallus.  Also,  a  close  relationship 
between  polynemids  and  phallostethids  has  not 
been  supported  (see  Relationships  of  Phallo- 
stethids to  other  Atherinomorphs). 

The  priapium  may  be  thought  of  as  having 
three  structural  and  functional  components:  sus- 
pensory, holding  or  clasping,  and  papillary  (Bai- 
ley 1936).  The  suspensory  component  consists 
of  expanded  anterior  pleural  ribs,  including  an- 
teriorly expanded  cleithra  in  Neostethus  and  Gu- 
laphallus, as  well  as  modified  pelvic  bones.  The 
holding  or  clasping  component  consists  of  the 
externalized  ctenactinia  and/or  toxactinium,  the 
movement  of  which  is  controlled  mainly  by  a  set 
of  longitudinal  muscles.  The  papillary  compo- 
nent consists  of  the  seminal  and  infrasulcar  pa- 
pillae and  papillary  bones  used  in  the  transfer  of 
sperm  bundles  from  males  to  females. 

To  describe  the  diverse  and  detailed  priapia 
of  phallostethids,  I  have  compiled  a  synonymy 
of  priapial  structures  (Table  3  and  below).  In 
addition,  I  propose,  where  possible,  the  homolog 
of  each  element  as  found  in  relatively  unmodified 
atherinoid  fishes.  Listed  is  the  name  of  a  struc- 
ture, first  and  subsequent  citations,  including 
misidentifications,  in  major  papers  on  phallo- 
stethids, brief  description,  probable  homolog, 
known  synonyms,  and  citations. 

antepleural  cartilage  or  bone  (Figs.  2,  3)— Regan 
(1916:7),  Roberts  (1971b:408),  Parenti  (1986a: 
228).  A  large  block  of  cartilage,  sometimes  par- 


tially or  fully  ossified,  that  lies  anterior  to  the 
distal  tips  of  the  first  pleural  ribs  in  males. 

aproctal  side  or  aspect— Regan  (1913:552,  1916: 
5),  Herre  (1 942: 1 38),  Parenti  (1986a:225).  That 
side  of  the  male's  body  without  the  anal  open- 
ing. Male  phallostethids  are  bilaterally  asym- 
metric such  that  the  anus  and  seminal  papilla 
are  offset  to  opposite  sides  of  the  body  (see 
Parenti  1986a). 

axial  bone-Regan  (1916:7),  Bailey  (1936:464), 
TeWinkel  (1939:61),  Villadolid  and  Manacop 
(1935:215).  An  elongate  bone  with  rodlike 
spine  and  broad,  dorsal  crest.  The  aproctal  and 
proctal  axial  bones  are  homologous  with  the 
pelvic  bones  (Regan  1913,  19 16;  Bailey  1936). 
aproctal  axial  bone  (Figs.  2-5)— Roberts 
(1971b:408-409).  The  axial  bone  on  the 
aproctal  side  of  the  body.  Synonyms:  an- 
terior infrasulcar  bone  (Regan   1916:8, 
Bailey  1936:466),  desmactinium  (Aurich 
1937:267),  ctenactinium  (misidentifica- 
tion,  Herre  1925:508,  1926:539;  Myers 
1937:142),  second  ctenactinium  (mis- 
identification,  Bailey  1936:463),  short 
ctenactinium  (misidentification,  Villado- 
lid and  Manacop  1935:195). 
proctal  axial  bone  (Figs.  3-5)  — Roberts 
(1971b:408-409).  The  axial  bone  on  the 
proctal  side  of  the  body,  with  which  the 
toxactinium  and  ctenactinia  articulate. 
Synonym:  axial  bone  (Regan  1916:7; 
Myers  1935:5;  Bailey  1936:464;  Parenti 
1986a:228,  1986b:308). 

ctenactinium— Regan  (19 13: 5 52,  1916:5),  Myers 
(1928:1,  1935:5),  Bailey  (1936:465-466), 
Manacop  (1936:376),  Aurich  (1937:264), 
TeWinkel  (1939:62),  Herre  (1939:140,  1942: 
138),  Woltereck(1942b:339),  Roberts  (197 la: 
6,  1971b:393),  Parenti  (1984:6,  1986b:309). 
One  of  two  curved,  rodlike  priapial  bones  that 
articulate  with  the  posterior  end  of  the  proctal 
axial  bone.  Homologous  with  a  pelvic  fin  ray 
(Bailey  1936). 

first  ctenactinium  (Figs.  3-5)— The  promi- 
nent, elongate  externalized  priapial  bone 
in  Neostethus  and  Gulaphallus.  Usually 
called  the  ctenactinium,  referred  to  herein 
as  the  first  ctenactinium  to  distinguish  it 
unambiguously  from  the  second  ctenac- 
tinium. Synonym:  toxactinium  (misiden- 


250 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 1 


TABLE  3.    Summary  of  names  of  priapial  structures  and 
synonyms.  See  text  for  definitions  and  references. 


TABLE  3.    Continued. 


Priapial  structure 

Synonyms 

Priapial  structure 

Synonyms 

antepleural  cartilage  or 
bone 
aproctal  side  or  aspect 
axial  bone 
aproctal  axial  bone 

proctal  axial  bone 

anterior  infrasulcar  bone,  des- 
mactinium,  ctenactinium, 
second  ctenactinium,  short 
ctenactinium 
axial  bone 

pulvinulus 
seminal  papilla 
toxactinium 

uncus 

shieldlike  pad,  pulvinular  pad 
genital  papilla,  penis 
outer  pulvinular  bone,  pul- 
vinular bone,  toxactinial 
bone,  inner  pulvinular 
bone,  modified  pulvinulus 

ctenactinium 
first  ctenactinium 
second  ctenactinium 


efferent  groove 

epididymis 

fringe 

glandular  groove 
infrasulcar  bones 
anterior  infrasulcar 

bone 

posterior  infrasulcar 
bone 

infrasulcar  prominence 

or  papilla 
keel 
longitudinal  muscles 

ctenactinial  muscle 

outer  muscle  of  aproctal 
side 

inner  muscle  of  proctal 
side 

muscle  of  proctal  side 

papillary  bones 
penial  bone 
basipenial  bone 
papillary  bone 
prepapillary  bone 
cristate  bone 

pleuro-priapial  muscle 
priapial  process  of  cleithra 
priapial  ribs 
aproctal  priapial  rib 

proctal  priapial  rib 

priapium 

proctal  side  or  aspect 

pulvinular  appendage 

pulvinular  spine 

pulvinular  bones 
inner  pulvinular  bone 
outer  pulvinular  bone 


toxactinium 
Priapklaue,  "pulvinular 

spine,"  anterior  infrasulcar 

bone 


terminal  coil  of  vas  deferens 
comblike  projections,  comb- 
like  cilia,  cilia,  comb 


aproctal  axial  bone 


embryonic  fin-fold,  fringe 


penial  bones 


first  pleural  ribs 

first  pleural  rib  on  aproctal 

side 
first  pleural  rib  on  proctal 

side 


toxactinium 


tification,  Herre  1925:508,  1926:539;  Vil- 
ladolid  and  Manacop  1935:200). 
second  ctenactinium  (Figs.  2,  4,  5,  9)— The 
shorter  of  the  two  externalized  priapial 
bones  that  may  have  serrations,  as  in 
Phallostethus,  and  N.  zamboangae.  Syn- 
onym: Priapklaue  (Aurich  1937:267), 
"pulvinular  spine"  or  tip  of  anterior  in- 
frasulcar bone  (misidentification,  Herre 
1942:138). 

efferent  groove— Regan  (1916:5).  A  groove  on 
the  aproctal  side  of  the  body,  posterior  to  the 
glandular  groove  and  separated  from  it  by  the 
seminal  papilla  and  infrasulcar  prominence  or 
papilla. 

epididymis  (Fig.  5) -Regan  (1913:555),  Roberts 
( 1 97 1  a:6),  Parenti  ( 1 986a:23 1 ).  The  coiled  vas 
or  ductus  deferens  enclosed  in  a  membranous 
sac  at  posterior  extent  of  the  priapium.  Syn- 
onym: terminal  coil  of  vas  deferens  (Regan 
1916:4). 

fringe  (Figs.  3,  13)-Regan  (1916:5),  Roberts 
(1971b:408).  Posterior  ends  of  pelvic  fin  rays 
that  extend  beyond  the  posteroventral  outline 
of  the  epididymis.  Synonyms:  comblike  pro- 
jections (Myers  1928:8;  Villadolid  and  Man- 
acop 1935:197),  comblike  cilia  (Herre  1939: 
140),  cilia  (Herre  1942:138),  comb  (Roberts 
1971b:408). 

glandular  groove— Regan  (1916:5).  A  groove  on 
the  aproctal  side  of  the  body  at  the  boundary 
between  the  priapium  proper  and  the  body 
wall. 

infrasulcar  bones— Regan  (1916:8).  Two  bones, 
referred  to  by  Regan  ( 1 9 1 6)  as  the  anterior  and 
posterior  infrasulcar,  that,  in  part,  support  the 
infrasulcar  and  seminal  papillae. 


PARENTI:  PHALLOSTETHID  FISHES 


251 


SCA 


RAD 


PEN 


ANT 


PAX 


APAX 


FIGURE  2.  Diagrammatic  representation  of  modified  pelvic  and  pectoral  fins  in  the  priapium  of  a  dextral  male,  Phenacostethus 
smithi  (CAS-SU  35957  and  modified  in  part  from  Bailey  1936),  slightly  exploded  view  from  proctal  side.  Anterior  to  left. 
Dashed  line,  outline  of  pulvinulus;  dotted  and  dashed  line,  anterior  extent  of  cleithrum.  Dense  stippling,  cartilage;  light  stippling, 
bone.  Abbreviations:  ant,  antepleural  bone  and  cartilage;  apax,  aproctal  axial  bone;  bp,  basipenial  bone;  cl,  cleithrum;  cte,  second 
ctenactinium;  fpr,  first  pleural  ribs;  pap,  papillary  bone;  pax,  proctal  axial  bone;  pen,  penial  bone;  pin,  posterior  infrasulcar 
bone;  pp,  prepapillary  bone;  ptt,  posttemporal  bone;  rad,  radials;  sea,  scapulocoracoid;  scl,  supracleithrum;  u,  uncus. 


anterior  infrasulcar  bone— Regan  (1916:8). 
See  aproctal  axial  bone. 

posterior  infrasulcar  bone  (Fig.  2)— Regan 
(19 16:8),  Bailey  (1936:467).  A  rod-shaped 
bone  at  the  base  of  the  aproctal  axial  bone 
with  a  laminar  projection  that  lies  ventral 
to  the  aproctal  axial  bone,  reaching  the 
base  of  the  infrasulcar  papilla  or  promi- 
nence. 

infrascular  prominence  or  papilla— Regan  (1916: 
5,  8),  Herre  (1939:141).  A  papilliform  projec- 
tion anterior  to  the  seminal  papilla,  supported 
internally  by  a  posterior  projection  of  the 
aproctal  axial  bone. 

keel-Regan  (1913:550),  Myers  (1935:5),  Rob- 
erts (1971b:410).  A  median,  abdominal, 
slightly  frayed,  fleshy  ridge  between  the  uro- 
genital  opening  and  anterior  rays  of  the  anal 
fin.  Homologous  with  embryonic  fin-fold 
(Roberts  1971b).  Synonym:  fringe  (alternate 


term  not  homologous  with  fringe  as  defined 
above,  Regan  1916:2;  Myers  1928:5). 

longitudinal  muscles  (Fig.  6)— Regan  (1916:13, 
fig.  10;  1916:14).  Four  fusiform  muscles— 
ctenactinial  muscle,  outer  muscle  of  the  aproc- 
tal side,  inner  muscle  of  the  aproctal  side,  and 
muscle  of  the  proctal  side— that  originate  on 
the  axial  bones  and  control  copulatory  move- 
ments of  the  priapium.  Precise  homology  with 
pelvic  fin  muscles  (Winterbottom  1974)  is  un- 
clear. 

ctenactinial  muscle  (Fig.  6)— Regan  (1916: 
14).  Largest  of  the  longitudinal  priapial 
muscles,  with  an  origin  at  the  anterior  end 
of  the  dorsal  crest  of  the  aproctal  axial 
bone  and  insertion  at  the  proximal  end  of 
the  ctenactinium. 

outer  muscle  of  the  aproctal  side  (Fig.  6)— 
Regan  ( 1 9 1 6: 1 4).  An  elongate  muscle  that 
originates  on  the  concave,  aproctal  side  of 
the  main  crest  of  the  proctal  axial  bone 


252 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 1 


SCL 


SCA 


RAD 


IPUL 


FIGURE  3.  Diagrammatic  representation  of  modified  pelvic  and  pectoral  fins  in  priapium  of  a  sinistral  male,  Neostethus 
lankesteri  (CAS-SU  67162).  Anterior  to  left.  Dense  stippling,  cartilage;  light  stippling,  bone.  Open  circles,  fibrous  connective 
tissue  of  pulvinular  appendage.  Dotted  line  approximates  outline  of  epididymis.  Abbreviations:  apax,  aproctal  axial  bone;  bp, 
basipenial  bone;  cl,  cleithrum;  cte,  second  ctenactinium;  ctn,  first  ctenactinium;  fpr,  first  pleural  ribs;  ipul,  inner  pulvinular  bone; 
opul,  outer  pulvinular  bone;  pap,  papillary  bone;  pax,  proctal  axial  bone;  pen,  penial  bone;  pfr,  pelvic  fin  rays;  ptt,  posttemporal 
bone;  rad,  radials;  sea,  scapulocoracoid;  scl,  supracleithrum. 


and  inserts  on  the  base  of  the  papillary 
bone. 

inner  muscle  of  the  proctal  side  (Fig.  6)— 
Regan  ( 1 9 1 6: 1 4).  An  elongate  muscle  that 
originates  at  the  junction  of  the  main  crest 
of  the  proctal  axial  bone,  internal  to  the 
outer  muscle  of  the  aproctal  side,  and  in- 
serts on  a  process  of  the  proximal  end  of 
the  toxactinium  or  outer  pulvinular  bone. 

muscle  of  the  proctal  side  (Fig.  6)— Regan 
(1916:14).  A  muscle  that  originates  at  the 
extreme  anterior  end  of  the  axial  bone, 
with  a  few  fibers  at  the  posterior  base  of 
the  toxactinium,  posteriorly  produced  into 
a  long  tendon  that  inserts  on  the  base  of 
the  papillary  bone. 

papillary  bones  (Figs.   2-4)— Regan  (1916:8), 

Parenti  (1984:6).  Five  small  posterior  priapial 

bones.  Synonym:  penial  bones  (Bailey  1936: 

467). 

penial  bone  (Figs.  2-4)- Bailey  (1936:467). 


A  thin,  laminar  bone  that  projects  into  the 
seminal  papilla. 

basipenial  bone  (Figs.  2-4)— Bailey  (1936: 
465,  468).  A  rodlike  bone  that  articulates 
with  the  base  of  the  penial  bone  and  pos- 
terior border  of  the  cristate  bone. 

papillary  bone  (Figs.  2-4)— Regan  (1916:8), 
Bailey  (1936:468),  Herre  (1939:143).  A 
slender,  curved  bone  that  supports  the 
seminal  papilla. 

prepapillary  bone  (Fig.  2)— Bailey  (1936: 
468).  A  slender,  rodlike  bone  that  artic- 
ulates proximally  with  the  basipenial  bone 
and  distally  with  the  papillary  bone. 

cristate  bone -Bailey  (1936:469).  A  flat- 
tened, platelike  bone  that  lies  between  the 
papillary  bone  and  the  aproctal  axial  bone 
and  projects  into  the  base  of  the  seminal 
papilla. 

pleuro-priapial  muscle  (Fig.  6) — Regan  (1916:12). 
A  thin  muscle  that  originates  on  the  enlarged 


PARENTI:  PHALLOSTETHID  FISHES 


253 


SCA 


RAD 


OPUL 


APAX 


PEN 


FIGURE  4.  Diagrammatic  representation  of  modified  pelvic  and  pectoral  fins  in  priapium  of  a  sinistral  male,  Gulaphallus 
mirabili?  (CAS-SU  38904  and  modified  in  part  from  Bailey  1936).  Anterior  to  left.  Dense  stippling,  cartilage;  light  stippling, 
bone.  Abbreviations:  apax,  aproctal  axial  bone;  bp,  basipenial  bone;  cl,  cleithrum;  ctn,  first  ctenactinium;  fpr,  first  pleural  ribs; 
opul,  outer  pulvinular  bone;  pap,  papillary  bone;  pax,  proctal  axial  bone;  pax-pro,  process  of  the  proctal  axial  bone;  pen,  penial 
bone;  ptt,  posttemporal  bone;  rad,  radials;  sea,  scapulocoracoid;  scl,  supracleithrum. 


transverse  process  of  the  third  or  fourth  ver- 
tebra on  the  proctal  side  and  inserts  on  the 
proctal  axial  bone  at  the  point  of  articulation 
of  the  first  ctenactinium. 

priapial  process  of  cleithra  (Figs.  3,  4)— Regan 
(1913:553-554).  Anterior  prolongation  of  the 
cleithra  in  Neostethus  and  Gulaphallus. 

priapial  ribs  (Figs.  2-4)-Bailey  (1936:463-464). 
The  first  pleural  ribs  on  the  third  or  fourth 
vertebra,  modified  to  support  the  priapium. 
aproctal  priapial  rib— Bailey  (1936:464).  The 
first  pleural  rib  on  the  aproctal  side  of  the 
body. 

proctal  priapial  rib— Bailey  (1936:463).  The 
first  pleural  rib  on  the  proctal  side  of  the 
body. 

priapium  (Figs.  2-5, 9)-Regan  (1 9 13:551,1916: 
4),  Myers  (1928:1),  Bailey  (1936:453),  Aurich 
(1937:263),  TeWinkel  (1939:59),  Woltereck 
(1942a:254, 1942b:338).Thesubcephaliccop- 


ulatory  organ  of  phallostethid  fishes  compris- 
ing a  bony  portion  used  to  hold  the  female 
during  mating,  a  fleshy  copulatory  organ  or 
seminal  papilla  through  which  sperm  bundles 


APROCTAL  AXIAL 
BONE 


SEMINAL  PAPILLA 


EPIDIDYMIS 


FIGURE  5.  Diagrammatic  representation  of  head  and  an- 
terior portion  of  body,  Gulaphallus  mirabilis,  sinistral  male 
(CAS  50720).  Anterior  to  left.  Dense  stippling,  first  ctenactin- 


254 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 1 


ROCTAL  TRANSVERSE  PROCESS 


VENTRAL  BODY  PROFILE 


APROCTAL  AXIAL  BONE 


'\ 

FIRST 

CTENACTINIUM 


CTM 


FIGURE  6.  Schematic  diagram  of  priapial  muscles  (modified  from  Regan  1916:fig.  10).  Anterior  to  right.  Stippling,  bone. 
Abbreviations:  ppm,  pleuro- priapial  muscle;  ctm,  ctenactinial  muscle;  mp,  muscle  of  proctal  side;  ima,  inner  muscle  of  aproctal 
side;  oma,  outer  muscle  of  aproctal  side. 


pass,  and  expanded  cleithra  and  pleural  ribs 
for  suspension. 

proctal  side  or  aspect— Regan  (1913:551,  1916: 
5),  Herre  (1 942: 1 38),  Parenti  (1 986a:233).  That 
side  of  the  body  with  the  anal  opening. 

pulvinular  appendage  (Fig.  3)— Regan  (1916:5, 
8),  Herre  (1942:137).  A  subconical,  com- 
pressed mass  of  fibrous  tissue  that  lies  just  pos- 
terior to  the  pulvinulus  and  is  supported  by 
the  inner  pulvinular  bone  in  Neostethus. 

pulvinular  spine  (Fig.  3)— Regan  (1916:5,  9).  A 
pointed,  curved  process  of  the  toxactinium 
(=outer  pulvinular  bone). 

pulvinular  bones  (Figs.  2-4)— Regan  (191 6:8-9). 
Two  bones,  termed  an  inner  and  an  outer  pul- 
vinular, that  articulate  with  the  anterior  end 
of  the  proctal  axial  bone, 
inner  pulvinular  bone  (Fig.  3) — Regan  (1916: 
8-9).  A  longitudinal,  chondral  bone  that 
lies  anteroventral  to  the  proctal  axial  bone, 
outer  pulvinular  bone  (Figs.  2-4)— Regan 
(1916:9).  See  toxactinium. 


pulvinulus  (Figs.  2, 9)— Regan  (1916:5, 8),  Myers 
(1928:6,  1935:5),  Villadolid  and  Manacop 
(1935:194),  Manacop  (1936:376),  Aurich 
(1937:264),  Herre  ( 1939:1 40),  TeWinkel(  1939: 
62),  Roberts  (197 la: 7).  Fibrous  tissue  sur- 
rounding the  inner  pulvinular  bone  and/or 
toxactinium  (=outer  pulvinular  bone)  cover- 
ing point  of  articulation  of  the  toxactinium, 
cleithra,  and  proctal  axial  bone.  Synonyms: 
shieldlike  pad  (Bailey  1936:454);  pulvinular 
pad  (Parenti  1986a:230). 

seminal  papilla  (Fig.  5)— Regan  (1916:5),  Herre 
(1939:143).  A  papilliform  projection  lying 
posterior  to  the  infrasulcar  prominence,  sup- 
ported internally  by  the  posterior  infrasulcar 
bone.  Synonyms:  genital  papilla  (TeWinkel 
1939:62),  penis  (Roberts  1971a:l;  Parenti 
1986a:228). 

toxactinium  (Figs.  2-4)-Regan  (1913:551), 
Myers  (1928:1),  Bailey  (1936:462),  Woltereck 
(1942b:339),  Herre  (1939:140,  1942:138), 
Roberts  (1971a:6),  Parenti  (1986a:228).  A 
rodlike,  longitudinal  bone,  the  posterior  end 


PARENTI:  PHALLOSTETHID  FISHES 


255 


of  which  movably  articulates  with  the  sym- 
physis  of  the  cleithra  and  anterior  end  of  the 
proctal  axial  bone.  In  Phallostethus  and  Phe- 
nacostethus,  the  anterior,  distal  end  curves  be- 
low the  head  towards  the  aproctal  side  of  the 
body.  Called  the  outer  pulvinular  bone  in  all 
other  phallostethids,  in  which  the  anterior  end 
articulates  with  the  anterior  end  of  the  urohyal 
and  the  first  basibranchial  bone.  Synonyms: 
outer  pulvinular  bone  (Regan  1916:9),  pulvin- 
ular bone  (Bailey  1936:463),  toxactinial  bone 
(Bailey  1 936:47 1),  inner  pulvinular  bone  (mis- 
identification,  Roberts  1971b:393),  modified 
pulvinulus  (Parenti  1984:6). 

uncus  (Fig.  2)-Bailey  (1936:465,  467).  A 
U-shaped  chondral  bone  that  articulates  with 
the  proctal  axial  bone  posteriorly. 

PHYLOGENETIC  ANALYSIS 

External  morphology  and  osteology  of  the 
priapium  has  been  reported  in  detail  for  a  variety 
of  phallostethid  species  (e.g.,  Regan  1913,  1916; 
Bailey  1936;  Aurich  1937;  TeWinkel  1939;  Rob- 
erts 197 la,  b;  and  Parenti  1984,  1986a).  How- 
ever, this  represents  the  first  attempt  to  sum- 
marize derived  priapial  and  other  characters 
within  all  phallostethid  fishes,  and  to  use  cor- 
roborated hypotheses  of  synapomorphies  to  di- 
agnose monophyletic  genera  and  groups  of  gen- 
era. 

Aurich  (1937:285)  perhaps  came  the  closest  to 
the  goal  of  diagnosing  groups  of  genera  by  tab- 
ulating identifying  characters  of  his  three 
subfamilies  (but  called  "Familien")— Phallo- 
stethinae,  Neostethinae,  and  Gulaphallinae— 
recognized  herein  at  the  tribal  level  with  modi- 
fied definition  and  composition.  Aurich  did  not 
treat  relationships  among  his  three  subfamilies, 
nor  did  he  indicate  which  of  the  listed  characters 
are  what  we  would  today  recognize  as  derived  or 
apomorphic. 

Herre  (1942)  provided  a  synopsis  of  all  genera 
and  species,  except  Phallostethus  and  Phenaco- 
stethus,  both  then  monotypic.  I  find  many  of 
Herre's  (1942)  characters,  especially  of  scales, 
unworkable  in  a  phylogenetic  analysis.  For  ex- 
ample, number  of  scales  on  the  opercle  was  used 
to  divide  Neostethus  species  into  subgenera 
Neostethus  and  Sandakanus.  Herre  (1942:153) 
described  several  new  species,  including  N.  pa- 
nayensis,  treated  as  closely  related  to  N.  zam- 
boangae  Herre,  1 942  and  N.  coronensis  Herre, 


1 942,  with  which  it  shares  two  scales  on  the  oper- 
cle. However,  in  priapial  morphology,  N.  panay- 
ensis  is  a  member  of  the  monophyletic  Gula- 
phallus  as  defined  below,  and  not  closely  related 
to  N.  zamboangae  or  TV.  coronensis  (=borneensis, 
herein). 

Roberts  (1971b:396)  published  a  diagram  in 
which  relationships  among  the  then  recognized 
nine  genera  were  summarized.  It  was  not  Rob- 
erts's  purpose  to  provide  a  phylogenetic  revision 
of  all  phallostethids;  however,  like  Herre,  Rob- 
erts depicted  the  monotypic  Manacopus  as  closely 
related  to  Neostethus;  Manacopus  is  considered 
a  synonym  of  Gulaphallus  herein. 

Monophyly  of  Phallostethidae 

As  stated  in  the  section  Relationships  of  Phal- 
lostethids to  other  Atherinomorphs,  phallosteth- 
id fishes  have  been  considered  monophyletic  since 
Regan's  (1916)  review  of  the  then  three  known 
species:  by  definition,  all  phallostethid  males  have 
a  priapium,  a  complex  structure  not  found  in  any 
other  group  of  fishes.  It  is  perhaps  impossible  to 
tabulate  all  the  characters  in  phallostethids  that 
have  undergone  transformation  from  the  state  in 
plesiomorphic  atherinoids.  Over  30  modifica- 
tions of  pelvic  and  pectoral  bones,  muscles,  and 
cartilage  can  be  identified  (see  Synonymy  and 
Homology  of  Priapial  Structures). 

Further,  presence  of  a  keel,  homologous  with 
the  ventral  embryonic  fin-fold  as  reported  by 
Takita  and  Nakamura  (1986)  in  larval  Hypoath- 
erina  bleekeri,  and  other  paedomorphic  charac- 
ters in  all  phallostethids  suggest  that  evolution 
of  a  priapium  is  correlated  with  maturation  at  a 
relatively  small  size.  Congruence  (Mickevich 
1978)  exists  between  patterns  of  loss  and  reduc- 
tion associated  with  small  size  and  development 
of  some  complex  priapial  characters,  especially 
within  phallostethins. 

Relationships  of  Subfamilies  and  Tribes 

The  following  classification  summarizes  phy- 
logenetic relationships  among  the  tribes  and 
subfamilies  of  phallostethids  I  recognize: 

Family  Phallostethidae 
Subfamily  Phallostethinae 

Tribe  Phallostethini 

Tribe  Neostethini 
Subfamily  Gulaphallinae 

Tribe  Gulaphallini 


256 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 1 


Phallostethini 


Neostethini  Gulaphallini 


Phallostethini  Neostethini  Gulaphallini 


FIGURE  7.  Two  alternative  cladograms  of  relationships 
among  the  three  phallostethid  tribes.  Characters  numbered  as 
in  Phylogenetic  Analysis. 


I  use  the  following  vernacular  terms:  phallo- 
stethins  for  Phallostethini,  neostethins  for  Neo- 
stethini, and  gulaphallins  for  Gulaphallini.  To 
avoid  confusion,  no  vernacular  is  used  for  Phal- 
lostethinae.  Monophyly  of  each  tribe  is  consid- 
ered below.  Two  alternative  relationships  are: 
Phallostethini  and  Neostethini  are  sister  tribes 
(Fig.  7a);  and  Neostethini  and  Gulaphallini  are 
sister  tribes  (Fig.  7b).  I  have  found  no  support 
for  the  hypothesis  that  Gulaphallini  and  Phal- 
lostethini are  sister  tribes;  that  possible  phylog- 
eny  will  not  be  discussed  further. 

The  first  alternative  is  supported  by  at  least 
four  derived  characters,  two  of  the  priapium,  one 
of  the  oral  jaws,  and  one  of  the  gill  arches.  Char- 
acters have  the  same  numbers  as  in  the  clado- 
grams of  Figures  7,  8,  and  11. 

1.  Second  ctenactinium  present  (Figs.  2,  3,  9). 
The  second  ctenactinium  is  a  small,  curved, 

externalized  priapial  bone  that  may  have  bony 
serrations  (Regan  1916:fig.  3).  It  is  present  in 
Neostethus  (Figs.  3,  9)  and  Phallostethus,  and  is 
rudimentary  in  Phenacostethns  (Fig.  2).  The  bone 
is  absent  in  Gulaphallus  (Fig.  4),  contrary  to  Au- 
rich  (1937)  who  misidentified  the  large,  partially 
externalized  papillary  bone  as  a  second  ctenac- 
tinium. 

2.  Shieldlike  pulvinulus  present  (Figs.  2,  9) 
and  outer  pulvinular  bone  robust  (Figs.  2,  3). 

The  oval  pulvinulus  of  Phallostethus,  Phena- 


costethus  (Fig.  2),  and  Neostethus  (Fig.  9)  forms 
a  shieldlike  covering  of  the  articulation  point  be- 
tween the  outer  pulvinular  and  proctal  axial 
bones.  The  outer  pulvinular  bone,  although  pres- 
ent in  gulaphallins  (Fig.  4),  is  larger  and  more 
robust  in  phallostethins  and  neostethins.  Bailey 
(1936)  and  Aurich  (1937)  reported  a  slender,  long 
pulvinulus  paralleling  the  longitudinal  ctenactin- 
ial  muscles  in  Gulaphallus.  However,  the  struc- 
ture in  Gulaphallus  appears  to  be  simply  a  fold 
of  skin  covering  the  muscles  and  is  viewed  here 
as  doubtfully  homologous  with  the  distinctive 
pulvinulus  of  neostethins  and  phallostethins. 

3.  Highly  protrusible  oral  jaws. 
Premaxillary  ascending  processes  are  long,  and 

a  short,  rodlike  block  (submaxillary  element  of 
Parenti  1984)  lies  between  the  maxilla  and  vo- 
mer  in  Neostethini  and  Phallostethini,  both 
characters  contributing  to  the  highly  protrusible 
oral  jaws  (Roberts  1971b;  Ivantsoffet  al.  1987). 
The  block  may  be  ossified  in  Phallostethus  and 
some  species  of  Neostethus;  otherwise  it  is  car- 
tilaginous. In  Gulaphallini,  there  is  a  thin  con- 
nective tissue  meniscus  between  the  maxilla  and 
vomer,  except  in  G.  panayensis  in  which  the  me- 
niscus is  enlarged. 

4.  Reduction  of  hyobranchial  apparatus. 
The  entire  hyobranchial  apparatus  is  reduced 

in  phallostethids  relative  to  other  atherinoids  (see 
Rosen  and  Parenti  1981).  Teeth  on  the  fourth 
ceratobranchial  toothplate  are  few  or  absent  in 
phalostethins  and  neostethins;  they  are  present 
in  outgroup  taxa  such  as  Dentatherina  merceri 
and  Pseudomugil  signifer,  as  well  as  in  Gula- 
phallus. 

The  alternative  that  Neostethini  and  Gula- 
phallini are  sister  tribes  is  supported  by  what  may 
be  described  as  one  complex  priapial  character. 

5.  Anterior  elongation  of  priapial  suspensory 
system,  including  anteriorly  expanded  cleithra 
and  transverse  processes  of  fourth  vertebrae  of 
males,  with  the  proctal  transverse  process  larger 
than  the  slightly  modified  aproctal  transverse 
process  (Figs.  3,  4),  and  elongate  first  ctenactin- 
ium, the  prominent  externalized  priapial  bone 
(Figs.  3,  4,  9). 

Parsimony  may  be  used  to  choose  the  first  phy- 
logenetic  hypothesis  (Fig.  7a)  over  the  second 
(Fig.  7b),  but  that  criterion  alone  is  not  satisfac- 
tory. Outgroup  comparison  is  of  limited  use  in 
deciding  polarity  of  priapial  structures,  and  like- 
wise, complex  priapial  characters  may  be  treated 


PARENTI:  PHALLOSTETHID  FISHES 


257 


as  one  or  several  characters.  I  choose  the  hy- 
pothesis that  Phallostethini  and  Neostethini  are 
sister  tribes  for  reasons  of  parsimony  and  also 
because  derived  characters  other  than  those  of 
the  priapium  are  shared.  This  hypothesis  re- 
quires that  anterior  elongation  of  the  priapium 
in  neostethins  and  gulaphallins  (character  5, 
above)  be  viewed  as  plesiomorphic  for  all  phal- 
lostethids. 

Monophyly  of  Tribes 

Phallostethini.— Four  characters  distinguish 
phallostethins  from  all  other  phallostethids. 

6.  A  large,  fleshy  seminal  papilla. 

The  seminal  papilla  is  enlarged  relative  to  the 
papilla  in  neostethins  or  gulaphallins  (Fig.  5). 
The  large  papilla  may  be  smooth  (Roberts  1 97 1  a: 
fig.  5)  or  ruffled  (Roberts  1971a:fig.  4).  An  en- 
larged seminal  papilla  refutes  the  hypothesis  that 
the  diminutive  phallostethins  can  be  described 
solely  as  paedomorphic  phallostethids. 

7.  The  curved  toxactinium  the  largest  exter- 
nalized bony  priapial  element  (Fig.  2). 

Treating  the  distinctive  curved  toxactinium  as 
a  derived  character  of  Phallostethus  and  Phena- 
costethus  is  consistent  with  the  interpretation 
(above)  that  the  elongate  ctenactinium  and  an- 
terior elongation  of  the  priapium  in  neostethins 
and  gulaphallins  are  primitive  for  phallostethids. 

8.  Extremely  small  and  delicate,  with  decid- 
uous scales. 

The  largest  phallostethid,  Gulaphallus  exi- 
mius,  reaches  a  recorded  37  mm  SL  (see  char- 
acter 47,  below).  TeWinkel  (1939)  gives  an  un- 
supported range  of  14-40  mm  SL.  Neostethins 
and  gulaphallins  generally  mature  at  over  20  mm 
SL;  the  exception,  N.  borneensis,  is  considered 
secondarily  reduced.  Small  adult  size  and  the 
deciduous  scales  of  phallostethins  are  probably 
paedomorphic  characters  correlated  with  char- 
acter 9. 

9.  A  translucent,  membranous  dome  on  dor- 
sal surface  of  head. 

The  membranous  dome  (Roberts  1971a:figs. 
2,  3)  on  the  dorsal  surface  of  the  head  may  best 
be  described  as  a  larval  characteristic  of  atheri- 
noids  (as  in  Hypoatherina  bleekeri;  see  Takita 
and  Nakamura  1986)  retained  in  adults. 

Neostethini.  — Two  characters  distinguish 
Neostethus  from  all  other  phallostethids. 

10.  An  inner  pulvinular  bone  (Fig.  3). 


The  inner  pulvinular  bone  is  a  large,  well-de- 
veloped chondral  bone  lying  just  posterior  to, 
and  articulating  with,  the  outer  pulvinular  bone. 
It  has  no  identified  homolog  in  phallostethins  or 
gulaphallins. 

1 1 .  A  high  number  of  thin,  bony  projections 
on  the  papillary  bone  (Fig.  3). 

The  papillary  bone  has  a  high  number  (re- 
portedly more  than  80;  Roberts  1971b)  of  thin, 
bony  projections  in  Neostethus  (Aurich  1937:fig. 
6).  These  are  the  projections  used  by  Myers  (1935) 
to  diagnose  a  new  genus,  Plectrostethus  (=Neo- 
stethus,  herein).  Myers  apparently  compared  his 
new  species  with  phallostethins  and  gulaphallins, 
which  lack  projections.  In  the  neostethins,  Neo- 
stethus thessa  and  N.  zamboangae,  the  papillary 
bone  has  only  rudimentary  projections. 

Gulaphallini.  —  Gulaphallus  species  are  togeth- 
er distinguished  from  all  other  phallostethids  by 
three  derived  characters. 

12.  A  perforated  gular  flap  of  skin  through 
which  the  anterior  end  of  the  first  ctenactinium 
may  pass  and  remain  while  at  rest  (Fig.  5;  Aurich 
1937:fig.  2). 

Neostethins  are  the  other  group  of  phallosteth- 
ids with  an  elongate  first  ctenactinium;  however, 
neostethins  do  not  possess  a  large  gular  flap 
through  which  the  anterior  end  of  the  ctenactin- 
ium may  pass.  This  character,  mentioned  and 
figured  by  Aurich  (1937:fig.  2),  is  uniquely  de- 
rived in  Gulaphallus. 

13.  The  aproctal  axial  bone  projects  beyond 
ventral  body  profile  (Figs.  4,  5). 

In  no  other  phallostethids  does  a  priapial  bone 
project  beyond  the  body  profile  and  away  from 
the  body  of  the  priapium. 

1 4.  A  small,  dorsal  crest  posteriorly  on  proctal 
axial  bone  (Fig.  4). 

In  other  phallostethids  (Figs.  2,  3)  the  proctal 
axial  bone  has  a  uniform  dorsal  outline. 


Species  Relationships  Within  Tribes 
Phallostethini 

Relationships  among,  and  definitions  of,  the 
four  phallostethin  species  Phallostethus  dun- 
ckeri,  and  Phenacostethus  smithi,  P.  posthon,  and 
P.  trewavasae  were  treated  in  detail  by  Parenti 
(1986a)  and  need  not  be  repeated  here.  Defining 
characters  of  each  species  are  listed  in  the  Sys- 
tematic Accounts. 


258 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 1 


40 
139 

138 
137 


>36 


33 

32 


291 

141     I 

1421 

135  1 

131     , 

28 

1 

134  1 

130    , 

27 

1 

1 

1 

^  

v  

1 

1 

i 

•      rb 

•  1 

I 

|             LJ 

An 

. 

r^     i   JL 

126 
125 
24 


FIGURE  8.  Cladogram  of  relationships  among  species  in  the 
Neostethini.  Solid  squares,  synapomorphies;  open  squares, 
symplesiomorphies.  Characters  numbered  as  in  Phylogenetic 
Analysis. 


Neostethini 

Hypothesized  relationships  among  the  ten 
Neostethus  species  are  depicted  in  Figure  8.  Neo- 
stethus  lankesteri  is  considered  to  be  the  most 
plesiomorphic  neostethin;  it  is  excluded  from  a 
monophyletic  group  comprising  Neostethus 
amaricola,  N.  villadolidi,  N.  robertsi,  N.  pala- 
wanensis,  N.  bicornis,  N.  borneensis,  N.  zam- 
boangae,  and  N.  thessa,  denned  by  two  shared 
derived  characters. 

15.  An  enlarged,  prominent,  second  ctenac- 
tinium. 

Presence  of  a  second  ctenactinium  is  treated 
as  a  derived  character  of  neostethins  and  phal- 
lostethins  (see  character  1,  above).  The  second 
ctenactinium  is  a  small  curved  bone  in  phallo- 
stethins  (Fig.  2)  and  N.  lankesteri  (Fig.  3).  It  is  a 
much  larger,  prominent,  externalized  priapial 
bone  in  all  other  species  of  Neostethus,  including 
N.  ctenophorus,  as  illustrated  by  Aurich  (1937: 
fig.  6).  Variation  in  shape  of  the  second  ctenac- 


tinium and  in  its  association  within  other  pria- 
pial structures  is  the  primary  source  of  infor- 
mation on  morphological  species  recognition 
within  neostethins. 

16.  A  short,  bladelike  aproctal  axial  bone. 
The  aproctal  axial  bone  is  nearly  as  long  as  the 

proctal  axial  bone  and  reaches  the  pulvinulus  in 
phallostethins  (Fig.  2),  Neostethus  lankesteri  (Fig. 
3),  N.  ctenophorus  (Aurich  1937:fig.  6),  and  Gu- 
laphallus  (Fig.  4).  In  all  other  species  of  Neos- 
tethus (Roberts  1971b:fig.  12),  the  aproctal  axial 
bone  is  shortened  and  somewhat  bladelike;  it 
does  not  reach  the  pulvinulus. 

Neostethus  ctenophorus,  of  which  no  speci- 
mens are  available,  is  treated  as  forming  an  un- 
resolved trichotomy  with  N.  lankesteri  and  all 
other  Neostethus  (Fig.  8).  I  divide  these  remain- 
ing species  of  Neostethus  into  two  monophyletic 
groups,  which  I  call  the  borneensis  and  the  bi- 
cornis groups,  diagnosed  as  follows: 

The  borneensis  group,  comprising  Neostethus 
borneensis,  N.  thessa,  and  N.  zamboangae,  is  di- 
agnosed by  two  shared  derived  characters  that 
could  be  treated  as  one  complex  priapial  char- 
acter. 

17.  Strongly  curved,  thin,  second  ctenactin- 
ium. 

The  second  ctenactinium  is  a  strongly  curved, 
thin  bone  (Aurich  1 937:fig.  4)  as  opposed  to  being 
a  stout  fin  ray  as  in  Neostethus  amaricola  and  N. 
villadolidi  (Fig.  9)  and  N.  bicornis  (Roberts  1 97  Ib: 
fig.  12). 

18.  Papillary  and  penial  bones  curved  and 
elongate,  lying  parallel  to  dorsal  surface  of  second 
ctenactinium. 

Position  of  the  papillary  and  penial  bones  just 
dorsal  to  the  second  ctenactinium  means  that  the 
seminal  papilla  (supported  by  the  papillary  and 
penial  bones)  empties  just  dorsal  to  the  second 
ctenactinium,  as  opposed  to  a  more  posterior 
position  in  those  phallostethids  with  smaller 
papillary  and  penial  bones. 

The  bicornis  group,  comprising  Neostethus  bi- 
cornis, N.  palawanensis,  N.  robertsi,  N.  amari- 
cola, and  N.  villadolidi,  is  diagnosed  by  one  com- 
plex, derived  character. 

19.  Both  the  proctal  and  aproctal  transverse 
processes  of  the  fourth  vertebra  in  males  are  en- 
larged and  oriented  anteriorly. 

It  was  concluded  (character  5,  above)  that  the 
anterior  elongation  of  the  entire  priapial  appa- 
ratus is  a  plesiomorphic  character  for  phallo- 
stethid  fishes.  This  character  complex  is  modi- 


PARENTI:  PHALLOSTETHID  FISHES 


259 


fied  in  the  bicornis  group  such  that  the  aproctal 
as  well  as  the  proctal  transverse  processes  of  the 
fourth  vertebra  in  males  are  drawn  so  far  ante- 
riorly that  the  anterior  tip  of  both  processes  lies 
parallel  to  the  first  vertebra. 

Within  the  borneensis  group,  Neostethus  bor- 
neensis  is  considered  plesiomorphic  to  N.  thessa 
and  N.  zamboangae,  diagnosed  as  sister  species 
by  sharing  two  derived  characters. 

20.  Anterior  aproctal  ossification  present. 

In  all  species  of  Neostethus  except  N.  lankes- 
teri,  the  aproctal  axial  bone  is  short  and  does  not 
reach  the  pulvinulus  (see  character  16,  above). 
In  N.  thessa  and  N.  zamboangae,  there  is  a  thin, 
elongate  ossification  just  anterior  to  the  aproctal 
axial  bone  but  separate  from  it.  Aurich  (1937: 
fig.  4)  figured  the  aproctal  axial  bone  and  anterior 
aproctal  ossification  as  one  long  bone.  A  similar 
bone  in  N.  bicornis  was  called  the  "pulvinular 
osselet"  by  Roberts  (1971b:fig.  12).  However, 
because  of  differences  in  size  and  shape  of  the 
bones,  I  do  not  consider  that  minute  ossification 
in  N.  bicornis  to  be  homologous  with  the  ossi- 
fication in  N.  thessa  and  N.  zamboangae. 

2 1 .  Reduction  in  number  of  papillary  projec- 
tions. 

The  thin,  elongate  papillary  and  penial  bones 
of  the  borneensis  group  (character  18,  above)  are 
even  further  reduced  within  N.  thessa  and  N. 
zamboangae  so  that  the  fine,  bony  projections 
of  the  papillary  bone  (Fig.  3)  are  few  or  absent. 

Within  the  bicornis  group,  I  recognize  a  tri- 
chotomy among  N.  bicornis  and  two  sister-group 
pairs,  N.  palawanensis  and  N.  robertsi,  and  N. 
villadolidi  and  N.  amaricola. 

Neostethus  palawanensis  and  N.  robertsi  share 
one  derived  character. 

22.  Claw-shaped,  anteriorly  bifurcated,  sec- 
ond ctenactinium  (Fig.  1 3). 

The  anterior  extent  of  the  second  ctenactinium 
in  all  other  phallostethids  that  have  the  bone  is 
either  blunt  (Fig.  3)  or  sharply  pointed  (Fig.  9). 
Hence,  the  claw-shaped,  anteriorly  bifurcated 
second  ctenactinium  of  N.  palawanensis  and  N. 
robertsi  (Fig.  1 3),  much  larger  and  better  devel- 
oped in  the  latter  species,  is  judged  to  be  uniquely 
derived  within  phallostethids. 

Neostethus  villadolidi  and  N.  amaricola  share 
one  derived  character. 

2  3 .  Sharply  pointed  second  ctenactinium  (Fig. 
9). 

Neostethus  villadolidi  and  N.  amaricola  are  two 
small,  narrow-bodied  phallostethid  species  dis- 


FIGURE  9.  Diagrammatic  representation  of  head  and  an- 
terior portion  of  body,  a.  Neostethus  villadolidi,  sinistral  male, 
CAS  60258,  b.  Neostethus  amaricola,  sinistral  male,  CAS-SU 
38898.  Anterior  to  left.  Arrow  points  to  second  ctenactinium. 


tinguished  from  all  other  phallostethids  by  their 
sharply  pointed,  either  anteriorly  or  dorsally  di- 
rected, second  ctenactinia. 

Autapomorphies  of  Each  Species  of  Neostethus 
Neostethus  lankesteri 

24.  Males  with  a  small,  slightly  pointed,  tri- 
angular, ventral  bony  projection  on  the  elongate 
first  ctenactinium  (Fig.  3). 

The  first  ctenactinium  is  an  elongate,  uniform- 
ly rod-shaped  bone  in  all  other  phallostethids  in 
which  it  occurs.  Neostethus  lankesteri  is  unique 
among  phallostethids  in  having  a  small,  slightly 
pointed,  triangular,  bony  projection  on  the  ven- 
tral surface  of  the  first  ctenactinium  (Fig.  3). 

25.  Distal  tips  of  the  first  pleural  ribs  project 
beyond  ventral  outline  of  priapium. 


260 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 1 


FIGURE  10.  Diagrammatic  representation  of  head  and  an- 
terior portion  of  body,  adult  female,  a.  Neostethus  thessa  (CAS- 
SU  36546),  b.  Neostethus  villadolidi(CAS  60258),  c.  Neostethus 
lankesteri  (CAS-SU  67 1 62),  d.  Neostethus  bicornis  (CAS  35783). 
Abbreviations:  a,  anus;  u,  urogenital  opening. 


The  first  pleural  ribs  in  all  phallostethid  males 
are  offset  to  the  proctal  side  of  the  priapium  (Fig. 
3).  In  TV.  lankesteri,  the  distal  tips  of  the  ribs 
project  beyond  the  ventral  outline  of  the  pria- 
pium when  it  is  viewed  from  the  proctal  side. 

26.  Females  have  a  large,  hooklike,  urogenital 
papilla  (Fig.  lOc). 

In  all  female  phallostethids,  the  anus  is  ante- 


rior to  the  urogenital  openings  (Fig.  10).  Most 
species  are  like  N.  thessa  (Fig.  lOa),  in  which 
females  have  no  anal  or  urogenital  papilla  or 
covering.  Adult  female  N.  lankesteri  are  readily 
distinguished  from  all  other  phallostethid  fe- 
males by  a  large,  hooklike  papilla  that  lies  pos- 
terior to  the  anal  opening  and  anterior  to  the 
urogenital  openings,  towards  which  it  curves  (Fig. 
lOc). 

Neostethus  bicornis 

21.  Mature  males  with  two,  elongate  ctenac- 
tinia. 

The  second  ctenactinium  of  TV.  bicornis  is  the 
longest  of  any  phallostethid,  at  least  one-half  the 
length  of  the  first  ctenactinium  in  adult  males 
(Roberts  1971b:fig.  12). 

28.  Immature  males  with  a  brown  blotch  on 
the  proctal  pelvic  rays  that  fades  with  growth  and 
maturity. 

Immature  males  can  be  readily  identified  as 
sinistral  or  dextral  (see  Parenti  1986a  and  char- 
acter 44,  below)  by  a  small  brown  blotch  present 
on  the  proctal  pelvic  fin  rays.  This  blotch  be- 
comes diffuse,  hence  less  recognizable,  with 
growth. 

29.  Females  with  a  fleshy,  hoodlike  flap  cov- 
ering the  urogenital  openings  (Fig.  1  Od). 

Adult  female  TV.  bicornis  are  readily  distin- 
guished from  all  other  phallostethids  by  the  pres- 
ence of  a  unique  fleshy,  hoodlike  flap  that  extends 
from  the  anus  to  the  urogenital  openings  (Fig. 
lOd).  Although  females  of  TV.  lankesteri  and  TV. 
bicornis  have  fleshy  coverings  of  the  urogenital 
openings,  I  do  not  recognize  this  as  a  shared 
derived  character  because  form  of  the  urogenital 
covering  is  different  in  the  two  species,  and  nu- 
merous other  characters  refute  a  close  relation- 
ship between  the  two. 

Neostethus  amaricola 

30.  Distal  tips  of  the  first  pleural  ribs  of  fe- 
males project  beyond  the  ventral  body  profile. 

Herre  ( 1 942: 1 48)  used  "Female  with  two  spurs 
or  papillae  on  sides  of  throat"  as  a  key  character 
of  TV.  amaricola.  The  spurs  or  papillae  are  the 
distal  tips  of  the  first  pleural  ribs  that  project 
beyond  the  ventral  body  profile  on  either  side  of 
the  urogenital  openings. 

31.  A  sharply  pointed,  anteriorly  directed, 
second  ctenactinium  (Fig.  9b). 

Males  of  TV.   amaricola  are  readily  distin- 


PARENTI:  PHALLOSTETHID  FISHES 


261 


guished  from  those  of  its  sister  species,  N.  vil- 
ladolidi, by  a  sharply  pointed  second  ctenactin- 
ium  that  is  directed  anteriorly  (Fig.  9). 

Neostethus  borneensis 

Neostethus  borneensis  is  a  relatively  short- 
bodied  species  that  is  distinguished  from  all  oth- 
er neostethins  by  two,  probably  correlated,  char- 
acters (see  G.  panayensis,  below,  for  a  similar 
definition). 

32.  Low  number  of  vertebrae  (31-32)  (Table 
2)  in  males  and  females. 

Vertebrae  in  other  species  of  Neostethus  range 
from  34  to  37  (Table  2),  in  Gulaphallus  from  35 
to  37,  except  for  G.  panayensis  in  which  the  num- 
ber of  vertebrae  is  also  reduced,  and  in  phallo- 
stethins  from  33  to  40  (Table  2).  Vertebrae  num- 
ber 40—43  in  Dentatherina  merceri  (fide  Patten 
and  Ivantsoff  1983:332),  and  29-30  in  Pseudo- 
mugil  signifer  (CAS  62565),  two  outgroup  species. 
The  plesiomorphic  vertebral  number  for  phal- 
lostethids  is  ambiguous.  I  treat  the  vertebral 
number  in  N,  borneensis  as  derived  because  it  is 
low  compared  to  all  other  neostethins  and  phal- 
lostethins. 

33.  Low  number  of  scales  in  a  lateral  series 
(25-27)  (Table  2)  in  males  and  females. 

Scales  in  a  lateral  series  in  other  species  of 
Neostethus  range  from  29  to  34,  and  in  Gula- 
phallus from  28  to  58  (Table  2).  Scales  in  a  lateral 
series  number  40-43  in  Dentatherina  merceri  (fide 
Patten  and  Ivantsoff  1983:332),  and  29-32  in 
Pseudomugil  signifer  (CAS  62565),  two  outgroup 
species.  The  plesiomorphic  number  of  scales  in 
a  lateral  series  in  phallostethids  is  ambiguous.  I 
hypothesize  that  the  number  in  TV.  borneensis  is 
derived  because  it  is  low  compared  to  outgroup 
species  and  other  phallostethids. 

Neostethus  villadolidi 

34.  A  fleshy  urogenital  protuberance  (Fig.  lOb) 
in  females. 

Adult  female  N.  villadolidi  have  a  fleshy  pro- 
tuberance just  posterior  to  the  anus  and  anterior 
to  the  urogenital  openings.  It  could  be  described 
as  an  incipient  or  rudimentary  papilla;  however, 
I  do  not  recognize  a  transition  series  from  the 
state  of  the  urogenital  papilla  in  TV.  lankesteri  or 
the  urogenital  flap  in  TV.  bicornis,  as  discussed 
above  for  character  29. 

35.  A  sharply  pointed,  dorsally  directed  sec- 
ond ctenactinium  (Fig.  9a). 


Male  specimens  of  TV.  villadolidi  are  readily 
distinguished  from  male  specimens  of  TV.  amari- 
cola,  its  proposed  sister  species,  by  a  sharply 
pointed  second  ctenactinium  that  is  directed  dor- 
sally (Fig.  9). 

Neostethus  zamboangae 

36.  A  small,  hooked,  second  ctenactinium  with 
two  to  seven  bony  serrations. 

Neostethus  zamboangae  adult  males  have  two 
to  seven  bony  serrations  on  the  ventral  surface 
of  the  hooked  second  ctenactinium.  Regan  (1913: 
figs.  1,  3)  illustrated  serrations  on  what  I  identify 
as  the  second  ctenactinium  in  Phallostethus 
dunckeri;  however,  I  do  not  propose  a  close  re- 
lationship between  N.  zamboangae  and  P. 
dunckeri  because  other  characters  lead  me  to 
conclude  that  they  are  members  of  different, 
monophyletic  tribes. 

Neostethus  thessa 

37.  Absence  of  first  dorsal  fin  rays  and  most 
of  the  first  dorsal  pterygiophore  in  males  and 
females. 

The  first  dorsal  fin  is  absent  in  three  phallo- 
stethid  species:  N.  thessa,  Phallostethus  dunckeri 
(Parenti  1986a),  and  Gulaphallus  bikolanus.  I 
view  this  absence  as  the  individual  loss  of  the 
first  dorsal  fin  in  each  species.  A  first  dorsal  fin 
is  plesiomorphic  for  atherinomorphs,  being  pres- 
ent in  outgroups  Bedotia,  Melanotaenia,  Hy- 
poatherina,  Dentatherina,  and  Pseudomugil  and 
all  other  phallostethids.  Each  of  the  three  phal- 
lostethid  species  that  lacks  a  first  dorsal  fin  is  a 
member  of  its  own  well-corroborated  monophy- 
letic group,  and  a  remnant  of  the  first  dorsal  pte- 
rygiophore is  present  in  some  specimens. 

38.  Oral  dentition  reduced  to  several  small 
unicuspid  teeth  on  premaxilla  and  dentary  in 
males  and  females. 

Oral  teeth,  in  one  or  two  rows  in  other  Neo- 
stethus species,  are  reduced  to  a  few  randomly 
distributed  unicuspid  teeth  in  TV.  thessa. 

39.  A  relatively  long  anal  fin  ( 1 7-22  rays)  (Ta- 
ble 2)  in  males  and  females. 

Anal  fin  rays  in  other  Neostethus  and  Gula- 
phallus number  from  1 3  to  18  (Table  2),  in  Phe- 
nacostethus  from  14  to  15  (Table  2).  Phalloste- 
thus dunckeri  is  the  one  other  phallostethid  species 
having  a  relatively  long  anal  fin  with  26-28  rays 
(Table  2).  In  one  outgroup  species,  Dentatherina 


262 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 1 


5O 


151       1 

152      1 

149      1 

154 

1 

146      1 

153 

i 

1 
r 

147 

4                r 

346 
]45 

]44 
143 

P 

c 

1  [ 

1 

14 

•  '3 
1112 


FIGURE  1 1 .  Cladogram  of  relationships  among  species  in 
the  Gulaphallini.  Solid  squares,  synapomorphies;  open  squares, 
symplesiomorphies.  Characters  numbered  as  in  Phylogenetic 
Analysis. 


merceri,  anal  fin  rays  number  from  1 5  to  17  (Pat- 
ten and  Ivantsoff  1983). 

40.  A  relatively  long  second  dorsal  fin  (8-10 
rays)  (Table  2)  in  males  and  females. 

Second  dorsal  fin  rays  in  other  Neostethus  range 
from  5  to  7,  in  Gulaphallus  from  6  to  8,  in  Phena- 
costethus  from  5  to  7,  and  in  Phallostethus,  8- 
10  (Table  2).  Absence  of  the  first  dorsal  fin  and 
increase  in  the  number  of  second  dorsal  fin  rays 
in  both  N.  thessa  and  P.  dunckeri  invites  spec- 
ulation that  the  two  dorsal  fins  have  become  con- 
fluent in  development.  However,  absence  of  the 
first  dorsal  fin  in  Gulaphallus  bikolanus  with  no 
increase  in  number  of  second  dorsal  fin  rays  re- 
futes this  as  a  general  phenomenon  in  phallo- 
stethids. 

Neostethus  palawanensis 

4 1 .  A  relatively  straight  first  ctenactinium  with 
a  broad,  fleshy  ventral  ridge. 

The  first  ctenactinium  is  the  prominent  exter- 
nalized priapial  bone  in  neostethins  and  gula- 
phallins  (see  character  5,  above).  It  is  typically  a 


long,  slender,  curved  bone  (Figs.  3,  4)  except  in 
TV.  palawanensis  in  which  it  is  relatively  straight 
and  has  a  thick,  fleshy  ventral  ridge  (Myers  1 935). 
I  consider  the  fleshy  ventral  ridge  uniquely  de- 
rived in  N.  palawanensis. 

Neostethus  robertsi 

42.  A  large,  well-developed,  claw-shaped 
second  ctenactinium  (Fig.  1 2). 

Neostethus  robertsi  and  N.  palawanensis  are 
considered  sister  species  because  they  share  the 
uniquely  derived  character  of  a  claw-shaped,  an- 
teriorly bifurcated,  second  ctenactinium  (see 
character  22,  above).  The  claw-shaped  second 
ctenactinium  is  much  larger  and  better  devel- 
oped in  N.  robertsi  than  in  N.  palawanensis. 

Neostethus  ctenophorus 

There  are  no  identifiable  autapomorphies  of 
TV.  ctenophorus  (see  Systematic  Accounts). 

Gulaphallini 

Relationships  among  the  five  Gulaphallus 
species  are  depicted  in  Figure  1 1 .  Four  species, 
G.  eximius,  G.  mirabilis,  G.  falcifer,  and  G.  bi- 
kolanus, are  considered  to  be  most  closely  related 
because  they  share  one  derived  character. 

43.  Large,  well-developed  aproctal  axial  bone 
(Fig.  4). 

The  aproctal  axial  bone  is  relatively  large  and 
has  a  well-developed  dorsal  and  ventral  ridge  in 
G.  eximius,  G.  mirabilis,  G.  falcifer,  and  G.  bi- 
kolanus (Fig.  4),  as  opposed  to  the  condition  in 
G.  panayensis  in  which  the  bone  is  thin  and  elon- 
gate. 

Gulaphallus  bikolanus  and  G.  eximius  are  con- 
sidered sister  species  because  they  share  one  de- 
rived character. 

44.  Adult  males  exclusively  dextral. 

All  adult  male  phallostethids  are  bilaterally 
asymmetric:  a  sinistral  male  has  the  seminal  pa- 
pilla offset  to  the  left  side  of  the  body  and  the 
anus  offset  to  the  right  (Parenti  1986a).  The  re- 
verse is  true  for  dextral  males.  Sinistral  and  dex- 
tral males  occur  in  about  equal  numbers  in  all 
phallostethid  species  except  the  sister  species 
Phenacostethus  trewavasae,  which  is  sinistral, 
Phenacostethus posthon,  which  is  dextral,  and  the 
two  species  of  Gulaphallus.  I  (Parenti  1986a)  re- 
viewed the  phenomenon  of  bilateral  asymmetry 


PARENTI:  PHALLOSTETHID  FISHES 


263 


in  phallostethids  and  speculated  that  the  Phen- 
acostethus  sister  species  evolved  from  an  ances- 
tral species  in  which  there  were  both  sinistral  and 
dextral  males  and  concluded  that  such  state- 
ments of  ancestry  are  untestable.  My  hypothesis 
that  G.  eximius  and  G.  bikolanus  are  sister  species 
because  males  of  each  are  exclusively  dextral  does 
not  refute  this  argument  about  speciation.  The 
proposal  requires  that  the  common  ancestor  of 
the  two  Gulaphallus  species  was  exclusively  dex- 
tral. If  that  ancestral  species  had  an  exclusively 
sinistral  sister  species,  it  is  unknown  to  us  now. 
Gulaphallus  mirabilis  and  G.  falcifer  are  con- 
sidered to  be  sister  species  because  they  share 
two  derived  characters. 

45.  Antepleural  element  ossified  in  males  and 
females. 

The  antepleural  cartilage  surrounds  the  distal 
tips  of  the  first  pleural  ribs.  It  is  ossified  in  G. 
mirabilis  and  G.  falcifer  and  cartilaginous  in  oth- 
er species  of  Gulaphallus. 

46.  A  brown  blotch  (concentration  of  mela- 
nophores)  at  midventral  surface  just  anterior  to 
anal  fin  in  males  and  females. 

Parenti  ( 1 986b)  used  this  pigmentation  pattern 
as  an  identifying  character  of  G.  falcifer.  How- 
ever, it  is  found  also  in  G.  mirabilis  and  hence 
used  as  a  derived  character  to  support  a  close 
relationship  between  the  two  species. 

Autapomorphies  of  Each  Species  of  Gulaphallus 
Gulaphallus  eximius 

47.  Relatively  large  body  size  and  scales  in  a 
lateral  series  more  than  50  (range  52-58)  (Table 
2)  in  males  and  females. 

Gulaphallus  eximius  is  the  largest  known  phal- 
lostethid,  reaching  a  recorded  37  mm  SL  (Herre 
1942).  Although  large  size  may  be  considered 
plesiomorphic  for  phallostethids,  G.  eximius  is 
perhaps  best  described  as  secondarily  large.  One 
possible  sister  taxon,  Dentatherina  merceri,  is 
relatively  small,  maturing  at  under  25  mm  SL 
(Patten  and  Ivantsoff  1983).  Correspondingly,  G. 
eximius  has  a  high  number  of  scales  in  a  lateral 
series,  ranging  from  52  to  58,  as  opposed  to  a 
range  of  25-34  for  all  other  neostethins  and 
gulaphallins  except  G.  mirabilis  (see  character 
50,  below).  Also,  outgroup  species,  D.  merceri 
and  Pseudomugil  signifer  have  fewer  scales  than 
does  G.  eximius  (see  character  33,  above). 


48.  A  brown  blotch  (concentration  of  mela- 
nophores)  surrounds  the  anus  and  urogenital 
openings  in  immature  females. 

This  pigmentation  pattern  is  not  found  in  oth- 
er phallostethids  nor  in  outgroup  species  and, 
therefore,  is  considered  uniquely  derived  in  G. 
eximius. 

49.  Rudimentary  pelvic  girdle  represented  by 
two  small  scale-like  discs  of  bone  in  adult  fe- 
males. 

Just  posterior  to  the  urogenital  openings  in 
females  is  a  pair  of  ossified  discs  inferred  to  be 
a  rudimentary  pelvic  girdle.  Similar  pelvic  girdle 
elements  are  found  in  adult  female  Phenacoste- 
thus  posthon  (see  Parenti  1986b:fig.  4). 

Gulaphallus  mirabilis 

50.  Scales  in  lateral  series  range  from  36  to  38 
(Table  2). 

The  high  number  of  scales  in  a  lateral  series 
in  G.  mirabilis  is  judged  to  be  an  independent 
increase  (see  character  47,  above). 

Gulaphallus  falcifer 

5 1 .  Adult  females  with  an  elongate  pelvic  fin 
ray  projecting  beyond  body  profile  on  either  the 
left  or  right  side  of  the  urogenital  opening. 

Parenti  (1986b)  described  development  of  the 
elongate  asymmetric  bony  element  in  females  of 
G.  falcifer,  concluding  that  it  is  homologous  with 
a  pelvic  fin  ray,  not  a  postcleithrum.  Gulaphallus 
falcifer  is  the  only  phallostethid  in  which  bilateral 
asymmetry  has  been  described  in  females. 

Gulaphallus  bikolanus 

52.  First  dorsal  fin  absent  in  males  and  fe- 
males. 

Absence  of  the  first  dorsal  fin  in  G.  bikolanus 
is  hypothesized  to  be  independent  of  its  absence 
in  N.  thessa  and  P.  dunckeri  (see  character  37, 
above)  for  reasons  of  parsimony. 

Gulaphallus  panayensis 

Gulaphallus  panayensis  is  a  relatively  small 
Gulaphallus  species  defined  by  two,  probably 
correlated,  characters  (see  N.  borneensis,  above, 
for  a  similar  definition). 

53.  A  low  number  of  vertebrae  (3 1  -3  3)  (Table 
2)  in  males  and  females. 


264 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 1 


The  low  number  of  vertebrae  in  G.  panayensis 
is  considered  a  reduction  from  a  higher,  plesio- 
morphic  number  for  phallostethids  (see  charac- 
ter 32,  above). 

54.  A  low  number  of  scales  in  a  lateral  series 
(28-30)  (Table  2)  in  males  and  females. 

The  low  number  of  scales  in  a  lateral  series  is 
considered  a  reduction  from  a  higher,  plesio- 
morphic  number  for  phallostethids  (see  charac- 
ter 33,  above). 


KEY  TO  PHALLOSTETHIDAE 

The  following  key  is  based  primarily  on  char- 
acters discussed  in  the  Phylogenetic  Analysis. 
However,  characters  used  to  identify  the  species 
of  Phallostethus  and  Phenacostethus  are  from 
Parenti(1986a). 

1A.  Second  ctenactinium  present  (Figs.  2,  3, 
9);  shieldlike  pulvinulus  present  (Figs.  2, 
9);  outer  pulvinular  bone  robust  (Figs. 
2,  3);  oral  jaws  highly  protrusible;  few  or 
no  teeth  on  fourth  ceratobranchial 
toothplate;  no  perforation  in  skin  of  gu- 
lar  region;  aproctal  axial  bone  does  not 
project  beyond  ventral  body  profile;  no 

dorsal  crest  on  proctal  axial  bone 

Phallostethinae  2A 

1 B.  No  second  ctenactinium;  no  pulvinulus; 
outer  pulvinular  bone  small;  oral  jaws 
not  highly  protrusible;  teeth  on  fourth 
ceratobranchial  toothplate;  perforated 
gular  flap  of  skin  through  which  anterior 
end  of  first  ctenactinium  may  pass  (Fig. 
5);  aproctal  axial  bone  projects  beyond 
ventral  body  profile  (Figs.  4,  5);  small, 
dorsal  crest  posteriorly  on  proctal  axial 
bone  (Fig.  4)  Gulaphallini  1 5  A 

2A.  Large,  fleshy,  smooth  or  ruffled,  seminal 
papilla;  large,  externalized,  curved  tox- 
actinium;  translucent,  membranous 
dome  on  dorsal  surface  of  head;  no  inner 
pulvinular  bone;  no  bony  projections  on 
papillary  bone Phallostethini  3A 

2B.  Seminal  papilla  small;  no  externalized 
toxactinium;  no  translucent,  membra- 
nous dome  on  dorsal  surface  of  head; 
inner  pulvinular  bone  present  (Fig.  3); 
high  number  of  thin,  bony  projections 
on  papillary  bone  (Fig.  3)  or  just  rudi- 
mentary projections Neostethini  6A 

3A.    Anal  fin  rays  26-28;  vertebrae  40;  sec- 


ond ctenactinium  serrated;  oral  jaws 
equal;  no  first  dorsal  fin;  second  dorsal 

fin  rays  8-10  

Phallostethus  dunckeri  (p.  267) 

3B.  Anal  fin  rays  14-15;  vertebrae  33-35; 
second  ctenactinium  greatly  reduced  and 
not  serrated;  lower  oral  jaw  projects  be- 
yond upper  oral  jaw;  first  dorsal  fin  ray 

1 ;  second  dorsal  fin  rays  5-7 

Phenacostethus  4A 

4A.  Distal  end  of  seminal  papilla  ruffled; 
slightly  curved  toxactinium  (Fig.  2),  pe- 

nial  bone  present  

Phenacostethus  smithi  (p.  268) 

4B.  Distal  end  of  seminal  papilla  smooth; 
greatly  curved  toxactinium;  penial  bone 
absent 5A 

5  A.    First  dorsal  fin  origin  posterior  to  base  of 
last  anal  fin  ray;  males  dextral;  eye-lens 
large  —  Phenacostethus  posthon  (p.  268) 

5B.  First  dorsal  fin  origin  opposite  midpoint 
of  anal  fin;  males  sinistral;  eye-lens  small 

to  minute    

Phenacostethus  trewavasae  (p.  268) 

6A.  Second  ctenactinium  small;  small, 
slightly  pointed,  ventral  bony  projection 
on  elongate  first  ctenactinium  (Fig.  3); 
posterior  tips  of  first  pleural  ribs  project 
beyond  body  of  priapium;  females  with 
large,  hooklike  genital  papilla  (Fig.  1  Oc) 
Neostethus  lankesteri  (p.  269) 

6B.  Second  ctenactinium  large;  ventral  mar- 
gin of  first  ctenactinium  smooth,  no 
pointed  projection;  posterior  tips  of  first 
pleural  ribs  do  not  project  beyond  body 
of  priapium;  females  with  urogenital  pa- 
pilla that  may  be  fleshy  but  does  not  form 
distinct  hook 7A 

7A.  Long  aproctal  axial  bone,  reaches  pul- 
vinulus . .  Neostethus  ctenophorus  (p.  272) 

7B.  Short,  bladelike  aproctal  axial  bone,  does 
not  reach  pulvinulus  8A 

8A.  Strongly  curved,  thin,  second  ctenactin- 
ium; papillary  and  penial  bones  curved 
and  elongate,  lying  parallel  to  dorsal  sur- 
face of  second  ctenactinium;  proctal 
transverse  processes  of  fourth  vertebra 
of  males  enlarged  and  oriented  anterior- 
ly   borneensis  group  9A 

8B.  Straight,  stout,  second  ctenactinium; 
papillary  and  penial  bones  short  and 
straight;  proctal  and  aproctal  transverse 
processes  of  fourth  vertebra  of  males  en- 


PARENTI:  PHALLOSTETHID  FISHES 


265 


larged  and  oriented  anteriorly 

bicornis  group  1 1A 

9A.  Anterior  aproctal  ossification  present; 
few  or  no  bony  projections  on  papillary 
bone;  vertebrae  35-37;  scales  in  lateral 
series  29-34 10A 

9B.  No  anterior  aproctal  ossification;  nu- 
merous bony  projections  on  papillary 
bone;  vertebrae  31-32;  scales  in  lateral 

series  25-27 

Neostethus  borneensis  (p.  270) 

1 0 A.  Small,  hooked,  second  ctenactinium  with 
2-7  bony  serrations;  first  dorsal  fin  rays 
2;  second  dorsal  fin  rays  5-6;  complete 
row  of  teeth  on  premaxilla  and  on  den- 

tary;  anal  fin  rays  1 5-1 7 

Neostethus  zamboangae  (p.  271) 

I  OB.   Small,  hooked  second  ctenactinium  with 

no  bony  serrations;  no  first  dorsal  fin; 
second  dorsal  fin  rays  8-10;  several  small 
unicuspid  teeth  on  premaxilla  and  on 

dentary;  anal  fin  rays  1 7-22 

Neostethus  thessa  (p.  272) 

II  A.  Mature  males  with  two  elongate  cte- 

nactinia;  immature  males  with  brown 
blotch  on  proctal  pelvic  rays;  females 
with  fleshy,  hoodlike  flap  covering  uro- 

genital  openings  (Fig.  lOd) 

Neostethus  bicornis  (p.  269) 

11B.  Mature  males  with  one  elongate  cte- 
nactinium; immature  males  with  clear 
pelvic  rays;  urogenital  openings  of  fe- 
males exposed,  not  covered  by  fleshy, 
hoodlike  flap 12A 

1 2 A.  Claw-shaped,  anteriorly  bifurcated,  sec- 
ond ctenactinium 1 3A 

12B.  Sharply  pointed  second  ctenactinium 
(Fig.  9) 14A 

13 A.  Straight  first  ctenactinium  with  broad, 
fleshy,  ventral  ridge;  small,  claw-shaped 

second  ctenactinium 

Neostethus  palawanensis  (p.  271) 

13B.  Slightly  bowed  first  ctenactinium  with 
smooth  ventral  outline,  no  fleshy  ridge; 
large,  claw-shaped  second  ctenactinium 
(Fig.  13) Neostethus  robertsi  (p.  272) 

1 4  A.  Distal  tips  of  first  pleural  ribs  of  females 
project  beyond  ventral  body  profile; 
sharply  pointed,  anteriorly  directed  sec- 
ond ctenactinium  (Fig.  9b);  no  fleshy 
protuberance  anterior  to  urogenital 

openings  of  females 

Neostethus  amaricola  (p.  270) 


1 4B.  Distal  tips  of  first  pleural  ribs  of  females 
do  not  project  beyond  ventral  body  pro- 
file; sharply  pointed,  dorsally  directed 
second  ctenactinium  (Fig.  9a);  fleshy 
protuberance  anterior  to  urogenital 

openings  in  females  (Fig.  lOb) 

Neostethus  villadolidi  (p.  271) 

1 5A.  Thin,  elongate  aproctal  axial  bone;  ver- 
tebrae 31-33;  scales  in  lateral  series  28- 
30 Gulaphallus  panayensis  (p.  275) 

15B.  Large,  well-developed  aproctal  axial 
bone  (Fig.  4);  35-37  vertebrae;  scales  in 
lateral  series  32  or  more 16A 

1 6A.  Males  dextral;  antepleural  element  car- 
tilaginous in  adults;  midventral  surface 
with  tiny,  scattered  melanophores  . .  1 7 A 

16B.  Males  sinistral  or  dextral;  antepleural 
element  ossified  in  adults;  brown  blotch 
at  midventral  surface  just  anterior  to  anal 
fin ISA 

17 A.  Scales  in  lateral  series  52-58;  brown 
blotch  surrounds  anus  and  urogenital 
openings  in  immature  females;  rudi- 
mentary pelvic  girdle  represented  by  two 
small  scalelike  discs  of  bone  in  adult  fe- 
males; first  dorsal  fin  present  with  rays 
1-2  Gulaphallus  eximius  (p.  274) 

1 7B.  Scales  in  lateral  series  32;  anus  and  uro- 
genital openings  in  females  surrounded 
by  tiny,  scattered  melanophores;  no  ru- 
dimentary pelvic  girdle  in  adult  females; 

no  first  dorsal  fin 

Gulaphallus  bikolanus  (p.  275) 

1 8 A.  Scales  in  lateral  series  36-38;  pelvic  rays 
do  not  project  beyond  ventral  body  pro- 
file   Gulaphallus  mirabilis  (p.  274) 

18B.  Scales  in  lateral  series  32;  an  elongate 
pelvic  fin  ray  projecting  beyond  body 
profile  on  either  the  left  or  right  side  of 
urogenital  opening  in  adult  females  . . . 
Gulaphallus  falcifer  (p.  274) 

CLASSIFICATION  OF  PHALLOSTETHIDAE 

The  following  classification  of  phallostethids 
reflects  some  conclusions  of  the  Phylogenetic 
Analysis.  All  higher  taxa  are  monophyletic. 
Species  are  listed  in  order  of  year  of  description. 

Family  Phallostethidae  Regan,  1913 
Subfamily  Phallostethinae  Regan,  1913 
Tribe  Phallostethini  new  status 
Genus  Phallostethus  Regan,  1913 
P.  dunckeri  Regan,  1913 


266 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 1 


Genus  Phenacostethus  Myers,  1 928 

P.  smithi  Myers,  1928 

P.  posthon  Roberts,  1971 

P.  trewavasae  Parent!,  1986 
Tribe  Neostethini  new  status 
Genus  Neostethus  Regan,  1916 

N.  lankesteri  Regan,  1916 

N.  bicornis  Regan,  1916 

N.  palawanensis  (Myers,   1935),  new 
combination 

TV.  amaricola  (Villadolid  and  Manacop, 
1935) 

N.  thessa  (Aurich,  1937),  new  combi- 
nation 

N.   ctenophorus  (Aurich,    1937),  new 
combination 

N.  borneensis  Herre,  1939 

TV.  villadolidi  Herre,  1942 

N.  zamboangae  Herre,  1 942 

N.  robertsi  new  species 
Subfamily  Gulaphallinae  Aurich,  1937 
Tribe  Gulaphallini  new  status 
Genus  Gulaphallus  Herre,  1925 

G.  eximius  Herre,  1925 

G.  mirabilis  Herre,  1925 

G.  bikolanus  (Herre,  1926),  new  com- 
bination 

G.  falcifer  Manacop,  1936 

G.  panayensis  (Herre,  1 942),  new  com- 
bination 

SYSTEMATIC  ACCOUNTS 

Family  Phallostethidae  Regan,  1913 
Type  genus  Phallostethus  Regan,  1913 

Phallostethinae  Regan,  1913  (as  subfamily  of  Cyprinodonti- 

dae) 
Neostethinae  Aurich,  1937  (as  subfamily  of  Phallostethidae, 

type  genus  Neostethus  Regan) 
Gulaphallinae  Aurich,   1937  (misprinted  Gulaphalinae,  as 

subfamily  of  Phallostethidae,  type  genus  Gulaphallus  Herre) 

DIAGNOSIS.— Small  to  minute,  laterally  com- 
pressed, nearly  transparent,  atherinomorph  fish- 
es, largest  size  recorded  37  mm  SL.  Anus  and 
urogenital  openings  anterior,  below  pectoral  fin 
base,  in  both  sexes.  A  median,  abdominal,  slight- 
ly frayed,  fleshy  ridge  or  keel  from  urogenital 
openings  to  anterior  rays  of  anal  fin.  First  dorsal 
fin  with  one  or  two  short  spines  or  thickened 
rays,  or  no  rays  but  pterygiophore  may  be  pres- 
ent. 

Copulatory  organ  of  males  a  priapium.  The 
following  characters  are  of  males,  unless  other- 


wise noted.  Main  bony  priapial  support  the  proc- 
tal  axial  bone.  Main  externalized  bone(s)  a  tox- 
actinium,  a  toxactinium  and  a  ctenactinium,  or 
one  or  two  ctenactinia.  Anterior  ramus  of  clei- 
thrum  elongate  anteriorly  or  not.  Postcleithra  ab- 
sent or  present  as  small  slivers  of  bone  in  both 
sexes.  First  pair  of  pleural  ribs  expanded  antero- 
posteriorly  and  dorsoventrally,  ventral  tips 
meeting  and  enclosed  in  block  of  cartilage  and/ 
or  bone,  the  antepleural  element.  Large,  fibrous 
pad,  the  pulvinulus,  if  present,  covers  point  of 
articulation  of  toxactinium,  proctal  axial  bone, 
and  anterior  extent  of  cleithra.  Fertilization  in- 
ternal, sperm  bundles  passed  to  females  through 
seminal  papilla,  which  may  be  large  or  small, 
smooth  or  ruffled,  and  supported  internally  by 
papillary  and  penial  bones.  Development  ovipa- 
rous; females  lay  fertilized  eggs. 

Males  bilaterally  asymmetric:  anus  and  sem- 
inal papilla  offset  to  opposite  sides  of  body,  proc- 
tal and  aproctal,  respectively.  Prominent  exter- 
nalized priapial  bones  arise  on  left  or  right  side 
of  body.  Proctal,  or  proctal  and  aproctal,  trans- 
verse processes  of  fourth  vertebrae  expanded  and 
oriented  anteriorly  or  not.  Females  bilaterally 
symmetric  except  those  of  Gulaphallus  falcifer, 
in  which  one  pelvic  ray  is  enlarged  on  either  left 
or  right  side  of  body  at  urogenital  openings.  Pel- 
vies  otherwise  absent  or  represented  by  bundle 
of  fin  rays,  called  fringe,  surrounding  epididymis. 
Small  postanal  papillae  may  contain  rudimen- 
tary pelvic  girdles  and  rays  in  adult  female  Phe- 
nacostethus posthon  and  Gulaphallus  eximius. 

Hyobranchial  apparatus  with  reductions  from 
typical  atherinoid  condition;  teeth  on  fourth  cer- 
atobranchial  toothplate  present  or  absent.  Bran- 
chiostegal  rays  4  to  6. 

DESCRIPTION.  — Background  uniformly  pale 
yellow  or  straw-colored  in  alcohol,  with  small, 
dark  melanophores  on  dorsal  surface  of  head  and 
body,  midlateral  intermuscular  septum,  along 
anal  fin  base,  dorsal  and  ventral  midline.  Seminal 
papilla  and  externalized  bones  of  priapium  may 
be  covered  with  scattered  melanophores.  Scales 
cycloid,  relatively  small  to  medium-sized,  and 
deciduous.  Dorsal  scales  with  faint  dark  margins. 
Eye-lens  large  to  minute.  Dorsal  surface  of  head 
with  translucent,  membranous  dome  or  not. 
Mouth  slightly  to  highly  protractile  with  thin  or 
no  meniscus,  cartilaginous  rod,  or  ossified  sub- 
maxillary  element  between  maxilla  and  vomer. 
Unicuspid  teeth  in  single  or  double  row,  with 
large  teeth  on  lateral  ramus  of  premaxilla,  or  few, 


PARENTI:  PHALLOSTETHID  FISHES 


267 


small  teeth.  Second  dorsal  fin  with  5-10  rays. 
Anal  fin  unmodified,  with  13-28  rays.  Pectoral 
fin  with  9-13  rays.  Caudal  fin  emarginate,  form- 
ing incipient  lobes.  Caudal  skeleton  with  two 
epurals,  autogenous  parhypural,  and  dorsal  and 
ventral  hypural  plate.  Vertebrae  31-40. 

REMARKS. -Regan  (19 13,  19 16),  Myers  (1928), 
Bailey  (1 936),  Herre(  1942),  and  Roberts  (1 97  lb) 
listed  both  diagnostic  and  descriptive  characters 
of  phallostethid  fishes.  This  diagnosis,  like  the 
ones  that  follow,  is  phylogenetic,  as  in  Weitzman 
and  Fink  (1985)  for  genera  of  xenurobryconin 
characid  fishes.  Only  characters  or  character  states 
hypothesized  to  be  derived  are  considered  di- 
agnostic of  taxa.  Comparisons  among  characters 
used  to  diagnose  species,  tribes,  and  subfamilies 
are  given  in  the  Phylogenetic  Analysis  and  dif- 
ferential diagnoses. 

Subfamily  Phallostethinae  Regan 

DIAGNOSIS.— Priapium  with  a  reduced  second 
ctenactinium,  and  a  prominent,  shieldlike  pul- 
vinulus  covering  articulation  point  of  robust  tox- 
actinium  and  proctal  axial  bone.  Jaws  of  males 
and  females  highly  protrusible  with  elongate  pre- 
maxillary  ascending  processes  and  distinct  sub- 
maxillary  cartilages  or  bones,  and  a  reduced  hyo- 
branchial  apparatus  with  few  or  no  teeth  on  fourth 
ceratobranchial  toothplate. 

COMPOSITION.— Two  tribes,  Phallostethini  and 
Neostethini,  both  family-group  names  used  for 
the  first  time  at  the  tribal  level. 

Tribe  Phallostethini  new  status 

DIAGNOSIS.— Males  with  a  large,  fleshy  sem- 
inal papilla  that  may  be  smooth  or  ruffled;  tox- 
actinium  relatively  straight  to  curved.  Adults  of 
both  sexes  extremely  small  and  delicate,  with 
deciduous  scales;  a  translucent,  membranous 
dome  on  dorsal  surface  of  head. 

COMPOSITION  AND  DISTRIBUTION.— Two  gen- 
era, Phallostethus  Regan  from  coastal  peninsular 
Malaysia,  and  Phenacostethus  Myers,  from 
coastal  peninsular  Thailand  and  Malaysia  and 
Sarawak,  as  diagnosed  below. 

Genus  Phallostethus  Regan,  1913 

Phallostethus  Regan,  1913:548-549  (type  species:  Phalloste- 
thus dunckeri  Regan,  1 9 1 3  by  monotypy). 

DIAGNOSIS.— As  for  P.  dunckeri  Regan,  1913, 
below. 


COMPOSITION  AND  DISTRIBUTION.— One  species 
from  brackish  water,  mouth  of  Muar  R.,  Johore, 
Malaysia  (see  Roberts  1971a:fig.  1). 

Phallostethus  dunckeri  Regan,  1913 
(Fig.  i) 

Phallostethus  dunckeri  Regan,  1913:548-555,  figs.  1-4  (origi- 
nal description,  mouth  of  Muar  R.,  Johore,  Malay  Penin- 
sula); 1916:1,  16-25,  figs.  12a,  13-15,  pis.  la,  Ilia,  IVa,  b 
(characters,  comparison  with  Neostethus).  Myers  1928:1-6 
(characters,  comparison  with  Phenacostethus);  1937:137 
(characters).  Herre  1939:140  (characters).  Roberts  1971a:l- 
20  (relationships).  Parenti  1984:1-12,  figs.  1-4  (osteology, 
relationships);  1986a:229-235,  tables  1,  2  (relationships, 
characters);  1986c:41  (characters). 

DIFFERENTIAL  DIAGNOSIS.  — Phallostethus 
dunckeri  has  an  anal  fin  longer  (26-28  rays,  as 
opposed  to  1 3-22)  and  vertebral  number  higher 
(40,  as  opposed  to  31-37)  than  any  other  phal- 
lostethid. Males  are  readily  separable  from  those 
of  Phenacostethus  by  a  serrated  second  ctenactin- 
ium. Both  sexes  distinct  from  Phenacostethus  by 
lacking  first  dorsal  fin,  and  a  second  dorsal  fin 
with  8-10  rays,  rather  than  5-7. 

MATERIAL.— Lectotype  (designated  by  Ladiges  et  al.  1958): 
ZMH  193,  dextral  3,  mouth  of  Muar  R.,  Johore,  Malaysia, 
1 902,  G.  Duncker,  coll.  Paralectotypes:  BMNH  1 9 1 3.5.24: 1 8- 
22  (5  spec.:  1  sinistral  <5,  3  9,  1  sex  undet.)  and  ZMH  194-195 
(19  spec.:  2  dextral  S,  14  9,  3  juv.),  collected  with  lectotype. 

REMARKS.— Ladiges  et  al.  (1958)  gave  lecto- 
type and  paratype  (=paralectotype)  status  to  ZMH 
193  and  194,  respectively.  Because  ZMH  195 
and  the  BMNH  specimens  are  of  the  original 
syntypic  series,  they  are  also  paralectotypes.  Pa- 
renti (1 986a)  called  all  these  specimens  syntypes. 


Genus  Phenacostethus  Myers,  1928 

Phenacostethus  Myers,  1928:6  (type  species:  Phenacostethus 
smithi  Myers,  1928  by  original  designation). 

DIFFERENTIAL  DIAGNOSIS.— A  lower  jaw  pro- 
truding beyond  upper  jaw  and  greatly  reduced 
second  ctenactinium  distinguishes  Phenacoste- 
thus from  all  other  phallostethids.  Anal  and  sec- 
ond dorsal  fin  rays  and  vertebrae  are  fewer  than 
in  Phallostethus.  First  dorsal  fin  with  one  ray  as 
opposed  to  first  dorsal  fin  absent  in  Phallostethus. 

COMPOSITION  AND  DISTRIBUTION.—  Phenaco- 
stethus smithi  Myers,  1 928  and  P.  posthon  Rob- 
erts, 197 la,  in  fresh  and  brackish  waters,  Thai- 
land and  peninsular  Malaysia;  and  P.  trewavasae 
Parenti,  1986a,  in  freshwater,  Sarawak,  Malay- 
sian Borneo. 


268 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 1 


Phenacostethus  smithi  Myers,  1 928 

(Figs.  1,  2) 

Neostethus  lankesteri  (non  Regan).  Smith  1927:353-355  (re- 
port of  specimens  from  Bangkok,  Thailand). 

Phenacostethus  smithi  Myers,  1 928:6-8,  figs.  1 ,  2  (original  de- 
scription, Bangkok,  Thailand).  Smith  1929:14  (characters, 
distribution).  Myers  1937:137-138  (characters).  Bailey  1936: 
470_474;  pis.  3,  4  (anatomy,  comparison  with  Gulaphallus). 
TeWinkel  1939:60-68,  figs.  1-5  (anatomy,  comparison  with 
Gulaphallus).  Herre  1939:140,  142  (characters,  listed).  Hubbs 
1944:73  (characters).  Smith  1945:475-476  (ecology).  Rob- 
erts 1971a:4-18,  figs.  1,  2,  4,  7  (mislabelled  as  fig.  6)  (anat- 
omy, relationships).  Parenti  1986a:229-234,  fig.  4a,  tables 
1,  2  (characters,  relationships). 

Phenacostethus  thai  Fowler,  1937:219-220,  figs.  189,  190 
(original  description,  Bangkok,  Thailand). 

DIFFERENTIAL  DIAGNOSIS.  — Unlike  other 
phallostethins,  males  of  Phenacostethus  smithi 
have  a  ruffled,  rather  than  smooth,  distal  end  of 
enlarged  seminal  papilla. 

MATERIAL.— Holotype:  AMNH  9247,  dextral  6,  13.5  mm 
SL,  Bangkok,  Thailand,  Dec.  1926,  H.  M.  Smith,  coll.  Para- 
types:  AMNH  9248  (1  2),  USNM  92979  (1 1  spec.:  4  sinistral 
S,  1  dextral  6,  5  2,  1  sex  undet.),  USNM  202564  (1  2),  USNM 
104417  (1  sinistral  3),  CAS-SU  26520  (8  spec.:  2  sinistral  3,  3 
dextral  6,  2  2,  1  juv.),  ANSP  108764  (ex.  SU  26520)  (2  spec.: 
1  dextral  3,  1  2)  and  USNM  92297  (2  spec.:  1  dextral  3,  1  2), 
collected  with  holotype. 

Holotype  and  paratypes  of  Phenacostethus  thai:  ANSP  51352, 
ANSP  51353-60  (9  spec.:  2  sinistral  3,  2  dextral  3,  5  2),  Bang- 
kok, Thailand,  2  Aug.  1923,  H.  M.  Smith,  coll. 

OTHER  MATERIAL. -THAILAND:  UMMZ  87345  (40  spec.: 
10  sinistral  3,  14  dextral  3,  1 5  2,  1  juv.),  MCZ  47055  (13  spec.: 
6  sinistral  3,  2  dextral  3,  5  2),  USNM  93506  (100+),  USNM 
93507  (50+),  USNM  93508  (36  spec.:  5  sinistral  3,  12  dextral 
3,  17  2,  2  juv.),  USNM  88668  (100+),  USNM  88667  (100+ 
spec.;  4  juv.  cleared  and  stained),  USNM  88659  (3  spec.:  1 
sinistral  3,  1  dextral  3,  1  2),  BMNH  1927.12.29:1-10(31  spec.: 
6  sinistral  3,  4  dextral  3,  102,  11  juv.;  2  cleared  and  stained), 
ZMA  100.652  (35),  and  CAS-SU  35957  (112  spec.:  5  cleared 
and  stained),  Bangkok,  and  MCZ  47299  (18  spec.;  1  sinistral 
3,  1  dextral  3  and  2  2  cleared  and  stained),  Chantaburi  City. 
MALAYSIA:  USNM  266124  (50+),  Trengganu;  USNM 
266125  (3  spec.:  1  sinistral  3,  2  2),  USNM  266126  (15  spec.: 
3  sinistral  3,  5  dextral  3,  7  2),  USNM  266127  (1  2),  Johore. 

REMARKS.  — In  the  original  description  of 
Phenacostethus  smithi,  Myers  (1928:8)  stated  that 
"...  nineteen  paratypes,  1 1  males  and  8  females, 
are  deposited  in  the  American  Museum  of  Nat- 
ural History,  the  United  States  National  Mu- 
seum, and  my  own  collection."  I  assume  that 
Myers's  "own  collection"  was  deposited,  in  part, 
in  the  Stanford  University  (SU)  collection,  now 
CAS-SU  26520  and  ANSP  108764.  However,  I 
have  no  reason  to  exclude  any  of  the  above 
USNM  paratypes  of  Phenacostethus  smithi  from 
type  status. 

In  his  description  of  Phenacostethus  thai,  Fow- 


lers (1937)  did  not  separate  what  he  called  the 
holotype  from  other  type  specimens  and  I  am 
unable  to  separate  that  specimen  from  the  rest 
of  the  type  material.  See  also  Bohlke  (1984: 140). 
DISTRIBUTION.— Coastal  fresh  and  brackish 
waters,  Thailand  and  peninsular  Malaysia  (Rob- 
erts 1971a:fig.  1). 

Phenacostethus  posthon  Roberts,  1971 
(Fig.  i) 

Phenacostethus  posthon  Roberts,  197 la:  12-1 7,  figs.  3,  5,  6 
(mislabelled  as  fig.  7)  (original  description,  Pungah  and  Lan- 
gu  Town,  Thailand).  Parenti  1986a:225-235,  fig.  5,  tables 
1,  2  (characters,  relationships);  1986b:308-309,  figs.  4,  5 
(anatomy). 

DIFFERENTIAL  DIAGNOSIS.— Phenacostethus 
posthon  is  unique  among  phallostethins,  having 
a  first  dorsal  fin  origin  posterior  to  base  of  last 
anal  fin  ray  rather  than  opposite  midpoint  of  anal 
fin  so  that  first  and  second  dorsal  fins  are  rela- 
tively close  together.  Males  are  dextral,  as  in  Gu- 
laphallus eximius  and  G.  bikolanus,  as  opposed 
to  sinistral,  as  in  P.  trewavasae,  or  sinistral  or 
dextral,  as  in  all  other  phallostethids. 

MATERIAL.— Holotype,  MCZ  47300, 3, 16.7  mm  SL,  Khlong 
Kla  Sohm,  about  15  km  SW  of  Pungah  Town,  Pungah  Prov., 
Thailand,  29  June  1970,  T.  R.  Roberts,  coll.  Paratypes:  MCZ 
47301  (2  3  cleared  and  stained),  MCZ  47301 A  (22  2),  MCZ 
4730 IB  (32  3),  same  data  as  holotype,  and  MCZ  47302  (20), 
Khlong  Langu  at  Langu  Town,  48  km  NW  of  Satul  Town, 
Thailand,  26  June  1970,  T.  R.  Roberts,  coll. 

OTHER  MATERIAL. -MALAYSIA:  USNM  229302  (182  spec.: 
82  3,  77  2,  23  juv.;  6  cleared  and  stained),  Muar  R.,  Johore. 

DISTRIBUTION.— Coastal  fresh  and  brackish 
waters,  Thailand  and  peninsular  Malaysia  (Rob- 
erts 1971a:fig.  1). 

Phenacostethus  trewavasae  Parenti,  1986 

(Fig.  1) 

Phenacostethus  trewavasae  Parenti,  1986a:226-229,  figs.  1-3, 
4b,  tables  1,  2  (original  description,  Baram  R.,  Sarawak, 
Malaysian  Borneo). 

DIFFERENTIAL  DIAGNOSIS.— Phenacostethus 
trewavasae  is  unique  among  phallostethids,  hav- 
ing a  minute  eye-lens,  and  males  sinistral  rather 
than  dextral,  as  in  P.  posthon,  Gulaphallus  exi- 
mius and  G.  bikolanus,  or  sinistral  or  dextral,  as 
in  all  other  phallostethids. 

MATERIAL.— Holotype:  ROM41826.3,  14.1  mmSL,Sungei 
Kejin  Tugang,  tributary  of  Sungei  Kejin,  Baram  R.,  Sarawak, 
Malaysia,  3  Aug.  1981,  D.  Watson,  coll.  Paratypes:  ROM  44289 
( 1 9),  ROM  CSS  1 2  (3  spec. :  2  3,  1  2,  cleared  and  stained),  ROM 
41827  (1),  collected  with  holotype.  ROM  41829  (1),  ROM 


PARENTI:  PHALLOSTETHID  FISHES 


269 


4 1830  (13),  CAS  55454  (5  spec.;  1  cleared  and  stained),  BMNH 
1 984.7. 1 2: 1-5  (5),  AMNH  55570  (5),  USNM  267266  (5),  Sun- 
gei  Kejin,  station  at  confluence  of  Kejin  Tugang  and  Kejin  R., 
Baram  R.,  Sarawak,  Malaysia,  1 1  Feb.  1980,  D.  Watson,  coll. 
OTHER  MATERIAL. -MALAYSIA:  ROM  41828  (1 1),  ROM 

44290  (29  spec.:  2  $  and  2  9  cleared  and  stained)  and  ROM 

44291  (50),  Sarawak. 

DISTRIBUTION.— Tributaries  of  Baram  R.,  Sa- 
rawak, Malaysian  Borneo  (Fig.  1). 

Tribe  Neostethini  new  status 

DIAGNOSIS.— Priapium  with  an  inner  pulvin- 
ular  bone,  and  thin  bony  projections  numbering 
80  or  more  on  papillary  bone. 

Genus  Neostethus  Regan,  1916 

Neostethus  Regan,  1916:1,  2  (type  species:  Neostethus  lankes- 
teri  Regan,  1916  by  original  designation,  use  of  "gen.  et  sp. 
n."  for  one  of  two  included  new  species). 

Ptectrostethns  Myers,  1935:5  (type  species:  Plectrostethus  pa- 
lawanensis  Myers,  1935,  by  original  designation). 

Ceratostethus  Myers,  1937:141  (type  species:  Neostethus  bi- 
cornis  Regan,  1 9 1 6,  by  original  designation). 

Solenophallus  Aurich,  1937:264  (unavailable;  no  type  species 
designated  of  the  two  treated  under  genus). 

Ctenophallus  Herre,  1939:144  (type  species:  Solenophallus 
ctenophorus  Aurich,  1937,  by  original  designation). 

Sandakanus  Herre,  1942:151  (type  species:  Neostethus  bor- 
neensis  Herre,  1 939,  by  original  designation,  as  a  subgenus 
of  Neostethus  Regan). 

Solenophallus  Herre  (ex  Aurich),  1953:242  (type  species:  So- 
lenophallus thessa  Aurich,  1 937,  by  original  designation). 

DIAGNOSIS.— As  for  the  tribe  Neostethini, 
above. 

COMPOSITION  AND  DISTRIBUTION.  — Ten 
species:  N.  lankesteri  Regan,  1916,  Thailand, 
peninsular  Malaysia,  Singapore,  and  Borneo;  N. 
bicornis  Regan,  1916,  peninsular  Malaysia,  Sin- 
gapore, Thailand,  Borneo,  and  Palawan,  Phil- 
ippines; N.  palawanensis  (Myers,  1935),  Pala- 
wan, Philippines;  N.  thessa  (Aurich,  1937), 
Mindanao,  Philippines;  N.  ctenophorus  (Aurich, 
1937),  Luzon,  Philippines;  N.  borneensis  Herre, 
1939,  Borneo  and  Coron,  Philippines;  N.  zam- 
boangae  Herre,  1942,  Mindanao  and  Luzon, 
Philippines;  N.  robertsinew  species,  Luzon,  Phil- 
ippines; N.  amaricola  (Villadolid  and  Manacop, 
1 935),  and  N.  villadolidi  Herre,  1 942,  throughout 
Philippines. 

Neostethus  lankesteri  Regan,  1916 

(Figs.  1,3,  10c) 

Neostethus  lankesteri  Regan,  1916:2-14,  figs.  1-10,  12B,  pis. 
IB,  2,  3A,  4C  (original  description,  Muar  R.,  Johore,  pen- 
insular Malaysia,  and  Singapore).  Weber  and  DeBeaufort 


1922:381-383,  fig.  103  (characters).  Myers  1928:8  (charac- 
ters). Smith  1929:13  (distribution).  Myers  1937:139  (listed). 
Herre  1939:143  (listed).  Roberts  1971a:8  (characters).  Ivan- 
tsoff  et  al.  1987:651-652,  figs.  2A-C,  3A  (anatomy,  com- 
parisons with  atherinoids). 

Neostethus  siamensis  Myers,  1937:139-140  (original  descrip- 
tion, Chantabun  R.,  Thailand).  Herre  1 939: 1 43  (listed).  Smith 
1945:477  (compiled).  Roberts  197  la:2-3  (collection  report). 
Parenti  1986c:39^»l,  fig.  1  (osteology,  histology). 

Neostethus  (Neostethus)  lankesteri.  Herre  1942: 149  (characters, 
relationships,  distribution). 

Neostethus  (Neostethus)  siamensis.  Herre  1 942: 1 50  (characters, 
probable  synonym  of  N.  lankesteri  Regan). 

DIFFERENTIAL  DIAGNOSIS.— Neostethus  lan- 
kesteri males  can  be  recognized  by  a  small,  slight- 
ly pointed,  triangular,  ventral  bony  projection  on 
elongate  first  ctenactinium,  and  posterior  tips  of 
first  pleural  ribs  projecting  beyond  body  of  pria- 
pium.  Females  are  unique  among  phallostethids 
having  a  large,  hooklike,  posteriorly  projecting, 
urogenital  papilla. 

MATERIAL.— Lectotype  (here  designated):  BMNH  1937.12.9: 
7,  dextral  3,  26.0  mm  SL,  Muar  R.,  peninsular  Malaysia,  18 
Feb.  1902,  G.  Duncker,  coll.  Paralectotypes:  BMNH  1937.12.9: 
8-10  (3  spec.:  1  dextral  3,  1  sinistral  6,  1  9)  Singapore,  G. 
Duncker,  coll.  Holotype  of  Neostethus  siamensis:  USNM 
102140,  adult  9,  28.7  mm  SL,  Chantabun  R.,  Thailand,  Apr. 
1933,  H.M.  Smith,  coll. 

OTHER  MATERIAL. -SINGAPORE:  BMNH  1970.7.22:71- 
77  (7  spec.:  1  sinistral  3,  1  dextral  3,  5  9),  R.  Tengah;  BMNH 
1970.7.22:78-79  (2  9),  R.  Serangoon;  RMNH  26425  (69  spec.: 
22  sinistral  3,  29  dextral  3,  18  9),  R.  Berih;  CAS-SU  67162 
(105  spec.;  5  cleared  and  stained),  R.  Selatar.  MALAYSIA: 
FMNH  51653  (1  dextral  3),  East  Coast  Residency,  CAS-SU 
61762  (10  spec.:  6  sinistral  3,  2  dextral  3,  2  9),  Sarawak.  BRU- 
NEI: CAS-SU  61763  (1).  INDONESIA:  AMS  1.19355-041  (26 
spec.:  7  sinistral  3,  7  dextral  3,  9  9,  3  juv.),  East  Kalimantan. 
THAILAND:  MCZ  76857  (375  spec.;  6  cleared  and  stained), 
near  Tha  Chalap  on  road  to  Chantaburi,  Chantaburi  Prov.; 
MCZ  47200  (3),  Satul  Prov.,  CAS  63194  (2),  Rayong  Prov.; 
CAS  63195  (17),  Chumporn  Prov. 

DISTRIBUTION.  — Coastal  localities  of  Thai- 
land, peninsular  Malaysia,  Singapore,  and  Bor- 
neo. 


Neostethus  bicornis  Regan,  1916 
(Figs.  1,  10d) 

Neostethus  bicornis  Regan,  1916:14-16,  fig.  11  (original  de- 
scription, Kuala  Langat,  peninsular  Malaysia).  Myers  1928: 
9  (characters).  Herre  and  Myers  1937:70  (ecology,  distri- 
bution). 

Ceratostethus  bicornis  (Regan).  Myers  1937:141-142  (rela- 
tionships, referral  to  new  genus).  Herre  1939:142  (distri- 
bution); 1942:145  (characters,  relationships,  distribution); 
1953:240  (synonymy,  distribution).  Roberts  1971a:6-10 
(characters).  Roberts  1971b:397^409,  figs.  2-12  (osteology, 
functional  anatomy).  Rosen  and  Parenti  1981:20,  fig.  19A 
(dorsal  gill  arch  anatomy). 


270 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 1 


DIFFERENTIAL  DIAGNOSIS.— Mature  males  have 
two  elongate  ctenactinia  rather  than  a  long  first 
ctenactinium  and  short  second  ctenactinium  as 
in  other  Neostethus.  Immature  males  have  a 
brown  blotch  on  proctal  pelvic  fin  rays  that  fades 
with  growth  and  maturity.  Females  have  a  fleshy, 
hoodlike  flap  that  covers  urogenital  openings. 

MATERIAL. -Lectotype  (here  designated):  BMNH  1937.12.9: 
4,  immature  sinistral  3,  20.9  mm  SL,  Kuala  Langat,  peninsular 
Malaysia,  G.  Duncker,  coll.  Paralectotypes:  BMNH  1 937. 12.9: 
5-6  (2  9,  16.7-19.1  mm  SL),  collected  with  lectotype. 

OTHER  MATERIAL. -THAILAND:  MCZ  47304  (142  spec.; 
2  9  cleared  and  stained);  MCZ  76856  (71  spec.;  1  sinistral  $ 
cleared  and  stained);  MCZ  47305  (93  spec.;  1  dextral  3  cleared 
and  stained);  MCZ  52364  (100+);  MCZ  47306  (85  spec.;  1 
dextral  3  and  1  9  cleared  and  stained);  MCZ  6008 1  (70),  near 
Tha  Chalap  on  road  to  Chantaburi,  Chantaburi  Prov.;  CAS 
63 1 96  (2),  Chumporn  Prov.  SINGAPORE:  BMNH  1970.7.22: 
69-70  (2  spec.:  1  dextral  3,  1  9),  R.  Poyen;  BMNH  1970.7.22: 
63-68  (6  spec.:  3  dextral  3,  3  9),  R.  Tengah;  CAS-SU  67161 
(10  spec.:  4  sinistral  3,  5  dextral  3,  1  juv.),  R.  Seletar;  CAS-SU 
35783  (51  spec.;  10  cleared  and  stained);  MNHN  38-14  (4), 
R.  Kranji;  CAS-SU  31 133  (1 1  spec.:  9  sinistral  3,  1  dextral  3, 
1  9);  USNM  102142  (8  spec.:  1  sinistral  3,  4  dextral  3,  3  9). 
BRUNEI:  CAS-SU  61764  (1).  MALAYSIA:  USNM  189012 
(5),  mangrove  swamps  on  island  opposite  Sandakan;  FMNH 
5 1 726-5 1 729  ( 1 73),  East  Coast  Residency,  Kinabatangan  Dist; 
and  private  collection  of  Maurice  Kottelat  (5),  Sungei  Seburut 
near  Matang,  Sarawak.  PHILIPPINES:  USNM  98839  (3)  and 
USNM  150632  (2),  Nakoda  Bay;  USNM  98840  (1),  Malam- 
paya  R.;  USNM  98838  (3),  Palawan.  No  locality:  MCZ  60049 
(50+). 

DISTRIBUTION.— One  of  the  most  widespread 
phallostethids;  coastal  localities  of  Thailand, 
peninsular  Malaysia,  Singapore,  Borneo,  and  Pa- 
lawan, Philippines. 

Neostethus  amaricola 
(Villadolid  and  Manacop,  1935) 
(Figs.  1,  9b) 

Gulaphallus  amaricola  Villadolid  and  Manacop,  1935:194- 
196,  pi.  1  (original  description,  Pasay,  Rizal  Prov.,  Luzon, 
Philippines). 

Neostethus  amaricola  (Villadolid  and  Manacop).  Myers  1937: 
139  (characters,  referral  of  G.  amaricola  to  Neostethus).  Herre 
1939:143  (distribution).  Hubbs  1944:73-74  (characters). 
Herre  1953:241  (distribution). 

Neostethus  (Neostethus)  amaricola.  Herre  1942:149  (charac- 
ters, relationships,  distribution). 

DIFFERENTIAL  DIAGNOSIS.  — Distal  tips  of  first 
pleural  ribs  of  females  of  Neostethus  amaricola 
project  beyond  ventral  body  profile.  Males  may 
be  recognized  by  a  sharply  pointed  second  cte- 
nactinium directly  anteriorly,  rather  than  dor- 
sally  as  in  N.  villadolidi. 


MATERIAL. -PHILIPPINES:  USNM  98835  (6  spec.:  1  si- 
nistral 3,  5  9;  13-20  mm  SL),  Ragay  R.,  Ragay  Gulf;  FMNH 
40771-^40785  (15  spec.:  2  sinistral  3,  5  dextral  3,  6  9,  2  juv.; 
13-21  mm  SL),  Navotas;  CAS-SU  32354  (3  spec.:  2  sinistral 
3,  1  9;  20-23  mm  SL),  Nasugbu;  CAS-SU  32926  (6  spec.:  1 
sinistral  3,  2  dextral  3,  3  9;  23-24  mm  SL),  Manila;  USNM 
98834  (2  spec.:  1  sinistral  3,  1  dextral  3),  Port  San  Vicente; 
USNM  98833  (3  sinistral  3),  Manila  Bay;  CAS-SU  38898  (200+ 
spec.;  1 3  cleared  and  stained),  Navotas;  CAS  50724  (266),  Nipa 
creek  S.  of  Masinloc,  Luzon.  CAS-SU  38899  (190  spec.;  5 
cleared  and  stained),  Mindoro. 

In  the  original  description,  Villadolid  and 
Manacop  (1935:195)  stated  that  there  were  to  be 
at  least  one  male  and  one  female,  as  well  as  sev- 
eral additional  cotypes  collected  in  Pasay,  Rizal 
Prov.,  Luzon,  in  the  zoological  museum  of  the 
College  of  Agriculture,  University  of  the  Phil- 
ippines. These  specimens  are  presumed  lost. 

DISTRIBUTION  AND  REMARKS.— A  Neostethus 
species  widespread  in  coastal,  brackish-water 
habitats  throughout  the  Philippines,  including 
Luzon  and  Mindoro.  Herre  (1942,  1953)  re- 
ported N.  amaricola  from  Leyte;  however,  I  have 
reidentified  two  USNM  lots  from  that  locality  as 
the  closely  related  N.  villadolidi. 


Neostethus  borneensis  Herre,  1939 
(Fig.  i) 

Neostethus  borneensis  Herre,  1939:143  (original  description, 

report  from  around  Sandakan  Bay,  North  Borneo);  1 940b: 

14-17,  pis.  10,  11  (description,  Kabili  R.,  Sandakan  Bay, 

Segaliud  R.,  North  Borneo). 
Neostethus  (Sandakanus)  borneensis.   Herre    1942:151-152 

(characters,  relationships,  distribution);  1953:242  (listed  as 

type  species  of  subgenus  Sandakanus). 
Neostethus  (Sandakanus)  coronensis  Herre,    1942:152-153 

(original  description,  Coron,  Busuanga,  Philippines);  1953: 

242  (distribution). 

Bohlke  (1953)  followed  Herre  (1942)  in  treat- 
ing Herre's  (1940b:14-17)  detailed  discussion, 
which  included  the  heading  "Neostethus  bor- 
neensis, new  species,"  as  the  original  description. 
However,  Herre's  1940b  publication  was  pre- 
dated by  a  compilation  and  brief  statement  of 
identifying  characters  of  phallostethid  species 
(Herre  1 939),  in  which  the  name,  characters,  and 
statement  of  distribution  are  recorded.  This  con- 
stitutes the  original  description  of  TV.  borneensis. 

DIFFERENTIAL  DIAGNOSIS.— Neostethus  bor- 
neensis is  a  robust,  relatively  short-bodied  species 
with  vertebrae  (31-32,  as  opposed  to  34-37),  and 
scales  in  a  lateral  series  (25-27,  as  opposed  to 
29-34),  fewer  than  in  congeners. 


PARENTI:  PHALLOSTETHID  FISHES 


271 


MATERIAL.— Holotype:  CAS-SU  33018,  sinistral<3, 16.5  mm 
SL,  Kabili  R.,  Sandakan  Dist,  North  Borneo,  30  Jan.  1937, 
A.  W.  C.  T.  Herre,  coll.  Paratypes:  CAS-SU  33022  (116  3;  2 
cleared  and  stained),  CAS-SU  33019  (45  2),  FMNH  51537- 
51538  (12  spec.:  6  S  and  6  2).  BMNH  1938.12.1:155-164  (19 
spec.:  8  3,  3  2,  8  juv.),  collected  with  holotype.  CAS-SU  69021 
(7),  CAS-SU  69022  (6  cleared  and  stained),  CAS-SU  33020 
(9)  Segaliud  R.,  North  (Malaysian)  Borneo,  A.  W.  C.  T.  Herre, 
coll.  Holotype  ofNeostethus  coronensis:  CAS-SU  36542,  dex- 
tral  6,  22.5  mm  SL,  Coron,  Busuanga,  Philippines,  28  June 
1 940,  A.  W.  C.  T.  Herre,  coll.  Paratypes  ofN.  coronensis:  CAS- 
SU  36543  (109  spec.:  23  sinistral  3,  25  dextral  3,  52  2,  9  juv.; 
3  cleared  and  stained)  collected  with  holotype. 

OTHER  MATERIAL. -MALAYSIA:  FMNH  51654  (16),  East 
Coast  Residency;  USNM  189015  (23),  mangrove  swamps  on 
island  opposite  Sandakan. 

DISTRIBUTION.— Borneo  and  Coron,  Busuan- 
ga, Philippines. 


Neostethus  villadolidi  Herre,  1 942 
(Figs.  1,  9a,  lOb) 

Neostethus  (Neostethus)  villadolidi  Herre,  1942:150  (original 
description,  Misamis  Oriental,  Mindanao,  and  Negros  Ori- 
ental, Philippines). 

Neostethus  villadolidi.  Herre  1953:241-242  (distribution). 

DIFFERENTIAL  DIAGNOSIS.—  Neostethus  villa- 
dolidi is  a  slender-bodied  species  closely  allied 
with  TV.  amaricola,  but  differing  from  it  and  all 
other  phallostethids  by  having  a  sharply  pointed, 
dorsally  directed,  second  ctenactinium,  and  a 
fleshy  urogenital  protuberance  in  females. 

MATERIAL.— Holotype:  CAS-SU  36537  (misprinted  23637 
in  original  description),  sinistral  S,  22.5  mm  SL,  mangrove 
swamp  next  to  Fishery  Experiment  Station,  Cagayan  de  Misa- 
mis, Misamis  Oriental  Prov.,  Mindanao,  Philippines,  20  Aug. 
1940,  A.  W.  C.  T.  Herre,  coll.  Paratypes:  CAS-SU  36538  (59 
spec.:  14  sinistral  3,  13  dextral  3,  26  2,  6  juv.),  collected  with 
holotype. 

OTHER  MATERIAL. -PHILIPPINES:  CAS-SU  52344  (36 
spec.;  4  cleared  and  stained),  CAS-SU  36566  (19  spec.:  2  si- 
nistral 3,  dextral  3,  1 1  2,  5  juv.),  and  CAS-SU  32355  (111  spec.: 
5  sinistral  3,  3 1  dextral  3,  58  2,  1 7  juv.;  7  cleared  and  stained), 
CAS  60258  (20  spec.:  10  3  and  10  2),  MNHN  38-15  (4),  ROM 
54986  ( 1 58),  ROM  54987  ( 1 3),  ROM  54988  (80),  ROM  50489 
(1),  ROM  51854  (426),  Negros  Oriental;  AMS  1.21938-016 
(39),  Mactan  Island,  Cebu;  CAS  54645  (12  spec.:  6  dextral  3, 
5  2,  1  juv.),  northern  end  of  Cuyo,  Palawan  Prov.;  USNM 
98836  (1  2,  21  mm  SL),  USNM  98837  (2  spec.:  1  2  and  1 
dextral  3),  Leyte. 

Herre  (1 942: 1 50)  mentioned  two  lots  from  Ne- 
gros Oriental  without  catalogue  numbers,  ques- 
tionably to  be  included  in  the  description  of  TV. 
villadolidi.  I  list  here  several  lots  from  Negros 
Oriental;  however,  like  Bohlke  (1953),  I  recog- 
nize none  as  type  material. 


DISTRIBUTION.  — Mindanao,  Cuyo,  Mactan, 
Negros  Oriental  and  Leyte,  Philippines. 

Neostethus  zamboangae  Herre,  1 942 
(Fig.  i) 

Neostethus  (Sandakanus)  zamboangae  Herre,  1942:153-154 
(original  description,  Zamboanga,  Mindanao,  Philippines); 
1953:242  (distribution). 

DIFFERENTIAL  DIAGNOSIS.— Males  of  Neoste- 
thus zamboangae  are  readily  distinguished  from 
congeners  by  two  prominent  ctenactinia:  an  elon- 
gate first  ctenactinium,  and  a  smaller,  hooked 
second  ctenactinium  with  from  two  to  seven  bony 
serrations.  Phallostethus  dunckeri  males  also  have 
a  serrated  second  ctenactinium;  however,  nu- 
merous derived  characters  that  distinguish  phal- 
lostethins  from  neostethins  support  a  conclusion 
that  these  similar  second  ctenactinia  are  inde- 
pendently derived. 

MATERiAL.-Holotype:  CAS-SU  36544,  dextral  3,  22  mm 
SL,  Zamboanga,  Mindanao,  Philippines,  6  Sep.  1940,  A.  W. 
C.  T.  Herre,  coll.  Paratypes:  CAS-SU  36545  (242  spec.:  30 
sinistral  3,  45  dextral  3,  1172,  50  juv.;  4  cleared  and  stained) 
collected  with  holotype. 

OTHER  MATERIAL. -PHILIPPINES:  CAS  50725  (141  spec.; 
4  cleared  and  stained);  UMMZ  211666  (37),  Sorsogon  Bay, 
Buhatan  R.,  5  km  E  of  Sorsogon  Town,  Luzon. 

DISTRIBUTION.  — Coastal  localities  of  Zam- 
boanga, Mindanao  and  Sorsogon,  Luzon,  Phil- 
ippines. 

Neostethus  palawanensis  (Myers,  1935), 
new  combination 

(Fig.  1) 

P lectrostethus palawanensis  Myers,  1935:5-6  (original  descrip- 
tion, mouth  of  the  Caiholo  R.,  Ulugan  Bay,  west  coast  of 
Palawan,  Philippines).  Myers  1937:141  (characters).  Herre 
1 939: 1 44  (distribution);  1 942: 1 55  (characters,  relationships, 
distribution).  Hubbs  1944:71,  73  (characters).  Herre  1953: 
242  (distribution). 

Plectrostethus  palawensis  (misspelling  of  P.  palawanensis  Myers). 
Ivantsoffet  al.  1987:653-654,  fig.  41  (anatomy,  comparisons 
with  atherinoids). 

DIFFERENTIAL  DIAGNOSIS.— Neostethus  pala- 
wanensis is  a  slender-bodied  species  readily  dis- 
tinguished by  a  relatively  straight  first  ctenactin- 
ium with  a  broad,  fleshy  ventral  ridge. 

MATERIAL.— Holotype:  USNM  93421,  sinistral  3,  22.5  mm 
SL,  mouth  of  Caiholo  R.,  Ulugan  Bay,  west  coast  of  Palawan, 
Palawan  Prov.,  Philippines,  29  Dec.  1908,  U.S.S.  "Albatross." 
Paratypes:  USNM  93422  (2,  19  mm  SL,  allotype  =  paratype); 
USNM  93423  (8  spec.:  5  dextral  3, 3  2),  collected  with  holotype. 


272 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 1 


OTHER  MATERIAL.  -  PHILIPPINES:  USNM  93424  (eggs  only) 
collected  with  holotype  and  paratypes  (no  type  status).  CAS 
63200  (17  spec.:  5  sinistral  6,  4  dextral  3,  8  9;  4  cleared  and 
stained),  Cuyo,  Palawan  Prov. 

DISTRIBUTION.— Coastal  localities  of  Palawan 
and  Cuyo,  Palawan  Prov.,  Philippines. 

Neostethus  thessa  (Aurich,  1937), 
new  combination 
(Figs.  1,  10a) 

Solenophallus  thessa  Aurich,  1937:264-272,  table  1,  figs.  1A, 
3-5,  71  (original  description,  Lake  Mainit,  Mindanao,  Phil- 
ippines). Herre  1939:142  (distribution);  1953:242  (valid  de- 
scription of  Solenophallus). 

DIFFERENTIAL  DIAGNOSIS.— Neostethus  thessa 
is  readily  distinguished  from  congeners  by:  no 
dorsal  fin  rays  (remnant  of  first  dorsal  pteryg- 
iophore  present  in  some  specimens);  oral  den- 
tition reduced  to  several  small,  unicuspid  teeth 
on  premaxilla  and  dentary;  anal  fin  with  1 7-22 
rays,  as  opposed  to  13-18;  and  second  dorsal  fin 
with  8-10  rays,  as  opposed  to  5-7. 

MATERIAL. -PHILIPPINES:  CAS-SU  36546  (46  spec.:  9  si- 
nistral 3,  9  dextral  6,  26  9,  2  juv.;  9  cleared  and  stained),  Lake 
Mainit,  Mindanao  (type  locality,  no  type  status). 

This  is  the  only  known  material  of  TV.  thessa. 
Aurich  (1937)  indicated  that  he  based  his  original 
description  on  at  least  one  male  (26  mm  SL)  and 
one  female  (29  mm  SL),  without  giving  place  of 
deposition  of  any  type  or  other  material.  No 
specimens  of  Neostethus  thessa  are  present  in 
collections  of  the  ZMB  (H.  J.  Paepke,  pers. 
comm.)  and  ZMH  (H.  Wilkens,  pers.  comm.). 
However,  I  do  not  designate  a  neotype  from  the 
CAS-SU  material,  even  though  it  is  from  the  type 
locality,  because  Aurich's  material  may  even- 
tually be  located  upon  a  thorough  search  of  pos- 
sible collections,  and,  to  my  knowledge,  there  has 
never  been  confusion  or  concern  among  ichthy- 
ologists over  identification  of  the  species  asso- 
ciated with  the  name  Solenophallus  thessa  Au- 
rich. 

DISTRIBUTION.  — Lake  Mainit,  Mindanao, 
Philippines  (Fig.  1). 

Neostethus  ctenophorus  (Aurich,  1937), 
new  combination 

(Fig.  1) 

Solenophallus  ctenophorus  Aurich,  1937:272-275,  table  l,figs. 
6,  7II-V  (original  description,  Laguna  de  Bay,  Luzon,  Phil- 
ippines). 

Ctenophallus  ctenophorus  (Aurich).  Herre  1939:144  (referral 


to  new  genus);  1942:154-155  (characters,  distribution,  re- 
lationships); 1953:240  (synonymy). 

REMARKS.— Herre  (1942:154)  diagnosed  Cte- 
nophallus as  distinct  ".  . .  from  other  phallos- 
tethids  by  the  presence  of  a  dense  fringe  along 
both  margins  of  the  open  groove  on  the  pria- 
pium."  This  statement  was  extracted  from  Au- 
rich's (1 937)  original  description  based  on  at  least 
one  male  (24  mm  SL)  and  one  female  (20  mm 
SL).  Place  of  deposition  of  type  or  other  material 
was  not  given  by  Aurich.  No  specimens  of  TV. 
ctenophorus  are  present  in  collections  of  the  ZMB 
(H.  J.  Paepke,  pers.  comm.)  or  the  ZMH  (H. 
Wilkens,  pers.  comm.). 

The  "dense  fringe"  is  formed  by  the  numerous 
bony  projections  of  the  papillary  bone,  diagnos- 
tic of  Neostethus.  The  same  character  was  also 
used  by  Myers  (1 935)  to  diagnose  his  new  genus, 
Plectrostethus,  literally  a  "quill  on  the  chest," 
placed  here  in  synonymy  of  Neostethus. 

I  have  examined  no  specimens  of  TV.  cteno- 
phorus; however,  without  this  species,  the  genus 
Neostethus  is  paraphyletic.  Therefore,  I  synon- 
ymize  Ctenophallus  with  Neostethus.  Meristic 
data  summarized  in  Table  2  are  from  Aurich 
(1937). 

DISTRIBUTION.— Laguna  de  Bay,  Luzon,  Phil- 
ippines. 

Neostethus  robertsi  new  species 

(Figs.  1,  12,  13) 

HOLOTYPE.— CAS  50723,  dextral  3,  22.3  mm  SL,  Calasiao 
R.,  about  12  km  N  of  San  Carlos  City,  Pangasinan  Prov., 
Luzon,  Philippines,  16  Mar.  1976,  T.  R.  Roberts,  coll. 

PARATYPES. -CAS  64254  (40  spec.:  12  sinistral  3,  9  dextral 
3,  10  9,  9  juv.  or  sex  undet.,  9.0-23.0  mm  SL;  1  sinistral  3,  1 
dextral  3  and  1  9  cleared  and  stained),  collected  with  holotype. 

DIFFERENTIAL  DIAGNOSIS.— Neostethus  rob- 
ertsi, like  N.  palawanensis,  has  a  second  cten- 
actinium  which  is  bifurcated  distally,  but  distin- 
guished from  that  species  by  having  a  larger, 
better-developed,  claw-shaped  second  ctenactin- 
ium,  and  a  slightly  bowed  second  ctenactinium 
that  lacks  a  fleshy,  ventral  ridge. 

DESCRIPTION.— A  small,  laterally  compressed 
species,  largest  size  recorded  23.0  mm  SL.  No 
vestigial  pelvic  fin  rays  or  bones  in  adult  females; 
males  with  pelvic  and  parts  of  pectoral  fin  mod- 
ified into  priapium  that  is  either  dextral  (holo- 
type and  eight  <5  paratypes)  or  sinistral  ( 1 3  <5  para- 
types).  Prominent  externalized  subcephalic  bone 
a  first  ctenactinium,  arising  on  either  left  or  right 


PARENTI:  PHALLOSTETHID  FISHES 


273 


FIGURE  12.    Neostethus  robertsi  new  species.  Holotype  (CAS  50723),  dextral  male,  22.3  mm  SL. 


side  of  body  and  articulating  with  posterior  ex- 
tent of  left  or  right  (proctal)  axial  bone,  curving 
gently  along  the  left  or  right  side  of  the  head  and 
the  tip  lying  just  ventral  to  lower  jaw.  First  cte- 
nactinium  smooth,  without  fleshy,  ventral  ridge. 
Large,  well-developed,  claw-shaped  second  cte- 
nactinium.  Fibrous  pulvinulus  lateral  to,  and 
covering  articulation  point  of,  inner  pulvinular 
and  proctal  axial  bones.  Pleural  ribs  of  fourth 
vertebra  in  males  expanded  anteroposteriorly, 
their  distal  tips  meeting  on  right  side  of  proctal 
axial  bone  in  sinistral  males  and  left  side  in  dex- 
tral males.  First  pleural  rib  on  third  vertebra  in 
females. 

Ventral  dermal  keel  extending  from  posterior 
extent  of  priapium  in  males  or  urogenital  open- 
ing in  females,  to  anal  fin  origin.  Scales  on  body 
of  moderate  size,  deciduous.  Color  in  alcohol  like 
most  other  phallostethids:  ground  coloration  very 
pale  yellow  or  light  brownish  (straw-colored). 
Dark  brown  melanophores  scattered  on  dorsal 
surface  of  head,  along  midlateral  intermuscular 
septum,  on  operculum  and  priapium,  and  along 
basal  portion  of  anal  fin,  dorsal  and  ventral  mid- 
line.  Dorsal  scales  with  faint  dark  margins.  Eye- 
lens  large. 

Skull  and  jaws  like  those  illustrated  for  Neo- 
stethus bicornis  by  Roberts  (1971b),  with  little 
modification.  Jaw  teeth  small,  conical,  in  a  single 
uneven  row  or  grouped  together  in  indistinct  rows. 

Caudal  skeleton  with  two  epurals,  autogenous 
parhypural,  and  a  dorsal  and  a  ventral  hypural 
plate.  Caudal  fin  rays  broken  in  nearly  all  spec- 
imens examined.  Pectoral  fin  narrow  and  elon- 
gate, with  10-1 1  rays.  Two  dorsal  fins,  the  first 
with  two  spines  or  thickened  rays  supported  by 
a  single  pterygiophore,  the  second  with  6  rays. 
Anal  fin  rays  17-18,  vertebrae  36,  scales  in  a 


lateral  series  30,  branchiostegal  rays  5  (see  Table 
2). 

DISTRIBUTION.  — Pangasinan  Prov.,  Luzon, 
Philippines  (Fig.  1). 

ETYMOLOGY.— The  specific  name,  robertsi,  in 
honor  of  Tyson  R.  Roberts,  collector  of  the  type 
series  and  student  of  phallostethid  fishes. 

Subfamily  Gulaphallinae  Aurich 
Tribe  Gulaphallini  new  status 

DIAGNOSIS.— Adult  males  with  a  perforated 
gular  flap  of  skin  through  which  anterior  end  of 
first  ctenactinium  projects  and  remains  while  at 
rest;  and  an  aproctal  axial  bone  that  projects  be- 
yond ventral  body  profile  with  a  small  postero- 
dorsal  crest. 

Genus  Gulaphallus  Herre,  1925 

Gulaphallus  Herre,  1925:508  (type  species:  Gulaphallus  exi- 
mius  Herre,  1925,  by  subsequent  designation  of  Myers  1928: 
9). 


FIGURE  13.  Diagrammatic  representation  of  head  and  an- 
terior portion  of  body,  Neostethus  robertsi,  dextral  male,  CAS 
50723.  Anterior  to  the  right.  Arrow  points  to  second  ctenac- 
tinium. 


274 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 1 


Mirophallus  Herre,  1926:539  (type  species:  Mirophallus  bi- 
kolanus  Herre,  1926,  by  monotypy). 

Acanthostethus  Herre,  1929:141-142  (type  species:  Gulaphal- 
lusfalcifer  Manacop,  1936,  by  original  designation;  preoc- 
cupied by  Acanthostethus  Smith,  1869  in  Hymenoptera). 

Manacopus  Herre,  1940a:141  (replacement  name  for  Acan- 
thostethus Herre,  1939,  and  therefore  taking  same  type  species: 
Gulaphallus  fakifer  Manacop,  1936). 

DIAGNOSIS.— As  for  the  tribe  Gulaphallini, 
above. 

COMPOSITION  AND  DISTRIBUTION.  — Five 
species,  four  from  freshwater  hill-streams,  Lu- 
zon, G.  eximius  Herre,  1925,  G.  mirabilis  Herre, 
1925,  G.  fakifer  Manacop,  1936,  and  G.  biko- 
lanus  (Herre,  1 926),  and  one  from  coastal  brack- 
ish to  marine  localities,  Panay,  G.  panayensis 
(Herre,  1942),  Philippines. 

Gulaphallus  eximius  Herre,  1925 
(Fig.  i) 

Gulaphallus  eximius  Herre,  1925:509-510,  pis.  1,  2;  figs.  1,  2 
(original  description,  Nueva  Vizcaya  Province,  Luzon,  Phil- 
ippines). Myers  1928:9  (characters).  Herre  1939:142  (char- 
acters, distribution).  Hubbs  1944:73  (characters).  Herre  1953: 
241  (distribution). 

DIFFERENTIAL  DIAGNOSIS.  — Gulaphallus  exi- 
mius, the  largest  phallostethid  species,  reaches 
37  mm  SL,  and  has  more  than  50  (range  52-58) 
scales  in  lateral  series,  as  opposed  to  25-38,  in 
all  other  phallostethids.  A  brown  blotch  sur- 
rounds anus  and  urogenital  openings  in  imma- 
ture females.  Rudimentary  pelvic  girdle  repre- 
sented by  two  small  scale-shaped  discs  of  bone 
in  adult  females.  Males  dextral  as  in  G.  bikola- 
nus,  but  unlike  that  species,  a  first  dorsal  fin  is 
present  in  both  sexes. 

MATERIAL.— Lecto type  (designated  by  BQhlke  1953):  CAS- 
SU  24474,  3  33.5  mm  SL,  creek,  Santa  Fe,  Nueva  Vizcaya 
Prov.,  Luzon,  Philippines,  1 8  May  1 924,  R.  C.  McGregor,  coll. 
Paralectotypes:  CAS-SU  18149  (3  spec.:  1  3,  2  9),  collected 
with  lectotye. 

OTHER  MATERIAL. -PHILIPPINES:  BMNH  1933.3.11.168 
(1  3,  32.2  mm  SL),  FMNH  40593-40596  (4  spec.:  1  3,  2  2,  1 
immature  3),  CAS-SU  38900  (72  spec.;  1 1  cleared  and  stained), 
CAS-SU  36832  (4  spec.:  1  3,  2  2,  1  juv.),  Luzon. 

DISTRIBUTION.  — Mountain  streams,  Luzon, 
Philippines. 

Gulaphallus  mirabilis  Herre,  1925 

(Figs.  1,4,5) 

Gulaphallus  mirabilis  Herre,  1 925:5 1 1 ,  pi.  2,  figs.  3-5  (original 
description,  Ibo  creek,  tributary  of  Angat  R.,  Bulacan  Prov- 
ince, Luzon,  Philippines).  Myers  1928:10  (characters).  Vil- 
ladolid  and  Manacop  1 9  3  5 : 1 96-2 1 7 ,  pis.  1  -4,  figs.  1  -3  (anat- 


omy,  development).  Bailey  1936:457-470,  pis.  1,  2 
(osteology).  Aurich  1937:275-279,  figs.  8, 9  (anatomy).  Myers 
1937:142  (characters).  TeWinkel  1939:63-67  (anatomy,  de- 
velopment). Herre  1939: 142  (characters,  distribution);  1942: 
144-145  (characters,  distribution,  relationships).  Hubbs  1944: 
73  (characters).  Herre  1953:241  (distribution).  Roberts  1971b: 
41 1  (characters).  Grier  et  al.  1980:332,  table  1  (testis  struc- 
ture). Rosen  and  Parenti  1981:20,  fig.  19B  (dorsal  gill  arch 
anatomy). 

DIFFERENTIAL  DIAGNOSIS.  — Gulaphallus  mi- 
rabilis is  readily  distinguished  from  congeners  by 
scales  in  lateral  series  ranging  from  36  to  38  as 
opposed  to  from  28  to  32,  or  more  than  50. 

MATERIAL. -PHILIPPINES:  Luzon:  ANSP  91045  (4  spec.: 
dextral  3,  9,  2  juv.,  12-25  mm  SL),  BMNH  1933.3.11.169- 
186  (18  spec.:  3  dextral  3,  2  sinistral  3,  1 1  9,  2  immature;  1 
dextral  $  and  1  9  cleared  and  stained),  FMNH  76769  (6  spec.: 
1  dextral  S,  5  juv.),  CAS  727  (1),  CAS  728  (1),  CAS  731-735 
(5),  CAS  740-743  (4),  CAS  747-751  (5),  CAS-SU  30720  (165), 
CAS-SU  38903  (64  spec.;  5  cleared  and  stained),  CAS-SU 
35959  (25),  CAS  54938  (2  spec.:  1  sinistral  6,  1  9),  CAS-SU 
35958  (121  spec.:  1 1  sinistral  3,  15  dextral  3,  32  2,  63  juv.), 
FMNH  46946  (9  spec.:  dextral  3,  4  2,  4  juv.),  USNM  94293 
(26  spec.:  8  dextral  3,  4  9,  14  juv.),  USNM  104412  (2),  MCZ 
33904  (10),  MNHN  27-192,  193  (25,  labelled  syntypes,  but 
not  from  type  locality),  MNHN  32-204  (4),  ZMA  100.651  (4 
spec.:  1  dextral  3,  1  9,  2  juv.,  17-29  mm  SL),  Molawin  Creek, 
CAS-SU  38209  (7),  CAS  50721  (279),  UMMZ  211664  (54), 
Olo  Creek,  CAS-SU  38901  (7),  CAS-SU  38902  (1  cleared  and 
stained),  Sison,  and  CAS  50719  (230),  Barabac  Creek,  Pan- 
gasinan  Prov.;  CAS-SU  38904  (43  spec.:  4  sinistral  3,  6  dextral 
3,  24  9,  9  juv.,  13-31  mm  SL;  2  dextral  3  and  2  9  cleared  and 
stained),  FMNH  40623^10627  (4  spec.:  2  sinistral  3, 2  9),  Santa 
Maria  R.;  MCZ  33904  (10),  Los  Banos,  Laguna  Prov.;  USNM 
197728  (31  juv.),  Central  Luzon;  CAS  50720  (23  spec.:  3  si- 
nistral 3,  2  dextral  3,  8  9,  10  juv.),  Porac  R.,  Pampanga  Prov. 

DISTRIBUTION  AND  REMARKS. —One  of  the  best- 
known  and  frequently  collected  phallostethid 
species  because  of  its  abundance  in  Molawin 
Creek,  which  flowed  through  the  campus  of  the 
College  of  Agriculture,  Luzon,  Philippines  (Herre 
1 942).  Known  also  from  other  localities  through- 
out Luzon. 

Gulaphallus  falcifer  Manacop,  1936 

(Fig.  i) 

Gulaphallus  falcifer  Manacop,  1936:375-379,  pis.  1,2  (original 
description,  Barrio  Laput,  Mexico,  Pampanga  Prov.,  and 
Nampicuan,  Nueva  Ecija,  Luzon,  Philippines). 

Acanthostethus  falcifer  (Manacop).  Herre  1939:141-143  (re- 
ferral to  new  genus). 

Manacopus  falcifer  (Manacop).  Herre  1940a:141  (referral  to 
new  genus);  1942:146-147  (characters,  distribution,  rela- 
tionships); 1953:241  (synonymy).  Parenti  1986b:306-309, 
figs.  1-3  (anatomy,  development). 

DIFFERENTIAL  DIAGNOSIS.— Gulaphallus  fal- 
cifer is  the  only  phallostethid  species  in  which 


PARENTI:  PHALLOSTETHID  FISHES 


275 


females  as  well  as  males  exhibit  bilateral  asym- 
metry: adult  females  with  an  elongate  pelvic  fin 
ray,  on  either  left  or  right  pelvic  girdle,  that  pro- 
jects beyond  body  profile  to  left  or  right  of  uro- 
genital  openings. 

MATERIAL. -PHILIPPINES:  Luzon:  CAS-SU  32356  (42 
spec.:  9  sinistral  6,  14  dextral  6,  10  sinistral  2,  3  dextral  2,  6 
juv.;  7  cleared  and  stained),  Barrio  Laput,  Pampanga  Prov. 
(type  locality).  No  locality:  CAS  63 1 98  (2  cleared  and  stained). 

REMARKS.— Manacop  (1936)  lists  the  follow- 
ing type  material:  Holotype:  Philippine  Fish  and 
Game  Administration  31778,  adult  <3,  27.5  mm 
SL,  gourami  pond  at  Barrio  Laput,  Mexico,  Pam- 
panga Prov.,  Luzon,  Philippines,  27  Aug.  1935, 
C.  Sandiko,  coll.  Allotype  (=paratype):  Philip- 
pine Fish  and  Game  Administration  3 1 779  (adult 
9  26.5  mm  SL).  Paratypes:  Philippine  Fish  and 
Game  Administration  31780  (17  spec.:  8  2  and 
9  <5).  This  material  is  presumed  unavailable  for 
study.  Bohlke  (1953)  did  not  designate  a  neotype 
from  the  CAS-SU  material. 

DISTRIBUTION.  — Barrio  Laput,  Pampanga 
Prov.,  Luzon,  Philippines. 

Gulaphallus  bikolanus  (Herre,  1926), 
new  combination 

(Fig.  1) 

Mirophallus  bikolanus  Herre,  1926:540-541,  pi.  3  (original 
description,  Lake  Bato,  Camarines  Sur  Province,  and  Lake 
Lanigay,  Albay  Province,  Luzon,  Philippines).  Myers  1928: 
10-1 1  (characters).  Aurich  1937:279-282,  fig.  10  (anatomy). 
Myers  1937: 142  (characters).  Herre  1939: 142  (distribution); 
1942:142  (characters,  distribution,  relationships).  Hubbs 
1944:73-74  (characters).  Herre  1953:241  (distribution).  Pa- 
renti  1986a:234,  table  2  (bilateral  asymmetry). 

DIFFERENTIAL  DIAGNOSIS.— Males  of  Gula- 
phallus bikolanus  are  dextral  as  in  G.  eximius, 
but  unlike  that  species,  both  males  and  females 
of  G.  bikolanus  lack  a  first  dorsal  fin  and  have 
32  scales  in  a  lateral  series,  as  opposed  to  more 
than  50. 

MATERIAL  AND  REMARKS. — Lectotype  (designated  by  Bohlke 
1953).  CAS-SU  24475,  immature  3,  19.5  mm  SL,  Lake  Bato, 
Camarines  Sur  Prov.,  Luzon,  Philippines,  30  Jan.  1926,  G.  A. 
Lopez,  coll.  Paralectotypes:  CAS-SU  18148  (2  9,  22-23  mm 
SL),  MNHN  27-194  (2,  21.1-23.8  mm  SL),  collected  with 
lectotype.  Bohlke  (1953)  was  apparently  unaware  of  the  MNHN 
specimens,  collected  with  the  syntypes,  which  also  have  para- 
lectotype  status. 

OTHER  MATERIAL. -PHILIPPINES:  AMNH  50592  (24  spec.: 
13  dextral  3,  9  9,  2  juv.),  CAS  50722  (194  spec.:  131  dextral 
3, 55  9, 8  juv.),  Guinobatan  R.;  CAS  53165  (70  spec.:  24  dextral 
3,  37  8,  9  juv.),  UMMZ  211665  (12  spec.:  5  dextral  3,  7  9), 
Cabangan  R.,  Albay  Prov.,  Bikol  Region,  Luzon. 


DISTRIBUTION.  — Camarines  Sur  and  Albay 
provinces,  Luzon,  Philippines. 

Gulaphallus  panayensis  (Herre,  1 942), 
new  combination 

(Fig.  1) 

Neostethus  (Sandakanus)  panayensis  Herre,  1942: 1 53  (original 
description,  Capiz  and  Estancia,  Panay,  Philippines);  1953: 
242  (distribution). 

DIFFERENTIAL  DIAGNOSIS.  — Gulaphallus  pa- 
nayensis is  a  small,  slender  species  known  only 
from  limited  type  material,  with  vertebrae  31- 
33,  as  opposed  to  35-37,  and  scales  in  lateral 
series  28-30,  as  opposed  to  32-58  in  congeners. 
The  largest  male  (holotype)  does  not  have  a  com- 
pletely developed  priapium;  however,  I  con- 
clude, primarily  from  examination  of  cleared  and 
stained  specimens,  that  panayensis  is  a  Gula- 
phallus. 

MATERIAL.— Holotype:  CAS-SU  36539,  dextral  3,  19.0  mm 
SL,  Capiz,  Panay,  Philippines,  3  Aug.  1940,  A.  W.  C.  T.  Herre, 
coll.  Paratypes:  CAS-SU  36540  (22  spec.:  1  sinistral  3, 3  dextral 
3,  18  2;  2  cleared  and  stained)  collected  with  holotype,  and 
CAS-SU  36541  (22  spec.:  5  dextral  3,  159,  2  juv.),  Estancia, 
Panay,  Philippines,  27  July  1940,  A.  W.  C.  T.  Herre,  coll. 

DISTRIBUTION.— Capiz  and  Estancia,  Panay, 
Philippines  (Fig.  1). 


CONCLUSIONS 

1 .  The  Phallostethidae  comprise  a  monophy- 
letic  group  of  1 9  species,  classified  in  four  genera, 
of  three   monophyletic  tribes— Phallostethini, 
Neostethini,  and  Gulaphallini.   Phallostethini 
(containing  Phallostethus  and  Phenacostethus) 
and  Neostethini  (containing  solely  Neostethus) 
are  sister  taxa  comprising  the  subfamily  Phallo- 
stethinae.  Subfamily  Gulaphallinae  is  coexten- 
sive with  tribe  Gulaphallini  and  genus  Gula- 
phallus. 

2.  Of  the  19  recognized  species,  one,  Neo- 
stethus robertsi,  from  Luzon,  Philippines,  is  de- 
scribed as  new.  Neostethus  siamensis  Myers,  1 937 
is  placed  in  synonymy  of  Neostethus  lankesteri 
Regan,  1916;  Neostethus  coronensis  Herre,  1 942 
is  placed  in  synonymy  of  Neostethus  borneensis 
Herre,  1939.  Five  new  combinations  are  pro- 
posed: Plectrostethus  palawanensis  Myers,  1935, 
Solenophallus  thessa  Aurich,  1937  and  S.  cteno- 
phorus  Aurich,  1937,  are  placed  in  Neostethus; 
Mirophallus  bikolanus  Herre,  1 926  and  Neoste- 


276 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  1 1 


thus  panayensis  Herre,  1 942  are  placed  in  Gu- 
laphallus. 

3.  All  known  names  proposed  for  parts  of  the 
bilaterally  asymmetric  priapium  are  listed  and 
defined;  synonyms  and  presumed  homologs  are 
identified.  Derived  characters  of  the  priapium 
provide  the  primary  information  for  species  group 
recognition.  However,  bilateral  asymmetry  in  fe- 
males of  one  species  (G.  falcifer),  elaborate  uro- 
genital  coverings  in  females  of  three  species  (N. 
lankesteri,  N.  borneensis,  and  N.  villadolidi),  and 
ontogenetic  changes  in  pigmentation  pattern  of 
the  anus  and  urogenital  openings  in  females  of 
one  species  (Gulaphallus  eximius)  indicate  that 
the  complex  evolution  of  the  male  priapium  has 
been  paralleled  by  numerous  modifications  in 
the  reproductive  system  of  females. 

ACKNOWLEDGMENTS 

This  study  could  not  have  been  completed 
without  the  generous  initial  support  of  NSF  grant 
BSR  83-15258,  which  is  gratefully  acknowl- 
edged. Additional  support  from  CAS,  including 
the  fine  curatorial  assistance  of  D.  Catania  and 
A.  Snyder,  and  NSF,  through  grant  BSR  87- 
0035 1 ,  facilitated  completion  of  the  project.  Dis- 
cussions with  T.  R.  Roberts  (CAS),  E.  Trewavas 
(BMNH),  and  R.  Winterbottom  (ROM)  contrib- 
uted to  my  understanding  of  phallostethid  anat- 
omy and  distribution.  W.  Ivantsoff  (Macquarie 
University)  kindly  donated  atherinoid  material 
to  CAS  and  discussed  atherinomorph  system- 
atics  with  me.  T.  R.  Roberts  read  and  com- 
mented on  an  earlier  version  of  the  manuscript. 
I  also  benefited  from  access  to  the  Genera  of 
Fishes  database  being  compiled  by  W.  N.  Esch- 
meyer  (CAS).  The  typescript  was  prepared  using 
a  personal  computer  purchased  in  part  with  a 
generous  donation  from  the  San  Francisco 
Aquarium  Society. 

The  following  provided  loans  of  specimens, 
information,  and/or  assistance  during  visits  to 
their  institutions:  M.  N.  Feinberg  and  G.  Nelson 
(AMNH),  J.  Paxton  (AMS),  B.  Chernoff  (ANSP), 
G.  J.  Howes  and  A.  Wheeler  (BMNH),  T.  Grande 
and  D.  J.  Stewart  (FMNH),  K.  E.  Hartel  and  K. 
F.  Liem  (MCZ),  M.-L.  Bauchot  and  M.  De- 
Soutter  (MNHN),  M.  van  Oijen  (RMNH),  E.  J. 
Crossman  and  R.  Winterbottom  (ROM),  R.  R. 
Miller  (UMMZ),  S.  Jewett  and  R.  P.  Vari 
(USNM),  H.  Nijssen  (ZMA),  H.  J.  Paepke  (ZMB), 
H.  Wilkens  (ZMH),  and  M.  Kottelat  (private  col- 
lection). 


Comments  of  several  reviewers  greatly  im- 
proved the  presentation  of  ideas. 

LITERATURE  CITED 

AURICH,  H.  1937.  Die  Phallostethiden  (Unterordnung  Phal- 
lostethoidea  Myers).  Intl.  Rev.  Ges.  Hydrobiol.  Hydrogr. 
34:263-286. 

BAILEY,  R.  J.  1936.  The  osteology  and  relationships  of  the 
phallostethid  fishes.  J.  Morph.  59:453-483. 

BOHLKE,  E.  B.  1984.  Catalog  of  type  specimens  in  the  ich- 
thyological  collection  of  the  Academy  of  Natural  Sciences 
of  Philadelphia.  Acad.  Nat.  Sci.  Philad.,  Spec.  Publ.  14:1- 
246. 

BOHLKE,  J.  1953.  A  catalogue  of  the  type  specimens  of  recent 
fishes  in  the  Natural  History  Museum  of  Stanford  Univer- 
sity. Stanford  Ichthyol.  Bull.  5:1-168. 

COLLETTE,  B.  B.,  G.  E.  McGowEN,  N.  V.  PARIN,  AND  S.  MITO. 
1984.  Beloniformes:  development  and  relationships.  Pp. 
334-354  in  Ontogeny  and  systematics  of  fishes.  H.  G.  Moser, 
W.  J.  Richards,  D.  M.  Cohen,  M.  P.  Fahay,  A.  W.  Kendall, 
Jr.,  and  S.  L.  Richardson,  eds.  Am.  Soc.  Ichthyol.  Herpetol. 
Spec.  Publ.  No.  1. 

DINGERKUS,  G.  AND  L.  D.  UHLER.  1977.  Enzyme  clearing  of 
alcian  blue  stained  whole  small  vertebrates  for  demonstra- 
tion of  cartilage.  Stain  Tech.  52:229-232. 

DUNCKER,  G.  1904.  Die  Fische  der  malayischen  Halbinsel. 
Mitt.  Nat.  Mus.  Hamburg  21:133-207. 

FOWLER,  H.  W.  1937.  Zoological  results  of  the  third  de- 
Schauensee  Siamese  Expedition,  Part  VIII— fishes  obtained 
in  1936.  Proc.  Philad.  Acad.  Nat.  Sci.  89:125-264. 

GRIER,  H.  J.  AND  B.  B.  COLLETTE.  1987.  Unique  spermato- 
zeugmata  in  testes  of  halfbeaks  of  the  genus  Zenarchopterus 
(Teleostei:  Hemiramphidae).  Copeia  1987:300-311. 

GRIER,  H.  J.,  J.  R.  LINTON,  J.  F.  LEATHERLAND,  AND  V.  L.  DE 
VLAMING.  1 980.  Structural  evidence  for  two  different  tes- 
ticular  types  in  teleost  fishes.  Am.  J.  Anat.  159:331-345. 

HENNIG,  W.  1966.  Phylogenetic  systematics.  Univ.  Illinois 
Press,  Urbana,  Illinois.  263  pp. 

HERRE,  A.  W.  C.  T.  1925.  Two  strange  new  fishes  from 
Luzon.  Philipp.  J.  Sci.  27:507-513. 

.  1 926.  Four  new  Philippine  fishes.  Philipp.  J.  Sci.  3 1 : 

533-543. 

.  1939.  The  genera  of  Phallostethidae.  Proc.  Biol.  Soc. 

Wash.  52:139-144. 

.  1940a.  Manacopus,  a  new  name  for  a  genus  of  Phal- 
lostethidae. Copeia  1940:141. 

.  1 940b.  New  species  of  fishes  from  the  Malay  Pen- 
insula and  Borneo.  Bull.  Raffles  Mus.  16:5-26. 

.  1942.  New  and  little  known  phallostethids,  with  keys 

to  the  genera  and  Philippine  species.  Stanford  Ichthyol.  Bull. 
2:137-156. 

.  1953.  Check  list  of  Philippine  fishes.  U.S.  Fish  Wildl. 

Ser.  Res.  Rep.  20. 

HERRE,  A.  W.  C.  T.  AND  G.  S.  MYERS.  1937.  A  contribution 
to  the  ichthyology  of  the  Malay  Peninsula.  Bull.  Raffles  Mus. 
13:5-75. 

HUBBS,  C.  L.  1 944.  Fin  structure  and  relationships  of  the 
phallostethid  fishes.  Copeia  1944:69-79. 

HUBBS,  C.  L.  AND  L.  C.  HUBBS.  1945.  Bilateral  asymmetry 
and  bilateral  variation  in  fishes.  Pap.  Mich.  Acad.  Sci.  Arts 
Let.  30:229-310. 

HUMPHRIES,  C.  J.  AND  L.  R.  PARENTI.  1986.  Cladistic  bio- 
geography.  Oxford  Monographs  on  Biogeography  No.  2. 
Clarendon  Press,  Oxford,  England.  98  pp. 


PARENTI:  PHALLOSTETHID  FISHES 


277 


I VANTSOFF,  W.,  B.  SAID,  AND  A.  WILLIAMS.  1987.  Systematic 
position  of  the  family  Dentatherinidae  in  relationship  to 
Phallostethidae  and  Atherinidae.  Copeia  1987:649-658. 

LADIGES,  W.,  G.  VON  WAHLERT,  AND  E.  MOHR.  1958.  Die 
Typen  und  Typoide  der  Fischsammlung  des  Hamburgischen 
Zoologischen  Staatsinstituts  und  Zoologischen  Museums. 
Mitt.  Hamburg  Zool.  Inst.  56:155-167. 

LEVTTON,  A.  E.,  R.  H.  GIBBS,  JR.,  E.  HEAL,  AND  C.  E.  DAWSON. 
1985.  Standards  in  herpetology  and  ichthyology:  part  I. 
Standard  symbolic  codes  for  institutional  resource  collec- 
tions in  herpetology  and  ichthyology.  Copeia  1985:802-832. 

MADDISON,  W.  P.,  M.  J.  DONOGHUE,  AND  D.  R.  MADDISON. 
1 984.  Outgroup  analysis  and  parsimony.  Syst.  Zool.  33:83- 
103. 

MANACOP,  P.  R.  1 936.  A  new  phallostethid  fish  with  notes 
on  its  early  development.  Philipp.  J.  Sci.  59:375-381. 

MICKEVICH,  M.  F.  1978.  Taxonomic  congruence.  Syst.  Zool. 
27:143-158. 

MYERS,  G.  S.  1928.  The  systematic  position  of  the  phallo- 
stethid fishes,  with  diagnosis  of  a  new  genus  from  Siam.  Am. 
Mus.Novit.  295:1-12. 

.  1935.  A  new  phallostethid  fish  from  Palawan.  Proc. 

Biol.  Soc.  Wash.  48:5-6. 

.   1937.  Notes  on  phallostethid  fishes.  Proc.  U.S.  Natl. 


Mus.  84:137-143. 

NELSON,  G.  1978.  Ontogeny,  phylogeny,  paleontology,  and 
the  biogenetic  law.  Syst.  Zool.  27:324-345. 

PARENTI,  L.  R.  1981.  A  phylogenetic  and  biogeographic  anal- 
ysis of  cyprinodontiform  fishes  (Teleostei,  Atherinomor- 
pha).  Bull.  Am.  Mus.  Nat.  Hist.  168:335-557. 

.  1 984.  On  the  relationships  of  phallostethid  fishes 

(Atherinomorpha),  with  notes  on  the  anatomy  of  Phallo- 
stethusdunckeriReg»n,  1913.  Am.  Mus.  Novit.  2779:1-12. 
-.  1986a.  Bilateral  asymmetry  in  phallostethid  fishes 


(Atherinomorpha),  with  description  of  a  new  species  from 
Sarawak.  Proc.  Calif.  Acad.  Sci.  44:225-236. 

.  1986b.  Homology  of  pelvic  fin  structures  in  female 

phallostethid  fishes  (Atherinomorpha,  Phallostethidae).  Co- 
peia 1986:305-310. 

.  1986c.  The  phylogenetic  significance  of  bone  types 

in  euteleost  fishes.  Zool.  J.  Linn.  Soc.  87:37-51. 

PATTEN,  J.  M.  AND  W.  IVANTSOFF.  1983.  A  new  genus  and 
species  of  atherinid  fish,  Dentatherina  merceri  from  the  west- 
em  Pacific.  Japan.  J.  Ichthyol.  29:329-339. 

PATTERSON,  C.  1982.  Morphological  characters  and  homol- 
ogy.  Pp.  21-74  in  Problems  of  phylogenetic  reconstruction. 
K.  A.  Joysey  and  A.  E.  Friday,  eds.  Academic  Press,  New 
York. 

REGAN,  C.  T.  1913.  Phallostethus  dunckeri,  a  remarkable  new 
cyprinodont  fish  from  Johore.  Ann.  Mag.  Nat.  Hist.  12:548- 
555. 

.  1916.  The  morphology  of  the  cyprinodont  fishes  of 

the  subfamily  Phallostethinae,  with  descriptions  of  a  new 
genus  and  two  new  species.  Proc.  Lond.  Zool.  Soc.  1916:1- 
26. 

ROBERTS,  T.  R.  197 la.  The  fishes  of  the  Malaysian  family 
Phallostethidae  (Atheriniformes).  Breviora  374:1-27. 

.   1971b.  Osteology  of  the  Malaysian  phallostethid  fish 


Ceratostethus  bicornis,  with  a  discussion  of  the  evolution  of 

remarkable  structural  novelties  in  its  jaws  and  external  gen- 

italia.  Bull.  Mus.  Comp.  Zool.  142:393-418. 
ROSEN,  D.  E.   1 964.  The  relationships  and  taxonomic  position 

of  the  halfbeaks,  killifishes,  silversides,  and  their  relatives. 

Bull.  Am.  Must.  Nat.  Hist.  127:217-268. 
ROSEN,  D.  E.  AND  L.  R.  PARENTI.    1981.    Relationships  of 

Oryzias,  and  the  groups  of  atherinomorph  fishes.  Am.  Mus. 

Novit.  2719:1-25. 
SMITH,  H.  M.    1927.  The  fish  Neostethus  in  Siam.  Science  65: 

353-355. 
.    1929.   Notes  on  some  Siamese  fishes.  J.  Siam  Soc. 

Nat.  Hist,  Suppl.  8:11-14. 
.    1945.   The  fresh-water  fishes  of  Siam,  or  Thailand. 


Bull.  U.S.  Natl.  Mus.  188:1-622. 

SPRINGER,  V.  G.  1983.  Tyson  belos,  new  genus  and  species 
of  western  Pacific  fish  (Gobiidae,  Xenisthminae),  with  dis- 
cussions of  gobioid  osteology  and  classification.  Smithson. 
Contrib.  Zool.  390:1-40. 

TAKITA,  T.  AND  K.  NAKAMURA.  1986.  Embryonic  develop- 
ment and  prelarva  of  the  atherinid  fish,  Hypoatherina  blee- 
keri.  Japan.  J.  Ichthyol.  33:57-61. 

TEWINKEL,  L.  E.  1939.  The  internal  anatomy  of  two  phal- 
lostethid fishes.  Biol.  Bull.  76:59-69. 

VILLADOLID,  D.  V.  AND  P.  R.  MANACOP.  (1934)  Issued  1935. 
The  Philippine  Phallostethidae,  a  description  of  a  new  species, 
and  a  report  on  the  biology  of  Gulaphallus  mirabilis  Herre. 
Philipp.  J.  Sci.  55:193-220. 

WEBER,  M.  AND  L.  F.  DEBEAUFORT.  1922.  Phallostethidae. 
Pp.  381-383  in  The  fishes  of  the  Indo-Australian  Archipel- 
ago IV.  Leiden,  The  Netherlands. 

WEITZMAN,  S.  H.  AND  S.  V.  FINK.  1985.  Xenurobryconin 
phylogeny  and  putative  pheromone  pumps  in  glandulocau- 
dine  fishes  (Teleostei:  Characidae).  Smithson.  Contrib.  Zool. 
421:1-121. 

WHITE,  B.  N.,  R.  J.  LAVENBERG,  AND  G.  E.  McGowEN.  1984. 
Atheriniformes:  development  and  relationships.  Pp.  355- 
362  in  Ontogeny  and  systematics  of  fishes.  H.  G.  Moser,  W. 
J.  Richards,  D.  M.  Cohen,  M.  P.  Fahay,  A.  W.  Kendall,  Jr., 
and  S.  L.  Richardson,  eds.  Am.  Soc.  Ichthyol.  Herp.  Spec. 
Publ.  No.  1. 

WHITMORE,  T.  C.,  ED.  1987.  Biogeographical  evolution  of 
the  Malay  Archipelago.  Oxford  Monographs  on  Biogeog- 
raphy  No.  4.  Clarendon  Press,  Oxford,  England.  147  pp. 

WILEY,  E.  O.  1981.  Phylogenetics;  the  theory  and  practice  of 
phylogenetic  systematics.  John  Wiley  and  Sons,  New  York. 
439  pp. 

WINTERBOTTOM,  R.  1 974.  A  descriptive  synonymy  of  the 
striated  muscles  of  the  teleostei.  Proc.  Acad.  Nat.  Sci.  Philad. 
125:225-317. 

WOLTERECK,  R.  1 942a.  Stufen  der  Ontogenese  und  der  Evo- 
lution von  Kopulationsorganen  bei  Neostethiden  (Perce- 
soces,  Teleostei).  Intl.  Rev.  Ges.  Hydrobiol.  Hydrogr.  42: 
253-268. 

.  1942b.  Neue  Organe,  durch  postembryonale 

Umkonstruktion  aus  Fischflossen  entstehend.  Intl.  Rev.  Ges. 
Hydrobiol.  Hydrogr.  42:317-355. 


CALIFORNIA  ACADEMY  OF  SCIENCES 

Golden  Gate  Park 
San  Francisco,  California  94 1 1 8 


PROCEEDINGS 

OF  THE 

CALIFORNIA  ACADEMY  OF  SCIENCES 


Vol.  46,  No.  12,  pp.  279-287,  3  figs. 


September  11,  1990 


NEW  AND  RECONSIDERED  MEXICAN  ACANTHACEAE.  IV. 

By 
Thomas  F.  Daniel 

Department  of  Botany,  California  Academy  of  Sciences,  Golden  Gate  Park, 
San  Francisco,  CA  94118 


ABSTRACT:  Dyschoriste  mcvaughii,  characterized  by  its  diminutive  habit,  red  corollas,  and  inconspicuously 
spurred  thecae,  is  newly  described  from  western  Jalisco.  The  monotypic  genera  Ixtlania  and  Tabascina  are 
considered  to  be  congeneric  with  Justicia  and  the  new  names,  /.  ixtlania  and  J.  tabascina  are  provided.  A 
new  combination,  Schaueria  parviflora,  is  proposed  for  the  species  resulting  from  the  taxonomic  merger  of 
Streblacanthus  parviflorus  and  Schaueria  calycobractea. 

Received  October  10, 1989.  Accepted  January  10, 1990. 


INTRODUCTION 

Ongoing  studies  of  the  more  than  350  species 
of  Acanthaceae  in  Mexico  continue  to  necessitate 
the  description  of  new  and  the  reevaluation  of 
previously  described  taxa.  In  this  report,  a  re- 
markable red-flowered  species  of  Dyschoriste  is 
described  for  the  first  time.  Recent  collections 
and  studies  of  two  monotypic  genera,  Ixtlania 
M.  E.  Jones  and  Tabascina  Baillon,  reveal  them 
to  be  congeneric  with  the  large  and  polymorphic 
genus  Justicia  L.  Schaueria  calycobractea  R.  Hil- 
senbeck  &  D.  Marshall,  a  species  recently  de- 
scribed from  Veracruz,  is  shown  to  be  correctly 
classified  generically  but  conspecific  with  the  pre- 
viously described  Streblacanthus  parviflorus 
Leonard  from  Guatemala  and  thus  in  need  of 
nomenclatural  renovation.  Detailed  descriptions 
are  provided  for  each  of  these  taxa. 

Dyschoriste  mcvaughii  T.  F.  Daniel,  sp.  nov. 

(Figures  1,  2f) 

TYPE.— MEXICO.  Jalisco:  between  Ayutla  and  Mascota  near 
summit  of  pass,  7-8  mi  NW  of  Los  Volcanes,  30  April  1951, 
R.  McVaugh  12187  (Holotype:  MICH!;  isotype:  US!). 


Herba  perennis  usque  ad  1  dm  alta.  Folia  sessilia  vel  sub- 
sessilia  petiolo  usque  ad  2  mm  longo,  elliptica  vel  obovata,  4- 
1 8  mm  longa,  3-7  mm  lata,  1 . 1-4-plo  longiora  quam  latiora. 
Dichasia  1-3-flora,  in  axillis  foliorum  distalium  sessilia  vel 
subsessilia.  Bracteolae  et  bracteolae  secundae  lineares  vel  lin- 
eares-ellipticae  vel  lineares-oblanceolatae,  5.5-1 1  mm  longae, 
1-3  mm  latae.  Calyx  9-12  mm  longus  lobis  tubo  1.4-3-plo 
longioribus.  Corolla  rubra,  34-42  mm  longa.  Stamina  9-14 
mm  longa,  thecae  1 .8-2.4  mm  longae,  basi  inconspicuae  cal- 
caratae  vel  muticae.  Stylus  27-38  mm  longus.  Capsula  7  mm 
longa,  glabra. 

Perennial  herb  from  woody  rhizome  to  1  dm 
tall,  with  numerous  woody  roots.  Stems 
subquadrate  to  quadrate-sulcate,  densely  pubes- 
cent with  straight  to  flexuose  eglandular  tri- 
chomes  0. 1-0.3(-0.5)  mm  long.  Leaves  sessile  or 
subsessile  with  petioles  to  2  mm  long,  blades 
elliptic  to  obovate,  4-18  mm  long,  3-7  mm  wide, 
1.1-4  times  longer  than  wide,  rounded  to  acute 
at  apex,  rounded  to  acute  to  cuneate  at  base, 
surfaces  pubescent  (especially  along  veins)  like 
stems  (although  the  trichomes  tending  to  be  more 
antrorse),  proximal  leaves  reduced  in  size,  1.5- 
3  mm  long.  Inflorescence  of  1-3 -flowered  di- 
chasia  borne  in  axils  of  distal  leaves  forming  a 
terminal  spicate  thyrse,  dichasia  sessile  or  sub- 


[279] 


280 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  12 


FIGURE  1.  Dyschoriste  mcvaughii.  a,  habit,  x  1.2  (McVaugh  12187);  b,  leaf,  x4  (Daniel  &  Bartholomew  4828);  c,  flowering 
node  with  leaf  removed  showing  3-flowered  dichasium,  x3  (McVaugh  12187);  d,  calyx,  x5  (McVaugh  12187);  e,  androecium 
in  opened  corolla,  x5  (Daniel  &  Bartholomew  4828);  f,  distal  portion  of  stamen,  x  12  (McVaugh  12187);  g,  distal  portion  of 
style  and  stigma,  x  15  (Daniel  &  Bartholomew  4828). 


DANIEL:  MEXICAN  ACANTHACEAE 


281 


FIGURE  2.  Scanning  electron  micrographs  of  pollen,  a,  Ruellia  petiolaris  (Nees)  T.  Daniel  (Daniel  &  Breedlove  4930);  b, 
Justicia  tabascina  (Cowan  2860),  polar  view;  c,  Justicia  ixtlania  (Daniel  2070),  equatorial  view;  d,  /.  tabascina  (Cowan  2860), 
equatorial  view  showing  surface  between  2  trema  regions;  e,  J.  tabascina  (Cowan  2860),  equatorial  view  showing  1  trema  region; 
f,  Dyschoriste  mcvaughii  (Daniel  4828),  equatorial  view;  g,  Schaueria  parviflora  (Trigos  198),  equatorial  view;  h,  S.  parviflora 
(Herrera  72),  subequatorial  view;  i,  S.  parviflora  (Contreras  11180),  equatorial  view.  Scale  in  c-i  same  as  in  b. 


282 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  12 


sessile  with  peduncles  to  1.5  mm  long.  Bractlets 
and  secondary  bractlets  linear  to  linear-elliptic 
to  linear-oblanceolate,  5.5-11  mm  long,  1-3  mm 
wide,  pubescent  like  leaves.  Calyx  9-12  mm  long, 
tube  3-5  mm  long,  hyaline  between  veins,  lobes 
subulate,  6-9  mm  long,  1 .4-3  times  longer  than 
tube,  aristate  and  becoming  stiff  apically,  pubes- 
cent like  leaves.  Corolla  red,  34-42  mm  long, 
externally  pubescent  with  flexuose  eglandular  tri- 
chomes,  tube  24-33  mm  long,  slightly  curved 
and  gradually  ampliate  distally  but  not  differ- 
entiated into  a  well-defined  throat,  upper  lip  9- 
1 1  mm  long  with  2  subcirculate  lobes  4.5-6  mm 
long,  4.5-7  mm  wide,  lower  lip  8.5-12  mm  long 
with  3  lobes  obovate-elliptic  to  obovate,  8-10 
mm  long,  5-7  mm  wide.  Stamens  didynamous, 
9-14  mm  long,  the  two  pairs  fused  for  1-2  mm 
at  base,  thecae  1.8-2.4  mm  long,  muticous  or 
with  blunt  to  pointed  appendages  0.05-0.1  mm 
long  at  base,  pollen  prolate,  3-colporate,  inter- 
colpal  regions  multi-striate  with  5-14  pseudo- 
colpi  of  irregular  lengths,  exine  minutely  verru- 
cate.  Style  27-38  mm  long,  pubescent  with 
eglandular  trichomes,  stigma  filiform,  1.3-1.5 
mm  long.  Capsule  ellipsoid,  7  mm  long,  gla- 
brous. Seeds  not  seen. 

DISTRIBUTION  AND  HABITAT.— Central  and 
western  Jalisco  (Fig.  3);  in  disturbed  areas  of  oak 
and  oak-pine  forest  at  elevations  from  1,600  to 
1,900m. 

PHENOLOGY.— Flowering:  March-June. 

Emery  Leonard  annotated  McVaugh's  collec- 
tion as  an  undescribed  species  of  Ruellia  L.  On 
the  basis  of  many  of  its  macromorphological  at- 
tributes, this  species  might  indeed  be  mistaken 
for  Ruellia.  Dyschoriste  Nees  is  usually  easily 
distinguished  from  Ruellia  by  the  presence  of 
conspicuous  pointed  appendages  at  the  base  of 
each  theca  which  are  not  known  among  Amer- 
ican Ruellia.  In  D.  mcvaughii  these  appendages 
are  inconspicuous  or  absent.  Characters  of  the 
pollen  and  calyx  preclude  the  placement  of  this 
species  in  Ruellia,  however.  Pollen  of  Ruellia 
(Fig.  2a)  is  spherical  and  three-porate  (Raj  1961, 
1973).  It  is  best  characterized  by  its  homobro- 
chate-reticulate  exine.  Pollen  of  Dyschoriste  (Fig. 
2f)  is  prolate  and  three-colporate.  The  exine  is 
verrucate  with  minute,  rounded  projections  and 
striate  with  multiple  pseudocolpi  between  the 
colpi.  In  Mexican  Dyschoriste,  the  calyx  tube  is 
often  prominently  angled  and  hyaline  between 
the  major  veins  which  extend  into  the  lobes.  The 


tube  commonly  ruptures  at  maturity  in  these  weak 
regions.  In  most  species  of  Dyschoriste,  the  calyx 
lobes  are  long  attenuate  to  aristate  and  become 
stiff  at  maturity.  Although  there  is  considerable 
variation  in  calyx  form  among  Mexican  species 
of  Ruellia,  none  have  a  partially  hyaline  tube 
with  stiff,  aristate  lobes.  In  features  of  both  pollen 
and  calyx  form,  the  plants  described  here  con- 
form to  other  species  of  Dyschoriste. 

The  reduced  thecal  appendages  of/),  mcvaughii 
are  an  unusual  feature  in  the  genus.  Anthers  of 
McVaugh  12187  vary  from  having  blunt  ap- 
pendages to  lacking  appendages  altogether.  Dan- 
iel and  Bartholomew  4828,  collected  in  the  same 
general  region,  has  anthers  with  the  pointed  ap- 
pendages typical  of  other  species  of  Dyschoriste. 
Elsewhere  in  the  genus,  variation  of  thecal  ap- 
pendages is  evident  among  the  Chiapas  collec- 
tions of  D.  ovata  (Cav.)  Kuntze  cited  by  Daniel 
(1986).  In  these  collections,  thecal  appendages 
vary  from  one  or  more  hairlike  projections  to  a 
single  stout  mucro. 

Kobuski  (1928)  revised  the  American  species 
of  Dyschoriste  and  recognized  2 1  species  as  oc- 
curring in  Mexico,  none  of  which  are  similar  to 
D.  mcvaughii.  Among  the  40  American  species 
he  treated,  only  D.  pringlei  Greenman  from  Ja- 
lisco has  corollas  similar  in  size  to  those  of  D. 
mcvaughii.  Recent  study  of  D.  pringlei  for 
McVaugh's  Flora  Novo-Galiciana  (Daniel,  un- 
published) reveals  that  this  species  differs  from 
D.  mcvaughii  by  having  conspicuously  flattened 
young  stems,  the  dichasia  crowded  at  or  near  the 
stem  apex  resulting  in  a  headlike  floral  cluster, 
and  blue  corollas.  A  closer  relative  is  undoubt- 
edly D.  rubiginosa  Ramamoorthy  &  Wassh.,  a 
species  with  large  reddish  corollas  recently  trans- 
ferred to  Dyschoriste  from  Hygrophila  R.  Br. 
(Ramamoorthy  and  Wasshausen  1985).  These 
two  species  can  be  distinguished  from  all  other 
Mexican  Dyschoriste  by  their  reddish  corollas 
and  they  can  be  distinguished  from  one  another 
by  the  following  couplet: 

Diminutive  perennial  herb  to  1  dm  tall;  distal 
leaves  1 .  1^4  times  longer  than  wide;  flowers 
1-3  in  leaf  axils;  bractlets  straight;  corolla 
34-42  mm  long;  thecal  appendages  absent 
or  up  to  0. 1  mm  long D.  mcvaughii 

Perennial  herb  or  shrub  to  1.5  m  tall;  distal 
leaves  5-15  times  longer  than  wide;  flowers 
more  than  3  (up  to  16)  in  leaf  axils;  bractlets 


DANIEL:  MEXICAN  ACANTHACEAE 


283 


D  DYSCHORISTE  MCVAUGHII 
A  JUSTICIA  IXTLANIA 

•  JUSTICIATABASCINA 

•  SCHAUERIA  PARVIFLORA 


FIGURE  3.    Map  showing  distribution  of  Dyschoriste  mcvaughii,  Justicia  ixtlania,  J.  tabascina,  and  Schaueria  parvijlora. 


usually  conspicuously  curved;  corolla  (20-) 
25-34  mm  long;  thecal  appendages  0.1-0.2 
mm  long D.  rubiginosa 

PARATYPES.— MEXICO.  Jalisco:  between  Ameca  and  Aten- 
guillo,  1 4.6-1 6.2  mi  W  of  Mixtlan,  T.  Daniel  &  B.  Bartholomew 
4828  (CAS);  Chiquilistlan,  M.  Jones  378  (POM,  US);  La  Pal- 
ma,  M.  Jones  s.n.  (POM). 

Justicia  ixtlania  T.  F.  Daniel,  nom.  nov. 
Ixtlania  acicularis  M.  E.  Jones,  Contr.  West.  Bot.  15:151.1 929. 

TYPE.— MEXICO.  Nayarit:  Ixtlan  [del  Rio],  19  February 
1927,  M.  E.  Jones  23534  (Holotype:  POM!,  photo  and  frag- 
ments: US!;  isotype:  MO!).  Not  /.  acicularis  Sw.  (1788). 

Erect  to  reclined  perennial  herb  to  3  dm  tall. 
Stems  subterete  to  subquadrate,  sulcate,  pubes- 
cent with  flexuose  to  recurved  eglandular  tri- 
chomes  0.5-1.5  mm  long,  the  trichomes  evenly 
disposed  or  usually  concentrated  in  2  vertical 
lines,  mature  stems  often  glabrate.  Leaves  short- 
petiolate  with  petioles  2-1 1  mm  long,  blades 
somewhat  coriaceous,  lanceolate  to  ovate  to  el- 
liptic (sometimes  narrowly  so)  to  subcirculate  to 
obovate,  12-85  mm  long,  1 1-47  mm  wide,  1.1- 
5  times  longer  than  wide,  rounded  to  acute  at 


apex,  acute  to  rounded  to  truncate  to  subcordate 
at  base,  surfaces  glabrous  or  pubescent,  margin 
entire  to  subcrenate,  white-callose.  Inflorescence 
of  axillary  and/or  terminal,  densely  bracteate, 
often  clustered  spikes  (or  spikelike  thyrses)  to  5 
cm  long,  rachis  pubescent  with  an  understory  of 
straight  glandular  and  eglandular  trichomes  to 
0. 1  mm  long  and  an  overstory  of  longer  flexuose 
eglandular  trichomes,  flowers  1  per  node.  Bracts 
sometimes  conduplicate,  ovate-lanceolate  to 
lance-subulate,  (3-)5-10  mm  long,  0.8-1.5  mm 
wide,  abaxial  surface  pubescent  with  an  under- 
story of  straight  glandular  and  eglandular  tri- 
chomes to  0. 1  mm  long  and  an  overstory  of  flex- 
uose eglandular  trichomes  0.2-1  mm  long,  margin 
subscarious  and  ciliate.  Bractlets  lanceolate  to 
lance-subulate,  (4-)5.5-9  mm  long,  0.8-1.5  mm 
wide,  pubescent  like  bracts.  Calyx  5-lobed,  6.5- 
9  mm  long,  lobes  divided  nearly  to  base,  lan- 
ceolate to  lance-subulate,  pubescent  like  bracts, 
margins  scarious,  posterior  lobe  reduced  in  size. 
Corolla  entirely  white  to  pink-purplish  with  pur- 
ple markings  on  lower  lip,  12-16  mm  long,  ex- 
ternally pubescent  with  straight  to  flexuose  eglan- 
dular trichomes  to  0.6  mm  long,  tube  8-1 1  mm 


284 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  12 


long,  slightly  ampliate  distally,  upper  lip  3-4  mm 
long,  bilobed  with  rounded  lobes  0.1-0.5  mm 
long  at  apex,  lower  lip  3.5-5  mm  long,  with  3 
subcirculate  to  obovate  lobes  3-4  mm  long,  2.5- 
3.5  mm  wide,  central  lobe  larger  than  lateral  lobes. 
Stamens  3.5-4  mm  long,  thecae  0.6-1  mm  long, 
unequally  inserted,  the  lower  (sometimes  slightly 
larger)  with  a  blunt  calcarate  basal  appendage 
0.2-0.3  mm  long,  pollen  prolate,  2-colporate  with 
each  colpus  flanked  by  2  pseudocolpi,  exine  re- 
ticulate. Style  9-1 1  mm  long,  glabrous,  stigma 
unlobed,  subcirculate,  oblique,  0.2-0.3  mm  long. 
Capsule  short-stipitate,  3-5  mm  long,  stipe  0.5- 
1.3  mm  long,  head  ellipsoid,  2.5-3.7  mm  long, 
externally  pubescent  with  downward  pointing  to 
appressedeglandulartrichomes  0.2-0.3  mm  long 
and  straight  glandular  trichomes  up  to  0. 1  mm 
long  (mostly  near  apex).  Seeds  4  per  capsule, 
somewhat  flattened,  subcordate  to  subcirculate 
to  elliptic  in  outline,  1-1.3  mm  long,  0.8-1 . 1  mm 
wide,  surface  papillose  (n  =  14,  Daniel  and  Bar- 
tholomew 4788). 

DISTRIBUTION  AND  HABITAT.— West-central 
Mexico  (Sinaloa,  Nayarit,  Jalisco,  and  Michoa- 
can) (Fig.  3);  along  streambanks  in  deciduous  or 
subdeciduous  forest  at  elevations  from  150  to 
1,250m. 

PHENOLOGY.— Flowering  and  fruiting:  March- 
April. 

Leonard  (1945)  accepted  Ixtlania  as  a  viable 
genus,  considered  a  probable  relationship  with 
Beloperone  Nees,  and  noted  a  superficial  resem- 
blance to  Chaetothylax  (erroneously  cited  as 
"Chaetochlamys")  umbrosus  Nees  of  South 
America.  Chaetothylax  Nees  differs  from  Ixtla- 
nia by  having  four  (vs.  five)  calyx  lobes,  dimor- 
phic thecae  with  the  lower  theca  reduced  or  ap- 
pearing abortive  and  separated  from  the  upper 
theca  by  about  1  mm  (vs.  homomorphic  thecae 
that  are  in  contact  with  one  another),  and  pu- 
bescent (vs.  glabrous)  seeds. 

In  a  recent  delimitation  and  infrageneric  clas- 
sification of  Justicia,  Graham  (1988)  included 
both  Beloperone  and  Chaetothylax  within  Jus- 
ticia. She  did  not  examine  material  or  discuss 
the  disposition  of  Ixtlania,  however.  On  the  basis 
of  the  diagnostic  characters  of  the  corolla  (i.e., 
rugulate)  and  androecium  (i.e.,  two  stamens  with 
bithecous  anthers,  the  lower  thecae  with  basal 
appendages),  it  is  clear  that  Ixtlania  is  likewise 
congeneric  with  the  large  and  polymorphic  Jus- 


ticia. The  chromosome  number  noted  above  for 
this  species  is  also  the  most  common  number  in 
Justicia  (Daniel  et  al.  1984,  in  press). 

Using  Graham's  (1988)  provisional  key  to  in- 
frageneric taxa  of  Justicia,  J.  ixtlania  appears  to 
be  most  closely  affiliated  with  sect.  Leucoloma 
V.  Graham  from  South  America.  Justicia  ixtla- 
nia differs  from  the  two  species  of  this  section 
by  the  appendages  of  the  lower  thecae  which  are 
more  than  one-fifth  of  the  length  of  the  thecae. 
It  resembles  Graham's  description  of  this  section 
in  most  other  aspects,  however,  including  the 
pollen  form  (Fig.  2c).  In  Justicia  two-colporate, 
four-pseudocolpate  pollen  is  known  only  in  sect. 
leucoloma  and  two  sections  restricted  to  the  Old 
World.  This  type  of  pollen  has  also  been  reported 
in  Mexacanthus  T.  Daniel  (Daniel  1981),  a 
monotypic  genus  of  Odontoneminae  endemic  to 
Mexico. 

Considerable  variation  in  corolla  color  was 
noted  within  living  populations  of  this  species. 
The  specimen  from  Michoacan  differs  from  those 
from  Jalisco,  Nayarit,  and  Sinaloa  by  its  narrow- 
er, lanceolate  to  narrowly  elliptic  to  oblanceolate 
leaves.  In  other  respects  it  is  identical  to  the  more 
northerly  collections. 

ADDITIONAL  SPECIMENS  EXAMINED.— MEXICO.  Jalisco:  vie. 
of  Arroyo  de  Chorillo  (tributary  of  Rio  de  Quimixto),  ca.  14 
km  SW  of  Puerto  Vallarta,  A.  Carter  &  F.  Chisaki  1249  (MEXU, 
MICH,  UC);  Mpio.  Cuautitlan,  1 .5-2  km  W  de  Tequesquitlan, 
R.  Cuevas  G.  1052  (CAS);  along  Hwy.  200  S  of  Puerto  Vallarta, 
7.6  mi  S  of  El  Tuito,  T.  Daniel  2070  (ASU,  CAS,  DUKE,  K, 
MEXU,  MO);  between  Tomatlan  and  Talpa  de  Allende,  ca. 
1 1  mi  N  of  Tomatlan,  T.  Daniel  2086  (ASU);  along  Hwy.  15 
ca.  1  mi  E  of  Nayarit  border,  T.  Daniel  &  B.  Bartholomew 
4788  (CAS);  Mpio.  La  Huerta,  Est.  Biol.  Chamela,  Arroyo 
Chamela,  E.  Lott  967  (ASU,  CAS,  DUKE,  MO);  Mpio.  Te- 
quila, Barranca  de  los  Tanques,  desv.  al  camino  de  San  Martin 
de  las  Canas,  L.  Puga  6114  (IBUG).  Michoacan:  Los  Chorros 
del  Varal,  Los  Reyes,  B.  Guillen  Jimenez  s.n.  (IBUG).  Nayarit: 
30  km  NE  of  Tepic,  2  km  E  of  Poicho titan,  M.  Baker  1104 
(ASU);  La  Bahada,  SE  of  San  Bias,  E.  Lehto  24208  (ASU); 
San  Felipe  to  Arroyo  San  Jose  del  Conde,  Y.  Mexia  1930  (A, 
CAS,  F,  GH,  MICH,  UC,  US);  Potrero  la  Taberna,  km  28  de 
Amatlan  de  Canas,  carr.  Tepic-Barranca  del  Oro,  B.  Ahuet 
Juan  R.  s.n.  (IBUG).  Sinaloa:  along  Hwy.  40  between  Villa 
Union  and  Cd.  Durango,  26  mi  SW  of  Tropic  of  Cancer,  T. 
Daniel  4019  (CAS);  near  Microondas  tower  near  Villa  Union, 
E.  Lehto  24273A  (ASU);  Mpio.  Concordia,  along  Hwy.  40  ca. 
5  mi  NE  of  Concordia,  A.  Sanders  et  al.  4996  (CAS);  Mpio. 
Cosala,  arroyo  NE  de  Guadalupe  de  Los  Reyes,  P.  Tenorio  L. 
et  al.  8345  (CAS). 

Justicia  tabascina  T.  F.  Daniel,  nom.  nov. 

Tabascina  lindenii  Baillon,  Hist.  PL  10:445.  1891.  (as  T.  lin- 
deni)  TYPE.— MEXICO.  Tabasco:  forets  de  Teapa,  October, 


DANIEL:  MEXICAN  ACANTHACEAE 


285 


J.  Linden  s.n.  (Holotype:  P!(  photographs:  CAS,  F).  Not  J. 
/iiwfe«/iHoullet(1870). 

Shrub  to  2.5  m  tall.  Young  stems  subquadrate 
to  quadrate,  evenly  pubescent  with  antrorse 
eglandular  trichomes  to  0.6  mm  long,  internodes 
often  constricted  just  above  nodes.  Leaves  pet- 
iolate  with  petioles  to  33  mm  long,  blades  ovate- 
elliptic  to  elliptic,  36-165  mm  long,  12-83  mm 
wide,  2-3  times  longer  than  wide,  acuminate  at 
apex,  acute  to  subattenuate  at  base,  surfaces  pu- 
bescent with  cauline  type  trichomes  mostly  or 
entirely  restricted  to  major  veins.  Inflorescence 
of  1-4  terminal  racemose  thyrse(s)  to  7.5  cm 
long,  flowers  paired  at  nodes,  sessile  to  short  (to 
1  mm)  pedicellate  in  axil  of  2  bractlets,  the  flow- 
ers and  bractlets  pedunculate  in  axil  of  a  bract 
with  peduncles  to  12  mm  long,  rachis  and  pe- 
duncles evenly  pubescent  with  flexuose  to  an- 
trorse eglandular  trichomes  to  0.6  mm  long. 
Bracts  often  caducous,  linear,  4-7  mm  long,  0.6- 
1.5  mm  wide,  pubescent  like  rachis.  Bractlets 
linear,  2.5-6  mm  long,  0.8-1  mm  wide,  pubes- 
cent like  rachis.  Calyx  5-lobed,  10-16  mm  long, 
tube  1-2  mm  long,  lobes  valvate  or  nearly  so, 
anterior  lobes  often  connivent  along  edges, 
asymmetrically  elliptic  to  lance-elliptic,  8-1 1  mm 
long,  2-3.2  mm  wide,  posterior  lobe  ovate-ellip- 
tic to  elliptic,  10-13  mm  long,  4-5  mm  wide,  all 
lobes  pubescent  with  straight  to  flexuose  eglan- 
dular trichomes  to  0.5  mm  long.  Corolla  yellow, 
24  mm  long,  externally  pubescent  with  glandular 
trichomes  0.05-0.2  mm  long,  tube  13  mm  long, 
upper  lip  1 1  mm  long,  entire,  internally  rugulate, 
lower  lip  10  mm  long  with  3  rounded  lobes  to  3 
mm  long  and  2.5  mm  wide.  Stamens  9  mm  long, 
filaments  glabrous,  thecae  subsagitate,  unequally 
inserted,  5.3  mm  long,  rounded  and  lacking  basal 
appendages,  pollen  prolate,  4-porate  (to  4-sub- 
colporate)  with  pores  in  a  trema  region  contain- 
ing 2  longitudinal  rows  of  circular  insulae,  exine 
reticulate.  Style  15-18  mm  long,  glabrous,  stigma 
more  or  less  capitate,  0.2  mm  long.  Capsule  19- 
22  mm  long,  glabrous,  stipe  8-9  mm  long,  head 
11-13  mm  long.  Seeds  4,  flattened,  subcirculate 
in  outline,  4.2-5  mm  long,  3.8-4.8  mm  wide,  the 
surfaces  smooth. 

DISTRIBUTION  AND  HABITAT.— South-central 
Tabasco  (Fig.  3);  in  lower  montane  rain  forest  at 
an  elevation  of  about  140  m. 

PHENOLOGY. —Flowering:  March  and  October; 
fruiting:  March. 


The  recent  rediscovery  of  plants  resembling 
the  type,  and  only  known  collection,  of  Tabas- 
cina  lindenii  allows  for  a  reevaluation  of  this 
monotypic  genus.  Cowan's  collection  was  made 
in  the  region  of  the  type  locality  and  contains 
fruits  which  were  unknown  to  Baillon  (1891). 
Lindau  (1895)  maintained  Tabascina  as  a  genus 
of  Justicieae,  differing  from  Justicia  by  its  broad, 
leaflike  (vs.  narrow)  calyx  lobes  and  flowers  in 
terminal  cymes  (vs.  flowers  solitary  or  in  spikes 
or  contracted  panicles).  The  rugulate  corolla,  an- 
droecium  of  two  bithecous  stamens,  and  pollen 
with  trema  regions  studded  with  circular  insulae 
(Fig.  2d,  e)  all  suggest  placement  of  this  species 
within  Justicia.  The  large,  foliaceous,  and  het- 
eromorphic  calyx  lobes  that  are  nearly  or  com- 
pletely valvate  constitute  an  unusual  feature  of 
the  species;  however,  among  Mexican  Justicia, 
J.  caudata  A.  Gray  also  has  valvate,  or  nearly 
valvate,  and  partially  connivent  calyx  lobes.  The 
inflorescence  of/,  tabascina  consists  of  a  single 
terminal  thyrse  (i.e.,  an  indeterminate  main  axis 
with  determinate  lateral  axes),  or  a  cluster  of 
these.  The  lateral  axes  consist  of  dichasia  in  which 
usually  only  a  single  peduncled  flower  develops. 
Additional  floral  buds  of  the  dichasia  are  evident 
on  the  youngest  portions  of  the  thyrse  but  are 
represented  only  by  prominent  scars  on  the  older 
portions.  In  view  of  the  diversity  of  calyx  forms 
and  inflorescence  types  documented  for  species 
of  Justicia,  these  characters  would  not  appear  to 
warrant  recognition  of  Tabascina  as  distinct  from 
Justicia. 

Graham  (1988)  noted  that  four-aperturate  pol- 
len has  been  reported  in  a  few  New  World  species 
of  Justicia  but  she  was  unable  to  confirm  its  oc- 
currence. Justicia  tabascina,  with  its  four-aper- 
turate pollen  (Fig.  2b),  can  not  be  readily  clas- 
sified in  any  of  the  infrageneric  taxa  recognized 
by  Graham  (1988).  Pollen  of  this  species  differs 
from  Graham's  (1988)  type  3  pollen,  which  is 
characteristic  of  section  Drejerella  (Lindau)  Gra- 
ham, only  by  the  presence  of  four  rather  than 
three  apertures.  Justicia  tabascina  exhibits  sev- 
eral macromorphological  features  (e.g.,  unap- 
pendaged  thecae)  that  exclude  it  from  this  sec- 
tion, however. 


ADDITIONAL  SPECIMEN  EXAMINED.— MEXICO.  Tabasco: 
Mpio.  Tacotalpa,  Cerro  de  Madrigal,  7  km  de  las  Est.  Tacotalpa 
hacia  Tapijulapa,  C.  Cowan  et  al.  2860  (CAS). 


286 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  12 


Schaueria  parviflora  (Leonard)  T.  F.  Daniel, 
comb.  nov. 

Streblacanthus  parviflorus  Leonard,  J.  Wash.  Acad.  Sci.  31: 
1941.  TYPE.— GUATEMALA.  Izabal:  Escoba,  across  bay 
(west)  from  Puerto  Barrios,  3  May  1939,  P.  Standley  72949 
(Holotype:  F!). 

Schaueria  calycobractea  Hilsenbeck  &  Marshall,  Brittonia  35: 
362.  1983.  TYPE.-MEXICO.  Veracruz:  Estacion  de  Biolo- 
gia  Tropical  Los  Tuxtlas,  22  July  1978,  D.  Marshall  1  (Ho- 
lotype: US  fide  Hilsenbeck  and  Marshall  1983  but  not  lo- 
cated there  in  1989;  isotypes:  MEXU,  TEX!,  US). 

Perennial  herb  to  shrub  to  2  m  tall.  Young 
stems  subquadrate  to  quadrate-sulcate  (often 
somewhat  compressed),  sparsely  pubescent  in  2 
decussate  lines  with  antrorse  to  flexuose  to  re- 
trorse  eglandular  trichomes  0.1-0.4  mm  long, 
soon  glabrate.  Leaves  petiolate  with  petioles  to 
80  mm  long,  (lance-ovate  to)  ovate  to  ovate- 
elliptic,  50-180  mm  long,  14-90  mm  wide,  2- 
3.6  times  longer  than  wide,  acuminate  to  subfal- 
cate  at  apex,  (subcordate  to)  rounded  to  acute 
and  usually  somewhat  asymmetric  at  base,  sur- 
faces glabrous  (or  pubescent  along  midvein  of 
adaxial  surface  with  cauline  type  trichomes).  In- 
florescence of  terminal,  sessile  to  pedunculate 
spikes  to  55  mm  long,  peduncles  to  1 5  mm  long, 
rachis  pubescent  with  antrorse  to  flexuose  to  re- 
trorse  eglandular  trichomes  0.1-0.2  mm  long 
either  in  2  lines  or  more  or  less  evenly  disposed, 
flowers  opposite  in  axils  of  bracts  (sometimes  in 
axils  of  distalmost  pair  of  leaves  as  well).  Bracts 
linear-lanceolate  to  lance-subulate,  12-22  mm 
long,  1.2-2.2  mm  wide,  abaxial  surface  glabrous 
to  very  sparsely  pubescent  with  antrorse  eglan- 
dular trichomes  0. 1  mm  long  (especially  distally) 
and  rarely  with  glandular  trichomes  0.05-0. 1  mm 
long  as  well,  margin  sparsely  ciliate  with  similar 
trichomes.  Bractlets  linear-setaceous  to  lance- 
subulate,  7-30  mm  long,  0.8-1  mm  wide,  pu- 
bescent like  bracts  although  always  with  glan- 
dular trichomes.  Calyx  5-lobed,  10-25  mm  long, 
lobes  linear-setaceous  to  subulate,  9-24  mm  long, 
0.5-1.5  mm  wide,  irregularly  unequal  (with  2  or 
more  lobes  up  to  25%  longer  and  50%  wider  than 
the  others),  9-24  times  longer  than  the  tube,  pu- 
bescent like  bractlets.  Corolla  yellow  (or  white 
in  Guatemala),  22-28  mm  long,  externally  pu- 
bescent with  eglandular  trichomes  0.1-0.2  mm 
long,  tube  14-18.5  mm  long,  ampliate  at  base 
and  apex,  upper  lip  4.5-10  mm  long,  emarginate 
with  2  rounded  lobes  0.3-0.7  mm  long,  lower  lip 
5.5-9.5  mm  long  with  3  lobes,  lateral  lobes  linear 
to  elliptic  to  oblanceolate-elliptic,  5-7  mm  long, 


1.2-2.8  mm  wide,  central  lobe  elliptic  to  oblan- 
ceolate,  5.5-7  mm  long,  1.5-3  mm  wide.  Sta- 
mens attached  3.5-5  mm  proximal  to  mouth  of 
corolla,  8-9  mm  long,  thecae  parallel,  equally 
inserted,  subequal,  1.4-2.5  mm  long,  pointed  at 
base,  pollen  subspheric,  3-colporate  with  each 
colpus  flanked  by  2  pseudocolpi,  pseudocolpi 
sometimes  fused  in  intercolpal  regions  forming 
a  pseudocolpal  ring,  exine  reticulate,  reticulum 
of  intercolpal  regions  sometimes  irregularly  di- 
vided or  fragmented.  Style  19-23  mm  long,  gla- 
brous, stigma  bilobed  with  lobes  0.2  mm  long. 
Capsule  clavate,  10.5-16  mm  long,  glabrous  (or 
pubescent  with  sparse  glands  to  0.05  mm  long 
in  Guatemala),  stipe  4.5-8.5  mm  long,  head  6- 
9  mm  long,  retinacula  2-2.8  mm  long.  Seeds  4, 
flat,  subcirculate  to  subcordate  in  outline,  3.5- 
5.2  mm  long,  3-4.9  mm  wide,  surfaces  and  mar- 
gin papillose. 

DISTRIBUTION  AND  HABITAT.— Southeastern 
Mexico  (southern  Veracruz)  and  east-central 
Guatemala  (Izabal)  (Fig.  3);  in  lowland  to  lower 
montane  rainforest  (primary  and  secondary 
growth)  from  near  sea  level  to  700  m. 

PHENOLOGY.  — Flowering:  January,  March- 
April,  July,  and  October;  fruiting:  March,  June- 
July,  and  October. 

In  the  protologue  of  Streblacanthus  parviflo- 
rus, Leonard  (1941)  noted  a  relationship  with  S. 
cordatus  Lindau,  a  species  of  southern  Central 
America.  Gibson  (1974)  noted  the  occurrence  of 
four  calyx  lobes  in  Streblacanthus  Kuntze  and 
maintained  S.  parviflorus  as  the  northernmost- 
occurring  species  in  the  genus.  In  1983,  Hilsen- 
beck and  Marshall  described  Schaueria  calyco- 
bractea from  Veracruz  and  noted  that  it  appears 
similar  to  Streblacanthus  parviflorus  from  Gua- 
temala. They  further  noted  that  Streblacanthus 
could  be  distinguished  from  Schaueria  Nees  by 
its  four-lobed  (vs.  five-lobed)  calyx  and  unequal- 
ly (vs.  nearly  equally)  inserted  thecae,  but  that 
these  generic  distinctions  had  not  always  been 
followed.  They  concluded  that  these  genera  are 
very  similar  and  may  not  deserve  separation.  I 
agree  with  them  that  further  study  of  the  generic 
limits  of  Schaueria  and  Streblacanthus  is  desir- 
able. I  also  concur  with  placement  of  the  plants 
from  the  vicinity  of  the  Estacion  de  Biologia 
Tropical  Los  Tuxtlas  in  Schaueria  based  on  the 
distinctions  provided  above.  However,  the  type 
of  Streblacanthus  parviflorus  and  other  collec- 


DANIEL:  MEXICAN  ACANTHACEAE 


287 


tions  from  Guatemala  likewise  have  five-lobed 
calyces  and  subequally  inserted  thecae.  There- 
fore, they  should  also  be  treated  in  Schaueria. 

Since  the  description  of  Schaueria  calycobrac- 
tea  (Hilsenbeck  and  Marshall  1983),  in  which 
two  collections  were  cited,  additional  collections 
of  this  species  have  been  made  in  the  vicinity  of 
San  Andres  Tuxtla  and  in  the  Uxpanapa  region 
about  1 50  km  to  the  southeast.  The  three  known 
collections  from  Guatemala  differ  from  the  Mex- 
ican collections  by  the  presence  of  sparsely  dis- 
tributed glandular  trichomes  (vs.  glabrous  in 
Mexico)  on  the  capsules  and  white  (vs.  yellowish 
in  Mexico)  corollas.  Given  the  overall  similarity 
in  all  other  characters,  these  distinctions  seem 
relatively  minor. 

Pollen  of  Schaueria  parviflora  from  Veracruz 
(Fig.  2g,  h)  is  similar  to  that  of  plants  from  Gua- 
temala (Fig.  2i)  and  agrees  with  a  description  of 
the  type,  S.  calycotricha  (Link  &  Otto)  Nees,  from 
South  America  (Raj  1961). 

ADDITIONAL  SPECIMENS  EXAMINED.— GUATEMALA.  Iza- 
bal:  El  Estor,  E.  Contreras  11180  (LL);  between  Escobas  and 
Montana  Escobas,  across  bay  from  Puerto  Barrios,  /.  Steyer- 
mark  39324  (F).  MEXICO.  Veracruz:  Mpio.  San  Andres  Tux- 
tla, cerca  de  Laguna  Escondida,  entre  Sontecomapan  y  Mon- 
tepio,  J.  Beaman  5763  (F);  Mpio.  San  Andres  Tuxtla,  Estacion 
de  Biologia  Tropical  Los  Tuxtlas,  J.  Calzada  764  (F,  US),  A. 
Gdmez-Pompa  4484  (F),  G.  Ibarra  M.  540  (MO),  G.  Ibarra 
M.  &  S.  Sanaca  C.  1776  (MO),  G.  Martinez  C.  1708  (CAS, 
F),  R.  C.  Trigos  198  (F,  US),  G.  Webster  20895  (TEX);  Mpio. 
Minatitlan,  Zona  de  Uxpanapa,  1-2  km  SE  of  La  Chinantla 
(17°15'30"N,  94°26'30"W),  T.  Daniel  &  T.  Wendt  5810  (CAS, 
CHAPA,  MEXU);  Mpio.  San  Andres  Tuxtla,  al  sur  de  la  Ebi- 
trolotu,  A.  Villegas  H.  72  (CAS,  F,  US);  Mpio.  Hidalgotitlan, 
desde  el  Poblado  6,  al  S  por  la  brecha  y  la  vereda  al  horcajo 
oriental  del  Rio  Cuevas,  17°15'N,  94°30'W,  T.  Wendt  et  al. 
2604  (CAS). 

ACKNOWLEDGMENTS 

I  am  grateful  for  the  financial  assistance  pro- 
vided by  the  National  Science  Foundation  (BSR- 


8609852),  the  American  Philosophical  Society, 
Arizona  State  University,  and  the  California 
Academy  of  Sciences.  Mary  Ann  Tenorio  pro- 
vided the  line  drawing  and  SEM  micrographs. 
My  field  work  in  Veracruz  was  made  possible  by 
Tom  Wendt.  Loans  and/or  other  courtesies  were 
provided  by  the  herbaria  cited  in  the  text. 

LITERATURE  CITED 

BAILLON,  H.  1891.  Histoire  des  plantes,  Vol.  10.  Librairie 
Hachette,  Paris. 

DANIEL,  T.  F.  1981.  Mexacanthus,  a  new  genus  of  Acantha- 
ceae  from  western  Mexico.  Syst.  Bot.  6:288-293. 

.  1986.  Acanthaceae.  Pp.  27-30  in  Listados  floristicos 

de  Mexico.  IV.  Flora  de  Chiapas.  D.  E.  Breedlove,  ed.  In- 
stituto  de  Biologia,  UNAM,  Mexico. 

DANIEL,  T.  F.,  B.  D.  PARFITT,  AND  M.  A.  BAKER.  1984.  Chro- 
mosome numbers  and  their  systematic  implications  in  some 
North  American  Acanthaceae.  Syst.  Bot.  9:346-355. 

DANIEL,  T.  F.,  T.  I.  CHUANG,  AND  M.  A.  BAKER.  In  Press. 
Chromosome  numbers  of  American  Acanthaceae.  Syst.  Bot. 

GIBSON,  D.  N.  1974.  Acanthaceae.  Pp.  328^61  in  Flora  of 
Guatemala  (Fieldiana,  Bot.  24  [pt.  10]).  P.  C.  Standley  et 
al.,  eds.  Field  Museum  of  Natural  History,  Chicago. 

GRAHAM,  V.  A.  1988.  Delimitation  and  infra-generic  clas- 
sification ofJusticia  (Acanthaceae).  Kew  Bull.  43:551-624. 

HILSENBECK,  R.  A.  AND  D.  L.  MARSHALL.  1983.  Schaueria 
calycobractea  (Acanthaceae),  a  new  species  from  Veracruz, 
Mexico.  Brittonia  35:362-366. 

KOBUSKI,  C.  E.  1928.  A  monograph  of  the  American  species 
of  the  genus  Dyschoriste.  Ann.  Missouri  Bot.  Gard.  15:9- 
91. 

LEONARD,  E.  C.  1941.  New  Acanthaceae  from  Guatemala. 
J.  Wash.  Acad.  Sci.  31:96-105. 

.  1945.  Acanthaceae.  P.  1 1 5  in  Mexican  phanerogams 

described  by  M.  E.  Jones  (Contr.  U.S.  Natl.  Herb.  29:87- 
116).  C.V.Morton,  ed. 

LINDAU,  G.  1895.  Acanthaceae.  Pp.  274-354  in  Die  Natttr- 
lichen  Pflanzenfamilien,  Vol.  4  (3b).  H.  G.  A.  Engler  and  K. 
A.  E.  Prantl,  eds.  Wilhelm  Englemann,  Leipzig. 

RAJ,  B.  1961.  Pollen  morphological  studies  in  the  Acantha- 
ceae. Grana  Palynol.  3:3-108. 

.  1973.  Further  contribution  to  the  pollen  morphology 

of  the  Acanthaceae.  J.  Palynol.  9:91-141. 

RAMAMOORTHY,  T.  P.  AND  D.  C.  WASSHAUSEN.  1985.  Anew 
name  in  Dyschoriste  (Acanthaceae).  Brittonia  37:358-359. 


CALIFORNIA  ACADEMY  OF  SCIENCES 

Golden  Gate  Park 
San  Francisco,  California  94 1 1 8 


PROCEEDINGS 

OF  THE 

CALIFORNIA  ACADEMY  OF  SCIENCES 


Vol.  46,  No.  13,  pp.  289-297,  5  figs.  September  11,  1990 


THREE  NEW  SPECIES  OF 
JUSTICIA  (ACANTHACEAE)  FROM  PANAMA 

By 

Thomas  F.  Daniel 

Department  of  Botany,  California  Academy  of  Sciences, 
Golden  Gate  Park,  San  Francisco,  California  94118 

and 
Dieter  C.  Wasshausen 

Department  of  Botany,  National  Museum  of  Natural  History, 
Smithsonian  Institution,  Washington,  D.C.  20560 


ABSTRACT:  Three  new  species  of  Justicia,  J.  fortunensis,  J.  veraguensis,  and  J.  readii,  are  described  from 
Panama.  The  three  species  are  illustrated  and  their  known  distributions  are  mapped.  A  key  to  these  and  a 
related  species  in  Panama,  /.  refractifolia,  is  provided.  A  chromosome  number  of  n  =  14  is  reported  for  J. 
fortunensis.  With  23  species  now  known  from  Panama,  Justicia  is  the  largest  genus  of  Acanthaceae  in  that 
country. 

Received  December  11,  1989.  Accepted  March  14,  1990. 

common  with  J.  refractifolia  (Kuntze)  Leonard 
INTRODUCTION  .  .         '  _/  A11\.  ... 

of  Costa  Rica  and  Panama.  All  four  have  axillary 

The  Acanthaceae  of  Panama  comprise  109  and  terminal  inflorescences  of  conspicuously  and 
species  in  34  genera  (Durkee  1988).  Justicia  is  densely  bracteate  spikes  with  the  bracts  and-co- 
the  largest  and  taxonomically  most  complex  ge-  rollas  less  than  20  mm  long.  These  characters 
nus  of  Acanthaceae  with  between  420  (Mabber-  distinguish  this  group  of  species  from  all  other 
ley  1987)  and  600  or  so  (Graham  1988)  species  Panamanian  Justicia.  A  key  to  these  four  species 
worldwide.  Durkee  (1988)  noted  the  occurrence  is  presented  below, 
of  2 1  species  of  Justicia  in  Panama.  Recent  ex- 
plorations in  the  species-rich  montane  regions  1.  Young  stems  pubescent  with  whitish  to  light 
near  Fortuna  Dam  and  on  Cerro  Tute  in  western          brown  trichomes  (and  bracts  not  drying  dark 
Panama  have  resulted  in  the  collection  of  two          red);  calyx  pubescent  with  eglandular  tri- 
undescribed  species  of  Justicia.  Another  species,          chomes;  filaments  pubescent, 
previously  known  only  from  Cerro  Jefe,  and          2.  Bracts  dark  pink  (often  drying  brown- 
treated  under  the  name  of  a  species  from  the  Old  ish);  corolla  8-10  mm  long,  externally 
World,  is  recognized  as  new  and  its  range  within  pubescent  with  glandular  and  eglandu- 
Panama  is  expanded.                                                         lar  trichomes;  filaments  proximally  pu- 

These  three  species  share  several  features  in  bescent  with  eglandular  trichomes,  dis- 

[289] 


290 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  13 


tally  glabrous;  capsule  pubescent  with 
glandular  and  eglandular  trichomes; 
seeds  smooth J.  fortunensis 

2.  Bracts  green  (drying  dark  green  or 
brown);  corolla  10-17  mm  long,  exter- 
nally pubescent  with  eglandular  tri- 
chomes; filaments  pubescent  with  glan- 
dular trichomes  throughout  length; 
capsule  pubescent  with  eglandular  tri- 
chomes; seeds  roughened  with  minute 
rounded  protrusions J.  refractifolia 

1.  Young  stems  glabrous  or  pubescent  with 
golden  trichomes  (if  trichomes  light  brown 
or  straw-colored  as  sometimes  in  J.  readii, 
then  bracts  drying  dark  red);  calyx  pubes- 
cent with  glandular  and  eglandular  tri- 
chomes; filaments  glabrous. 

3.  Young  stems  glabrous;  bracts  drying  pale 
to  dark  greenish;  bractlets  ovate-elliptic 
to  elliptic  to  oblanceolate  to  obovate; 
calyx  5-6  mm  long;  corolla  10-13  mm 
long,  externally  pubescent  with  glan- 
dular and  eglandular  trichomes;  seeds 
smooth /.  veraguensis 

3.  Young  stems  evenly  covered  with  a 
dense  layer  of  (flexuose  to)  antrorse  to 
antrorsely  appressed  golden  (to  light 
brown  or  straw-colored)  eglandular  tri- 
chomes; bracts  drying  dark  red;  bract- 
lets  linear-lanceolate  to  linear  to  linear- 
elliptic;  calyx  7.5-10  mm  long;  corolla 
15-19  mm  long,  externally  pubescent 
with  eglandular  trichomes;  seeds  papil- 
lose to  verrucose J.  readii 

Although  Leonard  (1958)  recognized  more  than 
70  species  of  Justicia  in  Colombia  and  Durkee 
(1986)  treated  27  species  from  Costa  Rica,  none 
of  the  taxa  from  these  adjacent  regions  appears 
to  be  conspecific  with  the  three  species  newly 
described  below.  Justicia  is  the  largest  genus  of 
Acanthaceae  in  Panama  with  23  species  now 
known  from  the  country. 

Justicia  fortunensis  sp.  nov. 

(Fig.  1) 

TYPE.— PANAMA.  Chiriqui:  vicinity  of  Fortuna  Dam,  along 
trail  across  valley  of  Rio  Homito,  12  March  1988,  T.  Daniel, 
F.  Almeda,  and  G.  McPherson  5550  (Holotype:  CAS!;  isotypes: 
DUKE!,  F!,  GH!,  K!,  MICH!,  MO!,  NY!,  PMA!,  US!). 

Suflrutex  vel  frutex  usque  ad  1  m  altus.  Caules  juniores 
trichomatibus  cglandulosis  antrorsis  (vel  retrorsis)  0.2-0. 5  mm 


longis  in  2  lineis  dispositis  pubescentes.  Folia  petiolata,  lam- 
inae lanci-ellipticae  vel  ovato-ellipticae  vel  ellipticae,  45-152 
mm  longae,  10-56  mm  latae,  2.1-5-plo  longiores  quam  lati- 
ores.  Inflorescentia  spicata  dense  bracteata,  rhachis  trichoma- 
tibus eglandulosis  pubescens.  Bracteae  atroroseae  saepe  in  sic- 
co  brunneolas  late  ellipticae  vel  subcirculares  vel  obovatae  7- 
13(-17)  mm  longae  6-11.5  mm  latae.  Bracteolae  ellipticae. 
Calyx  4.5-6.5  mm  longus  extus  eglandulosus.  Corolla  8-10 
mm  longa  extus  glandulosa.  Stamina  2.8-3  mm  longa  fila- 
mentis  pubescentibus.  Stylus  4.5-5  mm  longus.  Capsula  10 
mm  longa  extus  glandulosa.  Semina  2-3  mm  longa  2-2.5  mm 
lata  laevia. 

Perennial  herb  to  shrub  to  1  m  tall.  Young 
stems  subquadrate  to  quadrate,  pubescent  with 
antrorse  (or  retrorse),  whitish  to  light  brown  (and 
with  conspicuous  darker  brown  septae),  eglan- 
dular trichomes  0.2-0.5  mm  long,  mostly  con- 
centrated in  2  decussate  lines,  older  stems  be- 
coming glabrate.  Leaves  petiolate,  petioles  to  14 
mm  long,  pubescent  like  young  stems,  blades 
lance-elliptic  to  ovate-elliptic  to  elliptic,  45-152 
mm  long,  10-56  mm  wide,  2.1-5  times  longer 
than  wide,  acuminate  to  long-acuminate  at  apex, 
acute  to  attenuate  at  base,  surfaces  sparsely  pu- 
bescent with  cauline  type  trichomes  along  major 
veins,  otherwise  glabrous.  Inflorescence  of  pe- 
dunculate, densely  bracteate  spikes  to  60  mm 
long,  1 0-20  mm  in  diameter  near  midspike,  borne 
singly  (or  sometimes  2)  in  axils  of  distal  leaves, 
peduncles  to  45  mm  long,  pubescent  like  stems, 
rachis  more  or  less  evenly  pubescent  with  straight 
to  flexuose  to  antrorse,  eglandular  trichomes  0. 1- 
0.4  mm  long.  Bracts  dark  pink  (basal  pair  rarely 
somewhat  leaflike  and  partly  green),  drying  pink- 
ish or  brownish,  broadly  elliptic  to  subcirculate 
to  obovate,  7-1 3(-l  7)  mm  long,  6-11.5  mm  wide, 
rounded  or  obtuse  (and  often  apiculate)  to  acute 
at  apex,  either  abruptly  constricted  or  tapering 
into  a  short  (to  3  mm  long)  petiole  at  base,  abax- 
ial  surface  pubescent  with  prominent  or  incon- 
spicuous, straight  to  flexuose  eglandular  and  stip- 
itate  glandular  trichomes  to  0.5  mm  long,  margin 
prominently  ciliate  with  coarse,  conspicuously 
multi-celled,  eglandular  trichomes  to  1.4  mm 
long.  Bractlets  elliptic,  6-1 2  mm  long,  1.8-4  mm 
wide,  gradually  tapering  into  a  petiole  at  base, 
pubescent  like  bracts.  Calyx  4.5-6.5  mm  long, 
5-lobed,  lobes  dark  pink  (at  least  distally),  di- 
vided nearly  to  base  of  calyx,  lanceolate  to  sub- 
ulate, 0.5-0.8  mm  wide,  pubescent  with  eglan- 
dular trichomes.  Corolla  white  to  very  pale  pink, 
8-10  mm  long,  externally  pubescent  with  stipi- 
tate  glandular  and  eglandular  trichomes  0.1-0.3 


DANIEL  AND  WASSHAUSEN:  NEW  PANAMANIAN  JUSTICIA 


291 


2mm 


FIGURE  1.  a-d,  Justicia  fortunensis  (Daniel  et  al.  5550).  a,  Habit;  b,  Bract;  c,  Bract,  bractlets,  calyx,  and  corolla;  d,  corolla 
with  stamens,  e-h,  Justicia  readii  (Read  et  al.  85-6b).  e,  Habit;  f,  Bract;  g,  Bracts,  bractlet,  calyx,  and  young  corolla;  h,  corolla 
with  stamens. 


292 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  13 


\ 


20 


p  m 


FIGURE  2.  Camera-lucida  drawing  of  meiotic  chromo- 
somes of/. fortunensis (Daniel etal.  5500,  n  =  14)atmetaphase 
II. 


mm  long,  tube  4-5.5  mm  long,  internally  densely 
pubescent,  upper  lip  triangular,  internally  rugu- 
late,  3.5-4  mm  long  with  2  lobes  0.4-0.7  mm 
long,  lower  lip  4-4.5  mm  long  with  3  elliptic 
lobes  3-3.5  mm  long,  2-3.5  mm  wide,  central 
lobe  larger  than  lateral  lobes.  Stamens  attached 
near  apex  of  tube,  2.8-3  mm  long,  exserted  up 
to  2  mm  from  mouth  of  corolla,  filaments  1.5- 
2  mm  long,  pubescent  with  eglandular  trichomes 
proximally,  glabrous  distally,  thecae  superposed, 
the  upper  1.4-1.5  mm  long,  the  lower  1.5-1.8 
mm  long  (including  a  blunt  basal  appendage  to 
0.2  mm  long),  pollen  diporate  (to  subdicolporate) 
with  trema  region  traversed  by  2  rows  of  6-7 


insulae.  Style  4.5-5  mm  long,  distally  glabrous, 
proximally  pubescent  with  upward  pointing, 
eglandular  trichomes.  Capsule  1 0  mm  long,  pu- 
bescent with  glandular  and  eglandular  trichomes 
0.1-0.3  mm  long,  stipe  4  mm  long,  head  6  mm 
long.  Seeds  4  per  capsule,  flattened,  subcircular 
to  subelliptic  in  outline,  2-3  mm  long,  2-2.5  mm 
wide,  surface  smooth,  (n  =  14,  Daniel  et  al.  5550, 
Fig.  2) 

DISTRIBUTION  AND  PHENOLOGY.— Known  only 
from  Pacific  slopes  of  the  Cordillera  Central  in 
western  Panama  (north-central  Chiriqui)  in  the 
watershed  of  the  Rio  Chiriqui  (Fig.  3).  Plants 
occur  in  cloud  forests  at  elevations  from  about 
1025  to  1250  m.  Flowering:  January- April  and 
August;  fruiting:  January-April. 

Justicia  fortunensis  is  not  readily  identifiable 
with  any  subgeneric  taxa  using  the  keys  and  de- 
scriptions provided  by  Graham  (1988).  It  has 
pollen  ("Type  5,"  Fig.  4)  like  that  found  in  five 
sections  recognized  by  Graham.  The  chromo- 
some number  reported  here  for  the  type  is  the 
most  commonly  encountered  number  in  the  ge- 
nus and  occurs  in  various  subgeneric  taxa  (Daniel 
and  Chuang  1989;  Daniel  et  al.  1984,  1990). 

The  Fortuna  Dam  area,  the  only  region  from 
which  this  species  is  known,  has  received  con- 
siderable attention  by  plant  collectors  in  the  last 


•  FORTUNENSIS 
A   READII 

•  VERAGUENSIS 


too 


200  km 


FIGURE  3.    Map  showing  distribution  of  Justicia  fortunensis,  J.  readii,  and  J.  veraguensis. 


DANIEL  AND  WASSHAUSEN:  NEW  PANAMANIAN  JUSTICIA 


293 


FIGURE  4.  Scanning  electron  micrographs  of  pollen  ofJusticia.  a,  /.  readii  (Read  et  al.  85-6b),  view  of  "trema  region";  b,  J. 
readii  (Read  et  al.  85-6b),  view  between  2  "trema  regions";  c,  /.  veraguensis  (Croat  34153),  view  of  "trema  region";  d,  J. 
fortunensis  (Daniel  et  al.  5550),  view  of  "trema  region." 


294 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  13 


FIGURE  5.    Justicia  veraguensis  (Croat  34153).  a,  Habit;  b,  Bract;  c,  Bractlets,  calyx,  and  gynoecium;  d,  Corolla  with  stamens. 


few  years  and  has  yielded  a  wealth  of  new  taxa 
(Dwyer  1985). 

PARATYPES.— PANAMA.  Chiriqui:  Edwin  Fa- 
brega  Dam  and  Reserve  in  Fortuna  along  trail 
to  Rio  Hornito  above  Los  Planes,  F.  Almeda  et 
al.  6320  (CAS,  MO,  PMA,  US);  W  side  of  Rio 
Hornito,  W.  D'Arcy  16003  (MO);  La  Fortuna 
hydroelectric  project,  along  Chiriqui  River,  B. 
Hammel  2292  (MEXU);  between  Los  Planes  de 
Hornito  and  Fortuna  Lake,  8°41'N,  82°13'W,  R. 
Hampshire  &  C.  Whitefoord  701  (BM);  ca.  5  km 


E  of  Fortuna  Dam,  along  trail  crossing  Rio  Hor- 
nito, 8°45'N,  82°15'W,  S.  Thompson  5012  (CAS). 

Justicia  veraguensis,  sp.  nov. 

(Fig.  5) 

TYPE.— PANAMA.  Veraguas:  Cerro  Tute,  above  Santa  Fe 
beyond  Escuela  Agricola  Inter-americana,  1 .8  mi  beyond  fork 
in  rd.,  5  April  1976,  T.  Croat  34153  (Holotype:  NY!). 

Frutex  usque  ad  2  m  altus.  Caules  juniores  glabri.  Folia 
petiolata,  laminae  ovato-ellipticae  vel  ellipticae  65-165  mm 
longae  21-63  mm  latae  2.6-3.4-plo  longiorcs  quam  latiores. 


DANIEL  AND  WASSHAUSEN:  NEW  PANAMANIAN  JUSTICIA 


295 


Inflorescentia  spicata  dense  bracteata,  rhachis  trichomatibus 
eglandulosis  et  plerumque  glandulosis  pubescens.  Bracteae  in 
sicco  subviridulae  vel  atroviridulae  obovatae  vel  late  ellipticae 
vel  subcirculares  vel  deltatae  8-1 3  mm  longae  6-8.5  mm  latae. 
Bracteolae  ovato-ellipticae  vel  ellipticae  vel  oblanceolatae  vel 
obovatae.  Calyx  5-6  mm  longus  extus  glandulosus.  Corolla 
10-13  mm  longa  extus  glandulosa.  Stamina  5.5-6  mm  longa 
fi  la  mentis  glabris.  Stylus  8  mm  longus.  Capsula  9-1 1  mm  longa 
extus  glandulosa.  Semina  3-3.3  mm  longa  2.7-3.7  mm  lata 
laevia. 

Shrub  to  2  m  tall.  Young  stems  quadrate,  gla- 
brous, older  stems  glabrous.  Leaves  petiolate, 
petioles  to  50  mm  long,  glabrous,  blades  ovate- 
elliptic  to  elliptic,  65-165  mm  long,  21-63  mm 
wide,  2.6-3.4  times  longer  than  wide,  acuminate 
at  apex,  acute  to  subattenuate  at  base,  surfaces 
glabrous.  Inflorescence  of  axillary  and  terminal, 
pedunculate,  densely  bracteate  spikes  to  65  mm 
long,  13-21  mm  in  diameter  near  midspike,  pe- 
duncles to  15  mm  long,  glabrous  or  pubescent 
in  2  lines  with  retrorse  eglandular  trichomes  to 
0.3  mm  long,  rachis  pubescent  with  straight  to 
flexuose  eglandular  and  glandular  (sometimes  not 
evident)  trichomes  0.1-0.3  mm  long.  Bracts 
drying  pale  to  dark  greenish,  obovate  to  broadly 
elliptic  to  subcirculate  to  deltate,  8-13  mm  long, 
6-8.5  mm  wide,  rounded  to  acute  at  apex,  ses- 
sile or  petiolate  (petiole  to  2  mm  long),  con- 
stricted and/or  tapering  at  base,  abaxial  surface 
pubescent  with  straight  to  flexuose  eglandular 
and  glandular  trichomes  0. 1-0.3  mm  long  or  be- 
coming glabrate,  margin  ciliate  with  trichomes 
like  those  of  abaxial  surface  and  with  an  over- 
story  of  flexuose  eglandular  trichomes  to  1 .2  mm 
long.  Bractlets  ovate-elliptic  to  elliptic  to  oblan- 
ceolate  to  obovate,  6-10  mm  long,  1.7-2.7  mm 
wide,  tapering  and  often  petiolate  at  base,  pu- 
bescent like  bracts.  Calyx  5-6  mm  long,  5-lobed, 
lobes  divided  nearly  to  base  of  calyx,  lanceolate, 
0.7-1.7  mm  wide,  pubescent  like  bracts.  Corolla 
white  (to  greenish  white)  with  purplish  markings, 
10-13  mm  long,  externally  pubescent  with  flex- 
uose eglandular  and  glandular  trichomes  0. 1-0.3 
mm  long,  tube  6-6.5  mm  long,  internally  gla- 
brous, upper  lip  triangular,  internally  rugulate, 
4-7  mm  long,  emarginate  with  lobes  0.2-0.3  mm 
long,  lower  lip  4-7  mm  long  with  3  broadly  el- 
liptic to  subcirculate  lobes  2.5-4  mm  long,  2-4.5 
mm  wide,  central  lobe  larger  than  lateral  lobes. 
Stamens  attached  near  apex  of  tube,  5.5-6  mm 
long,  exserted  up  to  5  mm  beyond  mouth  of 
corolla,  filaments  4.5-5.2  mm  long,  glabrous, 
thecae  superposed,  the  upper  1.2-1.5  mm  long, 
the  lower  1.7-2  mm  long  (including  a  basal  spur 


0.3-0.4  mm  long),  pollen  diporate  (to  subdicol- 
porate)  with  trema  region  traversed  by  2  rows  of 
6-7  insulae.  Style  8  mm  long,  distally  glabrous, 
proximally  pubescent  with  eglandular  trichomes. 
Capsule  9-1 1  mm  long,  pubescent  with  glandular 
and  eglandular  trichomes  0. 1-0.2  mm  long,  stipe 
3.5-4  mm  long,  head  5.5-7  mm  long.  Seeds  4 
per  capsule,  flattened,  subcircular  in  outline,  3- 
3.3  mm  long,  2.7-3.7  mm  wide,  surface  smooth. 

DISTRIBUTION  AND  PHENOLOGY.— Known  only 
from  the  mountains  of  Veraguas  in  western  Pan- 
ama. The  type  and  one  paratype  were  collected 
on  the  Pacific  slopes  of  Cerro  Tute  in  the  Cor- 
dillera Central  of  northern  Veraguas  (Fig.  3).  We 
have  not  been  able  to  determine  the  exact  lo- 
cation in  Veraguas  of  the  other  paratype;  how- 
ever, T.  Croat  (personal  communication)  sug- 
gests that  the  Rio  Segundo  Brazo  is  the  second 
river  along  the  road  to  the  lowlands  on  the  Ca- 
ribbean slope  of  Cerro  Tute.  Plants  occur  on 
forested  slopes  and  on  rocks  in  watercourses  at 
elevations  from  630  to  1200  m.  Flowering: 
March-April  and  September;  fruiting:  April. 

Justicia  veraguensis  exhibits  a  suite  of  char- 
acters that  precludes  its  placement  into  any  of 
the  sections  recognized  by  Graham  (1988).  It  has 
pollen  (Fig.  4)  similar  to  that  of  J.  fortunensis, 
which  is  undoubtedly  a  close  relative. 

PARATYPES.— PANAMA.  Veraguas:  5  mi  W  of 
Santa  Fe  on  rd.  past  Escuela  Agricola  Alto  Pie- 
dra,  R.  Liesner  852  (GH,  MO);  Rio  Segundo 
Brazo,  P.  Maas  &  R.  Dressier  1660  (F,  U,  US). 

Justicia  readii,  sp.  nov. 

(Fig.  1) 

TYPE.— PANAMA.  Panama:  Cerro  Jefe,  5  January  1972, 
J.  Dwyer  9500  (Holotype:  MO!;  isotype:  NY!). 

Frutex  usque  ad  1.5  m  altus.  Caules  juniores  trichomatibus 
eglandulosis  aureis  (vel  pallidis  brunneolis  vel  stramineis)  (flex- 
uosis  vel)  antrorsis  vel  appressis  0.4-2  mm  longis  aequaliter 
pubescentes.  Folia  petiolata,  laminae  ellipticae  vel  obovatae 
45-185  mm  longae  17-78  mm  latae  2.3-3.8-plo  longiores  quam 
latiores.  Inflorescentia  spicata  dense  bracteata,  rhachis  tri- 
chomatibus glandulosis  et  eglandulosis  pubescens.  Bracteae 
virides  in  sicco  atrovinosae  obovatae  vel  subcirculares  vel  el- 
lipticae 8-16  mm  longae  4.5-11.5  mm  latae.  Bracteolae  line- 
ares-lanceolatae  vel  lineares  vel  lineares-ellipticae.  Calyx  7.5- 
10  mm  longus  extus  glandulosus.  Corolla  15-19  mm  longa 
extus  eglandulosa.  Stamina  5-6.5  mm  longa  filamentis  glabris. 
Stylus  11-16  mm  longus.  Capsula  9-12.5  mm  longa  extus 
glandulosa.  Semina  2.3-2.6  mm  longa,  2.3-2.5  mm  lata  papil- 
losa  vel  verrucosa. 

Shrub  to  1.5  m  tall.  Young  stems  subquadrate 
to  quadrate,  densely  and  evenly  pubescent  with 


296 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  13 


(flexuose  to)  antrorse  to  antrorsely  appressed, 
golden  (to  light  brown  or  straw-colored),  eglan- 
dular  trichomes  0.4-2  mm  long,  older  stems  re- 
maining pubescent.  Leaves  petiolate,  petioles  to 
30  mm  long,  pubescent  like  stems,  blades  elliptic 
to  obovate,  45-185  mm  long,  17-78  mm  wide, 
2.3-3.8  times  longer  than  wide,  acute  to  acu- 
minate at  apex,  acute  to  attenuate  at  base,  mid- 
vein  pubescent  like  stems,  surfaces  sparsely  pu- 
bescent with  similar  trichomes.  Inflorescence  of 
axillary  and  terminal,  pedunculate,  densely  brac- 
teate  spikes  to  75  mm  long,  12-21  mm  in  di- 
ameter near  midspike,  peduncles  to  1 4  mm  long, 
pubescent  like  stem,  rachis  pubescent  with 
straight  to  flexuose  (to  antrorsely  appressed,  es- 
pecially toward  base  of  spike)  eglandular  and 
glandular  trichomes  to  1  mm  long.  Bracts  green, 
drying  dark  red,  obovate  to  subcirculate  to  ellip- 
tic, often  somewhat  asymmetrical,  8-1 6  mm  long, 
4.5-11.5  mm  wide,  rounded  and  sometimes 
apiculate  to  acute  at  apex,  sessile,  rounded  or 
constricted  and  tapering  (to  subpetiolate)  at  base, 
abaxial  surface  pubescent  with  straight  to  flex- 
uose glandular  and  eglandular  trichomes  0.2-0.5 
mm  long  (sometimes  not  persisting  on  older 
bracts),  and  usually  with  coarser,  antrorse  eglan- 
dular trichomes  to  0.8  mm  long  as  well,  margin 
ciliate  with  flexuose  eglandular  (usually  coarse) 
and  glandular  trichomes  up  to  2  mm  long.  Bract- 
lets  linear-lanceolate  to  linear  to  linear-elliptic, 
8-13  mm  long,  1.5-2.1  mm  wide,  tapering  little, 
if  at  all,  and  sessile  at  base,  pubescent  like  bracts. 
Calyx  7.5-10  mm  long,  5-lobed,  lobes  divided 
nearly  to  base  of  calyx,  subulate,  0.5-0.8  mm 
wide,  pubescent  with  glandular  and  eglandular 
trichomes.  Corolla  whitish  to  lavender,  15-19 
mm  long,  externally  pubescent  with  flexuose 
eglandular  trichomes  to  0.6  mm  long,  tube  9-12 
mm  long,  internally  glabrous,  upper  lip  trian- 
gular, internally  rugulate,  4.5-8  mm  long,  emar- 
ginate  with  lobes  0.2-0.3  mm  long,  lower  lip  5- 
7.5  mm  long  with  3  subelliptic  lobes  2-4.5  mm 
long,  1.6-2.8  mm  wide,  central  lobe  larger  than 
lateral  lobes.  Stamens  attached  near  apex  of  co- 
rolla tube,  5-6.5  mm  long,  exserted  up  to  5.5 
mm  beyond  mouth  of  corolla,  filaments  4-5  mm 
long,  glabrous,  thecae  superposed,  the  upper  1 .2- 
1.5  mm  long,  the  lower  1 .5-2  mm  long  (including 
a  basal  spur  0.2-0.3  mm  long),  pollen  triporate 
(to  subtricolporate)  with  trema  region  traversed 
by  2  rows  of  7  insulae.  Style  11-16  mm  long, 
distally  glabrous,  proximally  pubescent  with 


eglandular  trichomes.  Capsule  9-12.5  mm  long, 
pubescent  with  glandular  and  eglandular  tri- 
chomes 0.1-0.4  mm  long,  stipe  3.5-6  mm  long, 
head  5-6.5  mm  long.  Seeds  4  per  capsule,  flat- 
tened, subcircular  in  outline,  2.3-2.6  mm  long, 
2.3-2.5  mm  wide,  surface  papillose  (when  im- 
mature) and  papillose  to  verrucose  (when  ma- 
ture). 

DISTRIBUTION  AND  PHENOLOGY.— Cloud  for- 
ests of  eastern  Panama  (Panama,  San  Bias,  and 
Darien)  at  elevations  from  500  to  1300  m  (Fig. 
3).  Flowering  November-January;  fruiting:  De- 
cember-January. 

The  type  and  most  other  collections  of  this 
species  have  golden,  antrorsely-appressed  cau- 
line  trichomes  to  0.8  mm  long,  relatively  small 
leaves  (blades  45  to  90  mm  long,  17  to  31  mm 
wide  with  petioles  to  1 7  mm  long),  and  relatively 
short  spikes  (to  55  mm  long)  with  small  bracts 
(8-12  mm  long,  4.5-7  mm  wide).  Four  collec- 
tions (Dressier  3221,  Folsom  et  al.  6296,  Maas 
&  Dressier  677,  and  Mori  &  Kallunki  3424)  differ 
most  conspicuously  from  those  specimens  re- 
sembling the  type  by  having  light  brown  to  straw- 
colored,  flexuose  to  antrorse-appressed  cauline 
trichomes  to  2  mm  long,  larger  leaves  (55  to  185 
mm  long,  18-78  mm  wide  with  petioles  to  30 
mm  long),  and  somewhat  longer  spikes  (to  75 
mm  long)  with  larger  bracts  (11  to  16  mm  long, 
5-1 1.5  mm  wide).  Although  these  two  forms  are 
readily  recognizable,  the  general  overlap  in  the 
character  states  noted  above  and  the  occurrence 
of  both  in  the  same  region  (i.e.,  Cerro  Jefe)  sug- 
gests that  they  be  treated  as  a  single  taxon  at  this 
time. 

Justicia  readii  also  cannot  be  classified  in  Gra- 
ham's (1988)  scheme.  It  has  trilaterally  sym- 
metric pollen  ("Type  2,"  Fig.  4)  which  is  char- 
acteristic of  Graham's  section  Tyloglossa  of  Africa 
and  Arabia.  Species  in  this  section  differ  from  J. 
readii  by  their  inflorescence  structure  and  bract 
form.  In  spite  of  the  obvious  difference  in  pollen, 
/.  readii  is  more  similar  to  /.  fortunensis  and  J. 
veraguensis  than  to  any  other  species  in  Panama. 

Durkee  (1978)  treated  this  species  as  /.  ad- 
hat oda  L.,  a  widely  cultivated  species  native  from 
the  Indian  subcontinent  to  southeastern  Asia. 
Justicia  adhatoda,  which  is  not  presently  known 
from  Panama,  can  be  distinguished  from  /.  readii 
by  its  longer  (23-30  mm)  corollas;  larger  (12-34 
mm  long  and  7-12  mm  wide)  and  eglandular 
bracts  that  do  not  dry  dark  red;  linear-elliptic, 


DANIEL  AND  WASSHAUSEN:  NEW  PANAMANIAN  JUSTICIA 


297 


eglandular,  and  wider  (1.5-2.3  mm)  calyx  lobes; 
longer  (3.5-4.5  mm)  thecae,  both  of  which  bear 
basal  spurs;  and  bilaterally  symmetric  pollen. 

Justicia  readii  has  been  rather  well  collected 
on  Cerro  Jefe,  a  mountain  northeast  of  Panama 
City  that  has  been  much  visited  since  it  became 
readily  accessible  to  plant  collectors  in  the  mid- 
1960's  (Dwyer  1967,  1985).  It  is  not  surprising 
that  the  species  was  recently  collected  along  the 
relatively  new  El  Llano-Carti  Road  in  the  Ser- 
rania  de  San  Bias.  There  is  a  continuous  ridge 
system  linking  Cerro  Jefe  with  the  Serrania  de 
San  Bias.  The  southeasternmost  station  for  this 
species,  in  the  Serrania  de  Pirre  in  Darien,  is 
rather  isolated  from  the  other  known  occur- 
rences. 

PARATYPES.— PANAMA.  Darien:  ridgetoparea 
N  of  Cerro  Pirre,  between  Cerro  Pirre  top  and 
Rancho  Plastico,  /.  Folsom  et  al.  6296  (MO). 
Panama:  Cerro  Jefe,  along  main  rd.  before  turn- 
off  to  summit,  T.  Croat  13054  (MO);  La  Eneida, 
region  of  Cerro  Jefe,  R.  Dressier  3  221  (MO);  Cer- 
ro Jefe,  /.  Dwyer  &  A.  Gentry  8489  (MO);  near 
top  of  Cerro  Jefe  to  1  mi.  beyond,  A.  Gentry  et 
al.  3440  (MO);  La  Eneida,  Cerro  Jefe  region,  P. 
Maas  &  R.  Dressier  677  (MO);  5-10  km  NE  of 
Altos  de  Pacora,  "Campamento  Quatro,"  Gor- 
gas  Memorial  Labs  yellow  fever  research  camp, 
S.  Mori  &  J.  Kallunki  3424  (MO,  NY);  Cerro 
Jefe,  near  jet.  of  rds.  to  Cerro  Jefe  and  Altos  de 
Pacora,  S.  Mori  et  al.  3776  (MO);  rd.  from  Cerro 
Jefe  to  Cerro  Azul,  E.  Tyson  5306  (MO);  Cerro 
Jefe  between  Cerro  Azul  and  La  Eneida,  ca.  1 5 
mi  NE  of  Panama  City,  R.  Wilbur  et  al.  15 54 IB 
(DUKE,  MO).  San  Bias:  Serrania  de  San  Bias, 
Nusagandi  along  El  Llano-Carti  Rd.,  R.  Read  et 
al.  85-6B  (US). 

ACKNOWLEDGMENTS 

We  are  grateful  to  Frank  Almeda  for  assistance 
in  the  field  and  for  making  a  special  attempt  to 
recollect  plant  material,  to  Gordon  McPherson 
for  assistance  in  the  field  and  for  facilitating  the 
senior  author's  stay  in  Panama,  Sue  Thompson 


for  providing  a  collection  of  one  species,  Tsan 
and  Fei-Mei  Chuang  for  making  the  chromo- 
some count,  Alice  Tangerini  for  skillfully  pre- 
paring the  line  drawings,  and  the  curators  of  the 
following  herbaria  for  loans  or  other  courtesies: 
BM,  CAS,  DUKE,  F,  GH,  MEXU,  MO,  NY,  U, 
US. 

RESUMEN 

Se  describen  e  ilustran  tres  especies  nuevas  de 
Justicia  de  Panama:  J.fortunensis,  J.  veraguensis, 
y  J.  readii.  Se  presenta  un  mapa  que  muestra  la 
distribucion  de  cada  una  de  las  especies.  Las  tres 
especies  nuevas  parecen  tener  afinidades  con  /. 
refractifolia  y  una  clave  de  estas  cuatro  especies 
es  presentada.  Se  reporta  un  numero  de  cro- 
mosomas  de  n  =  14  para  J.fortunensis.  Con  23 
especies  en  el  pais,  Justicia  es  el  genero  mas  gran- 
de  de  Acanthaceae  en  Panama. 

LITERATURE  CITED 

DANIEL,  T.  F.,  B.  D.  PARFITT,  AND  M.  A.  BAKER.  1984.  Chro- 
mosome numbers  and  their  systematic  implications  in  some 
North  American  Acanthaceae.  Syst.  Bot.  9:346-355. 

DANIEL,  T.  F.  AND  T.  I.  CHUANG.  1989.  Chromosome  num- 
bers of  some  cultivated  Acanthaceae.  Baileya  23:86-93. 

DANIEL,  T.  F.,  T.  I.  CHUANG,  AND  M.  A.  BAKER.  1990.  Chro- 
mosome numbers  of  American  Acanthaceae.  Syst.  Bot.  15: 
13-25. 

DWYER,  J.  D.  1967.  A  new  herbarium  in  the  Canal  Zone. 
Taxon  16:158-159. 

.  1985.  The  history  of  plant  collecting  in  Panama.  Pp. 

179-183  in  The  Botany  and  Natural  History  of  Panama:  La 
Botanica  e  Historia  Natural  de  Panama.  W.  D'Arcy  and  M. 
Correa  A.,  eds.  Monographs  in  Systematic  Botany  from  the 
Missouri  Botanical  Garden  10. 

DURKEE,  L.  H.  1978.  Acanthaceae.  Pp.  155-284,  in  Flora  of 
Panama.  R.  Woodson  et  al.,  eds.  Ann.  Missouri  Bot.  Gard. 
65. 

.  1986.  Acanthaceae.  Pp.  1-92  in  Flora  Costaricensis. 

W.  Burger,  ed.  Fieldiana,  Bot.  18. 

.    1988.    A  checklist  of  Acanthaceae  in  Costa  Rica, 


Nicaragua,  and  Panama.  Acanthus  3:3-4. 

GRAHAM,  V.  A.  1988.  Delimitation  and  infra-generic  clas- 
sification of  Justicia  (Acanthaceae).  Kew  Bull.  43:551-624. 

LEONARD,  E.  C.  1958.  The  Acanthaceae  of  Colombia,  III. 
Contr.  U.S.  Natl.  Herb.  31:323-781. 

MABBERLEY,  D.  1987.  The  Plant-Book.  Cambridge  Univ. 
Press,  Cambridge. 


CALIFORNIA  ACADEMY  OF  SCIENCES 

Golden  Gate  Park 
San  Francisco,  California  94 1 1 8 


PROCEEDINGS 

OF  THE 

CALIFORNIA  ACADEMY  OF  SCIENCES 


Vol.  46,  No.  14,  pp.  299-326,  12  figs.  September  11,  1990 


NEW  SPECIES  AND  NEW  COMBINATIONS  IN  BLAKEA  AND 

TOPOBEA  (MELASTOMATACEAE),  WITH  AN  HISTORICAL 

PERSPECTIVE  ON  GENERIC  LIMITS  IN 

THE  TRIBE  BLAKEEAE 

By 
Frank  Almeda 

Department  of  Botany,  California  Academy  of  Sciences,  Golden  Gate  Park, 
San  Francisco,  California  94118 


ABSTRACT:  Blakea  P.  Browne  and  Topobea  Aublet,  the  two  currently  recognized  genera  in  the  Blakeeae  have 
been  variously  combined  and  segregated  during  their  long  history.  A  synoptic  taxonomic  history  of  these 
genera  is  presented  with  special  attention  given  to  characters  used  to  combine  or  separate  them.  In  keeping 
with  traditional  generic  concepts,  Blakea  and  Topobea  are  maintained  and  a  revised  generic  key  is  provided 
to  accommodate  anomalous  or  unusual  Mesoamerican  species.  Descriptions,  diagnostic  illustrations,  discus- 
sions, and  distribution  maps  are  presented  for  four  new  species  of  Blakea  (B,  gregii,  B.  hammelii,  and  K. 
herrerae  from  Panama;  and  K.  scarlatina  from  Costa  Rica  and  Nicaragua)  and  three  new  species  of  Topobea 
(T.  fragrantissima,  T.  hexandra,  and  T.  suaveolens  from  Panama).  The  study  of  recently  collected  flowering 
material  of  B.  crassifolia  and  K.  parvifolia  necessitates  their  transfer  to  Topobea  and  the  adoption  of  a  new 
name,  T.  caliginosa,  for  the  narrow  Panamanian  endemic  previously  known  as  B.  micrantha. 

Received  February  2,  1990.  Accepted  March  14,  1990. 

The  striking  morphological  isolation  of  the 
INTRODUCTION  „,  ,  . ..  r  A   JT     ...... 

Blakeeae  within  the  family  and  the  homogeneity 

The  largely  pantropical  Melastomataceae  of  its  constituent  taxa  undoubtedly  account  for 

comprise  some  200  genera  and  approximately  the  fact  that  its  tribal  status  has  not  been  seriously 

4,800  species.  This  comparatively  large  natural  challenged  since  it  was  established  by  Bentham 

family  has  traditionally  been  divided  into  four-  and  Hooker  (1867).  However,  Blakea  P.  Browne 

teen  tribes  based  on  fruit  characters,  staminal  and  Topobea  Aublet,  the  two  currently  recog- 

features,  and  seed  morphology.  Of  the  eight  tribes  nized  genera  in  the  tribe,  have  been  variously 

restricted  to  the  western  hemisphere,  only  the  combined  and  segregated  during  their  long  his- 

Blakeeae  and  Miconieae  are  berry-fruited.  The  tory. 

Blakeeae  are  readily  recognized  by  their  axillary  Recent  fieldwork  done  in  connection  with  the 

flowers  that  are  individually  subtended  by  two  preparation  of  a  treatment  of  Melastomataceae 

pairs  of  decussate  bracts,  ovoid  to  pyramidal  for  Flora  Mesoamericana  has  resulted  in  the  dis- 

smooth  seeds,  and  wood  with  multiseriate  rays  covery  of  several  undescribed  species  in  the 

and  frequent  occurrence  of  druses  (Koek-Noor-  Blakeeae.  An  effort  to  devise  a  rational  basis  for 

man  et  al.  1979;  ter  Welle  and  Koek-Noorman  assigning  these  taxa  to  one  genus  or  the  other  has 

1981).  necessitated  another  look  at  generic  limits  in  the 

[299] 


300 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  14 


tribe.  I  recently  expressed  serious  reservations 
about  the  continued  recognition  of  Topobea  and 
even  suggested  that  it  be  submerged  in  the  older 
and  larger  genus  Blakea  (Almeda  1 989).  Addi- 
tional study  of  this  problem,  coupled  with  the 
paucity  of  information  on  pollen  morphology 
(Patel  et  al.  1 985)  and  chromosome  numbers  (Solt 
and  Wurdack  1980),  has  led  me  to  refrain  from 
merging  Blakea  and  Topobea  without  a  compre- 
hensive study  throughout  the  extensive  neotrop- 
ical range  of  these  genera. 

The  following  summary  provides  a  synoptic 
taxonomic  history  of  Blakea  and  Topobea  to- 
gether with  an  assessment  of  the  diagnostic  fea- 
tures used  to  combine  or  separate  them.  This  is 
followed  by  descriptions  of  four  new  species  of 
Blakea,  three  new  species  of  Topobea,  and  three 
generic  transfers.  Working  within  the  constraints 
of  a  geographically  denned  study,  my  decision 
to  recognize  Blakea  and  Topobea  is,  of  necessity, 
tempered  by  the  historical  consensus  of  opinion 
among  specialists  in  the  family  and  the  need  for 
more  data.  Nevertheless,  a  slight  modification  of 
original  generic  concepts  is  necessary  to  accom- 
modate the  accretion  of  new  and  unusual  Central 
American  species. 

SYNOPTIC  HISTORY  OF 
BLAKEA  AND  TOPOBEA 

The  genus  Blakea  was  first  defined  by  Patrick 
Browne  (1756)  for  a  Jamaican  species  that  was 
subsequently  given  the  binomial  B.  trinervia  by 
Linnaeus  (1759).  Browne's  circumscription  of 
Blakea  included  a  plate  and  a  brief  diagnosis 
describing  it  as  a  shrub  with  nitid  trinerved  el- 
liptic leaves,  6-merous  axillary  flowers,  and  tri- 
angular laterally  connate  anthers.  The  plate  is 
generalized  and  provides  no  details  of  anther 
morphology.  The  terse  description  is  sufficient, 
however,  to  preclude  confusion  with  the  allied 
genus  Topobea,  which  was  described  and  illus- 
trated by  Aublet  (1775)  from  material  collected 
in  French  Guiana.  Although  many  of  Aublet's 
published  drawings  are  mixtures  and  some  of  his 
descriptions  are  composites  of  discordant  ele- 
ments (Howard  1983),  both  the  plate  and  de- 
scription of  T.  parasitica  emphasize  the  oblong 
subulate  anthers  that  are  typical  of  so  many 
species  in  this  genus.  Curiously,  Aublet  over- 
looked or  chose  not  to  describe  and  illustrate  the 
broad,  gaping,  dorsally  inclined,  confluent  anther 
pores.  This  is  another  diagnostic  character  that 


figures  prominently  in  decisions  to  maintain  To- 
pobea as  a  distinct  genus.  In  choosing  the  epithet 
parasitica,  Aublet  was  responsible  for  perpetrat- 
ing a  misunderstanding  of  the  epiphytic  habit 
that  is  common  to  so  many  species  of  Blakea 
and  Topobea. 

The  younger  Linnaeus  (1781)  and  de  Jussieu 
(1789)  maintained  Blakea  and  Topobea  as  de- 
fined by  Aublet.  By  retaining  Blakea  quinque- 
nervia  Aublet  [=Bellucia  grossularioides  (L.) 
Triana]  in  Blakea,  it  is  clear  that  their  interpre- 
tation was  a  broad  one  that  included  at  least  two 
genera  now  relegated  to  two  different  berry-fruit- 
ed tribes. 

David  Don  (1823),  in  an  early  paper  devoted 
to  the  Melastomataceae,  was  the  first  to  part  ranks 
with  his  predecessors  by  merging  Topobea  and 
Blakea.  His  decision,  however,  was  based  on 
characters  other  than  anther  morphology.  The 
weakness  of  Don's  argument  is  best  conveyed  by 
his  own  words:  "Although  the  Topobea  of  Aublet 
recedes  somewhat  from  Blakea,  in  its  being  par- 
asitical; yet,  notwithstanding,  in  the  Lambertian 
Herbarium  are  several  unpublished  species,  from 
Don  Jose  Pavon,  natives  of  Peru,  and  not  par- 
asitical, which  agree  with  Topobea  in  every  es- 
sential point;  and  these  also,  accord  well  with 
Blakea,  except  in  having  four,  instead  of  six  scales, 
surrounding  the  calyx,  which,  however,  is  a  vari- 
able character;  and,  therefore  I  think  myself  jus- 
tified in  uniting  these  two  genera."  De  Candolle 
(1828)  also  adopted  this  interpretation  in  his 
Prodromus.  Like  Don  and  his  immediate  pre- 
decessors, de  Candolle's  broad  generic  concept 
of  Blakea  included  species  now  recognized  in 
Belinda. 

Naudin  (1852)  rejected  this  inclusive  dispo- 
sition and  emphasized  diversity  in  anther  mor- 
phology and  habit  as  reasons  for  resurrecting  To- 
pobea. He  provided  no  involved  explanation  to 
justify  his  decision,  yet  he  described  a  new  genus, 
Pyxidanthus,  to  accommodate  three  new  species 
that  he  considered  closely  related  to  Blakea.  The 
concept  of  Blakea  and  Topobea  espoused  by 
Naudin  was  faithfully  followed  by  Bentham  and 
Hooker  (1867)  without  comment. 

Triana  (1871),  as  astute  specialist  in  the  Me- 
lastomataceae, synonymized  Pyxidanthus  under 
Blakea  along  with  Valdesia  Ruiz  and  Pavon, 
another  minor  generic  segregate.  Pyxidanthus, 
Valdesia,  and  the  subsequently  described  Ama- 
raboya  Linden  are  now  universally  accepted  syn- 
onyms of  Blakea.  As  recognized  by  Triana  (1 87 1), 


ALMEDA:  NEW  BLAKEA  AND  TOPOBEA 


301 


Blakea  included  those  species  with  short,  oval 
or  elliptic,  apically  blunt  or  rounded  anthers  with 
two  minute  well-separated  pores.  Topobea  in- 
cluded those  species  with  linear-oblong  or  subu- 
late anthers  with  two  confluent  dorsally  inclined 
pores.  Baillon  ( 1879),  who  was  noted  for  his  broad 
view  of  generic  limits,  was  unimpressed  by  these 
differences  in  anther  morphology.  He  returned 
Topobea  to  the  synonymy  of  Blakea  but  provid- 
ed no  convincing  reasons  aside  from  a  comment 
describing  the  anthers  of  Topobea  as  more  elon- 
gate and  narrower  than  those  in  Blakea.  For  a 
botanist  who  excelled  in  character  analysis,  Bail- 
Ion's  brief  treatment  of  these  genera  reflects  a 
superficial  understanding  of  the  anther  differ- 
ences between  them. 

The  most  comprehensive  and  perhaps  most 
influential  accounts  of  generic  limits  in  the  Blake- 
eae  are  those  of  Cogniaux  (1888,  1891).  Cog- 
niaux  was  undoubtedly  influenced  by  Triana. 
Unlike  Triana,  however,  he  was  the  first  to  pro- 
vide the  following  explicit  key  clarifying  the  dis- 
tinctions between  Blakea  and  Topobea  as  he 
understood  them: 

A.  Staminum  filamenta  crassiuscula,  anthe- 
rae  breves,  obtusae  a  latere  compressae, 
connective  crassisimo,  postice  calcarato  .... 
Blakea 

B.  Staminum  filamenta  filiformia;  antherae 
lineari  vel  oblongo-subulatae,  rostratae, 
connective  mediocri,  interdum  ecalcarato 
Topobea 

These  differences  amply  characterized  the 
species  then  known  to  Cogniaux,  but  they  cannot 
be  used  for  a  strict  characterization  of  the  two 
genera  at  present.  The  short,  obtuse,  laterally 
compressed  anthers  of  Blakea  (Fig.  2  A,  C,  E;  Fig. 
3A,  E)  versus  the  linear  to  oblong-subulate  ros- 
trate anthers  of  Topobea  (Fig.  IE,  F;  2B;  3B,  C) 
are  the  only  modal  contrasts  that  appear  to  be 
valid  for  distinguishing  the  majority  of  species 
now  assigned  to  these  genera.  The  differences  in 
filament  thickness  and  presence  or  absence  of 
dorsal  appendages  on  the  connective  are  incon- 
sistent criteria  for  differentiating  Blakea  and  To- 
pobea. Like  his  predecessors,  Cogniaux  neglected 
to  emphasize  that  Blakea  has  anthers  with  two 
typically  minute  and  well-separated  pores  and 
Topobea  has  two  (rarely  one)  broad  approximate 
or  confluent  dorsally  inclined  pores.  The  generic 
concepts  promoted  by  Cogniaux  have  been 
adopted  by  all  subsequent  specialists  (Almeda, 


in  press;  Gleason  1935,  1958;  Wurdack  1973, 
1980)  and  other  authors  of  major  regional  treat- 
ments of  the  Melastomataceae  (Standley  1924, 
1 938;  Standley  and  Williams  1 963),  except  Mac- 
bride  (1941).  Invoking  a  rationale  similar  to  that 
of  Baillon,  Macbride  combined  Blakea  and  To- 
pobea and  commented  on  the  inconsistency  of 
the  filament  character  used  as  a  diagnostic  fea- 
ture by  Cogniaux. 

In  a  paper  describing  many  new  species  of 
Blakea  and  Topobea,  Gleason  (1945)  sympa- 
thized with,  but  rejected,  Macbride's  view.  The 
close  relationship  between  these  two  genera  is 
readily  apparent  in  their  many  parallel  and  over- 
lapping variations  in  habit,  leaves,  indument, 
floral  bracts,  petals,  calyx  lobes,  ovary  cell  num- 
ber, and  seeds.  Although  Baillon  (1879)  and 
Macbride  (1941)  intimated  that  anther  charac- 
ters provide  insufficient  grounds  on  which  to  base 
the  recognition  of  genera,  Gleason  (1945,  1947), 
who  defended  their  use,  pointed  out  that  speci- 
mens can  be  sorted  precisely  into  one  genus  or 
the  other  when  staminal  material  is  available. 

According  to  Patel  (in  litt,  18  July  1985),  pre- 
liminary results  of  a  numerical  taxonomic  anal- 
ysis of  pollen  (using  SEM  data)  in  the  Melasto- 
mataceae reveal  that  Blakea  and  Topobea  sort 
out  into  different  clusters.  Because  pollen  of  Me- 
lastomataceae is  diverse  exomorphically  (Patel 
et  al.  1985),  an  expanded  survey  is  needed  to 
extend  and  refine  these  observations. 

Until  more  information  about  micromor- 
phological  characters  such  as  chromosome  num- 
bers, pollen  morphology,  and  chemistry  is  avail- 
able, it  seems  preferrable  to  adhere  to  traditional 
generic  concepts  until  we  can  rest  decisions  on  a 
broader  data  base  than  was  available  to  our  pre- 
decessors. 

This,  however,  does  not  ease  the  current  dif- 
ficulty of  assigning  the  few  unusual  species  to  one 
genus  or  the  other.  In  the  Mesoamerican  region, 
at  least  four  different  species  groups  are  problem- 
atic in  determining  where  to  draw  the  line  be- 
tween Blakea  and  Topobea.  I  will  consider  the 
relationships  among  Blakea  wilburiana  Almeda 
and  allied  species  in  a  future  paper.  The  other 
three  species  groups  are  mentioned  here  because 
at  least  one  species  from  each  is  newly  described 
below. 

The  first  of  these,  Blakea  gregii,  belongs  to  a 
group  of  five  closely  related  species  characterized 
by  nectar-producing  pendant  flowers  with  green 
petals.  The  anthers  of  B.  gregii  are  elliptic-oblong 


302 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  14 


FIGURE  1 .  Representative  stamen  morphology  in  Blakea  and  Topobea.  A,  B.  wilburiana  Almeda;  B,  B.  austin-smithii  Standley; 
C,  B.  purpusii  Brandegee;  D,  T.  crassifolia  (Almeda)  Almeda;  E,  T.  pittieri  Cogn.;  F,  T.  brenesii  Standley.  (A  from  Almeda  et 
al  6525,  CAS;  B  from  Almeda  et  al.  2145,  CAS;  C  from  Breedlove  &  Almeda  47747,  CAS;  D  from  de  Nevers  5943,  CAS;  E 
from  Almeda  et  al.  3161,  CAS;  F  from  Wilbur  et  al.  15809,  DUKE.) 


and  laterally  compressed  with  two  dorsally  in- 
clined approximate  pores  (Fig.  4G).  Anther  con- 
formation is  like  that  of  Blakea  but  the  dorsal 
inclination  of  the  apical  pores  is  more  like  that 


encountered  in  Topobea.  The  pore  width,  how- 
ever, is  narrower  than  that  typically  found  in 
many  species  of  the  latter  genus.  In  deciding  to 
place  this  species  in  Blakea,  I  was  influenced  by 


ALMEDA:  NEW  BLAKEA  AND  TOPOBEA 


303 


FIGURE  2.  Representative  stamen  morphology  in  Blakea  and  Topobea.  A,  B.  elliptica  (Gleason)  Almeda;  B,  T.  calycularis 
Naudin;  C,  B.  woodsonii  Gleason;  D,  T.  storkii  Standley;  E,  B.  tuberculata  J.  D.  Smith.  (A  from  Almeda  et  al.  6261,  CAS;  B 
from  Breedlove  50940,  CAS;  C  from  Almeda  et  al.  3046,  CAS;  D  from  Almeda  et  al.  2884,  CAS;  E  from  Almeda  &  Nakai  4529, 
CAS.) 


the  somewhat  more  typical  Blakea-type  anthers 
of  its  close  relatives,  B.  austin-smithii  Standley 
(Fig.  IB)  and  B.  purpusii  Brandegee  (Fig.  1C). 
Topobea  hexandra  has  been  the  most  difficult 


to  place  genetically.  It  is  part  of  an  unusual  as- 
semblage in  the  Blakeeae  distinguished  by  small 
sessile  or  short-pedicellate  flowers  with  only  six 
antesepalous  stamens.  Its  close  relatives  include 


304 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  14 


C   I 


FIGURE  3.  Representative  stamen  morphology  in  Blakea  and  Topobea.  A,  B.  litoralis  L.  O.  Williams;  B,  T.  parvifolia  (Gleason) 
Almeda;  C,  T.  albertieae  Wurdack;  D,  B.  hirsuta  Triana  var.  rotundata  Markgraf;  E,  B.  anomala  J.  D.  Smith.  (A  from  Almeda 
&  Nakai  4189,  CAS;  B  from  Almeda  et  al.  5898,  CAS;  C  from  Almeda  et  al.  6153,  CAS;  D  from  Lugo  4140,  CAS;  E  from 
Almeda  2655,  CAS.) 


T.  cordata  Gleason,  T.  caliginosa  Almeda,  and 
T.  crassifolia  (Almeda)  Almeda.  Stamens  of  T. 
cordata  were  unknown  to  Gleason  (1950)  when 
he  proposed  it  as  a  new  species.  His  decision  to 
place  it  in  Topobea  was  based  on  its  superficial 
resemblance  to  some  sessile-leaved  species  of  that 
genus.  Although  each  of  the  species  in  this  com- 
plex is  known  from  very  few  flowering  collec- 


tions, available  anther  material  exhibits  inter- 
esting variations  on  the  Topobea  theme.  With 
the  exception  of  T.  hexandra,  all  have  anthers 
with  two  broad  confluent  pores.  In  T.  cordata, 
the  anthers  are  oblong-subulate  (5-5.5  mm  long) 
and  the  confluent  pores  are  somewhat  ventrally 
inclined.  In  T.  caliginosa  and  T.  crassifolia  the 
anthers  are  linear-oblong  but  shorter  (2-2.5  mm) 


ALMEDA:  NEW  BLAKEA  AND  TOPOBEA 


305 


with  truncate  to  somewhat  dorsally  inclined  pores 
(Fig.  1 D).  The  anthers  of  T.  hexandra  are  also 
short  (2  mm)  and  oblong,  but  they  are  laterally 
compressed  and  the  two  apical  pores,  although 
broad,  are  ventrally  inclined  and  separated  from 
one  another  (Fig.  1 1 F).  In  the  context  of  varia- 
tion exhibited  by  its  close  relatives,  the  sum- 
mation of  anther  characters  in  this  species  argues 
for  placement  in  Topobea.  Without  these  in- 
sights, one  might  be  inclined  to  assign  it  to  Blak- 
ea. 

The  last  of  the  problematic  species  to  be  con- 
sidered here  is  T.  suaveolens  Almeda.  This  species 
is  unusual  in  having  fragrant  pendant  flowers 
with  brightly  colored  petals  and  connivent  an- 
thers that  form  a  ring  around  the  exserted  style 
(Fig.  1 2C).  The  essentially  elliptic  shape  and  con- 
spicuous lateral  compression  of  the  anthers  (Fig. 
12E)  are  reminiscent  of  Blakea-lype  anthers  but 
the  broad  solitary  dorsally  inclined  apical  pore 
is  a  feature  known  only  in  Topobea.  Again,  my 
generic  placement  of  this  species  is  influenced  by 
its  close  relationship  to  T.  fragrantissima  Al- 
meda (Fig.  10),  a  species  with  androecial  char- 
acters that  can  only  be  attributed  to  Topobea. 

Interestingly,  the  anomalous  androecial  mor- 
phology in  the  cases  discussed  above  all  appear 
to  represent  derived  syndromes.  Each  is  associ- 
ated with  a  unique  pollen  presentation  mecha- 
nism (T.  suaveolens),  a  specialized  pollinator 
spectrum  (see  discussion  under  B.  gregii),  or  great 
reduction  in  flower  size  (T.  hexandra).  Blakea, 
with  about  1 00  described  species,  and  Topobea, 
with  over  60  species,  are  large  and  diverse,  closely 
related  genera.  If  more  unusual  anther  forms  come 
to  light,  it  may  become  increasingly  difficult  to 
separate  them  without  establishing  arbitrary  lim- 
its or  fueling  counterarguments  to  submerge  To- 
pobea. For  floristic  purposes,  I  find  it  expedient 
to  recognize  both  genera.  This  is  done  for  con- 
venience and  to  preclude  extensive  nomencla- 
tural  changes  pending  a  comprehensive  study  of 
the  tribe.  It  is  not  done  out  of  a  firm  conviction 
that  generic  status  is  necessarily  the  best  way  to 
reflect  relationships  based  on  currently  perceived 
differences.  Accordingly,  the  following  revised 
key  to  these  genera  is  offered  in  an  attempt  to 
incorporate  the  most  obvious  Mesoamerican 
anomalies  discovered  to  date: 

Stamens  1 2  in  number;  anthers  2-pored,  oval, 
oblong,  or  elliptic,  compressed  laterally, 
bluntly  obtuse  or  broadly  rounded  at  the 


summit  with  two  typically  well-separated 

(often  minute)  apical  pores Blakea 

Stamens  6,  8,  or  12  in  number;  anthers  1 -pored 
or  2-pored,  linear-oblong  to  oblong-subu- 
late or  rostrate,  usually  not  compressed  lat- 
erally (if  conspicuously  compressed  then 
1  -pored  or  dorso-basally  appendiculate)  with 
approximate  or  confluent  dorsally  inclined 
pores  (if  pores  are  prominently  inclined  ven- 
trally then  flowers  are  hexandrous  or  pedi- 
cels are  beset  with  spreading  brown  hairs  1- 
2.5  mm  long) Topobea 

NEW  TAXA  AND  NEW  COMBINATIONS 
Blakea  gregii  Almeda,  sp.  nov. 

(Fig.  4) 

TYPE:  — PANAMA.  Chiriqui:  Cerro  Pate  Macho,  windswept 
ridge,  8°49'N,  82°24'W,  elev.  2,100  m,  17  Jan.  1986,  deNevers 
&  McPherson  6840  (holotype:  CAS!  isotypes:  MEXU!,  MO!, 
PMA!,  US!). 

Frutex  ca.  2—4  m  altus.  Ramuli  primum  quadrangulati  de- 
mum  teretes  sicut  petioli  laminarum  subtus  pedicelli  bracte- 
aeque  pilis  1-2.5  mm  longis  induti.  Petioli  1.8-3.5  cm  longi; 
lamina  9.3-14.5  x  5-9  cm  elliptica  vel  elliptico-ovata  apice 
acuminata  basi  obtusa  vel  rotundata  vel  late  acuta,  5-nervata 
vel  5-plinervata,  coriacea  et  integra,  nervis  secundariis  3-6  mm 
inter  se  distantibus.  Flores  6-meri  penduli  in  quoque  nodo 
superiore  plerumque  1-4,  pedicellis  1.2-2  cm  longis;  bracteae 
exteriores  7-8  x  5-8  mm  ellipticae  vel  ovato-ellipticae  ad 
basim  paulo  (2  mm)  coalitae;  bracteae  interiores  omnino  libe- 
rae  5-6  x  4-7  mm  ovatae  vel  suborbiculares  apice  truncato- 
rotundato.  Hypanthium  (ad  torum)  6  mm  longum  extus  gla- 
brum  vel  sparse  subamorpho-furfuraceum;  calycis  tubus  2-3 
mm  longus,  lobis  2-4  mm  longis  extus  sparsiuscule  strigosis. 
Petala  13-14  x  9-11  mm  obovata  apice  rotundato  sparse  ca- 
duceque  glanduloso-ciliolata.  Filamenta  6-7  mm  longa;  an- 
therae  4-4.5  x  2  cm  subulatae  lateraliter  non  cohaerentes, 
dorsaliter  biporosae;  connectivum  dorsaliter  ca.  0.5  mm  supra 
thecarum  basim  inconspicue  tuberculatum.  Stylus  1.2-1.3  cm; 
ovarium  6-loculare  et  Vi  inferum  apice  glabro  (cono  et  collo 
non  evoluto). 

Terrestrial  shrubs  2-4  m  tall.  Young  inter- 
nodes  quadrate,  becoming  rounded  with  age. 
Distal  branchlets,  vegetative  buds,  petioles,  and 
pedicels  densely  covered  with  a  brown  indument 
of  appressed,  ±  curved  subulate  hairs  mostly  1- 
2.5  mm  long.  Mature  leaves  of  a  pair  essentially 
equal  in  size;  petioles  1.8-3.5  cm  long;  blades 
coriaceous,  9.3-14.5  cm  long  and  5-9  cm  wide, 
elliptic  to  elliptic-ovate,  apex  acuminate,  base 
obtuse  to  rounded  but  sometimes  varying  to 
broadly  acute,  margin  entire  and  somewhat  revo- 
lute  when  dry,  5 -nerved  or  5-plinerved,  the  in- 
nermost pair  of  primaries  diverging  from  the  me- 
dian nerve  4-6  mm  above  the  blade  base,  the 


306 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  14 


2.6cm 


B 


FIGURE  4.  Blakea  gregii  Almeda.  A,  habit;  B,  representative  leaf  with  enlargement  (abaxial  surface);  C,  representative  flower 
(natural  posture)  with  petals  and  stamens  removed;  D,  outer  floral  bract  (abaxial  surface);  E,  inner  floral  bract  (abaxial  surface); 
F;  petals  (adaxial  surface);  G,  stamens,  lateral  view  (left),  ventral  view  (middle),  dorsal  view  (right).  (A  from  Almeda  et  al.  6137, 
CAS;  B-G  from  the  holotype.) 


ALMEDA:  NEW  BLAKEA  AND  TOPOBEA 


307 


elevated  network  of  transverse  secondaries  spaced 
3-6  mm  apart  at  the  widest  portion  of  the  blade, 
the  young  unexpanded  leaves  beset  with  a  de- 
ciduous indument  of  white  wooly  hairs  but  typ- 
ically becoming  glabrous  adaxially  at  maturity 
and  moderately  to  copiously  covered  abaxially 
with  smooth  appressed  to  ±  curved  brown  hairs 
1.5-2.5  mm  long.  Flowers  pendant,  solitary, 
paired,  or  in  fascicles  of  4  in  the  leaf  axils  of 
distal  branches;  pedicels  1.2-2  cm  long.  Floral 
bracts  thick  and  semisucculent,  sessile  and  en- 
tire; outer  bracts  connate  basally  for  2  mm,  7-8 
x  5-8  mm,  elliptic  to  elliptic-ovate,  apex  acute, 
densely  covered  abaxially  with  appressed  to  in- 
curved smooth  brown  hairs;  inner  bracts  free  to 
the  base,  5-6  x  4-7  mm,  ovate  to  subrotund, 
apex  rounded  to  truncate  varying  to  retuse  with 
a  bluntly  acute  median  lobe,  sparingly  to  mod- 
erately beset  with  appressed  brown  hairs  that  are 
typically  concentrated  around  or  restricted  to  the 
median  apex.  Hypanthia  (at  anthesis)  campan- 
ulate  to  suburceolate,  6  mm  long  to  the  torus  and 
7-8  mm  in  diameter,  glabrous  or  sparingly  brown 
furfuraceous  toward  the  base.  Calyx  tube  2-3 
mm  long;  calyx  lobes  erect,  2-4  mm  long  and  4- 
5  mm  wide  basally,  ovate  to  deltoid-ovate  with 
a  blunt  callose-thickened  tooth  at  the  abaxial  apex 
of  each  lobe,  roughened  along  interlobe  sinuses, 
adaxially  glabrous,  abaxially  covered  with  a  sparse 
indument  of  tardily  deciduous  short  incurved 
brown  hairs  that  are  restricted  to  or  largely  con- 
centrated near  the  apex.  Petals  6,  glabrous,  con- 
nivent  to  somewhat  imbricate  when  fully  ex- 
panded, 13-14  x  9-11  mm,  green,  obovate, 
apically  rounded,  entire  but  irregularly  glandu- 
lar-ciliate.  Stamens  1 2,  isomorphic,  free  and  en- 
circling the  exserted  style;  filaments  complanate 
and  glabrous,  6-7  mm  long;  anthers  4-4.5  mm 
long,  2  mm  wide,  elliptic-oblong,  laterally  com- 
pressed, each  anther  tipped  with  two  dorsally 
inclined  approximate  pores;  connective  dilated 
dorso-basally  ca.  0.5  mm  above  the  filament  in- 
sertion into  a  blunt  callose  spur.  Ovary  Vi  infe- 
rior, 6-celled,  glabrous  and  lobulate  at  the  apex 
but  not  expanded  into  a  cone  or  collar.  Style 
straight,  glabrous,  1.2-1.3  cm  long,  conspicu- 
ously exserted  beyond  the  petals  and  calyx  lobes; 
stigma  truncate.  Berry  globose,  about  6  mm  long 
and  8  mm  in  diameter.  Seeds  clavate  to  narrowly 
pyriform,  ca.  1  mm  long,  beige  with  a  smooth 
glossy  testa  and  a  conspicuous  lateral  raphe. 
PHENOLOGY.— Flowering  specimens  have  been 


collected  in  January;  the  only  fruiting  specimen 
was  collected  in  March. 

DISTRIBUTION.— A  little-collected  cloud  forest 
or  elfin  forest  species  known  only  from  Chiriqui 
Province,  Panama  in  the  region  extending  from 
Cerro  Horqueta  to  Cerro  Pate  Macho  at  1,800- 
2,200  m  (Fig.  5). 

ADDITIONAL  SPECIMENS  EXAMINED.— PANAMA.  Chiriqui: 
S  slopes  of  Cerro  Horqueta,  21  Jan.  1971,  Wilbur  et  al.  13486 
(CAS,  DUKE,  MO);  Cerro  Pate  Macho,  8°49'N,  82°24'W,  13 
Mar.  1988,  Almeda  et  al.  6137  (CAS,  CR,  MO,  NY,  PMA, 
TEX). 

Blakea  gregii  is  closely  related  to  four  other 
species  characterized  by  cryptic,  pendant  flowers 
with  green,  connivent  petals  and  nocturnal  or 
crepuscular  nectar  production  (Almeda  1980, 
198  la).  The  green-flowered  species  are  typically 
cloud  forest  hemiepiphytes  that  germinate  ter- 
restrially, ascend  nearby  trees  by  adventitious 
roots,  and  ultimately  become  epiphytic  by  losing 
root  contact  with  the  soil.  Both  B.  chlorantha 
and  B.  gregii,  however,  are  often  conspicuous 
terrestrial  shrubs  on  windswept  ridges  of  the  con- 
tinental divide  where  the  vegetation  is  common- 
ly reduced  to  elfin  woodland. 

Intensive  field  studies  by  Lumer  (1981)  and 
Lumer  and  Schoer  (1986)  have  shown  that  B. 
austin-smithii,  B.  chlorantha,  and  B.  penduliflora 
are  visited  and  pollinated  by  six  species  of  ro- 
dents belonging  to  four  genera:  Oryzomys,  Pero- 
myscus,  Reithrodontomys,  and  Scotinomys.  Lu- 
mer and  Schoer  (1986)  suggest  that  rodents  may 
indeed  play  a  significant  role  as  pollinators  of 
flowering  plants  in  the  harsh  windy  cloud  forest 
environment.  Nothing  is  known  about  the  pol- 
lination of  B.  gregii,  but  similarities  in  floral  mor- 
phology and  habitat  suggest  that  small  rodents 
may  also  play  an  important  role  in  the  repro- 
ductive biology  of  this  species. 

To  facilitate  comparison,  diagnostic  characters 
of  the  five  species  of  Blakea  with  pendant  green 
flowers  are  summarized  in  the  following  key: 

1 .  Leaf  blades  modified  adaxially  at  the  pet- 
iole-laminar junction  into  flap-like  saccate 
pouches,  the  basal  and  lateral  margins  of 
which  are  free  from  but  conspicuously  de- 
current  on  the  petiole. 
2.  Pubescence  of  uppermost  internodes 
consisting  of  hairs  1-2  mm  long;  leaf 
blades  bluntly  denticulate  and  revolute 
on  drying;  outer  floral  bracts  linear-Ian- 


308 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  14 


A   B.  gregii 

•  B.  scarlatina 

•  T.  fragrantissima 


-ioc 


100  200  km 


FIGURE  5.    Distributions  of  Blakea  gregii,  B.  scarlatina,  and  Topobea  fragrantissima. 


ceolate,  1 5-2 1  mm  long,  equaling  or  ex- 
ceeding the  calyx  lobes  at  anthesis;  Cos- 
ta Rica  (Cerro  Chompipe,  Volcan  Barva, 

and  Volcan  Irazu) 

_ B.  austin-smithii  Standley 

2.  Pubescence  of  uppermost  internodes 
consisting  of  hairs  mostly  less  than  0.5 
mm  long;  leaf  blades  entire  and  revolute 
on  drying;  outer  floral  bracts  ovate  to 
elliptic-ovate,  5.5-9  mm  long,  conspic- 
uously shorter  than  the  calyx  lobes  at 
anthesis;  Costa  Rica  (Cordillera  de  Ti- 
laran) B.  chlorantha  Almeda 

Leaf  blades  not  modified  adaxially  at  the 
petiole-laminar  junction  into  flap-like  sac- 
cate pouches. 

3.  Older  petioles,  pedicels,  and  floral  bracts 
essentially  glabrous;  outer  floral  bracts 
13-17  mm  wide  on  fruiting  hypanthia; 
Costa  Rica  (Cordillera  de  Talamanca 

and  Volcan  Barva) 

B.  penduliflora  Almeda 

3.  Older  petioles,  pedicels,  and  floral  bracts 
moderately  to  copiously  covered  with 
appressed  or  spreading  brown  hairs; 
outer  floral  bracts  5-9  mm  wide  on 
fruiting  hypanthia. 
4.  Uppermost  internodes  and  elevated 


primary  nerves  on  lower  leaf  sur- 
faces moderately  to  densely  covered 
with  spreading  brown  hairs  mostly 
less  than  0.5  mm  long;  pedicels  (20-) 
28-38  mm  long;  outer  floral  bracts 
13-20  mm  long;  Mexico  (Chiapas) 
and  western  Guatemala  (Quezalten- 

ango  and  San  Marcos) _ 

B.  purpusii  Brandegee 

4.  Uppermost  internodes  and  elevated 
primary  nerves  on  lower  leaf  sur- 
faces densely  covered  with  ap- 
pressed brown  hairs  1-2.5  mm  long; 
pedicels  1 2-20  mm  long;  outer  floral 
bracts  7-8  mm  long;  western  Pana- 
ma (Chiriqui) B.  gregii  Almeda 

Among  the  green-flowered  species,  B.  gregii  is 
probably  most  closely  related  to  B.  chlorantha. 
They  are  separable  even  in  sterile  condition  be- 
cause B.  gregii  lacks  the  flap-like  saccate  pouches 
at  the  petiole-laminar  junction  that  are  typical 
of  B.  chlorantha.  The  latter  also  differs  consis- 
tently in  a  number  of  other  characters.  Pubes- 
cence of  the  upper  cauline  internodes  consists  of 
±  flattened  subulate  hairs  mostly  less  than  0.5 
mm  long  and  the  floral  bracts  are  beset  with  a 
mixture  of  spreading  shaggy  hairs  (these  mostly 


ALMEDA:  NEW  BLAKEA  AND  TOPOBEA 


309 


less  than  0.5  mm  long)  and  appressed  stellate 
hairs.  In  addition,  the  pedicels  of  B.  chlorantha 
are  shorter  (0.4-1  cm),  the  calyx  lobes  are  tri- 
angular instead  of  bluntly  rounded,  and  the  pet- 
als are  about  half  the  size  (6-8.5  x  5-6  mm)  of 
those  in  B.  gregii. 

This  species  is  named  for  Greg  de  Nevers 
(195 5-),  a  student  of  Arecaceae,  who  collected 
the  type  and  many  other  important  additions  to 
the  melastome  flora  of  Mesoamerica  in  the  course 
of  his  extended  field  work  in  Panama  from  1984 
to  1986. 


Blakea  hammelii  Almeda,  sp.  nov. 

(Fig.  6) 

TYPE.— PANAMA.  Chiriqui:  3.5  mi.  NE  of  Boquete,  end  of 
rd.  on  slope  S  of  Rio  Palo  Alto,  17  Nov.  1978,  Hammel  5688 
(holotype:  CAS!;  isotype:  MO). 

Frutex  epiphyticus  vel  terrestris.  Petioli  0.7-1.7  cm  longi; 
lamina  3.5-7.5  x  1.6-4.5  cm  elliptica  vel  elliptico-ovata  apice 
caudata  vel  cuspidata  basi  acuta  ad  maturitatem  subcoriacea 
et  glabra,  5-nervata,  nervi  in  axillis  acarodomatiis  instructi, 
nervis  secundariis  0.25  mm  inter  se  distantibus.  Flores  6-meri 
in  quoque  nodo  superiore  singuli  vel  bini,  pedicellis  2.6-3.5 
cm  longis,  bracteae  omnino  liberae;  bracteae  exteriores  3.5-5 
x  2-3  mm  ovatae  vel  ovato-ellipticae  extus  sparse  subamor- 
pho-furfuraceae  demum  glabratae;  bracteae  interiores  3-4  x 
3-3.5  mm  oblongo-ovatae  extus  sparse  inconspicueque  fim- 
briolato-ciliolatae.  Hypanthium  (ad  torum)  3-4  mm  longum 
extus  glabrum;  calycis  tubus  1-1.5  mm  longus,  lobis  1-1.5  x 
2.5-3. 5  mm.  Petala  glabra  1.4-1.5  x  0.8-0. 9  cm  obovata  apice 
rotundato.  Filamenta  4.5-5.5  mm  longa;  antherae  3.5-4  x  1- 
1.5  mm  oblongae  lateraliter  non  cohaerentes  apicaliter  bipo- 
rosae;  connectivum  dorsaliter  ca.  1-1.5  mm  supra  thecarum 
basim  tuberculatum.  Stylus  10-1 1  mm;  ovarium  5-6-loculare 
et  %  inferum,  cono  glabro  (collo  non  evoluto). 

Epiphytic  or  terrestrial  shrubs  or  small  trees 
2.5-6  m  tall.  Young  vegetative  buds  deciduously 
furfuraceous,  otherwise  glabrous,  the  distal 
branchlets  glabrous  and  bluntly  quadrate  with 
thickened  interpetiolar  lines  or  ridges.  Mature 
leaves  of  a  pair  equal  to  unequal  in  size,  glabrous 
throughout;  petioles  0.7-1 .7  cm  long;  blades  sub- 
coriaceous,  3.5-7.5  cm  long  and  1.6-4.5  cm  wide, 
elliptic  to  elliptic-ovate,  apex  caudate  to  cuspi- 
date, base  acute,  margin  entire,  5 -nerved  abax- 
ially  with  well-developed,  irregularly  ruptured 
domatia  formed  in  the  angles  between  the  me- 
dian vein  and  each  of  the  two  proximal  lateral 
veins,  the  striolate  transverse  secondary  veins 
spaced  mostly  less  than  0.25  mm  apart  at  the 
widest  portion  of  the  blade.  Flowers  erect,  soli- 
tary or  paired  in  leaf  axils  of  distal  branches; 
pedicels  2.6-3.5  cm  long,  glabrous.  Floral  bracts 


sessile,  entire,  and  free  to  the  base;  outer  bracts 
3.5-5  x  2-3  mm,  ovate  to  ovate-lanceolate  or 
elliptic-ovate,  apex  bluntly  acute  to  rounded, 
sparingly  and  deciduously  brown  furfuraceous- 
puberulent;  inner  bracts  3-4  x  3-3.5  mm,  ob- 
long-ovate, apex  rounded,  essentially  glabrous 
but  deciduously  fimbriolate-puberulent  on  the 
margins.  Hypanthia  (at  anthesis)  campanulate, 
3-4  mm  long  to  the  torus  and  6-7  mm  in  di- 
ameter, glabrous.  Calyx  tube  1-1.5  mm  long;  ca- 
lyx lobes  1-1.5  mm  long  and  2.5-3.5  mm  wide 
basally,  broadly  ovate  with  a  blunt  callose-thick- 
ened  tooth  at  the  abaxial  apex  of  each  lobe,  mar- 
gin entire,  glabrous  on  both  surfaces.  Petals  6, 
glabrous,  1.4-1.5  x  0.8-0.9  cm,  white  flushed 
with  pink  unilaterally  on  the  abaxial  surface,  ob- 
ovate,  apically  rounded,  entire.  Stamens  1 2,  iso- 
morphic,  free  and  declined  to  one  side  of  the 
flower  opposing  the  style;  filaments  complanate 
and  glabrous,  4.5-5.5  mm  long;  anthers  3.5—4 
mm  long,  1-1.5  mm  wide,  yellow,  linear-oblong 
in  dorsal  and  ventral  view,  narrowly  ovoid  in 
profile  view,  truncate  at  the  apex  with  two  some- 
what dorsally  inclined  pores;  connective  slightly 
thickened  dorsally  and  dilated  into  a  blunt  spur 
ca.  1-1.5  mm  above  the  filament  insertion.  Ovary 
%  inferior,  5-6-celled,  glabrous  at  the  summit 
which  is  distended  into  a  prominent  cone  that 
becomes  rounded  with  age.  Style  declinate  and 
somewhat  sigmoid,  glabrous,  10-11  mm  long; 
stigma  truncate.  Berry  globose,  5-6.5  mm  long 
and  4-7  mm  in  diameter.  Seeds  irregularly  pryi- 
form  to  pyramidate,  0.5-1  mm  long,  beige  with 
a  smooth  testa  and  conspicuous  lateral  raphe. 

PHENOLOGY.— Flowering  specimens  have  been 
collected  from  January  through  April  and  in  No- 
vember; the  only  known  fruiting  specimens  were 
collected  in  January  and  April. 

DISTRIBUTION. — Local  and  uncommon  in  cloud 
forests  on  the  slopes  of  Cerro  Pate  Macho  and 
along  the  Rio  Palo  Alto  in  Chiriqui  Province, 
Panama  at  1,600-1,900  m  (Fig.  8). 


ADDITIONAL  SPECIMENS  EXAMINED.  — PANAMA.  Chiriqui: 
S  slopes  of  Cerro  Pate  Macho,  along  trail  E  of  Jaramillo  Arriba, 
8°49'N,  82-23' W,  12  Apr.  1984,  Churchill  &  Kuijt  5103  (CAS); 
trail  to  Cerro  Pate  Macho,  8°49'N,  82°24'W,  17  Jan.  1986,  de 
Nevers  &  McPherson  6830  (CAS);  back  side  of  mountain,  Bo- 
quete, 20  Mar.  1977,  Folsom  2185  (CAS);  along  trail  between 
N  fork  of  Rio  Palo  Alto  and  Cerro  Pate  Macho,  ca.  6  km  NE 
of  Boquete,  8°48'N,  82°23.5'W,  6  Feb.  1986,  Grayum  et  al. 
6371  (CAS);  road  along  Rio  Palo  Alto  ca.  3  km  NE  of  Boquete 
to  the  end,  15  Apr.  1982,  Huft  1855  (CAS);  Rio  Palo  Alto,  28 
Apr.  1983,  Schmalzel  1478  (CAS). 


310 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  14 


FIGURE  6.  Blakea  hammelii  Almeda.  A,  habit;  B,  representative  leaf  with  enlargement  showing  domatium  (abaxial  surface); 
C,  representative  flower  (natural  posture);  D,  petal  (adaxial  surface);  E,  stamens,  lateral  view  (left),  dorsal  view  (middle),  ventral 
view  (right);  F,  berry  with  persistent  decussate  bracts;  G,  seeds.  (A,  D,  E  from  Churchill  &  Kuijt  5103;  B  from  Huft  1855;  C 
from  de  Nevers  &  McPherson  6830;  F,  G  from  Schmalzel  1478.) 


ALMEDA:  NEW  BLAKEA  AND  TOPOBEA 


311 


Because  of  its  elliptic  to  elliptic-ovate  leaves 
with  finely  striolate  secondary  veins,  erect  long- 
pedicellate  flowers,  and  white  petals  that  are 
flushed  with  pink  abaxially,  B.  hammelii  is  readi- 
ly mistaken  for  B.  pauciflora  Gleason,  to  which 
it  seems  most  closely  related.  The  two  differ  pri- 
marily in  androecial  details  and  the  nature  of  the 
calyx.  In  B.  pauciflora  the  anthers,  which  are 
fused  laterally  for  most  of  their  length,  are  conic 
to  rounded  apically  with  two  minute  subterminal 
pores  positioned  on  the  ventral  side  of  the  anther 
sacs.  Many  specimens  of  B.  pauciflora  and  B. 
hammelii  are  collected  only  in  fruiting  condition, 
but  the  flaring,  essentially  truncate,  calyx  of  the 
former  provides  a  consistent  character  by  which 
the  two  species  can  be  distinguished. 

Although  the  domatia  of  B.  hammelii  and  B. 
pauciflora  are  superficially  similar  to  the  pocket 
domatia  ofClidemia  hammelii  Almeda  (Almeda 
1 989),  they  appear  to  be  modified  versions  of  the 
pit  or  marsupiform  (pocket-shaped)  domatia  de- 
scribed and  illustrated  by  Jacobs  (1 966)  and  Stace 
(1965),  respectively.  The  term  domatium  was 
coined  by  Lundstrom  (1887)  when  he  hypothe- 
sized a  mutualistic  relationship  between  mites 
and  plants  bearing  these  structures.  Because  of 
insufficient  field  study  and  the  lack  of  experi- 
mental data,  the  functional  significance  of  foliar 
domatia  has  remained  conjectural  for  over  a  cen- 
tury. Based  in  part  on  Lundstrom's  seminal  pa- 
per, Pemberton  and  Turner  (1989)  recently  pre- 
sented new  data  to  "support  the  hypothesis  of  a 
widespread  facultative  mutualism  in  which  leaf 
domatia  serve  as  shelters  and  nurseries  for  ben- 
eficial mites,  which  in  turn  reduce  the  number 
of  phytophagous  arthropods  and  pathogens  using 
the  plants."  Interestingly,  Pemberton  and  Turner 
found  three  genera  of  predaceous  mites  in  32% 
of  the  domatia  of  an  unidentified  Panamanian 
species  ofBlakea  examined  specifically  for  their 
study. 

This  species  is  named  for  Barry  E.  Hammel 
(1946-),  avid  student  of  the  Mesoamerican  flora, 
who  collected  the  type  of  this  and  several  other 
recently  described  species  of  Melastomataceae. 

Blakea  herrerae  Almeda,  sp.  nov. 

(Fig.  7) 

TYPE.  — PANAMA.  Comarca  de  San  Bias:  El  Llano-Carti 
road  at  about  km  1 9.  Ina  Igar  trail  in  the  vicinity  of  Nusagandi, 
elev.  350  m,  1  Feb.  1989,  Almeda  et  al.  6507  (holotype:  CAS!; 
isotypes:  AAU!,  BM!,  BR!,  CR!,  DUKE!,  F!,  G!,  MA!,  MEXU!, 
MICH!,  MO!,  NY!,  P!,  PMA!,  TEX!,  US!,  WIS!). 


Frutex  epiphyticus.  Ramuli  subquadrangulares  demum  te- 
retes  primum  sicut  folia  novella  obscure  furfuracei  mox  glabrati . 
Petioli  1-2  cm  longi;  lamina  6.5-1 1  x  3.2-5.8  cm  clliptica  vel 
elliptico-obovata  apice  caudata  vel  cuspidata  basi  acuta, 
3-nervata  vel  3-plinervata,  subcoriacea  et  integra,  nervis  se- 
cundariis  0.25  mm  inter  se  distantibus.  Flores  6-meri  in  quo- 
que  nodo  superiore  l-2(-3),  pedicellisO.5-1.2  cm  longis,  brac- 
teae  omnino  liberae;  bracteae  exteriores  2.5-5  x  1.5-2.5  mm 
oblongo-lanceolatae  acutae  primum  extus  sparse  vel  modice 
subamorpho-furfuraceae  demum  glabratae;  bracteae  interiores 
2-4  x  2-3  mm  ellipticae  vel  ovato-ellipticae  apice  late  acuto 
extus  glabrae.  Hypanthium  (ad  torum)  3-4  mm  longum  extus 
sparsiuscule  caduceque  subamorpho-furfuraceum;  calycis  tu- 
bus  1.5-2.5  mm  longus,  lobis  1-2  mm  longis.  Petala  6-7  x  3- 
5  mm  obovato-oblonga  apice  rotundato.  Filamenta  3-4  mm 
longa;  antherae  2.5-3  x  0.75-1  mm  oblongae  inter  se  late- 
raliter  cohaerentes  subapicaliter  minute  biporosae;  connecti- 
vum  nee  prolongatum  nee  appendiculatum.  Stylus  6-7  mm; 
ovarium  6-loculare  et  omnino  inferum  apice  glabro  (cono  et 
collo  non  evoluto). 

Epiphytic  shrubs  2-4  m  tall.  Distal  branchlets 
subquadrate,  becoming  rounded  with  age.  Veg- 
etative buds  and  very  young  leaves  deciduously 
scurfy-puberulent.  Mature  leaves  of  a  pair  equal 
to  somewhat  unequal  in  size,  glabrous  through- 
out; petioles  1-2  cm  long;  blades  subcoriaceous, 
6.5-1 1  cm  long  and  3.2-5.8  cm  wide,  elliptic  to 
elliptic-obovate,  apex  caudate  to  cuspidate  or 
rarely  varying  to  acuminate,  base  acute,  margin 
entire,  3-nerved  or  3-plinerved  with  an  addi- 
tional submarginal  pair  of  inconspicuous  veins, 
the  striolate  transverse  secondary  veins  spaced 
mostly  0.25  mm  apart  at  the  widest  portion  of 
the  blade.  Flowers  erect,  solitary  or  paired,  rarely 
in  fascicles  of  three  in  each  leaf  axil  of  the  distal 
branches;  pedicels  0.5-1.2  cm  long,  sparingly 
scurfy-puberulent  to  glabrous.  Floral  bracts  ses- 
sile, entire,  and  free  to  the  base;  outer  bracts  2.5- 
5  x  1.5-2.5  mm,  oblong-lanceolate,  apex  acute, 
sparingly  to  moderately  scurfy-puberulent  to  gla- 
brate;  inner  bracts  2—4  x  2-3  mm,  elliptic  to 
elliptic-ovate,  apex  acute,  essentially  glabrous  but 
fimbriolate-puberulent  on  the  margins.  Hypan- 
thia  (at  anthesis)  campanulate,  3—4  mm  long  to 
the  torus  and  3-4  mm  in  diameter,  sparingly  and 
deciduously  scurfy-puberulent.  Calyx  tube  1.5- 
2.5  mm  long;  calyx  lobes  1-2  mm  long  and  2.5- 
3  mm  wide  basally,  triangular  to  triangular-ovate, 
margin  entire  or  deciduously  ciliate,  glabrous  on 
both  surfaces.  Petals  6,  glabrous,  6-7  x  3-5  mm, 
pale  greenish-white,  oblong-obovate,  apically 
rounded,  entire.  Stamens  12,  isomorphic;  fila- 
ments 3-4  mm  long,  complanate,  glabrous  and 
declined  to  one  side  of  the  flower  opposing  the 
style;  anthers  2.5-3  mm  long,  0.75-1  mm  wide, 
yellow,  laterally  connate  for  their  entire  length, 


312 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  14 


B 


FIGURE  7.  Blakea  herrerae  Almeda.  A,  habit;  B,  C,  representative  leaves  (abaxial  surface)  showing  variation  in  shape  and 
enlargement  (Code  population)  showing  pubescence  details;  D,  representative  flower  and  bud  (natural  posture)  with  enlargement 
showing  hypanthial  pubescence;  E,  petal  (adaxial  surface);  F,  berry  with  persistent  decussate  bracts;  G,  seeds.  (A-E  from  the 
holotype;  F,  G  from  Sytsma  et  al.  2603.) 


ALMEDA:  NEW  BLAKEA  AND  TOPOBEA 


313 


A    B.  hammelii 

•  B.  herrerae 

•  T.  hexandra 
T.  suaveolens 


0  100  200  km 


-10° 


FIGURE  8.    Distributions  of  Blakea  hammelii,  B.  herrerae,  Topobea  hexandra,  and  T.  suaveolens. 


linear-oblong,  with  two  minute  pores  positioned 
ventrally  just  below  the  truncate  apex;  connec- 
tive slightly  thickened  dorsally  but  not  prolonged 
or  appendaged  at  the  filament  insertion.  Ovary 
completely  inferior,  6-celled,  glabrous  at  the 
summit  surrounding  the  stylar  scar  but  not  dis- 
tended into  a  prominent  cone  or  collar.  Style 
somewhat  declinate  and  slightly  incurved  dis- 
tally,  glabrous,  6-7  mm  long;  stigma  truncate. 
Berry  globose  and  pink  at  maturity,  9-10  mm 
long  and  10-12  mm  in  diameter.  Seeds  clavate 
to  narrowly  pyriform  or  pyramidate,  1-1.5  mm 
long,  beige  with  a  smooth  testa  and  conspicuous 
lateral  raphe. 

PHENOLOGY.— Flowering  and  fruiting  occur 
sporadically  throughout  the  year. 

DISTRIBUTION.— Low  rainforests  from  the  Ca- 
ribbean slope  of  central  Panama  (Code)  east  to 
the  Llano-Carti  road  in  the  Nusagandi  region 
(Comarca  de  San  Bias)  at  100-400  m  (Fig.  8). 

ADDITIONAL  SPECIMENS  EXAMINED.  — PANAMA.  Code:  be- 
tween La  Junta  and  Limon,  5  hr.  walk  N  of  Alto  Calvario,  1 1 
Oct.  1977,  Folsom  5878  (CAS);  trail  between  Rio  Blanco  and 
Continental  Divide  N  of  El  Cope  and  El  Petroso  sawmill, 
8°38'N,  80°36'W,  14  Dec.  1980,  Sytsma  et  al.  2603  (CAS). 
Comarca  de  San  Bias:  Llano-Carti  road,  km  1 6,  trail  to  creek 
on  Atlantic  drainage,  2  Feb.  1989,  Almeda  et  al.  6523  (CAS, 
PMA);  Nusagandi,  Llano-Carti  road,  28  Jul.  1984,  de  Nevers 
&  Todzia  3540  (CAS);  along  Llano-Carti  road,  W  of  road  1 3.8 


km  to  15.8  km  from  Interamerican  Hwy.,  9°19'N,  78°55'W, 
24  Aug.  1984,  de  Nevers  et  al.  3746  (CAS);  Llano-Carti  road, 
13.8  km  to  19  km  from  Interamerican  Hwy.,  9°19'N,  78°55'W, 
3  Sep.  1 984,  de  Nevers  &  Porras  3822  (CAS);  Llano-Carti  road, 
km  19.1,  9°19'N,  78°56'W,  9  Nov.  1984,  de  Nevers  &  Herrera 
4242  (CAS);  Llano-Carti  road,  km  19.1,  9°19'N,  78°55'W,  1 1 
Mar.  1985,  de  Nevers  &  Herrera  5100  (CAS).  Comarca  de  San 
Bias/Panama  Border:  on  Kuna  divide  trail  N  of  Llano-Carti 
road,  9°20'N,  79°00'W,  16  Dec.  1987 ',  McPherson  11 879  (MO). 
Panama:  Llano-Carti  road,  9°16'N,  78°58'W,  12  Sep.  1980, 
Sytsma  1077  (CAS);  headwaters  of  Rio  Chagres,  Rio  Esperan- 
za,  and  Rio  Piedras,  9°20'N,  79°20' W,  1 7  Oct.  1 984,  de  Nevers 
et  al.  4088  (CAS). 

First  collected  in  1977,  Blakea  herrerae,  is 
another  species  that  has  been  confused  with,  and 
is  probably  derived  from,  B.  paudflora.  Both 
species  have  finely  striolate  secondary  foliar  ve- 
nation and  laterally  fused  anthers  that  are  de- 
clined to  one  side  of  the  flower  to  form  a  semi- 
circle. Blakea  paudflora  differs  from  B.  herrerae 
in  a  number  of  floral  characters.  The  pedicels  of 
B.  paudflora  are  longer  (2.5-5.3  cm),  the  calyx 
is  essentially  truncate  and  flaring,  the  style  is 
longer  (9-12  mm)  and  flexuous,  and  the  larger 
petals  (13- 19  x  8- 10  mm)  are  white  flushed  with 
pink  instead  of  pale  greenish- white.  Blakea  her- 
rerae is  also  distinctive  in  that  some  mature  leaves 
on  all  collections  examined  have  cryptic  tufts  of 
short  stiff  hairs  on  the  lower  surface  in  the  angles 
between  the  median  vein  and  each  of  the  two 


314 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  14 


proximal  lateral  veins.  These  may  well  function 
as  acarodomatia,  like  the  concentrated  foliar  hair 
tufts  now  known  to  harbor  beneficial  mites  in 
many  plant  groups  (Pemberton  and  Turner  1 989). 

There  is  some  variation  in  the  size  and  shape 
(Fig.  7B,  C)  of  mature  leaves,  but  none  of  this 
can  be  correlated  with  geography,  elevation,  or 
other  diagnostic  characters.  Despite  the  limited 
range  of  this  species,  there  is  some  variation  that 
appears  to  be  geographically  correlated.  The  two 
cited  collections  from  Code  Province  west  of  the 
Canal  Area  (Fig.  8)  differ  from  populations  east 
of  the  Canal  in  having  elevated  primary  leaf  veins 
below  that  are  sparsely  beset  with  a  mixture  of 
scurfy  hairs  and  spreading  glandular  hairs.  The 
lower  leaf  surfaces  of  these  two  collections  are 
also  unusual  in  having  a  liberal  scattering  of  mi- 
nute red  glands  on  and  between  the  striolate  sec- 
ondary veins.  More  material  from  Code  and  oth- 
er intervening  areas  is  needed  to  better  assess  the 
significance  of  these  differences  in  pubescence. 

A  population  from  Cerro  Jefe,  Panama  (rep- 
resented by  D'Arcy  12181,  McPherson  7435,  and 
Sytsma  1414,  all  CAS),  is  similar  to  B.  herrerae 
in  foliar  venation,  details  of  the  floral  bracts,  and 
anther  morphology.  These  collections  differ  from 
other  cited  material  in  having  thicker  leaves, 
larger  petals  (12-15  x  8-9  mm)  that  are  pink, 
and  longer  styles  (1 1-12  mm  long).  Because  plants 
of  this  population  differ  from  typical  specimens 
of  Blakea  herrerae  in  characters  that  are  com- 
monly diagnostic,  they  are  excluded  from  my 
circumscription  of  this  species  pending  addi- 
tional study. 

Field  study  of  this  species  in  the  Nusagandi 
region  revealed  an  unusual  method  of  vegetative 
reproduction  that  to  my  knowledge  has  never 
been  reported  for  an  epiphytic  member  of  the 
Melastomataceae.  Blakea  herrerae  is  commonly 
a  massive  rainforest  epiphyte  with  lax  spreading 
branches  that  skirt  host  trees.  In  two  individuals 
examined,  the  outermost  branches  were  attached 
to  adjacent  trees  by  clinging  adventitious  roots. 
One  individual  had  colonized  four  neighboring 
host  trees  in  this  fashion. 

Blakea  herrerae  also  appears  to  be  unique 
among  the  erect-flowered  species  of  the  genus  in 
having  greenish-white  flowers  that  produce  a 
pleasant  spicy  fragrance  but  lack  nectar.  The  fra- 
grance is  reminiscent  of  that  produced  by  neo- 
tropical orchids  that  are  pollinated  by  male  eu- 
glossine  bees.  The  few  species  of  Blakea  and 
Topobea  with  showy  white  and/or  pink  flowers 


that  have  been  studied  in  the  field  are  buzz-pol- 
linated by  many  species  of  bees  (including  two 
euglossine  genera)  seeking  pollen  as  a  floral  re- 
ward (Lumer  1981;  Renner  1989).  Several  eu- 
glossine genera  are  among  the  common  buzz- 
pollinating  bees  in  the  neotropics  (Buchmann 
1983;  Roubik  1989).  However,  only  female  eu- 
glossines  are  known  to  collect  pollen  from  neo- 
tropical Melastomataceae  (Renner  1 989).  In  the 
male  euglossine  pollination  syndrome  of  the  Or- 
chidaceae,  the  flowers  produce  no  nectar  and  the 
pollen  is  not  available  as  a  food  source.  It  will 
be  interesting  to  learn  if  euglossine  bees  buzz- 
pollinate  the  poricidal  anthers  of  B.  herrerae  and 
if  its  floral  fragrance  serves  only  as  a  secondary 
attractant,  as  suggested  by  Renner  (1989)  for 
melastomes  in  general.  In  addition  to  the  Or- 
chidaceae,  floral  fragrance  collection  by  male 
euglossines  has  been  found  in  some  species  of 
Anthurium  and  Spathiphyllum  (both  in  the  Ara- 
ceae),  one  species  of  Gloxinia  (Gesneriaceae), 
Cyphomandra  (Solanaceae),  Dalechampia  (Eu- 
phorbiaceae),  and  a  scattering  of  species  in  other 
families  of  flowering  plants  (Ackerman  1986; 
Williams  1983).  Because  pollen  presentation  in 
Cyphomandra  provides  a  parallel  to  the  porici- 
dally  dehiscent  anthers  of  Blakea  herrerae,  ad- 
ditional fieldwork  is  needed  to  document  details 
of  its  pollination  biology. 

This  species  is  named  for  Heraclio  Herrera 
(195 9-),  a  student  of  the  Panamanian  flora  who 
kindly  directed  me  to  a  flowering  population  of 
this  species  at  Nusagandi,  Panama. 

Blakea  scarlatina  Almeda,  sp.  nov. 

(Fig.  9) 

TYPE.— COSTA  RICA.  Alajuela:  20-30  km  SE  of  Cataratas 
de  San  Ramon,  elev.  2,150-2,500  ft.  (655-762  m),  20  Mar. 
1978,  Almeda  et  al.  4308  (holotype:  CAS!). 

Frutex  epiphyticus  ca.  4  m  altus.  Ramuli  primum  subquad- 
rangulati  demum  teretes  sicut  folia  novella  pedicelli  hypanthia- 
que  sparse  vel  densi  furfuracei  demum  glabrati.  Petioli  1.7- 
4.5(-6.3)  cm  longi;  lamina  8.5-19.7  x  5.4-10  cm  elliptica  vel 
elliptico-ovata  apice  acuminata  basi  acuta  vel  obtusa  vel  ro- 
tundata,  3-nervata,  coriacea  et  integra,  nervis  secundariis  1-2 
mm  inter  se  d istanti bus.  Flores  6-meri  in  quoque  nodo  superio- 
re  singuli,  pedicellis  1-1.8  cm  longis;  bracteae  exteriores  2-2.7 
x  2.2-3.6  cm  late  ovatae  vel  suborbiculares  ca.  1-1.5  cm  coa- 
litae  apice  truncato-rotundato  plerumque  mucronato;  bracteae 
interiores  1-2  cm  omnino  coalitae.  Hypanthium  (ad  torum) 
1-1.3  cm  longum;  calycis  tubus  5-6  mm  longus,  lobis  5-10 
mm  longis.  Petala  3-4.5  x  2.5-4.3  cm  obovata  vel  cuneata 
apice  truncate  vel  rotundato.  Filamenta  9-13  mm  longa;  an- 
therae  6-8  x  4-6  mm  inter  se  non  cohaerentes,  poris  duobos 
minutis  terminalibus;  connectivum  ad  basim  dorsaliter  dente 


ALMEDA:  NEW  BLAKEA  AND  TOPOBEA 


315 


FIGURE  9.  Blakea  scarlatina  Almeda.  A,  habit;  B,  representative  leaf  (abaxial  surface)  with  enlargement  showing  pubescence 
detail;  C,  outer  floral  bracts;  D,  inner  floral  bracts  enveloping  hypanthium;  E,  hypanthium  (top  view)  showing  ovary  summit, 
torus  and  calyx  lobes;  F,  petal  (adaxial  surface);  G,  stamens,  lateral  view  (left),  ventral  view  (right).  (A,  B  from  the  holotype; 
C-G  from  Hammel  9297,  CAS.) 


3-5  mm  armatum.  Stylus  2-2.1  cm;  ovarium  6-iocuiare  et  buds,  juvenile  leaves,  pedicels,  and  young  hy- 

omnino  inferum  cono  5  mm  longo  (coiio  inciuso)  giabro.  panthia  sparsely  to  densely  furfuraceous-puber- 

Massive  epiphytic  shrubs  to  4  m  tall  with  coarse  ulent  but  becoming  glabrate  with  age.  Mature 

spreading  branches  0.5-8  m  long.  Vegetative  leaves  of  a  pair  essentially  equal  or  only  slightly 


316 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  14 


unequal  in  size;  petioles  1.7-4.5(-6.3)  cm  long; 
blades  thick  and  coriaceous,  8.5-19.7  cm  long 
and  5.4-10  cm  wide,  elliptic  to  elliptic-ovate, 
apex  acuminate,  base  acute  to  obtuse  or  rounded, 
margin  entire,  3 -nerved  with  an  additional  pair 
of  inconspicuous  intramarginal  veins,  the  trans- 
verse secondary  veins  mostly  1-2  mm  apart  at 
the  widest  portion  of  the  blade,  essentially  gla- 
brous but  whitish  puncticulate  above,  sparsely 
furfuraceous-puberulent  on  the  elevated  primary 
veins  below  or  almost  glabrous.  Rowers  erect, 
solitary  in  the  axils  of  uppermost  leaves;  pedicels 
somewhat  compressed,  1-1.8  cm  long.  Floral 
bracts  thick  and  coriaceous,  sessile  and  entire; 
outer  bracts  connate  basally  for  about  half  their 
length,  2-2.7  x  2.2-3.6  cm,  broadly  ovate  to 
suborbicular,  apex  rounded  to  subtruncate,  typ- 
ically mucronate,  deciduously  furfuraceous-pu- 
berulent abaxially;  inner  bracts  connate  for  their 
entire  length  to  form  a  shallowly  lobed,  glabrous, 
bowl-like  collar  1-2  cm  high  that  envelops  the 
hypanthium  but  is  usually  concealed  by  the  outer 
bracts.  Hypanthia  (at  an  thesis)  campanulate,  1- 
1.3  cm  long  to  the  torus  and  1.3-1.6  cm  in  di- 
ameter. Calyx  tube  5-6  mm  long;  calyx  lobes 
erect  or  slightly  spreading,  5-10  mm  long  and 
9-14  mm  wide  basally,  triangular-ovate  and 
apiculate,  margin  entire,  moderately  to  sparingly 
furfuraceous-puberulent  abaxially.  Petals  6,  fleshy 
and  glabrous  when  fresh,  3-4.5  x  2.5-4.3  cm, 
brilliant  red  but  white  for  the  basal  '/4  of  their 
length,  obovate  to  cuneate,  apically  truncate  to 
broadly  rounded,  entire.  Stamens  12,  isomor- 
phic,  free  but  connivent  and  forming  a  somewhat 
declinate  nearly  semicircular  ring  around  the  style; 
filaments  complanate,  white,  fleshy  and  gla- 
brous, 9-13  mm  long;  anthers  6-8  mm  long  and 
4-6  mm  wide,  yellow,  oblong-ovoid,  laterally 
compressed  and  tipped  with  two  minute  pores; 
connective  thickened  and  prolonged  dorsally  at 
the  base  into  an  acute  spur  3-5  mm  long  that 
assumes  a  pseudoterminal  position  because  of 
the  pronounced  adaxial  tilt  of  each  anther  apex. 
Ovary  completely  inferior,  6-celled,  glabrous  at 
the  gently  fluted  summit  and  dilated  apically  into 
a  cone  and  stylar  collar  5  mm  long.  Style  decli- 
nate and  incurved  distally,  glabrous,  2-2.1  cm 
long  and  0.5-1  mm  wide,  stigma  truncate  to  cap- 
itellate.  Mature  berry  and  seeds  not  seen. 

PHENOLOGY.— Flowering  specimens  have  been 
collected  in  every  month  except  December  and 
January;  of  the  nearly  40  collections  available 
for  this  study  none  had  mature  fruits. 


DISTRIBUTION. —Widespread  but  local  and  un- 
common in  wet  evergreen  forests  from  southern 
Nicaragua  (Zelaya)  south  to  Costa  Rica  from  Par- 
que  Nacional  Rincon  de  la  Vieja  (Guanacaste) 
and  the  Cordillera  de  Tilaran  (Alajuela)  south- 
east to  the  vicinity  of  Turrialba  (Cartago)  from 
sea  level  to  1 ,450  m  (Fig.  5). 

ADDITIONAL  SPECIMENS  EXAMINED.— COSTA  RICA.  Ala- 
juela: 3  km  N  of  Rio  Cataratas  on  road  to  Bajo  Rodriquez,  23 
Feb.  1 978,  Almeda  &  Nakai  3884  (CAS,  CR);  1 8-35  km  NNW 
of  San  Ramon  on  road  to  Cataratas,  21  Mar.  1986,  Almeda  el 
al.  5652  (CAS);  La  Palma  de  San  Ramon,  6  Nov.  1923,  Brenes 
3920  (F);  La  Palma  de  San  Ramon,  29  Sep.  1925,  Brenes  4466 
(F);  La  Palma  de  San  Ramon,  24  Oct.  1926,  Brenes  5031  (F); 
La  Palma  de  San  Ramon,  24  Nov.  1926,  Brenes  5115  (F);  La 
Palma  de  San  Ramon,  24  Nov.  1926,  Brenes  5138  (F);  La 
Palma  de  San  Ramon,  22  Aug.  1927,  Brenes  5671  (F);  La 
Palma  de  San  Ramon,  5-8  Aug.  1935,  Brenes  20627  (F);  Re- 
serva  Forestal  de  San  Ramon  above  and  in  valley  of  Rio  San 
Lorencito,  10°13'N,  84°37'W,  12-14  Mar.  1987,  Burger  et  al. 
12105  (CAS,  F);  Monteverde  Reserve,  Atlantic  slope,  Rio  Pe- 
nas Blancas  valley,  19  Mar.  1986,  Haber  1519  (CAS);  San 
Carlos,  margin  of  Rio  Penas  Blancas  at  elevation  of  San  Pedro, 
29  Jun.  1985,  Haber  &  Bella  1744  (CAS);  San  Carlos,  Penas 
Blancas,  9  Jul.  1985,  Haber  &  Bella  1873  (CAS);  Rio  Penas 
Blancas,  San  Carlos,  1 1  Aug.  1 985,  Haber  &  Bella  2306  (CAS); 
finca  Don  Bolivar  Ruiz  on  road  between  the  Reserva  de  San 
Ramon  station  and  the  road  to  Palmarena,  Feb.  1987,  Herrera 
508  (CAS);  6  km  S  of  Ciudad  Quesada,  near  Rio  La  Vieja,  10 
Sep.  1967,  Lent  1265  (F);  8  km  NE  of  Villa  Quesada,  near 
Artezalea,  17  Feb.  1966,  Molina  et  al.  17258  (F);  near  San 
Isidro  de  Penas  Blancas,  22  Oct.  1965,  Schnell  231  (F);  La 
Marina,  near  Ciudad  Quesada,  23  Oct.  1965,  Schnell  243  (F); 
Villa  Quesada,  21  Feb.  1939,  Smith  HI 603  (F);  5.8  km  N  of 
Cataratas  de  San  Ramon,  1 1  Apr.  1976,  J.  &  K.  Utley  4606 
(CAS,  DUKE).  Cartago:  Instituto  Interamericano  de  Ciencias 
Agricolas,  Turrialba,  4  Oct.  1950,  Ledn  2790  (MO);  2.5  km  E 
of  Tuis  along  banks  of  Rio  Tuis,  9  Mar.  1978,  Utley  6006 
(CAS,  DUKE).  Cartago/San  Jose  Border:  Estacion  Carrillo, 
Parque  Nacional  Braulio  Carrillo,  28  Jul.  1984,  Zamora  & 
Elizondo  672  (CAS).  Guanacaste:  Parque  Nacional  Rincon  de 
la  Vieja,  Hacienda  Santa  Maria,  SE  of  Mirador  on  road  to 
Volcan  Santa  Maria,  10°48'N,  85°19'W,  Herrera  &  Robles  756 
(CAS,  MO).  Heredia:  Finca  La  Selva  on  Rio  Puerto  Viejo,  just 
E  of  its  junction  with  the  Rio  Sarapiqui,  along  Far  Loop  Trail, 
1,000  m  S,  21  Jul.  1980,  Hammel  9297  (CAS,  DUKE,  F); 
Finca  La  Selva  on  Rio  Puerto  Viejo  just  E  of  its  junction  with 
Rio  Sarapiqui,  ridge  above  Q.  Esquina,  30  Aug.  1981,  Smith 
135  (CAS,  DUKE).  Limon:  hills  2  airline  km  SSE  of  Islas 
Buena  Vista  in  Rio  Colorado,  10°40'N,  83°40'W,  13-14  Sep. 
1986,  Davidse  &  Herrera  31127  (CAS);  Hacienda  Tapezco- 
Hda.  La  Suerte,  29  airline  km  W  of  Tortuguero,  10°30'N, 
83°47'W,  28  Aug.  1979,  Davidson  &  Donahue  8913  (CAS); 
path  beyond  Rio  Sucio,  Braulio  Carrillo,  May  1984,  Gomez 
et  al.  22760  (CAS);  ridge  between  Rio  Chirripo  and  Q.  El 
Molinete,  10°12'N,  83°54'W,  19  Jul.  1984,  Grayumetal.  3553 
(CAS);  Cerro  Coronel,  E  of  Laguna  Danto,  10°41'N,  83°38'W, 
1 5-20  Sep.  1986,  Stevens  &  Montiel 24614  (CAS).  Puntarenas: 
Monteverde  community,  windbreak  next  to  road,  10°20'N, 
84°50'W,  22  Jul.  1986,  Haber  5729  (CAS).  NICARAGUA. 
Zelaya:  4  km  N  of  Nueva  Guinea,  between  Rio  Plata  and  San 
Antonio,  1 1°44'N,  84°26'W,  7  Sep.  1983,  Nee  27883  (CAS). 


ALMEDA:  NEW  BLAKEA  AND  TOPOBEA 


317 


Blakea  scarlatina  is  unique  among  its  Central 
American  congeners  in  having  large  flowers  (8- 
9  cm  across  at  anthesis)  with  brilliant  red  petals 
that  are  white  for  the  basal  '/4  of  their  length.  For 
an  obligate  epiphyte  this  species  has  a  compar- 
atively broad  geographic  and  elevational  distri- 
bution; however,  all  available  collections  exhibit 
remarkable  uniformity  in  floral  and  vegetative 
features. 

Blakea  scarlatina  was  first  collected  by  A.  M. 
Brenes  in  1 923.  Many  additional  specimens  have 
accumulated  in  herbaria  since  that  time,  but  most 
have  repeatedly  been  misidentified  as  B.  gran- 
diflora  Hemsley.  Blakea  grandiflora,  which  is 
known  only  from  higher  elevations  ( 1 ,400-2,350 
m)  in  Costa  Rica,  differs  in  having  5-plinerved 
leaves  with  an  additional  intramarginal  pair  of 
veins,  longer  pedicels  (3-5  cm),  flowers  borne  in 
clusters  of  2-4  in  the  upper  leaf  axils,  petals  that 
are  white  flushed  with  pink  distally,  and  anther 
connectives  that  are  modified  dorso-basally  into 
short  deflexed  spurs.  Vegetatively,  B.  scarlatina 
has  an  aspect  most  reminiscent  of  B.  cuneata 
Standley,  a  little-collected  species  of  Belize,  Gua- 
temala, and  Honduras.  Striking  similarities  in 
foliar  morphology  and  pubescence  details  ini- 
tially led  me  to  assign  many  collections  of  B. 
scarlatina  to  B.  cuneata.  Study  of  the  type  of  B. 
cuneata  and  more  recent  collections,  however, 
leaves  no  doubt  that  these  two  allopatric  species 
are  readily  separated  by  a  number  of  diagnostic 
characters.  In  B.  cuneata  there  is  a  consistent 
tendency  for  the  elliptic  leaves  to  be  narrower 
(3.5-6  cm),  and  the  free  portions  of  the  outer 
floral  bracts  are  elliptic-lanceolate  and  bluntly 
acute  at  the  apex.  Blakea  cuneata  also  differs  in 
having  smaller  petals  (1.8-2.1  x  1.5-2  cm)  that 
are  pink,  shorter  filaments  (5-6  mm),  and  linear- 
oblong  anthers  that  are  short  (5-5.5  mm),  erect, 
and  unappendaged  dorso-basally. 

In  anther  morphology  B.  scarlatina  most 
closely  approaches  B.  tuberculata].  D.  Smith  and 
B.  cuatrecasii  Gleason.  Blakea  tuberculata  also 
has  anthers  that  are  connivent  and  free,  but  they 
form  a  complete  circle  around  the  straight  style. 
Blakea  cuatrecasii  has  stamens  that  form  a  nearly 
circular  ring  opposing  the  declinate  style,  but  the 
anthers  are  laterally  fused  for  nearly  half  of  their 
length.  Because  both  of  these  species  differ  from 
B.  scarlatina  in  so  many  other  characters,  I  see 
no  compelling  reason  to  suggest  that  they  form 
a  particularly  close  alliance  based  solely  on  an- 
droecial  morphology. 


The  flowers  of  B.  scarlatina  are  some  of  the 
most  spectacular  among  neotropical  epiphytes. 
They  are  also  some  of  the  most  enigmatic  when 
attempting  to  make  inferences  about  likely  pol- 
linators. The  casual  observer  might  be  inclined 
to  suggest  bird  pollination  because  of  the  brilliant 
red  coloration  of  the  petals.  This  seems  highly 
unlikely.  I  have  found  no  detectable  nectar  in 
any  flowers  examined  in  the  field  during  morning 
and  late  afternoon  hours.  These  observations  are 
supported  by  Stein  and  Tobe  (1989),  who  found 
no  anatomical  basis  for  nectar  production  in  this 
species  (referred  to  as  Blakea  sp.  nov.). 

Although  pollinators  of  B.  scarlatina  remain 
unknown,  a  color  slide  photograph  in  the  frag- 
ment packet  of  Burger  et  al.  12105  shows  the 
brown  anther  discoloration  (bruising)  typically 
found  on  yellow-anthered  melastome  species  that 
have  been  visited  by  buzz-pollinating  bees.  Field 
studies  will  be  needed  to  determine  the  pollinator 
spectrum  of  this  species  and  to  evaluate  conflict- 
ing reports  on  the  nature  of  its  floral  fragrance. 
The  flowers  of  B.  scarlatina  have  been  described 
as  having  a  musky  odor  (Haber  1519),  a  distinct 
delicate  fragrance  (Almeda  et  al.  5652),  and  a 
strange,  unpleasant  odor  (Nee  27883). 

The  name  for  this  species  is  derived  from  scar- 
latinus,  Latin  for  scarlet,  in  reference  to  the  vivid 
red  petal  color. 

Topobea  caliginosa  Almeda,  nom.  nov. 
Blakea  micrantha  Almeda,  Rhodora  82:614.  1980. 

TYPE.  — PANAMA.  Veraguas:  Cerro  Tute  ca  10  km  NW  of 
Santa  Fe  on  ridgetop  in  cloud  forest  above  1 ,000  m,  1 9  Jun. 
1975,  Mori  6765  (holotype:  CAS!;  isotype:  MO!). 

A  study  of  anther  morphology  in  recently  col- 
lected flowering  material  of  this  species  has  led 
to  the  conclusion  that  it  is  closely  related  to  T. 
cordata,  T.  crassifolia,  and  T.  hexandra.  This  has 
necessitated  a  reassessment  of  the  proper  generic 
placement  for  this  specialized  group  of  species. 
For  reasons  discussed  in  the  synoptic  history  at 
the  beginning  of  this  paper,  it  seems  appropriate 
to  transfer  Blakea  micrantha  to  Topobea.  Adop- 
tion of  a  nomen  novum  is  needed  because  the 
epithet  micrantha  is  pre-empted  in  Topobea  (Al- 
meda 1981b:307).  For  additional  comments  and 
an  enumeration  of  its  diagnostic  features,  see  the 
discussion  under  T.  hexandra.  The  new  name 
for  this  species  is  derived  from  caligo,  Latin  for 
fog,  mist,  or  darkness,  in  reference  to  its  fog- 
shrouded  habitat  at  the  summit  of  Cerro  Tute. 


318 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  14 


Topobea  crassifolia  (Almeda)  Almeda,  comb, 
nov. 

Blakea  crassifolia  Almeda,  Rhodora  82:612.  1980. 

TYPE:— PANAMA.  Code:  La  Mesa  above  El  Valle  in  forest 
on  both  sides  of  junction  with  road  to  Cerro  Pilon,  ca  800  m, 
21  Jul.  1974,  Croat  25430  (holotype:  CAS!;  isotypes:  MO!, 
US!). 

The  rationale  for  transferring  B.  micrantha  to 
Topobea  also  applies  to  B.  crassifolia.  Among 
the  four  hexandrous  species,  T.  crassifolia  is  most 
like  T.  cordata  in  having  ovate  to  elliptic-ovate 
leaves  and  a  4-celled  ovary.  The  leaves  of  T. 
crassifolia  are  petiolate  whereas  those  of  T.  cor- 
data are  invariably  sessile.  Anther  differences  be- 
tween these  two  species  are  discussed  above  un- 
der the  synoptic  history.  With  a  distribution  that 
extends  to  Costa  Rica,  T.  crassifolia  is  the  only 
hexandrous  species  known  to  occur  outside  of 
Panama. 

Topobea  fragrantissima  Almeda,  sp.  nov. 

(Fig.  10) 

TYPE:— PANAMA.  Chiriqui:  vicinity  of  Fortuna  Dam,  along 
trail  across  valley  of  Rio  Homito,  elev.  1,100-1,250  m,  12 
Mar.  1988,  Almeda  et  al.  6086  (holotype:  CAS!;  isotypes:  CR!, 
F!,  MO!,  PMA!,  TEX!,  US!). 

Frutex  epiphyticus  vel  terrestris.  Petioli  0.5-3  cm  longi;  lam- 
ina 1.5-5.5  x  1.6-3  cm  elliptica  vel  elliptico-obovata  apice 
acuminata  basi  acuta  ad  maturitatem  glabra,  3-nervata  vel 
3-plinervata,  subcoriacea  et  integra  nervis  secundariis  0.25  mm 
inter  se  distantibus.  Flores  6-meri  in  quoque  nodo  superiore 
singuli  vel  bini,  pedicellis  2-3  cm  longis,  bracteae  omnino 
liberae  ellipticae  vel  obovatae;  bracteae  exteriores  5-11  x  3- 
5  mm  apice  rotundato;  bracteae  interiores  4-6  x  4-5  mm  apice 
rotundato.  Hypanthium  (ad  torum)  3-4  mm  longum  extus 
glabrum;  calycis  tubus  1  mm  longus,  lobis  1  mm  longis.  Petala 
1.2-1.4  x  1-1.2  cm  obovato-oblonga  apice  rotundato.  Fila- 
menta  5-6  mm  longa;  antherae  3.5-5  x  1  mm  oblongae  la- 
teraliter  non  cohaerentes,  poro  unico  dorsaliter  inclinato;  con- 
nectivum  ad  basim  dorsaliter  dente  0.25  mm  longo  descendenti 
armatum.  Stylus  1-1.1  cm;  ovarium  4-loculare  et  lh  inferum 
apice  glabro  (cono  et  collo  non  evoluto). 

Epiphytic  or  terrestrial  shrubs  or  small  trees 
1.5-4  m  tall.  Distal  branchlets  subquadrate  and 
glabrous  with  interpetiolar  ridges  or  lines.  Veg- 
etative buds  and  young  leaves  sparingly  and  de- 
ciduously  lepidote-furfuraceous.  Mature  leaves 
of  a  pair  equal  to  somewhat  unequal  in  size,  gla- 
brous throughout;  petioles  0.5-3  cm  long;  blades 
subcoriaceous,  1.5-5.5  cm  long  and  1.6-3  cm 
wide,  elliptic  to  elliptic-obovate,  apex  acumi- 
nate, base  acute,  margin  entire,  3-nerved  or 
3-plinerved  abaxially  with  an  additional  sub- 
marginal  pair  of  inconspicuous  veins,  the  striol- 
ate  transverse  secondary  veins  spaced  mostly  0.25 


mm  apart  at  the  widest  portion  of  the  blade. 
Flowers  erect,  solitary  or  paired  in  leaf  axils  of 
distal  branches;  pedicels  2-3  cm  long,  glabrous. 
Floral  bracts  sessile,  entire  and  free  to  the  base; 
outer  bracts  5-11  x  3-5  mm,  elliptic  or  rarely 
varying  to  obovate,  glabrous,  apex  rounded;  in- 
ner bracts  4-6  x  4-5  mm,  obovate,  glabrous, 
apex  broadly  rounded.  Hypanthia  (at  anthesis) 
campanulate,  3-4  mm  long  to  the  torus  and  4- 
5  mm  in  diameter,  glabrous.  Calyx  tube  1  mm 
long;  calyx  lobes  1  mm  long  and  1-1.5  mm  wide 
basally,  ovate  to  deltoid-ovate  with  a  blunt  cal- 
lose-thickened  tooth  on  the  abaxial  apex  of  each 
lobe,  margin  entire  but  sometimes  roughened 
along  interlobe  sinuses,  glabrous  on  both  sur- 
faces. Petals  6,  glabrous,  1.2-1.4  x  1-1.2  cm, 
white  flushed  with  pink  unilaterally,  obovate, 
apically  rounded,  entire.  Stamens  12,  isomor- 
phic,  free  and  strongly  declined  to  one  side  of 
the  flower  opposing  the  style;  filaments  com- 
planate  and  glabrous,  5-6  mm  long;  anthers  3.5- 
5  mm  long,  1  mm  wide,  yellow,  linear-oblong 
and  tipped  with  a  solitary  dorsally  inclined  pore; 
connective  dilated  dorso-basally  into  a  deflexed 
spur  0.25  mm  long.  Ovary  Va  inferior,  4-celled, 
glabrous  at  the  summit  but  not  distended  into  a 
cone  or  collar.  Style  declinate  and  somewhat  sig- 
moid,  glabrous,  10-11  mm  long;  stigma  punc- 
tiform.  Berry  globose,  7-10  mm  long  and  10  mm 
in  diameter.  Seeds  clavate  to  narrowly  pyriform 
or  pyramidate,  1  mm  long,  beige  with  a  smooth 
testa  and  conspicuous  lateral  raphe. 

PHENOLOGY.— Flowering  and  fruiting  speci- 
mens have  been  collected  in  January,  March, 
April,  and  July. 

DISTRIBUTION.— Local  and  uncommon  in  cloud 
forests  from  the  Boquete  region  of  western  Pan- 
ama (Chiriqui)  to  the  slopes  bordering  the  Rio 
Hornito  above  Los  Planes  east  to  Cerro  Colorado 
at  1,000- 1,300m  (Fig.  5). 

ADDITIONAL  SPECIMENS  EXAMINED.— PANAMA.  Chiriqui: 
Edwin  Fabrega  Dam  and  Reserve  along  trail  to  Rio  Hornito 
above  Los  Planes,  8°45'N,  82°15'W,  18  Jan.  1989,  Almeda  et 
al.  6309  (CAS,  MO,  PMA);  Monte  Rey,  above  Boquete,  21 
Jul.  1971,  Croat  &  Porter  15692  (CAS,  MO);  trail  to  Zarzo, 
between  Los  Planes  de  Hornito  and  Fortuna  Lake,  8°41'N, 
82°13'W,  Hampshire  &  W hitefoord  689  (BM,  CAS);  ca.  5  km 
E  of  Fortuna  Dam  along  trail  crossing  Rio  Hornito,  8°45'N, 
82°15'W,  26  Apr.  1988,  Thompson  5021  (CAS,  CM).  Chiriqui/ 
Bocas  del  Tore  Border:  windswept  cloud  forest  off  the  road  to 
Cerro  Colorado,  26  Jan.  1989,  Almeda  et  al.  6418  (CAS,  CR, 
DUKE,  MO,  PMA,  US). 

Topobea  fragrantissima  is  distinguished  by  its 
glabrous  leaves  with  finely  striolate  secondary 


ALMEDA:  NEW  BLAKEA  AND  TOPOBEA 


319 


FIGURE  10.  Topobea  fragrantissima  Almeda.  A,  habit;  B,  representative  leaf  (abaxial  surface);  C,  D,  enlargement  of  foliar 
venation  (abaxial  surface);  E,  petal  (adaxial  surface);  F,  stamens,  ventral  view  (left),  lateral  view  (middle),  dorsal  view  (right); 
G,  berry  with  persistent  decussate  bracts;  H,  outer  floral  bract  (adaxial  surface);  I,  inner  floral  bract  (adaxial  surface);  J,  seeds. 
(A-I  from  the  holotype;  J  from  Croat  &  Porter  15692,  CAS.) 


venation,  long-pedicellate,  erect  flowers,  linear- 
oblong  anthers  with  solitary  dorsally  inclined 
pores,  and  a  4-celled  ovary.  Although  it  has  been 
described  as  an  epiphytic  shrub  or  small  tree,  my 


field  observations  and  the  label  information  on 
most  collections  suggest  that  this  species  typi- 
cally becomes  a  small,  free-living  tree.  The  style 
in  this  species  is  decimate  like  many  other  species 


320 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  14 


of  Topobea,  but  the  posture  and  orientation  of 
its  androecium  is  unlike  that  of  other  congeners 
with  free  anthers.  In  the  erect-flowered  species 
of  Topobea  with  free  anthers,  the  filaments  are 
declined  to  one  side  of  the  flower  opposing  the 
style  and  the  anthers  are  typically  erect  or  some- 
what incurved  apically.  In  T.fragrantissima  this 
orientation  is  taken  to  such  an  extreme  that  both 
filaments  and  anthers  lie  flat  on  the  surface  of 
the  petals  to  form  a  semicircular  configuration 
(Fig.  10A).  In  this  unusual  androecial  orientation 
the  ventral  surface  of  the  anther  faces  upward 
and  the  dorsal  surface  (including  the  dorsally  in- 
clined pore)  face  downward  toward  the  petal  sur- 
face. 

Field  observations  of  pollination  in  this  species 
should  be  of  special  interest  because  this  pollen 
presentation  mechanism  is  coupled  with  the  pro- 
duction of  a  perfume-like  fragrance.  The  unusual 
anther  posture  and  orientation  of  the  pore  may 
require  some  peculiar  manipulative  behavior  on 
the  part  of  bees  if  this  species  is  buzz-pollinated 
like  the  majority  of  species  in  the  family  with 
poricidally  dehiscent  anthers. 

Topobea  fragrantissima  is  similar  to  T.  suave- 
olens  (also  described  below)  in  a  number  of  char- 
acters, and  it  seems  likely  that  these  two  species 
are  closely  related.  The  shape  and  venation  of 
mature  leaf  blades  are  nearly  identical.  In  the  few 
known  collections  of  T.  suaveolens,  however,  fo- 
liar pocket  domatia  (Fig.  1 2B)  are  formed  abax- 
ially  in  the  angle  between  the  median  vein  and 
the  innermost  pair  of  lateral  veins;  domatia  are 
lacking  in  T.  fragrantissima.  The  differences  be- 
tween these  species  are  most  readily  apparent 
when  fresh  flowers  are  examined.  In  T.  suaveo- 
lens the  connivent  petals  give  the  pendant  flowers 
a  bell-like  conformation,  the  free  stamens  form 
a  ring  around  the  straight  style,  and  the  ovary  is 
6-celled.  The  most  distinctive  feature  shared  by 
the  two  species  is  the  solitary  anther  pore.  Within 
the  Blakeeae,  this  derived  character  is  known  in 
other  species.  Topobea  acuminata  Wurdack,  T. 
caudata  Wurdack,  T.  dodsonorum  Wurdack,  and 
T.  pittieri  Cogn.  are  a  few  notable  examples.  The 
fact  that  solitary  anther  pores  occur  in  more  than 
one  species  complex  within  Topobea  suggests  that 
this  character  has  arisen  more  than  once  in  the 
course  of  its  evolutionary  history. 

The  name  for  this  species  is  derived  fromfra- 
gro,  Latin  for  sweet  smelling,  in  reference  to  the 
perfume-like  fragrance  of  fresh  flowers. 


Topobea  hexandra  Almeda,  sp.  nov. 

(Fig.  11) 

TYPE.— PANAMA.  Panama:  Cerro  Jefe,  along  summit  road 
and  along  trail  into  the  Chagres  Valley,  elev.  ca.  900  m,  19 
Feb.  1988,  Almeda  et  al.  5837  (holotype:  CAS!;  isotypes:  CR!, 
DUKE!,  F!,  MO!,  NY!,  PMA!,  TEX!,  US!). 

Frutex  hemiepiphyticus.  Ramuli  primum  sulcato-quadran- 
gulati  demum  teretes  glabri  (in  nodis  caduce  puberuli  pilis 
castaneis  ca.  0.5-1  mm  longis).  Petioli  5-14  mm  longi;  lamina 
2-3.9  x  1.1-3  cm  subrotundata  vel  elliptico-ovata  apice  ro- 
tundato  vel  obtuso  basi  late  acuta  vel  obtusa  vel  rotundata, 
3-nervata  coriacea  et  integra,  nervis  secundariis  nervulisque 
invisis.  Flores  6-meri  sessiles  vel  subsessiles  in  quoque  nodo 
superiore  singuli  vel  bini;  bracteae  omnino  liberae;  bracteae 
exteriores  5-6.5  x  3-5  mm  ovatae  vel  ovato-ellipticae  apice 
obtuso  vel  mucronato;  bracteae  interiores  4-5  x  4-6  mm  late 
ovatae  vel  suborbiculares  apice  rotundato.  Hypanthium  (ad 
torum)  3  mm  longum  extus  sparsiuscule  caduceque  stellulato- 
furfuraceum;  calycis  tubus  1  mm  longus,  lobis  2  mm  longis. 
Petala  6.5-7  x  4  mm  obovato-elliptica  apice  obtuso.  Fila- 
menta  3  mm  longa;  antherae  6,  ca.  2  x  1  mm  oblongae  inter 
se  non  cohaerentes,  ventraliter  biporosae;  connectivum  ad  ba- 
sim  dorsaliter  dente  0.25  mm  descendenti  armatum.  Stylus 
5.5  mm;  ovarium  2-loculare  et  omnino  inferum  apice  glabro 
(cono  et  collo  non  evoluto). 

Hemiepiphytic  shrubs  to  1  m  tall  adhering  to 
the  bark  of  host  trees  by  nodal  and  internodal 
adventitious  roots.  Distal  branchlets  quadrate  to 
quadrisulcate,  glabrous  or  sparsely  beset  with 
spreading,  deciduous,  glandular  hairs  1-2  mm 
long  like  the  young  petioles  and  upper  and  lower 
surfaces  of  the  juvenile  leaves;  older  branches 
rounded  with  leaf  scars  that  are  typically  swollen 
and  nodular  in  appearance.  Uppermost  nodes 
copiously  beset  with  brown  spreading  hairs.  Veg- 
etative buds  copiously  covered  with  a  deciduous 
brown  stellate-lepidote  indument.  Leaves  of  a 
pair  equal  in  size,  glabrous  throughout;  petioles 
5-14  mm  long;  mature  blades  coriaceous,  2-3.9 
cm  long  and  1.1-3  cm  wide,  suborbicular  to  el- 
liptic-ovate, apex  rounded  or  varying  to  obtuse, 
base  obtuse  to  rounded,  rarely  varying  to  acute, 
margin  entire,  3 -nerved,  often  with  an  additional 
intramarginal  pair  of  depressed  veins.  Flowers 
erect,  solitary  or  paired  in  the  leaf  axils  of  distal 
branches,  sessile  or  subsessile  with  short  (to  1 
mm)  ill-defined  pedicels  formed  by  the  com- 
pressed bases  of  the  outer  floral  bracts.  Floral 
bracts  thick  and  semisucculent,  sessile,  entire  and 
free  to  the  base,  sparingly  stellulate-furfuraceous 
abaxially;  outer  bracts  5-6.5  x  3-5  mm,  con- 
cave, ovate  to  elliptic-ovate,  apex  obtuse  to 
bluntly  mucronate;  inner  bracts  4-5  x  4-6  mm, 
broadly  ovate  to  suborbicular,  apex  rounded. 
Hypanthia  (at  anthesis)  narrowly  campanulate, 


ALMEDA:  NEW  BLAKEA  AND  TOPOBEA 


321 


VJ 


D 


FIGURE  11.  Topobea  hexandra  Almeda.  A,  habit;  B,  representative  leaf  (abaxial  surface),  C,  enlargement  of  distal  node 
showing  spreading  hairs;  D,  representative  flower  (natural  posture);  E,  petal  (abaxial  surface);  F,  stamens,  ventral  view  (left), 
lateral  view  (right);  G,  berry  with  persistent  floral  bracts  removed;  H,  abaxial  surface  of  inner  floral  bract  (left),  abaxial  surface 
of  outer  floral  bract  (right);  I,  seeds.  (A-H  from  the  holotype;  I  from  Witherspoon  8552.) 


322 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  14 


3  mm  long  to  the  torus  and  4-5  mm  in  diameter, 
sparingly  stellate-lepidote  or  stellulate-furfura- 
ceous.  Calyx  tube  1  mm  long;  calyx  lobes  erect, 
2  mm  long  and  2-2.5  mm  wide  basally,  ovate  to 
deltoid-ovate,  entire  but  irregularly  roughened 
along  interlobe  sinuses,  sparingly  stellate-lepi- 
dote. Petals  6,  liberally  covered  with  disc-shaped 
hyaline  processes,  6.5-7  x  4  mm,  pink,  elliptic- 
obovate,  apically  obtuse,  entire.  Stamens  6,  iso- 
morphic  and  free;  filaments  complanate  and  gla- 
brous, 3  mm  long,  declinate  but  incurved  distally 
bringing  the  anthers  to  an  incurved  position  op- 
posing the  style;  anthers  2  mm  long,  1  mm  wide, 
pale  yellow,  turning  brownish  orange  with  age, 
oblong  with  two  ventrally  inclined  pores  at  the 
broadly  rounded  apex;  connective  slightly  thick- 
ened and  dilated  dorsally  at  the  filament  inser- 
tion into  a  short  blunt  spur  up  to  0.25  mm  long. 
Ovary  completely  inferior,  2-celled,  sparingly 
beset  with  stellate-lepidote  hairs,  glabrous  at  the 
summit  surrounding  the  stylar  scar  but  not  dis- 
tended into  a  prominent  cone  or  collar.  Style 
declinate,  incurved  distally,  glabrous,  5.5  mm 
long;  stigma  punctiform.  Berry  globose,  5-6  mm 
long  and  4-7.5  mm  in  diameter.  Seeds  bluntly 
deltoid,  1-1.5  mm  long,  beige  with  a  smooth 
glossy  testa  and  conspicuous  lateral  raphe. 

PHENOLOGY.  — The  only  known  flowering 
specimens  were  collected  in  February;  fruiting 
collections  have  been  made  in  February,  Septem- 
ber, October,  and  December. 

DISTRIBUTION. —Endemic  to  the  low  cloud  for- 
ests on  Cerro  Jefe  in  central  Panama  at  900- 
1,000  m  (Fig.  8). 

ADDITIONAL  SPECIMENS  EXAMINED.— PANAMA.  Panama: 
summit  and  S  facing  slopes  of  Cerro  Jefe,  9  Feb.  1978,  Almeda 
&  Nakai  3459  (CAS);  Cerro  Jefe,  along  trail  on  ridge  running 
NE  from  summit,  18  Dec.  1974,  Mori  &  Kallunki  3755  (MO); 
Cerro  Jefe,  29  Oct.  1980,  Sytsma  2007  (MO);  Cerro  Jefe,  road 
leading  N  from  summit,  26  Sep.  1975,  /.  T.  &  F.  WUherspoon 
8552  (MO). 

Collections  of  T.  hexandra  are  few  despite  the 
fact  that  it  is  one  of  the  most  common  shrubby 
epiphytes  in  the  forest  at  the  summit  of  Cerro 
Jefe.  In  the  field,  fertile  material  is  evidently 
overlooked  without  the  aid  of  binoculars,  be- 
cause the  small  flowers  are  produced  on  upper- 
most branchlets  positioned  high  on  host  trees 
where  access  to  sunlight  is  optimal. 

Topobea  hexandra,  together  with  T.  cordata, 
T.  caliginosa  and  T.  crassifolia,  form  a  closely 
related  species  group  best  interpreted  as  a  spe- 


cialized evolutionary  line  within  the  genus.  They 
are  distinguished  from  other  congeners  by  a  com- 
bination of  specialized  features  associated  with 
great  reduction  in  flower  size.  These  are:  (1)  flow- 
ers sessile  or  short-pedicellate  (2-3  mm)  at  an- 
thesis;  (2)  diminutive  petals  (4. 5-9  x  1.5-6  mm); 
(3)  each  flower  has  only  six  stamens,  each  of 
which  is  attached  to  the  torus  opposite  a  calyx 
lobe;  and  (4)  the  ovary  is  2-celled  or  4-celled. 
The  stamen  number  in  T.  crassifolia  and  T.  cali- 
ginosa was  unknown  to  me  when  I  described 
these  taxa,  because  all  available  study  material 
had  flowers  with  detached  stamens  (Almeda 
1980).  Recent  collections  of  these  two  species 
and  T.  cordata  clearly  show  that  they  are  all  hex- 
androus.  The  reduction  in  stamen  number  ex- 
hibited by  this  group  of  species  is  otherwise  un- 
known in  the  tribe. 

Among  its  close  allies,  T.  hexandra  is  most 
similar  to  T.  caliginosa.  Both  species  have 
3-nerved  leaves,  translucent  petals  that  have  a 
liberal  scattering  of  hyaline  disc-shaped  gland- 
like  processes,  and  2-celled  ovaries.  Topobea  ca- 
liginosa differs  in  having  glabrous  uppermost 
nodes,  oblanceolate  to  spatulate,  basally  atten- 
uate leaf  blades,  confluent  anther  pores,  and  un- 
appendaged  anther  connectives. 

Within  its  restricted  range,  T.  hexandra  shows 
inconstancy  in  characters  of  the  indument.  Distal 
branchlets  can  either  be  glabrous  or  sparsely  be- 
set with  spreading  glandular  hairs.  When  present 
these  hairs  are  found  on  distal  internodes,  peti- 
oles of  juvenile  foliage,  and  on  upper  and  lower 
surfaces  of  some  young  leaves,  but  most  of  the 
hairs  appear  to  fall  away  with  age.  All  specimens 
of  the  type  collection  are  devoid  of  glandular 
hairs.  Mori  &  Kallunki  3755  and  WUherspoon 
8552  are  beset  with  hairs  as  described  above,  but 
Almeda  &  Nakai  3459,  which  consists  of  two 
branches,  has  hairs  on  one  branch  but  not  on  the 
other.  Unfortunately,  T.  hexandra  is  known  from 
too  few  collections  to  determine  which,  if  any, 
of  these  forms  constitutes  the  prevalent  condi- 
tion. 

The  epithet  for  this  species  is  derived  from  the 
Greek  words  hex,  six,  and  andros,  male,  in  ref- 
erence to  the  6-stamened  flowers. 

Topobea  parvifolia  (Gleason)  Almeda,  comb.  nov. 

Blakea  parvifolia  Gleason,  Phytologia  3:357.  1950. 

TYPE.— PANAMA.  Code:  crest  of  Cerro  Pajita,  El  Valle  de 
Anton,  1,100  m,  Allen  3761  (holotype:  NY;  isotype:  MO!). 


ALMEDA:  NEW  BLAKEA  AND  TOPOBEA 


323 


In  the  protologue,  Gleason  described  the  an- 
thers ofBlakea  parvifolia  as  acute  and  emargin- 
ate  at  the  tip  with  each  anther  sac  opening  by  a 
separate  dorso-terminal  pore  (Fig.  3B).  Because 
anthers  of  this  kind  can  only  be  interpreted  as 
those  of  Topobea,  Gleason  appears  to  have  erred 
in  placing  this  species  in  Blakea  and  subsequent- 
ly maintaining  it  in  that  genus  for  his  treatment 
of  the  family  in  the  Flora  of  Panama  (Gleason 
1958).  Gleason's  brief  discussion  in  the  proto- 
logue also  stated  that  "the  one  flower  remaining 
had  lost  part  of  its  petals  and  stamens  but  there 
was  no  evidence  that  there  had  been  more  than 
four  of  the  one  and  eight  of  the  other."  My  study 
of  this  species  in  the  field  and  herbarium  shows 
that  it  has  stamens  that  are  coherent  in  a  ring, 
as  noted  by  Gleason,  but  the  flowers  are  consis- 
tently 6-merous  with  1 2  stamens. 

Despite  the  heightened  collecting  activity  in 
Panama  during  the  past  decade,  T.  parvifolia  is 
still  known  only  from  the  windswept  slopes  and 
ridges  in  the  vicinity  of  El  Valle  de  Anton. 

Topobea  suaveolens  Almeda,  sp.  nov. 
(Fig.  12) 

TYPE:  PANAMA.  Veraguas:  along  trail  to  summit  of  Cerro  Tute 
about  '/2  mile  above  the  Escuela  de  Agriculture  Alto  Piedra 
near  Santa  Fe,  elev.  900-1,100  m,  29  Jan.  1989,  Almeda  et  al. 
6484  (holotype:  CAS!;  isotypes:  AAU!,  BM!,  BR!,  CR!,  DUKE!, 
F!,  MEXU!,  MICH!,  MO!,  NY!,  PMA!,  TEX!,  US!). 

Arbor  epiphytica  4  m.  Ramuli  primum  sulcato-quadran- 
gulati  demum  teretes  glabri;  linea  interpetiolaris  paulo  elevata 
evoluta.  Petioli  1.1-1.8  cm  longi;  lamina  3.6-5.3  x  1.6-2.5 
cm  ell i plica  apice  acuminata  vel  caudato-acuminata  basi  acuta, 
3-plinervata,  nervi  in  axillis  acarodomatiis  instructi,  subcori- 
acea  et  integra,  nervis  secundariis  0.25  mm  inter  se  distantibus. 
Flores  6-meri  penduli  in  quoque  nodo  superiore  singuli,  pedi- 
cellis,  2.3-3  cm  longis;  bracteae  omnino  liberae;  bracteae  ex- 
teriores  4.5-7.5  x  3-4  mm  ellipticae  vel  ovato-ellipticae  apice 
acuto  vel  rotundato;  bracteae  interiores  4-4.5  x  4  mm  ovatae 
vel  suborbiculares  apice  rotundato.  Hypanthium  (ad  torum)  4 
mm  longum  extus  glabrum;  calycis  tubus  1 .5  mm  longus,  lobis 
1  mm  longis.  Petala  1.2-1.5  x  0.9-1.1  cm  obovata  apice  ro- 
tundato. Filamenta  2.5-3.5  mm  longa;  antherae  2.5  x  1  mm 
oblongae  inter  se  lateraliter  non  cohaerentes,  poro  unico  dor- 
saliter  inclinato;  connectivum  nee  prolongatum  nee  appendi- 
culatum.  Stylus  8.5-9  mm;  ovarium  6-loculare  et  '/2  inferum 
apice  glabro  (cono  et  collo  non  evoluto). 

Epiphytic  trees  to  4  m  tall,  often  obscuring  and 
overtaking  the  crowns  of  host  trees.  Distal 
branchlets  quadrate  to  quadrisulcate  and  gla- 
brous with  well-defined  interpetiolar  ridges  or 
lines.  Vegetative  buds  deciduously  lepidote-fur- 
furaceous.  Mature  leaves  of  a  pair  equal  or  slight- 
ly unequal  in  size,  glabrous  throughout;  petioles 
1.1-1.8  cm  long;  blades  subcoriaceous,  3.6-5.3 


cm  long  and  1.6-2.5  cm  wide,  elliptic,  apex  acu- 
minate to  caudate-acuminate,  base  acute,  margin 
entire,  3-plinerved  with  an  additional  incon- 
spicuous pair  of  submarginal  veins  and  numer- 
ous striolate  transverse  secondary  veins  spaced 
mostly  0.25  mm  apart  at  the  widest  portion  of 
the  blade,  pocket  domatia  typically  formed  abax- 
ially  in  the  angle  between  the  median  vein  and 
the  two  proximal  lateral  veins.  Flowers  pendant 
and  solitary  in  the  leaf  axils  of  uppermost 
branches;  pedicels  2.3-3  cm  long,  glabrous.  Flo- 
ral bracts  sessile,  glabrous,  entire,  and  free  to  the 
base;  outer  bracts  4.5-7.5  x  3-4  mm,  elliptic  to 
elliptic-ovate,  apex  bluntly  acute  to  rounded;  in- 
ner bracts  4-4.5  x  4  mm,  ovate  to  suborbicular, 
apex  broadly  rounded.  Hypanthia  (at  anthesis) 
campanulate,  4  mm  long  to  the  torus  and  5.5-6 
mm  in  diameter,  glabrous.  Calyx  tube  1.5  mm 
long;  calyx  lobes  1  mm  long  and  4  mm  wide 
basally,  broadly  ovate  to  deltoid-ovate  with  a 
blunt  callose-thickened  tooth  on  the  abaxial  apex 
of  each  lobe,  margin  entire,  glabrous  on  both 
surfaces.  Petals  6,  glabrous,  connivent  to  some- 
what imbricate  and  bell-like  when  fully  expand- 
ed, 1.2-1.5  x  0.9-1.1  cm,  white  flushed  with 
dark  pink  along  a  broad  continuous  margin,  ob- 
ovate,  apically  rounded,  entire.  Stamens  12,  iso- 
morphic,  free  and  encircling  the  exserted  style; 
filaments  complanate  and  glabrous,  2.5-3.5  mm 
long;  anthers  2.5  mm  long,  1  mm  wide,  yellow, 
laterally  compressed,  oblong  in  ventral  view  and 
narrowly  ovoid  in  profile  view  with  a  shallow 
dorso-basal  depression  at  the  filament  insertion, 
tipped  with  a  solitary,  dorsally  inclined  pore  0.75 
mm  in  diameter;  connective  simple.  Ovary  Vz 
inferior,  6-celled,  dilated  at  the  glabrous  summit 
into  a  smooth  gentle  dome  1-1.5  mm  high.  Style 
straight,  glabrous,  8.5-9  mm  long;  stigma  trun- 
cate. Berry  globose,  5-6  mm  long  and  6-7  mm 
in  diameter.  Seeds  narrowly  and  irregularly  pyr- 
iform,  0.75-1  mm  long,  pale  brown  with  a  smooth 
testa  and  prominent  lateral  raphe. 

PHENOLOGY.— The  three  known  collections  of 
this  species  which  are  in  flower  and  fruit  were 
collected  in  January,  February,  and  March. 

DISTRIBUTION.— Known  only  from  the  cloud 
forests  of  Cerro  Tute  in  west-central  Panama  at 
850-1, 100m  (Fig.  8). 

ADDITIONAL  SPECIMENS  EXAMINED.— PANAMA.  Veraguas: 
vicinity  of  Cerro  Tute,  along  trail  to  summit,  08°30'N,  8 1°07'W, 
19  Mar.  1987,  McPherson  10654  (CAS);  near  Cerro  Tute- 
Arizona,  above  Santa  Fe  and  Alto  de  Piedra,  8°30'N,  8  TIO'W, 
5  Feb.  1988,  McPherson  12043  (CAS). 


324 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  14 


FIGURE  12.  Topobea  suaveolens  Almeda.  A,  habit;  B,  representative  leaf  (abaxial  surface)  with  enlargement  showing  do- 
matium;  C,  flower  (natural  posture);  D,  petal  (adaxial  surface);  E,  stamens,  lateral  view  (left),  dorsal  view  (middle),  ventral  view 
(right);  F,  berry  with  persistent  decussate  bracts;  G,  seeds.  (A-F  from  the  holotype;  G  from  McPherson  10654.) 


ALMEDA:  NEW  BLAKEA  AND  TOPOBEA 


325 


Topobea  suaveolens  is  unusual  in  having  flow- 
ers that  are  bell-like  and  pendant  with  free  con- 
nivent  anthers  that  form  a  ring  around  the  exsert- 
ed  straight  style.  This  floral  syndrome  is  not  found 
in  the  other  species  of  Topobea  with  brightly 
colored  petals.  Topobea  suaveolens  is  clearly  a 
sister  species  of  T.fragrantissima  (also  described 
above).  In  addition  to  differences  noted  in  the 
discussion  of  the  latter,  T.  suaveolens  is  distinc- 
tive in  at  least  two  other  floral  features.  Its  fila- 
ments are  conspicuously  broadened  at  the  base 
(Fig.  1 2E),  and  its  shorter,  laterally  compressed 
anthers  are  narrowly  ovate  in  profile  view  and 
have  a  conspicuous  dorso-basal  depression. 

The  strong  vegetative  similarities  between  T. 
suaveolens  and  T.fragrantissima  can  make  iden- 
tification of  fruiting  specimens  difficult.  Al- 
though T.  suaveolens  is  known  from  few  collec- 
tions, even  sterile  material  can  be  distinguished 
by  the  peculiar  foliar  pocket  domatia  that  are 
formed  abaxially  in  the  angle  between  the  mid- 
vein  and  the  two  proximal  lateral  veins  (Fig.  1 2B). 

Perhaps  the  most  remarkable  feature  of  this 
species  is  its  agreeable  floral  fragrance.  I  de- 
scribed it  as  fruity  (Almeda  6484),  but  a  more 
telling  comparison  likens  it  to  grape-flavored 
chewing  gum  (McPherson  12043).  The  name  for 
this  species  is  derived  from  suavis,  Latin  for  sweet, 
in  reference  to  the  sweet-smelling  flowers. 

ACKNOWLEDGMENTS 

This  study  was  supported,  in  large  part,  by 
U.S.  National  Science  Foundation  Grant  BSR 
8614880  (Rora  Mesoamericana)  and  the  G. 
Lindsay  Field  Research  Fund  of  the  California 
Academy  of  Sciences.  For  field  assistance  and/ 
or  special  technical  support  in  the  herbarium  I 
thank  Bruce  Bartholomew,  Thomas  F.  Daniel, 
Greg  de  Nevers,  Gordon  McPherson,  Kei  Nakai, 
Orbelia  R.  Robinson,  and  Colleen  Sudekum.  For 
logistical  support  in  the  field,  special  thanks  go 
to  the  Museo  Nacional  de  Costa  Rica,  the  Mis- 
souri Botanical  Garden,  the  Organization  for 
Tropical  Studies,  and  the  Smithsonian  Tropical 
Research  Institute.  I  am  also  grateful  to  John  J. 
Wurdack  for  a  review  of  the  manuscript;  Varsha 
C.  Patel  for  information  on  pollen  morphology; 
Ellen  del  Valle  for  the  line  drawings;  Gina  Umana 
Dodero  for  reviewing  the  Spanish  summary;  and 
the  curators  and  staffs  of  the  following  herbaria 
(acronyms  fide  Holmgren  et  al.  1981)  for  loans, 
gifts,  or  special  assistance  during  study  visits: 


BM,  BR,  CM,  CR,  DUKE,  F,  K,  MEXU,  MO, 
NY,  P,  PMA,  TEX,  US,  WIS. 

RESUMEN 

Blakea  y  Topobea,  los  unices  generos  en  la 
tribu  Blakeeae,  estan  intimamente  relacionados, 
pero  las  anteras  de  Topobea  tienen  tecas  lineal- 
oblongas,  con  uno  o  dos  poros  dorso-apicales  o 
dos  poros  ventro-apicales.  Topobea  es  un  genero 
con  mas  de  60  especies  descritas,  incluido  en 
Blakea  solamente  por  Don,  Baillon  y  Macbride, 
pero  es  muy  util  reconocerlo  hasta  que  dispon- 
gamos  de  una  monografia.  Se  presenta  una  his- 
toria  taxonomica  de  los  dos  generos  con  un  co- 
mentario  sobre  las  caracteristicas  de  las  anteras 
empleado  originalmente  por  Browne,  Aublet, 
Cogniaux,  y  otros  botanicos.  Las  diferencias  ge- 
nericas  a  menudo  se  han  alterado  un  poco  para 
reflejar  mejor  las  caracteristicas  inusuales  de  al- 
gunas  especies  que  se  encuentran  en  las  monta- 
nas  de  Costa  Rica  y  Panama.  Se  provee  una  clave 
neuva  para  los  dos  generos.  Se  describen  cuatro 
especies  nuevas  de  Blakea  (B.  gregii,  B.  ham- 
melii,  y  B.  herrerae  de  Panama;  y  B.  scarlatina 
de  Costa  Rica  y  Nicaragua),  y  tres  especies  nue- 
vas de  Topobea  (T.fragrantissima,  T.  hexandra, 
y  T.  suaveolens  de  Panama)  y  se  transfieren  tres 
especies  de  Blakea  a  Topobea.  Estos  cambios  se 
basan  en  caracteristicas  de  las  anteras.  Se  pro- 
veen  descripciones,  ilustraciones  y  discusiones 
sobre  las  afinidades  entre  las  especies  nuevas. 
Ademas  se  presentan  notas  sobre  polinizacion  y 
mapas  de  distribucidn. 

LITERATURE  CITED 

ACKERMAN,  J.  D.    1986.  Coping  with  the  epiphytic  existence: 

pollination  strategies.  Selbyana  9:52-60. 
ALMEDA,  F.    1 980.   Central  American  novelties  in  the  genus 

Blakea  (Melastomataceae).  Rhodora  82:609-615. 
.    1 98 1  a.  Blakea  penduliflora  (Melastomataceae):  a  new 

green-flowered  species  from  Costa  Rica.  Brittonia  32(4):508- 

511. 
.    1981b.    New  and  reconsidered  species  of  Miconia 


(Melastomataceae)  from  Costa  Rica  and  Panama.  Proc.  Cal- 
if. Acad.  Sci.  42(10):303-314. 
.     1989.    Five  new  berry-fruited  species  of  tropical 


American  Melastomataceae.  Proc.  Calif.  Acad.  Sci.  46(5): 
137-150. 

.  In  press.  Melastomataceae.  In  Flora  de  Nicaragua. 

W.  D.  Stevens,  ed.  Missouri  Botanical  Garden,  St.  Louis. 

AUBLET,  J.  B.  C.  F.  1775.  Histoire  des  plantes  de  la  Guiane 
francoise.  London  and  Paris.  4  vo\s.  (Topobea,  1:476,  t.  189.) 

BAILLON,  H.  E.  1879.  Melastomacees.  Hist.  PI.  7:1-65.  (En- 
glish transl.  7:1-65.  1881.) 

BENTHAM,  G.  AND  J.  D.  HOOKER.  1867.  Melastomaceae. 
Genera  plantarum  l(2):725-773. 


326 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  14 


BROWNE,  P.    1756.   The  civil  and  natural  history  of  Jamaica 

in  three  parts.  London.  (Blakea,  p.  323,  t.  34.) 
BUCHMANN,  S.  L.   1 983.  Buzz  pollination  in  angiosperms.  Pp. 

73-113  in  Handbook  of  experimental  pollination  biology. 

C.  E.  Jones  and  R.  J.  Little,  eds.  Van  Nostrand  Reinhold 

Co.,  Inc.,  New  York,  New  York. 
CANDOLLE,  A.  P.  DE    1828.    Melastomaceae.  In  Prodromus 

systematis  naturalis  regni  vegetabilis  3:99-202. 
COGNIAUX,  A.    1888.   Melastomaceae  (Blakeae).  In  C.  F.  P. 

von  Martius,  Flora  Brasiliensis  14(4):558-564. 
.   1891.  Melastomacees.  In  A.  and  C.  de  Candolle,  eds. 

Monographiae  phanerogamarum  7: 1-1 256.  G.  Masson,  Paris. 
DON,  D.    1823.    An  illustration  of  the  natural  family  called 

Melastomataceae.  Mem.  Wern.  Nat.  Hist.  Soc.  4:276-329. 
GLEASON,  H.  A.   1935.  Melastomaceae.  In  A.  Pulle,  ed.  Flora 

of  Surinam  3:178-281.  J.  H.  de  Bussy  Ltd.,  Amsterdam. 
.    1945.    On  Blakea  and  Topobea.  Bull.  Torrey  Bot. 

Club  72(4):385-393. 
.   1 947.  Additional  notes  on  Blakea  and  Topobea.  Phy- 

tologia  2:279-281. 
.     1950.    Observations  on  tropical  American  mela- 

stomes.  Phytologia  3:345-360. 
.    1958.   Melastomataceae.  In  R.  E.  Woodson,  Jr.  and 

R.  W.  Schery,  eds.  Flora  of  Panama.  Ann.  Missouri  Bot. 

Gard.  45:203-304. 
HOLMGREN,  P.  K.,  W.  KEUKEN,  AND  E.  K.  SCHOFIELD.    1981. 

Index  Herbariorum.  Part  1, 7th  ed.  The  herbaria  of  the  world. 

Regnum  Veg.  106:1-152. 
HOWARD,  R.   1983.  The  plates  of  Aublet's  Histoire  des  plantes 

de  la  Guiane  francoise.  J.  Arnold  Arbor.  64:225-292. 
JACOBS,  M.    1966.    On  domatia— the  viewpoints  and  some 

facts.  I,  II,  III.  Proc.  Koninkl.  Nederl.  Akad.  Wetensch.  C69: 

275-316. 
JUSSIEU,  A.  L.  DE.    1789.   Genera  plantarum.  Paris. 

KOEK-NOORMAN,  J.,  P.  HOGEWEG,   W.   H.   M.  VAN  MAANEN, 

AND  B.  J.  H.  TER  WELLE.  1979.  Wood  anatomy  of  the 
Blakeeae  (Melastomataceae).  Acta  Bot.  Neerl.  28:21^*3. 

LINNAEUS,  C.  1759.  Systema  naturae.  Vol.  2,  10th  ed.  Stock- 
holm. (Facsimile  ed.  by  J.  Cramer,  Weinheim,  1964.) 

LINNAEUS,  C.,  VON  FILIUS  1781.  Supplementarum  plantarum. 
Braunschweig. 

LUMER,  C.  1981.  Rodent  pollination  of  Blakea  (Melasto- 
mataceae) in  a  Costa  Rican  cloud  forest.  Brittonia  32(4): 
512-517. 

LUMER,  C.  AND  R.  D.  SCHOER.    1986.    Pollination  of  Blakea 


austin-smithii  and  B.  penduliflora  (Melastomataceae)  by  small 
rodents  in  Costa  Rica.  Biotropica  18(4):363-364. 

LUNDSTROM,  A.  N.  1 887.  Pflanzenbiologische  studien  II.  Die 
anpassungen  der  pflanzen  an  thiere,  I.  Von  domatia.  Nova 
Acta  Regiae  Soc.  Sci.  Upsal.  (Ser.  3)  13(10):l-88. 

MACBRIDE,  J.  F.  1 94 1 .  Melastomataceae.  In  Flora  of  Peru. 
Field  Mus.  Nat.  Hist.,  Bot  Ser.  13:249-521. 

NAUDIN,  C.  1852.  Melastomacearum  monographicae  de- 
scriptionis.  Ann.  Sci.  Nat.  Ill,  18:85-154. 

PATEL,  V.  C,  J.  J.  SKVARLA,  AND  P.  H.  RAVEN  1985.  Pollen 
characters  in  relation  to  the  delimitation  of  Myrtales.  Ann. 
Missouri  Bot.  Gard.  71:858-969. 

PEMBERTON,  R.  W.  AND  C.  E.  TURNER  1989.  Occurrence  of 
predatory  and  fungivorous  mites  in  leaf  domatia.  Amer.  J. 
Bot.  76(1):  105- 11 2. 

RENNER,  S.  S.  1 989.  A  survey  of  reproductive  biology  in 
neotropical  Melastomataceae  and  Memecylaceae.  Ann.  Mis- 
souri Bot.  Gard.  76:496-518. 

ROUBIK,  D.  W.  1989.  Ecology  and  natural  history  of  tropical 
bees.  Cambridge  University  Press,  Cambridge,  England. 

SOLT,  M.  L.  ANDj.J.  WURDACK  1980.  Chromosome  numbers 
in  the  Melastomataceae.  Phytologia  47:199-220. 

STAGE,  C.  A.  1965.  Cuticular  studies  as  an  aid  to  plant  tax- 
onomy. Bull.  Brit.  Mus.  (Nat.  Hist.),  Bot.  4(1):  1-78. 

STANDLEY,  P.  C.  1 924.  Melastomaceae.  In  Trees  and  shrubs 
of  Mexico.  Contr.  U.S.  Natl.  Herb.  23(4):  1046-1074. 

.    1938.   Melastomaceae.  In  Flora  of  Costa  Rica.  Field 

Mus.  Nat.  Hist.,  Bot  Ser.  18(3):783-845. 

STANDLEY,  P.  C.  AND  L.  O.  WILLIAMS.  1963.  Melastomaceae. 
In  Flora  of  Guatemala.  Fieldiana,  Bot.  24:407-525. 

STEIN,  B.  A.  AND  H.  TOBE.  1989.  Floral  nectaries  in  Melas- 
tomataceae and  their  systematic  and  evolutionary  implica- 
tions. Ann.  Missouri  Bot.  Gard.  76:519-531. 

TRIANA,  J.  1871.  Les  Melastomacees.  Trans.  Linn.  Soc.  Lon- 
don 28:1-188. 

WELLE,  B.  J.  H.  TER  AND  J.  KOEK-NOORMAN.  1981.  Wood 
anatomy  of  the  neotropical  Melastomataceae.  Blumea  27: 
335-394. 

WILLIAMS,  N.  H.  1983.  Floral  fragrances  as  cues  in  animal 
behavior.  Pp.  50-72  in  Handbook  of  experimental  polli- 
nation biology.  C.  E.  Jones  and  R.  J.  Little,  eds.  Van  Nos- 
trand Reinhold  Co.,  Inc.,  New  York,  New  York. 

WURDACK,  J.  J.  1973.  Melastomataceae.  In  T.  Lasser,  ed. 
Flora  de  Venezuela  8:1-819. 

.   1980.  Melastomataceae.  In  G.  Harling  and  B.  Sparre, 

eds.  Flora  of  Ecuador  13: 1^06. 


CALIFORNIA  ACADEMY  OF  SCIENCES 

Golden  Gate  Park 
San  Francisco,  California  941 18 


PROCEEDINGS 

OF  THE 

CALIFORNIA  ACADEMY  OF  SCIENCES 


Vol.  46,  No.  15,  pp.  327-335,  4  figs. 


September  11,  1990 


THREE  NEW  SPECIES  OF  CONOSTEGIA 

(MELASTOMATACEAE:  MICONIEAE)  FROM 

SOUTHERN  CENTRAL  AMERICA 

By 
Frank  Almeda 

Department  of  Botany,  California  Academy  of  Sciences,  Golden  Gate  Park, 
San  Francisco,  California  94118 

ABSTRACT:  Three  new  species  of  Conostegia,  a  well-defined  berry-fruited  genus  of  Melastomataceae  with  a 
center  of  diversity  in  Costa  Rica  and  Panama,  have  been  collected  in  little-explored  regions  of  southern  Costa 
Rica  and  western  Panama  during  the  past  decade.  Descriptions,  discussions,  diagnostic  illustrations,  pheno- 
logical  notes,  and  a  distribution  map  are  provided  for  Conostegia  fragrantissima,  C.  muriculata,  and  C. 
orbeliana.  Attention  is  drawn  to  the  need  for  recording  field  data  on  taxonomically  useful  floral  and  fruit 
characters. 

Received  February  22, 1990.  Accepted  April  10,  1990. 


INTRODUCTION 

Conostegia,  a  neotropical  genus  of  about  50 
species  is  one  of  approximately  30  berry-fruited 
genera  assigned  to  the  New  World  tribe  Mico- 
nieae.  Unlike  some  genera  in  this  large  tribe  of 
1,800-1,900  species,  Conostegia  is  a  natural  as- 
semblage readily  denned  by  its  terminal  inflo- 
rescence, calyptriform  calyx  that  is  circumscissle 
at  or  near  the  torus,  and  isomorphic  stamens 
with  lorate  anthers  and  unappendaged  connec- 
tives. Conostegia  is  also  noteworthy  among  neo- 
tropical Melastomataceae  because  most  (but  not 
all)  of  its  species  have  flowers  that  are  anisomer- 
ous  and  typically  pleiostemonous.  Among  other 
genera  in  the  Miconieae  only  a  few  species  of 
Clidemia  and  Miconia  have  flowers  with  unequal 
numbers  or  proportions  of  serially  homologous 
structures. 

Variation  in  several  species  of  Conostegia  re- 
mains imperfectly  understood  because  of  inad- 
equate geographic  sampling  and  fragmentary 
preservation  of  floral  characters  on  herbarium 


specimens.  Petal  and  stamen  numbers,  the  pos- 
ture and  orientation  of  the  style,  stigma  mor- 
phology, and  ovary  cell  number  provide  useful 
information  for  the  delimitation  of  species  in 
Conostegia.  These  characters  are  often  difficult 
to  evaluate  after  specimens  have  been  pressed 
and  dried.  Ideally,  notes  on  these  features  should 
be  recorded  in  the  field  when  the  preservation  of 
flowers  and  fruits  in  a  liquid  medium  is  not  fea- 
sible. 

The  three  species  described  below  all  come 
from  rich,  but  little-known,  areas  of  Costa  Rica 
and  Panama.  Over  33  species  of  Conostegia  oc- 
cur in  this  small  isthmian  region,  making  it  the 
center  of  diversity  for  the  genus  and  a  promising 
area  for  continued  botanical  exploration  in  the 
Mesoamerican  region. 

Conostegia  fragrantissima  Almeda,  sp.  nov. 

Fig.  1 

TYPE.— PANAMA.  Bocas  del  Toro.  Fortuna  Dam  area,  along 
continental  divide  trail  bordering  Chiriqui  Province  at 
8°45'04"N,  82°15'04"W,  1,200-1,300  m,  10  Mar.  1988,  Al- 


[327] 


328 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  15 


FIGURE  1.  Conostegia  fragrantissima  Almeda.  A,  habit,  x  ca.  l/y,  B,  representative  leaf  (abaxial  surface),  x  1;  C,  flower  bud 
when  fresh,  x  ca.  4;  D,  young  berry  with  detached  calyptriform  calyx,  x  5;  E,  flower  bud  when  dry,  x  ca.  4;  F,  representative 
flower  showing  reflexed  petals,  declinate  style  and  stigma  enlargement  (right),  x  ca.  3;  G,  petal  (adaxial  surface),  x  ca.  5;  H, 
stamens,  ventral  view  (left)  and  lateral  view  (right),  x  ca.  8;  I,  seeds,  x  ca.  30.  (A-I  from  the  holotype.) 


meda  et  al.  6064  (holotype:  CAS!;  isotypes:  AAU!,  BM!,  BR!,       1-2.6  cm  longi;  lamina  4.5-9.5  x    1.5^4.2  cm  elliptica  vel 
CR!,  DUKE!,  F!,  MEXU!,  MO!,  NY!,  PMA!,  TEX!,  US!).  elliptico-lanceolata  apice  acuminata  vel  caudato-acuminata  basi 

acuta,  3(-5)-plinervata,  membranacea  et  Integra  supra  glabra, 

Arbor  4-11  m.  Ramuli  primum  quadrangulati  demum  tere-      subtus  sparsiuscule  caduceque  lepidota.  Inflorescentia  4-7.5 
tes  glabri;  linea  interpetiolaris  paulo  elevata  evoluta.  Petioli      cm  longa  multiflora;  pedicellis  (ad  anthesim)  2.5-7  mm  longi; 


ALMEDA:  NEW  SPECIES  OF  CONOSTEGIA 


329 


C.  fragrantissima 
C.  muriculata 
C.  orbeliana 


too         200  km 


FIGURE  2.    Distributions  of  Conostegia  fragrantissima,  C.  muriculata,  and  C.  orbeliana. 


bracteolis  0.25-1.5  mm  longis  mox  caducis.  Alabastra  matura 
acuminata  vel  apiculata  5-7  x  2-3  mm;  calyx  calyptriformis 
3-4.5  mm  longus.  Petala  6-7,  glabra  5-7  x  3.5-4.5  mm,  obo- 
vata  apice  irregulariter  rotundato.  Stamina  15-19,  isomor- 
phica  glabra;  filamenta  2-2.5  mm  longa;  antherarum  thecae 
1.5-2  x  0.5  mm  oblongae,  poro  ventraliter  inclinato;  connec- 
tivum  nee  prolongatum  nee  appendiculatum.  Stylus  5-5.5  mm 
glaber;  stigma  capitatum  1.5  mm  diam.;  ovarium  6-loculare 
et  omnino  inferum  apice  glabro  (cono  et  collo  non  evoluto). 

Trees  4-1 1  m  tall.  The  distal  branchlets  gla- 
brous and  bluntly  quadrate  with  darkened  inter- 
petiolar  lines  or  ridges.  Leaves  of  a  pair  equal  to 
somewhat  unequal  in  size;  petioles  1-2.6  cm  long; 
blades  membranaceous,  4.5-9.5  cm  long  and  1.5- 
4.2  cm  wide,  elliptic  to  elliptic-lanceolate,  apex 
acuminate  to  caudate-acuminate,  base  acute, 
margin  entire,  3(-5)-plinerved,  the  innermost  pair 
of  elevated  primaries  diverging  from  the  median 
nerve  in  opposite  fashion  (l-)3-5  mm  above  the 
blade  base,  glabrous  above,  sparsely  and  decid- 
uously  furfuraceous-lepidote  below.  Inflores- 
cence a  terminal  paniculiform  dichasium  4-7.5 
cm  long  typically  branching  above  the  node  ini- 
tiating the  inflorescence,  the  rachis  inconspic- 
uously and  deciduously  furfuraceous-lepidote; 
bracteoles  sessile  to  subsessile,  early  deciduous, 
narrowly  triangular  to  subulate  or  linear-oblong, 
0.25-1.5  mm  long,  0.25  mm  wide,  glabrous, 
margin  entire.  Pedicels  2.5-7  mm  long,  decidu- 


ously furfuraceous-lepidote.  Flower  buds  ob- 
long-ellipsoid, becoming  horizontally  constrict- 
ed at  the  torus  and  truncate  at  the  base  when 
dry,  5-7  mm  long,  2-3  mm  wide,  smooth  and 
glabrous  or  sparsely  furfuraceous-lepidote,  the 
calyptriform  calyx  3-4.5  mm  long,  acuminate  to 
apiculate;  torus  glabrous  adaxially.  Petals  6-7, 
glabrous,  white  with  a  horizontal  red  band  near 
the  base,  obovate,  reflexed,  irregularly  lobed  api- 
cally,  5-7  mm  long  and  3.5-4.5  mm  wide  dis- 
tally.  Stamens  15-19,  gently  declined  to  one  side 
of  the  flower  opposing  the  style;  filaments  gla- 
brous, complanate,  2-2.5  mm  long;  anthers  1.5- 
2  mm  long,  0.5  mm  wide,  orange,  linear-oblong, 
truncate  to  broadly  rounded  at  the  apex  with  a 
somewhat  ventrally  inclined  terminal  pore  0.25 
mm  in  diam.;  connective  thickened  dorsally  but 
not  prolonged  or  appendaged  at  the  filament  in- 
sertion. Ovary  (at  anthesis)  completely  inferior, 
6-celled,  ovoid,  the  apex  glabrous,  smooth  and 
lacking  an  elevated  cone  or  collar.  Style  decli- 
nate,  glabrous,  5-5.5  mm  long;  stigma  capitate, 
1.5  mm  in  diam.  with  6-7  papillose  lobes.  Berry 
depressed-globose,  3  mm  long  and  4  mm  in  diam. 
Seeds  oblong,  oblong-ovoid,  or  narrowly  pyram- 
idate,  0.5-0.75  mm  long,  beige,  smooth  and  pol- 
ished on  the  convex  face. 
PHENOLOGY.— Flowering  specimens  have  been 


330 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  15 


collected  in  the  months  of  February  through  May; 
fruiting  specimens  have  been  collected  in  Feb- 
ruary and  April. 

DISTRIBUTION.— Local  in  wet  evergreen  and 
elfin  forests  of  western  Panama  from  Cerro  Pate 
Macho  (Chiriqui)  to  the  Cerro  Colorado  region 
(Bocas  del  Toro)  at  1,200-2,000  m  (Fig.  2). 

ADDITIONAL  SPECIMENS  EXAMINED.  — PANAMA.  Bocas  del 
Toro:  Fortuna  Dam  area,  along  continental  divide  trail, 
8°45'04"N,  82°15'04"W,  10  Mar.  1988,  Almeda  et  al.  6068 
(CAS);  Fortuna  Dam  region  near  trail  along  continental  divide, 
8°45'N,  82°15'W,  11  Feb.  1986,  McPherson  8406  (CAS);  vi- 
cinity of  Cerro  Colorado  on  trail  along  creek  8.6  miles  from 
Camp  Chami,  8°35'N,  81°45'W,  14  Apr.  1986,  McPherson 
8900  (CAS);  vicinity  of  Cerro  Colorado  mine  above  San  Felix, 
8°35'N,  8°50'W,  26  Jan.  1988,  McPherson  12024  (CAS).  Chi- 
riqui: SE  slopes  and  summit  of  Cerro  Pate  Macho,  4  km  NE 
of  Boquete,  26  May  1981,  Sytsma  et  al.  4884  (CAS). 

This  species  is  distinguished  by  a  combination 
of  indument  and  floral  characters.  The  lower  leaf 
surfaces,  branchlets  of  the  inflorescence,  pedi- 
cels, and  distal  cauline  internodes  are  sparsely 
beset  with  deciduous  furfuraceous-lepidote  hairs. 
Although  many  specimens  superficially  appear 
glabrous  to  the  unaided  eye,  the  inconspicuous 
indument  is  readily  detected  on  one  or  all  of  the 
above-mentioned  structures  when  examined  with 
a  dissecting  microscope.  Aside  from  the  decli- 
nate  style  and  lobulate  depressed-capitate  stig- 
ma, most  of  the  distinctive  floral  features  are  not 
readily  gleaned  from  herbarium  specimens  with- 
out good  field  notes.  The  flowers  are  fragrant,  the 
anthers  are  bright  orange  in  color,  and  each  con- 
spicuously reflexed  white  petal  has  a  red  hori- 
zontal band  near  the  base.  This  distinctive  petal 
coloration  is,  to  my  knowledge,  otherwise  un- 
known in  Conostegia. 

The  closest  relative  of  C.  fragrantissima  is 
probably  C.  pittieri  Cogn.,  a  species  known  only 
from  Costa  Rica  and  Nicaragua.  They  both  have 
elliptic  to  elliptic-lanceolate,  caudate-acuminate, 
plinerved  leaves  and  flowers  with  similar  petal 
and  stamen  numbers.  Conostegia  pittieri  is  gla- 
brous throughout,  and  its  leaves  are  prevailingly 
undulate-denticulate.  It  also  differs  from  C.  fra- 
grantissima in  having  longer  floral  buds  (9-13 
mm),  larger  petals  (9-13  x  5-8  mm),  more  ovary 
cells  (8-10),  and  a  crateriform  capitate  stigma. 

In  Gleason's  treatment  of  Conostegia  for  the 
Flora  of  Panama  (Gleason  1958),  C.  fragrantis- 
sima keys  closest  to  C.  montana  (Sw.)  D.  Don 
ex  DC.,  a  widespread  species  ranging  from  south- 
ern Mexico  (Chiapas)  and  the  West  Indies  (How- 


ard 1989)  south  through  Central  America  to  Co- 
lombia, Venezuela,  and  Ecuador.  Conostegia 
fragrantissima  shares  many  features  with  C. 
montana  including  foliar  shape,  petal  and  sta- 
men numbers,  and  the  number  of  ovary  cells, 
but  the  latter  has  stellate  hairs  on  the  inflores- 
cence rachis,  pedicels,  and  elevated  primary  veins 
on  lower  leaf  surfaces.  In  addition,  the  flowers 
of  C.  montana  produce  no  detectable  odor,  the 
pedicels  are  shorter  (1  mm),  the  petals  are  com- 
pletely white,  the  anthers  are  pale  yellow,  and 
the  style  (when  dry)  is  abruptly  expanded  just 
below  the  subtruncate  unlobed  stigma.  The  style 
and  stigma  enlargement  of  C.  fragrantissima  (Fig. 
IF)  was  drawn  from  liquid-preserved  flowers. 
Upon  drying,  the  style  retains  its  linear  outline, 
but  the  capitate  stigma  becomes  gently  fluted  and 
bluntly  obconic  in  profile  view. 

The  fragrant  flowers  of  C.  fragrantissima  are 
unusual  but  not  unique  among  its  congeners.  The 
flowers  of  C.  pittieri  are  also  fragrant  (fide  Lumer 
1148,  CAS),  but  information  of  this  kind  is  rarely 
reported  on  herbarium  labels.  In  choosing  the 
epithet  for  this  species,  which  is  derived  from 
fragro,  Latin  for  sweet  smelling,  I  draw  attention 
to  the  very  fruity  fragrance  of  the  fresh  flowers. 

Conostegia  muriculata  Almeda,  sp.  nov. 

(Fig.  3) 

TYPE.  —  PANAMA.  Bocas  del  Toro:  Above  Chiriqui  Grande, 
1 0  road-miles  from  the  continental  divide  and  2  miles  along 
road  to  the  east,  8°55'N,  82°10'W,  300  m,  6  Aug.  1988, 
McPherson  12836  (holotype:  CAS!;  isotypes:  CR!,  DUKE!, 
MO!,  PMA!,  US!). 

Frutex  vel  arbor  parva  2.5-3.5  m.  Ramuli  primum  acute 
quadrangulati  demum  teretes  glabri;  linea  interpetiolaris  ob- 
scure evoluta.  Petioli  1.5-7  cm  longi;  lamina  8-25.5  x  4.5- 
14  cm  elliptica  vel  obovato-elliptica  apice  acuminata  basi  acu- 
ta,  5-nervata  vel  5-7-plinervata,  membranacea  et  integra,  su- 
pra glabra,  subtus  sparsiuscule  caduceque  lepidota.  Inflores- 
centia  12-30  cm  longa  (pedunculo  4-10  cm  longo  incluso)  laxa 
multiflora;  pedicellis  (ad  anthesim)  0.5-3  mm  longis;  bracteolis 
0.5-1  mm  longis  mox  caducis.  Alabastra  matura  acuminata 
6-10  x  3.5-4.5  mm;  calyx  calyptriformis  3-5  mm  longus. 
Petala  5,  glabra,  5-6  x  4_5(_6.5)  mm,  obovata  apice  rotun- 
dato.  Stamina  (8-)9-10,  isomorphica  glabra;  filamenta  2.5-3 
mm  longa;  antherarum  thecae  2-3  x  0.75  mm  oblongae,  poro 
ventraliter  inclinato;  connectivum  nee  prolongatum  nee  ap- 
pendiculatum.  Stylus  3-4  mm  glaber;  stigma  capitatum  0.75 
mm  diam. :  ovarium  6-loculare  et  omnino  inferum  apice  glabro 
(cono  et  collo  non  evoluto). 

Shrubs  or  small  trees  2.5-3.5  m  tall.  The  distal 
branchlets  glabrous  with  interpetiolar  lines, 
quadrate  and  carinate  on  the  angles  but  becom- 


ALMEDA:  NEW  SPECIES  OF  CONOSTEGIA 


331 


B 


FIGURE  3.  Conostegia  muriculata  Almeda.  A,  habit,  x  ca.  Vi;  B,  representative  leaf  (abaxial  surface),  x  '/2;  C,  flower  bud 
(left)  and  bud  with  detached  calyptriform  calyx  and  floral  parts  removed  (right),  x  4;  D,  representative  flower  showing  spreading 
petals,  arching  style,  and  stigma  enlargement  (right),  x  5;  E,  petal  (adaxial  surface),  x  6;  F,  stamens,  ventral  view  (left)  and 
lateral  view  (right),  x  7;  G,  seeds,  x  26.  (A-F  from  the  holotype;  G  from  Almeda  et  al.  6328.) 


332 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  15 


ing  rounded-quadrate  with  age.  Leaves  of  a  pair 
equal  or  unequal  in  size  (if  unequal  then  the 
smaller  one  2/3  to  3/5  the  size  of  the  larger  one); 
petioles  1.5-7  cm  long,  sparsely  to  moderately 
lepidote  when  young;  blades  membranaceous,  8- 
25.4  cm  long  and  4.5-14  cm  wide,  elliptic  to 
elliptic-obovate,  apex  acuminate,  base  acute, 
margin  entire,  5 -nerved  or  inconspicuously  5-7- 
plinerved  with  a  prominulous  network  of  sec- 
ondary and  higher  order  veins,  glabrous  above, 
sparsely  and  deciduously  lepidote  below.  Inflo- 
rescence a  terminal  paniculiform  dichasium  1 2- 
30  cm  long  on  a  well-defined  arching  or  deflexed 
peduncle  mostly  3.5-10  cm  long,  the  rachis  gla- 
brous; bracteoles  sessile,  early  deciduous,  nar- 
rowly triangular  to  subulate,  0.5-1  mm  long, 
0.25-0.5  mm  wide,  sparingly  lepidote  to  glabrate 
abaxially,  margin  entire.  Pedicels  0.5-3  mm  long, 
sparsely  to  moderately  lepidote.  Flower  buds  ob- 
long-ellipsoid, 6-10  mm  long  3.5-4.5  mm  wide, 
sparsely  to  moderately  lepidote  basally,  the  ca- 
lyptriform  calyx  3-5  mm  long,  acuminate;  torus 
glabrous  adaxially  and  callose  thickened  along 
the  adjacent  circumscissle  line.  Petals  5,  gla- 
brous, obovate,  reflexed  to  widely  spreading, 
gently  lobed  to  somewhat  oblique  apically,  5-6 
mm  long,  4-5(-6.5)  mm  wide  distally.  Stamens 
(8-)9-10;  filaments  glabrous,  complanate,  2.5-3 
mm  long;  anthers  2-3  mm  long,  0.75  mm  wide, 
pale  yellow,  linear-oblong  with  a  ventrally  in- 
clined terminal  pore  0.25  mm  in  diam.;  connec- 
tive thickened  dorsally  but  not  prolonged  or  ap- 
pendaged  at  the  filament  insertion.  Ovary  (at 
anthesis)  completely  inferior,  6-celled,  ovoid,  the 
apex  glabrous,  smooth,  and  lacking  an  elevated 
cone  or  collar.  Style  conspicuously  curved  dis- 
tally, glabrous,  3-4  mm  long;  stigma  capitate, 
0.75  mm  in  diameter,  copiously  papillose  but 
lacking  well-defined  lobes.  Berry  purple  at  ma- 
turity, depressed-ovoid,  5-6  mm  long  and  5-7 
mm  in  diameter.  Seeds  narrowly  pyriform,  0.75 
mm  long,  yellowish-white,  irregularly  muriculate 
on  the  convex  face. 

PHENOLOGY.— Flowering  specimens  have  been 
collected  in  April,  July,  August,  and  September; 
specimens  in  young  fruit  have  been  collected  in 
September,  but  mature  fruiting  collections  have 
been  made  only  in  January. 

DISTRIBUTION.  — A  local  and  uncommon 
understory  tree  of  wet  evergreen  forests  from 
southeastern  Costa  Rica  (Limon)  southeast  to  the 
Caribbean  slopes  of  Panama,  from  Bocas  del  Toro 
province  to  Cerro  Tute  (Veraguas)  and  the  region 


north  of  El  Cope  (Code)  from  90-1,200  m  (Fig. 
2). 

ADDITIONAL  SPECIMENS  EXAMINED.— COSTA  RICA.  Limon: 
Hitoy  Cerere  Biological  Reserve,  SW  of  Valle  La  Estrella  along 
Rio  Cerere  to  ca.  1  km  upstream  from  Quebrada  Barrera, 
9°40'30"N,  83°02'W,  31  Jul.  1985,  Grayum  &  Hammel  5773 
(CAS).  PANAMA.  Bocas  del  Toro:  above  Chiriqui  Grande  on 
a  side  road  ca.  10  miles  below  the  continental  divide  ca.  2'/2 
miles  E  on  that  road,  8°55'N,  82°10'W,  19  Jan.  1989,  Almeda 
et  al  6328  (CAS,  CR,  MO,  NY,  PMA).  Code:  New  Works,  7 
km  N  of  El  Cope,  18  Aug.  1977,  Folsom  4938  (CAS,  PMA); 
forest  on  continental  divide  above  El  Cope,  8°38'N,  80°38'W, 
27-29  Apr.  1985,  Hammel  13663  (CAS).  Veraguas:  Cerro  Tute, 
trail  past  agricultural  school  near  Santa  Fe,  17  Sep.  1979,  An- 
tonio 1840  (CAS). 

Specimens  of  C.  muriculata  were  first  collected 
in  1977  but  the  distinctiveness  of  this  species 
became  evident  only  after  study  of  the  flowering 
collection  designated  here  as  the  type.  The  dis- 
tinguishing features  of  C.  muriculata  are  its 
quadrate  carinate  distal  branchlets;  sparingly 
lepidote  indument  on  lower  leaf  surfaces,  pedi- 
cels, and  flower  buds;  distinctly  pedunculate  pan- 
iculiform dichasia;  and  muriculate  seed  coat. 
Collectors  have  described  the  petals  of  this  species 
as  deep  blue  (Folsom  4938),  pale  blue- violet 
(Grayum  &  Hammel  5773),  lavender  (Mc- 
Pherson  12836),  and  purple  (Antonio  1840, 
Hammel  13663).  All  of  these  collectors,  except 
Folsom  and  McPherson,  have  also  noted  that  the 
inflorescence  is  pendant  or  deflexed.  Because  these 
characters  are  unusual  in  Conostegia  and  poten- 
tially diagnostic,  additional  field  study  is  needed 
to  determine  whether  they  are  consistent. 

The  close  relatives  of  C.  muriculata  among 
described  congeners  are  not  readily  apparent.  It 
superficially  resembles  C.  tenuifolia  J.  D.  Smith 
and  C.  rhodopetala  J.  D.  Smith.  The  former,  de- 
scribed from  Costa  Rica  (Smith  1899)  but  now 
also  known  from  Nicaragua  and  Panama,  differs 
in  having  narrower  leaves  (3.6-7.4  cm)  with 
abruptly  caudate-acuminate  apices,  ovoid  buds 
that  are  conspicuously  subtruncate  at  the  base 
(when  dry),  16-22  stamens  per  flower,  an  ovary 
with  10-12  locules,  and  seeds  with  a  smooth 
testa.  In  having  oblong-ellipsoid  acuminate  flow- 
er buds  and  a  6-celled  ovary,  C.  muriculata  is 
also  similar  to  the  Costa  Rican  endemic  C.  rho- 
dopetala. In  the  field,  C.  rhodopetala  is  readily 
distinguished  from  other  species  of  Conostegia 
by  the  bright  pink  coloration  of  inflorescence 
branches,  flower  buds,  and  petals.  The  most  no- 
table characters  separating  it  from  C.  muriculata 


ALMEDA:  NEW  SPECIES  OF  CONOSTEGIA 


333 


include  its  sparse  and  deciduously  stellulate-fur- 
furaceous  indument  on  the  inflorescence  rachis, 
longer  petals  (8-10  mm),  larger  number  of  sta- 
mens per  flower  (12-17),  and  smooth  seeds. 

The  name  for  this  species  is  the  diminutive  of 
muricatus,  Latin  for  short  and  tubercular  ex- 
crescences. It  emphasizes  the  finely  sculptured 
seeds.  The  seeds  of  C.  polyandra  Benth.  are  de- 
picted as  muriculate  in  the  plate  included  with 
the  protologue  (Bentham  1 844).  However,  an  ex- 
amination of  several  recent  collections  of  C.  po- 
lyandra from  throughout  its  range  reveals  that 
its  seeds  are  completely  smooth. 

Conostegia  orbeliana  Almeda,  sp.  nov. 

(Fig.  4) 

TYPE.— PANAMA.  Bocas  del  Toro:  vicinity  of  Cerro  Col- 
orado mine  above  San  Felix,  along  trails  N  of  road  along 
continental  divide,  8°35'N,  81°50'W,  1,500  m  26  Jan.  1988, 
McPherson  12014  (holotype:  CAS!;  isotypes:  CR!,  DUKE!, 
MEXU!,  MO!,  PMA!,  TEX!,  US!). 

Arbor  6-7  m.  Ramuli  quadrangulati  demum  teretes  sicut 
foliorum  subtus  venae  primarieae  inflorescentia  hypanthiaque 
dense  pilis  stellatis  et  pilis  stipitato-stellatis  induti.  Petioli  1.2- 
3  cm  longi;  lamina  5-1 1.2  x  3-5.5  cm  elliptica  vel  elliptico- 
ovata,  apice  acuminata  vel  cuspidata,  basi  acuta,  3-5-pliner- 
vata,  membranacea  et  obscure  serrulata  vel  subintegra.  Inflo- 
rescentia 4-7(-10)  cm  longa  laxa  pauciflora;  pedicellis  (ad  an- 
thesim)  8-20  mm  longis;  bracteolis  1-2  mm  longis  mox  caducis. 
Alabastra  matura  acuta  vel  acuminata  12-15  x  9-11  mm; 
calyx  calyptriformis  6-8  mm  longus.  Petala  8,  glabra,  11-15 
x  9-13  mm,  obovata  apice  rotundato  vel  rotundato-subtrun- 
cato.  Stamina  19-22,  isomorphica  glabra;  filamenta  5-5.5  mm 
longa;  antherarum  thecae  3-3.5  x  0.75  mm  oblongae,  poro 
dorsaliter  inclinato;  connectivum  nee  prolongatum  nee  appen- 
diculatum.  Stylus  7  mm  glaber,  stigmate  subpeltato  centre 
concavo  costato;  ovarium  12-13-loculare  et  omnino  inferum, 
collo  2  mm  alto  glabro  paulo  costulato. 

Trees  6-7  m  tall.  Uppermost  internodes  quad- 
rate becoming  rounded  with  age,  moderately  to 
copiously  covered  with  a  mixture  of  sessile-stel- 
late and  short  stipitate-stellate  hairs.  Leaves  of 
a  pair  equal  to  somewhat  unequal  in  size;  petioles 
1.2-3  cm  long;  blades  membranaceous,  5-11.2 
cm  long  and  3-5.5  cm  wide,  elliptic  to  elliptic- 
ovate,  apex  acuminate  to  cuspidate,  base  acute, 
margin  inconspicuously  serrulate  to  subentire,  3- 
5-plinerved,  the  innermost  pair  of  elevated  pri- 
maries diverging  from  the  median  nerve  in  op- 
posite or  subalternate  fashion  5-9  mm  above  the 
blade  base,  glabrous  above  at  maturity,  moder- 
ately covered  with  a  mixture  of  sessile-stellate 
and  short  stalked-stellate  hairs  on  the  elevated 
primary  and  secondary  veins  below.  Inflores- 


cence a  terminal  paniculiform  dichasium  4-7 
(-10)  cm  long,  the  rachis  copiously  beset  with 
sessile  stellate  and  short  stalked  stellate  hairs; 
bracteoles  sessile,  early  deciduous,  subulate  to 
linear-oblong,  1-2  mm  long,  0.25-0.5  mm  wide 
with  pubescence  like  that  of  the  rachis  on  the 
abaxial  surface,  margin  entire.  Pedicels  8-20  mm 
long,  copiously  beset  with  sessile-stellate  and 
short,  stalked,  stellate  hairs.  Rower  buds  ellip- 
soid, 12-15  mm  long,  9-11  mm  wide,  coarsely 
verrucose  for  the  basal  half  of  their  length  and 
moderately  covered  with  tardily  deciduous  stel- 
late hairs  throughout,  the  calyptriform  calyx  6- 
8  mm  long,  acute  to  acuminate;  torus  glabrous 
adaxially.  Petals  8,  glabrous,  connivent,  white, 
obovate,  widely  spreading,  apically  rounded  to 
subtruncate  with  irregularly  lobed  undulate  mar- 
gins, 11-15  mm  long,  9-13  mm  wide  distally. 
Stamens  19-22,  typically  positioned  in  a  ring 
around  the  style;  filaments  glabrous,  complanate, 
5-5.5  mm  long;  anthers  3-3.5  mm  long,  0.75 
mm  wide,  yellow,  linear-oblong,  truncate  to 
broadly  rounded  at  the  apex  with  a  somewhat 
dorsally  inclined  terminal  pore  0.25  mm  in  di- 
ameter; anther  connective  thickened  dorsally  but 
not  prolonged  or  appendaged  at  the  filament  in- 
sertion. Ovary  (at  anthesis)  completely  inferior, 
12- 13 -celled,  depressed-globose,  the  apex  gla- 
brous, gently  fluted  or  costulate  with  an  elevated 
stylar  collar  2  mm  long.  Style  straight,  glabrous 
7  mm  long;  stigma  flared,  subpeltate  and  crater- 
iform  with  12-13  longitudinal  riblike  lobes.  Ma- 
ture berry  and  seeds  not  seen. 

PHENOLOGY. —The  two  known  collections,  both 
of  which  are  in  flower,  were  made  in  January. 

DISTRIBUTION.— A  local  and  uncommon  ele- 
ment of  cloud  forest  vegetation  on  slopes  and 
valleys  in  the  vicinity  of  Cerro  Colorado  mine 
above  San  Felix  in  western  Panama  at  1,450- 
1,500m  (Fig.  2). 


ADDITIONAL  SPECIMENS  EXAMINED.— PANAMA.  Bocas  del 
Toro/Chiriqui  Border:  slopes  and  valleys  of  Cerro  Colorado 
region,  27  Jan.  1989,  Almeda  et  al.  6445  (CAS,  MO,  NY, 
PMA). 

Judging  from  the  few  specimens  collected,  this 
species,  which  was  first  gathered  in  1988,  is  ap- 
parently rare  and  localized.  Conostegia  orbeliana 
has  a  combination  of  features  that  readily  set  it 
apart  from  its  congeners.  The  uppermost  cauline 
internodes,  pedicels,  and  inflorescence  rachis  are 
mostly  densely  beset  with  a  mixture  of  sessile 


334 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46,  No.  15 


B 


FIGURE  4.  Conostegia  orbeliana  Almeda.  A,  habit,  x  ca.  'A,  B,  representative  leaf  (abaxial  surface)  with  enlargement  (right), 
x  ca.  %;  C,  flower  bud  (left)  and  bud  with  detached  calyptriform  calyx  and  floral  parts  removed  (right),  x  ca.  1  lh;  D,  representative 
flower  showing  spreading  petals,  straight  style,  and  stigma  enlargement  (right),  x  ca.  4/5;  E,  petals  (adaxial  surface),  x  3;  F, 
stamens,  ventral  view  (left)  and  lateral  view  (right),  x  6.  (A-C  from  the  holotype;  D-F  from  Almeda  et  al.  6445.) 


stellate  and  short-stalked  stellate  hairs.  Its  in-  Other  notable  features  of  C.  orbeliana  are  the 

conspicuously  serrulate  to  subentire  leaves  are  coarsely  verrucose  hypanthial  cups  (Fig.  4C)  and 

unusually  small  for  a  species  with  comparatively  the  dorsally  inclined  anther  pore  (Fig.  4F). 

large  flowers  that  measure  ca.  3  cm  in  diameter.  Overlapping  similarities  in  foliar  pubescence, 


ALMEDA:  NEW  SPECIES  OF  CONOSTEGIA 


335 


petal  and  stamen  numbers,  style  posture,  stigma 
morphology,  and  the  number  of  ovary  cells  sug- 
gest that  C.  orbeliana  is  closely  related  to  C.  vol- 
canalis  Standley  &  Steyerm.  of  southern  Mexico 
(Chiapas,  Guerrero,  Jalisco),  Guatemala,  Hon- 
duras, and  Nicaragua.  I  define  C.  volcanalis  in  a 
broad  sense,  perhaps  too  broadly,  to  accom- 
modate what  I  interpret  as  regional  variation  in 
leaf  and  petal  size,  bud  shape,  and  foliar  pubes- 
cence density.  Some  populations,  for  example, 
have  leaves  that  are  nearly  glabrous  below.  Oth- 
ers have  lower  leaf  surfaces  beset  with  a  mixture 
of  sessile  stellate  and  long-stalked  stellate  hairs, 
whereas  others  have  only  long-stalked  stellate 
hairs  mostly  confined  to  the  elevated  primary 
and  secondary  veins.  Even  in  its  broadest  cir- 
cumscription, C.  volcanalis  is  readily  separated 
from  C.  orbeliana  by  its  larger  (12-27  x  6-15 
cm)  mostly  undulate-dentate  leaves,  longer  pet- 
ioles (3.5-6.7  cm),  smaller  floral  buds  (8-10  mm), 
and  completely  smooth  pedicels  and  hypanthial 
cups. 

This  species  is  named  for  Orbelia  R.  Robin- 
son, friend  and  fellow  student  of  the  Melasto- 
mataceae,  in  grateful  recognition  of  all  the  vol- 
unteer technical  and  research  assistance  she  has 
rendered  in  the  herbarium  of  the  California 
Academy  of  Sciences  during  the  past  decade. 

ACKNOWLEDGMENTS 

This  study  was  supported,  in  part,  by  U.S.  Na- 
tional Science  Foundation  Grant  BSR  8614880 
(Flora  Mesoamericana)  and  the  Nathan  Jay  and 
Virginia  Friedman  Fund  of  the  California  Acad- 
emy of  Sciences.  For  special  technical  and/or  field 
assistance  during  the  course  of  this  study  I  thank 
Thomas  F.  Daniel,  Gerrit  Davidse,  Greg  de  Nev- 
ers,  Gordon  McPherson,  and  Orbelia  R.  Rob- 
inson. My  special  thanks  go  to  the  Missouri  Bo- 
tanical Garden  and  the  Smithsonian  Tropical 
Research  Institute  for  logistical  support  in  the 


field;  Ellen  del  Valle  for  the  diagnostic  illustra- 
tions; Gina  Umana  Dodero  for  reviewing  the 
Spanish  summary;  and  the  curators  and  staffs  of 
the  following  herbaria  (acronyms  fide  Holmgren 
et  al.  1 98 1 )  for  loans,  gifts,  and/or  assistance  dur- 
ing study  visits:  CR,  DUKE,  F,  MO,  PMA,  US. 

RESUMEN 

Conostegia  es  un  genero  natural  de  amplia  dis- 
tribution pero  con  gran  concentration  en  Costa 
Rica  y  Panama,  33  de  sus  50  especies  occurren 
alii.  La  relativa  inaccesibilidad  a  ciertas  areas 
tales  como  las  selvas  de  la  parte  sur  de  Costa 
Rica  (Limon)  y  de  las  provincias  de  Bocas  del 
Toro  y  Chiriqui  en  Panama,  se  reflejan  en  el 
numero  de  especies  recolectados  hasta  el  presen- 
te.  El  numero  de  los  petalos  y  estambres,  la  po- 
sition del  estilo,  la  forma  y  tamano  del  estigma, 
y  el  numero  de  loculos  en  el  ovario  son  a  menudo 
caracteres  taxonomicos  muy  utiles  en  Conoste- 
gia. Se  proveen  descripciones,  illustraciones  y 
discusiones  referentes  a  las  afinidades  de  tres  es- 
pecies nuevas:  Conostegia  fragrantissima  y  C. 
orbeliana  de  Panama  y  C.  muriculata  de  Costa 
Rica  y  Panama.  Tambien  se  presenta  un  mapa 
de  distribution  para  todas  ellas. 

LITERATURE  CITED 

BENTHAM,  G.  1844.  The  botany  of  the  voyage  of  H.M.S. 
Sulphur,  under  the  command  of  captain  Sir  Edward  Belcher 
. . .  during  the  years  1836-1842.  London. 

GLEASON,  H.  A.  1958.  Melastomataceae.  In  Flora  of  Panama. 
R.  E.  Woodson,  Jr.  and  R.  W.  Schery,  eds.  Ann.  Missouri 
Bot.  Gard.  45:203-304. 

HOLMGREN,  P.  K.,  W.  KEUKEN,  AND  E.  K.  SCHOFIELD.  1981. 
Index  Herbariorum.  Part  1, 7th  ed.  The  herbaria  of  the  world. 
Regnum  Veg.  106:1^52. 

HOWARD,  R.  A.  1 989.  Melastomataceae.  In  Flora  of  the  Less- 
er Antilles.  5(2):532-579.  Arnold  Arboretum,  Harvard  Uni- 
versity, Jamaica  Plain,  Massachusetts. 

SMITH,  J.  D.  1 899.  Undescribed  plants  from  Guatemala  and 
other  Central  American  republics  XXI.  Bot.  Gaz.  27:331- 
339. 


CALIFORNIA  ACADEMY  OF  SCIENCES 

Golden  Gate  Park 
San  Francisco,  California  94118 


INDEX  TO  VOLUME  46 

(Compiled  by  Tomio  Iwamoto) 
New  names 


Blakea  fuchsioides 
Blakea  gregii 
Blakea  hammelii 
Blakea  herrerae 
Blakea  scarlatina 
Calicina  arida 
Calicina  basalta 
Calicina  conifera 
Calicina  diminua 
Calicina  dimorphica 
Calicina  galena 
Calicina  mesaensis 
Chiloglanis  harbinger 
Chiloglanis  niger 
Chiloglanis  polypogon 
Chiloglanis  reticulatus 
Chiloglanis  sanagaensis 
Clidemia  hammelii 
Conostegia  fragrantissima 
Conostegia  muriculata 
Conostegia  orbeliana 
Drapetis  arnaudi 
Drapetis  cerina 
Drapetis  destitute 
Drapetis  Solaris 
Drapetis  torulosa 
Dyschoriste  mcvaughii 
Hologr aphis  caput-medusae 


Holographis  leticiana 
Holographis  tolantongensis 
Justicia  fortunensis 
Justicia  ixtlania 
Justicia  readii 
Justicia  tabascina 
Justicia  veraguensis 
Miconia  calocoma 
Miconia  dissitiflora 
Miconia  grayumii 
Miconia  ibarrae 
Miconia  iteophylla 
Miconia  ligulata 
Miconia  peltata 
Neostethus  robertsi 
Pachycara  crossacanthum 
Pachycara  gymninium 
Pachycara  lepinium 
Pachycara  mesoporum 
Pachycara  pammelas 
Pachycara  rimae 
Pachycara  shcherbachevi 
Pachycara  sulaki 
Tococa  croatii 
Topobea  caliginosa 
Topobea  fragrantissima 
Topobea  hexandra 
Topobea  suaveolens 


New  names  in  boldface  type 


Abies  concolor  10 

Acacia  greggi  1 2 

Acanthaceae  73,  79,  81,  279,  289-290,  297 

Acanthodactylus  sp.  197 

Acanthostethus  274 

falcifer  274 
Adenostoma 

fasciculatum  12 

sparifolium  12 
Aegotheles  bennettii  196 
Aegothelidae  196 
Agama  agama  194 
Agamidae  195,  197 
Agave 

aurea  12 

shawii  12 

sp.  27,  29 

spp.  12-13 


Amaraboya  300 
Amblyarrhena  146,  148 
Aneides  47 

lugubris  48,  58 
Anguidae  3 1 
Anniella 

geronimensis  31-32,  45-46,  48,  59 

pulchra  32,  44,  48 
Anniellidae  31 
Anoplopoma  fimbria  222 
Anthurium  314 
Antimora  93 

microlepis  93 

rostrata  93 

Aphelandra  scabra  79 
Aplonis  sp.  196 
Araceae  150,  314 
Archoleptoneta  131 


[337] 


338 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46 


Arctostaphylos 

glauca  12 

patula  12 

pringlei  12 

pungens  12 
Arizona 

elegans  45,  49 

elegans  eburnata  33 

elegans  occidentalis  33 

elegans  pacata  33 
Artemisia  californica  1 2 
Assam  iidae  123 
Atherinidae  247-248 
Atherinomorpha  243,  247-248,  254 
Atopochilus  152,  163-164,  176 

savor gnani  153,  164 

sp.  undet.  a  153,  163 

sp.  undet.  b  153,  163 
Austrolycichthys  221,  223 

brachycephalus  240 

Sagarins 

bagarius  163 

yarrelli  163 

Banksula96,  107,  122,  131 
Batrachoseps  16,  60-61,  95,  129-130 

attenuatus  129 

pacificus  16,  59 

pacificus  major  15-16,  44,  48 

pacificus  spp.  15-16,  44,  48 
Bedotia24&,  261 

sp.  245 
Bellucia  300 

grossularioides  300 
Beloperone  284 
Betula  glandulosa  188 
Biantidae  123 
fl/pes  60 

biporus  48-49,  52,61 
Blabomma  131 

137,  140,  299-311,  323,  325 

anomala  304 

austin-smithii  302-303,  307-308 

chlorantha  307-309 

crassifolia299,  318 

cuatrecasii  3 1 7 

c  lineal  a  317 

elliptica  303 

fuchsioides  137-140 

grandiflora  3 1 7 

gregii  299,  301,  305-309,  325 

hammelii  299,  309-31 1,  313,  325 

herrerae299,  311-314,  325 

hirsuta  var.  rotundata  304 

litoralis  304 

micrantha299,  317 

parvifolia  299,  322-323 

pauciflora  311,  313 


penduliflora  307-308 
purpusii  302,  308 
quinquenervia  300 

scarlatina  299,  308,  314-315,  317,  325 
trinervia  300 
tuberculata  303,  317 
wilburiana  301-302 
woodsonii  303 
Blakeeae  299,  301,  303 

Boidae  32 

193-194,  198-203 

angulata  194,  198,  201 

blandingi  194,  197-201,  203 

ceylonensis  194,  198,  200-201 
195, 200-201,  203 
n  195,  198-201,  203 

dendrophila  193,  195,  198-200,  202-203 

dightoni  195,  201 

drapiezii  195,  201,  203 

w  196,  200-201,  203 
196,  201 

hexagonatus  197 

irregularis  193-194,  196,  198-203 

jaspidea  197,  201,  203 

kraepelini  197-198,  201,  203 

multifasciata  197 

multomaculata  197,  200-201 

nigriceps  197,  201,  203 

nuchalis  195 

ocellata  197,  201,  203 

ochracea  197-198,  201 

pulverulenta  197-198,  201,  203 

schultzei  197-198,  203 

trigonata  197-199,  201,  203 
Brotulinae  226 
Bufo  19 

alvarius  63 

boreas  19,  38,  44,  47^18,  58 

boreas  halophilus  1 7 

cognatus  63 

microscaphus  19,  44,  48 

microscaphus  calif ornicus  18 

punctatus  20,  45,  48 

woodhousei  63 
Bufonidae  17 
Buparellus  122 
Bupares  122 


jamaicensis  27 

Caddidae  134 

Caddo  pepper ella  134 

Calicina  95-102,  112,  117,  121-123,  126,  129-135 
arida95,  100,  111-112,  124-126,  130-133 
antfa  subgroup  99-100,  111-112,  122,  124,  127, 

129,  132 
basalta95,  116,  119,  126,  128-129,  132-133 


INDEX 


339 


breva95,  116,  119,  126,  128-129,  132-133 
cloughensis  95,  111-112,  124-126,  130-134 
conifera95,  105,  126,  132-133 
digita  95,  106,  122,  125-126,  130-134 
digita  group  99,  106,  113,  122-125,  129,  132,  134 
digita  subgroup  99,  106-107,  122-123,  127,  129, 

132 

diminua95,  99,  120-121,  125-126,  130-133 
dimorphica  95,  116-117,  119,  126,  128,  132-133 
ensata95, 101, 114,  117, 125-126, 128-129, 132- 

133 

galena  95,  109,  126,  130-134 
kaweahensis  95,  108,  124-127,  130,  132-135 
kaweahensis  group  1 24 
kaweahensis  subgroup  99,  107-108,  122,  132 
keenea95,  110,  126,  130,  132-133 
macula  95,  114,  117,  119,  126,  128,  132-133 
mariposa  95,  99-102,  105-106,  108-111,  122- 

126, 132-134 
mariposa  group  99- 100,  105,  113,  122-123,  125, 

129-132 

mariposa  subgroup  99,  105,  124,  127,  132 
mesaensis95,  101,  116,  119,  126,  128,  132-133 
minor95,  101,  113-114,  116,  125-127, 130, 132- 

133 

minor  subgroup  105,  113,  122,  124,  132 
morroensis95,  110,  126,  129-130,  132-133,  135 
palapraeputia  95,  99,  102,  112-113,  122,  124- 

127,  130,  132-135 

palapraeputia  group  100,  112,  123,  132 
palapraeputia  subgroup  100,  112,  124,  132 
piedra95,  114,  116,  125-126,  128-129,  132-133, 

135 
polina95,  116-117,  125-126,  128-129,  132-133, 

135 

sequoia  95,  120-121,  125-126,  130-133,  135 
sequoia  subgroup  105,  120-122,  124-125,  127, 

132 

serpentinea95,  102,  114,  116,  122,  126,  128-134 
serpentinea  group  99,  100,  113,  116-117,  122- 

123,  125,  129-130,  132,  134 
serpentinea  subgroup  105,  114,  119,  122,  124- 

125,  127-130,  132 
sierra95,  106-107,  126,  130-134 
topanga95,  102,  109-110,  126-127,  129-134 
topanga  subgroup  100,  109-110,  124,  132 
yosemitensis  95,  105,  126,  132-133 
Calicina  species  group  99 
Callisaurus  52 

draconoides  23,  45^6,  48^9,  62-63 
Calloselasma  rhodostoma  195 
Calotes  199 

cristatelus  195 
sp.  194-195,  197 
versicolor  195,  197 
Calymmaria  131 
Carlowrightia  arizonica  19 
Cassia  foment osa  192 


Caudata  15 
Ceanothus 

cordulatus  12 

greggi  12 

leucodermis  12 
Ceratostethus  244,  269 

bicornis  245,  269 
Cericidium  microphyllum  12 
Chaetochlamys  284 
Chaetothylax  284 

umbrosus  284 
Charasia  dorsal  is  195 
Chilmeniscus  complex  62 
Chiloglanis  151-166,  168-169,  175-176 

batesii  151-163,  165,  168-169,  171-172 

benuensis  151,  153,  157,  164,  167-168 

brevibarbis  152,  155,  163-164 

cameronensis  151,  153,  157,  162-163,  165,  168, 
176 

congicus  152,  156-157,  163 

deckenii  152,  163-164 

disneyi  151-159,  161-163,  165,  168,  172-173 

harbinger  151-153,  155,  157,  161-163,  165,  175 

lamottei  157 

macropterus  162 

marlieri  157 

micropogon  151-152,  155,  162,  169,  172 

neumanni  152 

niger  151-163,  165,  173-174 

niloticus  153,  162 

normani  162 

occidentalis  153,  160,  162-163,  172 

polyodon  153,  157,  163 

polypogon  151-164,  167-168 

reticulatus  151-160,  162-166,  168 

sanagaensis  151-163,  165,  169-170 

sardinhai  163 

savorgnani  163 

voltae  151,  153-155,  161,  163-166,  168 
Chilomeniscus  52 
Chilomeniscus 

cinctus  33-34,  45-46,  48-49,  63 

punctissimus  48 

savagei  48 

stramineus  48 
Chilomeniscus  complex  48 
Chilopsis  linearis  12,14 
Chionactis 

occipitalis  44,  48,  63 

occipitalis  annulata  34 
Chiropodomys  gliroides  197 
Chrysemys  scripta  48-49,  6 1 
Chrysopelea  ornata  195 
Clemmys  marmorata  22,  44,  47-48,  58 
Clidemia  137,  327 

costaricensis  142 

hammelii  137,  140-143,311 

reitziana  142 


340 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46 


Clusiaceae  143 
Cnemidophorus  52 

ceralbensis  48,  60 

deppei  complex  48 

hyperythrus  30-31,  45-46,  48,  60,  62 

hyperythrus  beldingi  30 

hyperythrus  hyperythrus  30 

hyperythrus  schmidti  30 

labialis  30,  45^6,  48,  59,  62 

multiscutatus  3  1 

tigris  45 

tigris  multiscutatus  30 

tigris  tigris  30 
Cobitis  taenia  158 
Coleonyx  52 

switaki  23,  48,  62 

variegatus  45 

variegatus  abbotti  23 

variegatus  peninsularis  23 

variegatus  variegatus  23 
Coluber 

"lateralis  complex"  58 

aurigulus  48-49 

barbouri  48-49 

flagellum  34,  45,  49 

flagellum  fuliginosus  34 

flagellum  piceus  34 

lateralis  44,  48^49 

later  alis  complex  48 

lateralis  lateralis  34 
Colubridae  33,  193,  195,  198 
Conostegia  327,  330,  332,  335 

cinnamomea  148 

fragrantissima  327-330,  335 

montana  330 

muriculata  327,  329-332,  335 

orbeliana  327,  329,  333-335 

pittieri  330 

polyandra  333 

rhodopetala  332 

tenuifolia  332 

volcanalis  335 
Coryphaenoides  armatus  93 
Coturnix  chinensis  196 
Crocidura  sp.  1  96 
Crocodylus  6  1 
Crossopalpus  179-180,  188 
Crossostomus  chilensis  238 
Crotalus  52 


insularis  46 

mitchellii  45-46,  48-49,  62-63 

mitchellii  mitchellii  39 

mitchellii  pyrrhus  39 

ndjCT-  45-^6,  48,  62 


afrax  afro*  38 
flfrax  complex  48 
catalinensis  48,  52 
cerastes  44,  48,  63 
cerastes  laterorepens  38 
39,  45-46,  48,  62 

39 
enyofurvus  39 


tortugensis  48 

vinrfts  44,  48-49 

v/rafts  Ae//eri  39-40 
Crotaphopeltis  200 
Crotaphytus  52 

insularis  45,  48,  62 

insularis  vestigium  24 
Ctenophallus  269,  272 

ctenophorus  245,  272 
Ctenosaura  hemilopha  48^9,  52,  60-61 
Cyphomandra  314 
Cyprinodontidae  247-248,  266 


Dalechampia  3 1 4 
Dentatherina  248,  261 

merceri  244-245,  256,  261-263 
Diadophis 

punctatus  44,  48 

punctatus  similis  34 
Dicliptera  resupinata  79 
Dipsadoboa  200 
Dipsosaurus  52 

rforsafo  45,  49,  62-63 

dorsalis  dorsalis  24 
Diptera  179 
Draco  sp.  1 94 
Drapetis  179,  188 

(Crossopalpus)  discalis  \  84 

(Eudrapetis)  discalis  1 84 

arcuata  179 

arnaudi  179,  181-183 

assimilis  191 

assimilis  group  179-181 

cerina  179,  181-184 

destituta  179,  181-182,  184-185 

</woi/w  179-182,  184-186,  188 

divergens  180 

rfwVfaa  180 

exi'/w  180 

infumata  179,  181-182,  185-186,  188,  190 

ingrica  179 

latipennis  179-182,  188-190 

Ha/oz  188 

na/ca  group  1 80-1 8 1 

simulans  179 

Solaris  179,  181-182,  190-191 

torulosa  179-182,  191-192 
Dudleyaspp.  12-13 
Dyschoriste  279,  282 

mcvaughii  279-283 


INDEX 


341 


pringlei  282 
rubiginosa  282-283 

Echinocactus  sp.  29 
Elaphe  rosaliae  34-35,  48,  62 
Elaphropeza  179-180 
Elapidae  198 
ElgariaSS,  61-62 

cedrosensis  48-49 

kingi  48,  62 

multicarinata  44,  48-49 

multicarinata  webbi  3  1 

panamintina  48,  62 

paucicarinata  48-49,  62 
Elgaria  complex  48 
Elytraria  imbricata  79 
Emoia  cyanura  196 
Empididae  179 
Emydidae  22 

Ensatina  eschscholtzii  47-48,  58 
Eonycteris  spelea  195,  199 
Ephedra  californica  1  2 
Eptesicus  tenuipennis  1  94 
Eridiphas  60 

sleveni  52 

s/ev/m'48^9,  61 
Eriogonum  fasciculatum  12 
Euchilichthys  152,  163-164,  176 

dybowskii  153,  163-164 

royauxi  153,  163 
Eudrapetis  179-180,  188 

discalis  184 
Eumeces  58,  61 

#/&?/*/  44,  48 

gilberti  rubricaudatus  29 

lagunensis  48-49 

skiltonianus  29-30,  44,  48-49 

skiltonianus  complex  48 

skiltonianus  skiltonianus  29 
Euphorbiaceae  314 

Fa/co  sparverius  30 


mutilata  196-197 
sp.  196 


acanthodes  12 
sp.  22 
Fourquiera  splendens  1  2 


chenopodiafolia  12 
dumosa  12 
Fuchsia  139 

Gallusgallus  195 

Gamasida  143 

Gambelia 

wislizenii  45 
wizlizenii  49 
wizlizenii  copei  24 
wizlizenii  wizlizenii  24 


195 
vittatus  196 
Gekkonidae  23,  195 
Geochelone  6  1 
Gesneriaceae  314 
Gloxinia  314 
Gochnatia  hypoleuca  75 


modestus  196 
sp.  195 

Gopherus  agassizii  63 
Graffenrieda  gracilis  216 
Gulaphallinae  244,  254,  266,  273 
Gulaphallini  244,  254,  256-257,  264,  273-275 
Gulaphallus  243-245,  249,  253-254,  256-258,  261- 

263,  266,  268,  273,  275-276 
amaricola  270 
bikolanus  244-247,  261-263,  265-266,  268,  274- 

275 
ex/mms  244-247,  257,  262-263,  265-266,  268, 

273-276 

/afa/er  244-247,  262-263,  265-266,  274,  276 
mirabilis  244-247,  249,  253,  262-263,  265-266, 

274 

panayensis  244-247,  256,  261-266,  274-275 
Gymnelus  93 


Henrya  insularis  79 
Heteromeles  arbutifolia  12 
Holographs  73,  76-77,  81 

anisophylla  73,  75 

argyrea  73,  79 

caput-medusae  73-74,  76-79 

ehrenbergiana  74 

hintonii  74,  79 

ilidfolia  74 

Ieticiana73,  75,  77,  80-81 

/?a///Vfo  74-75 

par  ay  ana  73-75,  81 

peloria  73-74 

pueblensis  73-74 

tamaulipica  74 

tolantongensis  73-77 

virgata  74 

websteri  73-74 
Hoplocephalus  202 
Hygrophila  282 
58,61 

arenicola  49 

cadaverina  20,  22,  38,  44,  48^9 

exima  49 

microscaphus  49 

regilla22,  44,  48-^9,  61 


342 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46 


regilla  deserticola  22 

regilla  hypochondriaca  21-22 
Hylidae  20 
Hypoatherina  26 1 

bleekeri  254,  257 

ovolaua  246 
Hypsiglena  52 

torquata  35,  45,  48 

torquata  deserticola  35 

torquata  klauberi  35 

Iguanidae  23 
Imantodes  202 

cenchoa  202 
Ischyropsalidae  134 
Ixtlania  279,  284 

acicularis  283 

Japalura  variegata  197 
Juniperus  75 

calif ornica  12 
Jwtf/da  279,  284-285,  289-297 

acicularis  283 

adhatoda  296 

caudata  285 

fortunensis  289-293,  295-297 

ixtlania  279,  281,  283-284 

lindenii  285 

readii  289-293,  295-297 

refractifolia  289-290,  297 

tabascina  279,  281,  283-285 

veraguensis  289-290,  292-297 

sect.  Drejerella  285 

sect.  Leucoloma  284 

sect.  Tyloglossa  296 

Laceraria  146,  148 

Lacertidae  197 

Laelapidae  143 

Lamprolepis  smaragdinum  195-196 

Lampropeltis  52 

catalinensis  48 

g#M/MS  35,  45,  48 

getulus  californiae  35 

getulus  complex  48 

getulus  yumensis  35 

w'nVfa  48 

pyromelana  49 

zomtfa  35,  44,  48-49,  59 

zonata  agalma  35 
Laniatores  95 
Larrea  tridentata  12 
Leandra  216 

consimilis209,  216 
Leiocassis  162-163 
Leptotyphlopidae  32 
Leptotyphlops  52 

,  45,  48,  52 


humilis  cahuilae  32 

humilis  humilis  32 
Libocedrus  decurrens  1 2 
Lichanura  52 

roseofusca  32 

trivirgata  32,  45^6,  48-49,  62-63 

trivirgata  gracia  32-33 

trivirgata  roseofusca  32-33 

trivirgata  trivirgata  32-33 
Lycenchelys  86 

"D"  91,  235 

"E"  88,  234 

crassiceps  222,  228 

monstrosa  238 
Lycodapus  mandibularis  93 
Lj*ofes  82,  221,228 

atlanticus  228 

brachycephalus  82,  222-223,  240 

crassiceps  228 

diapterus  93 

macrops  228 

mucosus  228 

sp.  88,  91,  228,  234-235 
Lycodryas  202 

Manacopus  244,  254,  274 

./&/«/«•  245,  274 

MayAira  bulbiceps  82,  84,  223,  227 
Melanostigma  226 
Melanotaenia  248,  261 

qffinis  246 
Melastomataceae  137,  209,  219,  299-301,  311,  314, 

327 

Mexacanthus  284 
A//a»iifl  137,  150,  209,  214,  216,  327 

arbor icola  139 

Irenes/7  2 11-2 12 

calocoma  137,  144-146,  148 

centrodesma  148 

consimilis  2 1 6 

dissitiflora  137,  146-148 

fulvostellata  214 

grayumii  209-2 12 

ibarrae209,  212-214 

iteophylla  209,  214-217,  219 

Hgulata209,  212,  216-217 

peltata209,  216-219 

rupticalyx  146 

wagneri  146 

Miconieae  150,  216,  327 
Micrurus  fulvius  198 
Mirophallus  274 

bikolanus  245,  274-275 
Mochokidae  151,  164 
Molossidae  194 
Monochaetum  148 
Motacillasp.  195,  197 
Motacillidae  195,  197 


INDEX 


343 


Mugiloidea  247 
Muridae  195,  197 
Mus  musculus  196 
Muscicapidae  197 
Mystus  163 

Nectarina  famosa  194 

Nemastomatidae  134 

Neostethidae  243 

Neostethinae  254,  266 

Neostethini  243,  254,  256-257,  264,  266,  269,  275 

Neostethus  243,  249,  253-254,  256-257,  259,  261- 

262,  266-267,  269-270,  272,  275 
(Neostethus)  254 
(Neostethus)  amaricola  245,  270 
(Neostethus)  lankesteri  245,  269 
(Neostethus)  siamensis  245,  269 
(Neostethus)  villadolidi  245,  271 
(Sandakanus)  254,  270 
(Sandakanus)  borneensis  245,  270 
(Sandakanus)  coronensis  245,  270 
(Sandakanus)  panayensis  245,  275 
(Sandakanus)  zamboangae  245,  271 
amaricola  243, 245-247, 258-26 1 , 265-266, 269- 

271 

bicornis  243-247,  258-261,  265-266,  269,  273 
borneensis  244-245, 247, 254, 257-259, 26 1 , 263- 

266, 269-270,  275-276 
coronensis  254,  271,  275 
ctenophorus  244-247,  258,  262,  264,  266,  269, 

272 
lankesteri  243,  245-247,  252,  258-261,  264-266, 

268-269,  275-276 
palawanensis  244-247,  258-259,  262,  265-266, 

269,  271-272 
panayensis  254,  275-276 
robertsi  243-247,  258-259,  262,  265-266,  269, 

272-273,  275 
siamensis  269,  275 
thessa  244-247,  257-266,  269,  272 
villadolidi  244-247,  258-261,  265-266,  269-271, 

276 
zamboangae  244-247,  250,  254,  257-259,  261, 

265-266,269,271 
Neozoarcinae  226 
Nerodia  60 

valida  48^9,  52 
\allida  61 

Odontonema  auriculatum  79 
Odontoneminae  284 
Olneyatesota  12,  28 
Ololaelaps  137 

sp.  143 

Onagraceae  139 
Opiliones95,  134 
Opuntia 

littoralis  12 


prolifera  12 

sp.  27 

spp.  12 

Orchestina  131 
Orchidaceae  314 
Oryzomys  307 

Pachycara  82, 84-89, 91, 221, 223, 226-227, 231, 233, 
235,  238-240 

brachycephalum  222,  227,  240-241 

bulbiceps  84-87,  92-93,  221-222,  224-225,  227- 
228,  241 

crassiceps  222-223,  227-230 

crossacanthum  221,  224,  226-227,  230 

gymninium  82,  84,  86-89,  92-93,  222,  225,  227, 
234-235 

lepinium  82,  84,  86-87,  90-91,  93,  222,  227,  235 

mesoporum  221,  224-226,  235,  237,  241 

obesa  84,  222-223,  227-228 

obesum  82,  84,  86 

pammelas  221,  227,  235,  239 

rimae  221,  223,  225-226,  233,  235-236,  241 

shcherbachevi  221-222,  226,  233-234,  241 

sp.  85 

sulaki  221,  225,  227,  231,  238,  241 

suspectum  84,  86,  88-89,  222,  226,  235 

suspectus  87 
Pachycarichthys  223 
Pachycereus  pringlei  12,  22 
Pachychara  223 

obesa  84 

Pachycormus  discolor  1 2 
Pangasius  163 
Parabeloniscus  122 
Parabrotulidae  226 
Parabuteo  unicinctus  21 
Passer  sp.  195 
Passeridae  195 
Pelobatidae  16 
Pentanychidae  134 
Peramelidae  196 
Percesoces  247 
Peromyscus  307 

sp.  34 
Petrosaurus  24,  52 

mearnsi  24-25,  45^6,  48 

repens  48 

thalassinus  48 
Petrosaurus  complex  48,  62 
Phalangodidae  95-96,  98 
Phalangodinae  98,  123 

Phallostethidae  243,  246-248,  254,  264-266,  275 
Phallostethinae  243,  254,  256,  264-267 
Phallostethini  243,  247,  254,  256-257,  264-265,  267, 

275 

Phallostethus  243-244,  250,  254,  256-257,  262,  264- 
267,  275 

dunckeri  243,  245-247,  257,  262,  265,  267,  271 


344 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46 


Phasianidae  195-196 
Phenacostethini  264 

Phenacostethus243,  249,  254,  256-257,  262-264,  267, 
275 

position  243-247,  257,  262-264,  267-268 

smithi  243,  245-247,  249,  251,  257,  262,  264, 
267-268 

thai  268 

trewavasae  243-247,  257,  262,  264,  267-268 
Philautus  sp.  195 
Phrynosoma  52 

cerroense  48 

coronatum  25,  46,  48,  60,  62 

coronatum  complex  48 

coronatum  schmidti  25 

coronatus  45 

mcallii  25,  48^9 

platyrhinos  25,  44,  48 

platyrhinos  calidiarum  25 
Phucocoetes  suspectus  82,  88,  235 
Phyllodactylus  52,  60 

angelensis  48 

apricus  48 

bugastrolepis  48 

homolepidurus  nolascoensis  48 

nocticolus  23,  45-46,  48 

nocticolus  nocticolus  23 

nocticolus  sloani  23 

partidus  48 

santacruzensis  48 


tt'  48 

Phyllodactylus  complex  62 

Phyllorhynchus  52 

decurtatus  45,  48^19,  62-63 
decurtatus  decurtatus  35 
decurtatus  perkinsis  35 

Physocarpus  188 

Pilocosta  2  19-220 

campanensis  2  1  9-220 


2  19-220 
oerstedii  219 

oerstedii  campanensis  209,  219 
Pwzws  131 

P//2WS 

Jeffrey  i  12 
lambertiana  12 
monophylla  12 
murrayana  10 
pinceana  75 
quadrifolia  12 
Pituophis 

melanoleucus  45,  49 
melanoleucus  affinis  36 
melanoleucus  annectens  36 
melanoleucus  bimaris  36 


Platanus  190 

racemosa  12 
Plectrostethus  257,  269,  272 

palawanensis  245,  269,  271,  275 

palawensis  27  1 
Plesiositagra  cucullatus  194 
Plethodontidae  15 
Polynemoidea  247 
Popw/us 

balsamifera  188 

fremontia  12 

sp.  35 

tremuloides  10 
Praomys  sp.  197 
Prosopis  juliflora  12,  14 
Psammodynastes  pulverulentus  195 
Pseudemys  script  a  52 
Pseudomugil  26  1 

signifer246,256,26\,  263 
Pseustes  poecilonotus  202 
Pteropodidae  195 
Ptychosoma  122 
Pyxidanthus  300 


dumosa  12 
opaca  75 
spp.  12 
turbinella  12 

22 

awrora  22,  44,  47,  58 
aurora  draytoni  22 
boylii  22,  44,  48 
limnocharis  194-195 
sp.  196 
Ranidae  195-196 


exulans  196 

rattus  196 

sp.  195 

Reithrodontomys  307 
Rhacophoridae  195 
Rhamnus  californica  1  2 
Rhinocheilus 

lecontei  45 

lecontei  lecontei  36 

leconti  49 
.R/zws  ovata  1  2 
7?osa  californica  1  2 
RuelliaZl,  282 

albiflora  79 

petiolaris  28  1 


horsfieldi  195 
Salientia  16 
Sa&cspp.  12,  14 


INDEX 


345 


Salvadora 

hexalepis  36,  45,  49 
hexalepis  hexalepis  36-37 
hexalepis  klauberi  36-37 
hexalepis  virgultea  37 

Salvia 

apiana  12 
pachyphylla  12 

Sandakanus  269 


angustus  48,  52 

grandaevus  48 

spp.  49 

Sa/or  complex  48,  61 
Sauna  23 
Sauromalus  52 

ater  48 

australis  25,  48 

hispidus  48 

klauberi  48 

o&esttf  25-26,  45^6,  48,  63 

obesus  obesus  25-26 

slevini  48 

var/ws  48 

Sauromalus  complex  48,  62 
Saxicoloides  fulica  197 
Scaphiella  131 
Scaphiopus  47 

couchii  16,  48 

hammondii  16,  44,  48 
Sceloporus  52 

graciosus  26-27,  44,  48 

graciosus  vandenburghianus  26 

hunsakeri  48 

/«#/  48 

magister  26-27,  45 

magister  complex  48,  57,  62 

magister  lineatulus  48 

magister  monserratensis  48 

magister  rufidorsum  26-27,  48 

magister  uniformis  26-27,  49 

magister  zosteromus  48 

occidentalis  26-27,  44,  48 

occidentalis  biseriatus  27 

orcMtt/  27,  45,  48 

orcutti  complex  48,  62 
Schaphiopus  couchii  45 
Schaueria  286-287 

calycobractea  279,  286-287 

calycotricha  287 

cordatus  286 

parviflora  279,  281,  283,  286-287 
Scincidae  29 
Scotinomys  307 
Scotolemon  122 

balearicus  122 

espanoli  122 
Serpentes  32 


Stem*  106 

Siphonoglossa  8 1 

Sitalces  californica  95-96 

Sitalcina  95-98,  106,  122,  131,  134-135 

borregoensis  97 

californica  97-98 

cloughensis  1 1 2 

cockerelli  121,  131,  135 

digitus  106 

ensata  1 1 7 

kaweahensis  108 

keenea  1 10 

macula  1 1 7 

madera  97 

mariposa  98,  105 

minor  1 1 3 

morroensis  1 10 

palapraeputia  112 

p/«/ra  1 1 6 

polina  1 1 6 

scapula  97 

sequoia  121 

serpentinea  1 1 6 
106 
6reva  1 1 9 

swra  97 

topanga  109 

yosemitensis  105 
Solanaceae  314 
Solenophallus  269,  272 

ctenophorus  269,  272,  275 

fAassa  245,  269,  272,  275 
Sonora  52 

bancroftae  37 

mosaueri  37 

semiannulata  37 

semiannulata  bancroftae  45-46,  48 

semiannulata  mosaueri  48 

semiannulata  semiannulata  45,  49 
Spathiphyllum  314 
Stenandrium  73 
Streblacanthus  286 

parviflorus  279,  286 
Sturnidae  196 
Synodontis  155,  164 

Tabascina  279,  285 

lindenii  284 
Tachydromiinae  179 
Tachydromus  sexlineatus  195 
Tadarida  sp.  194 
Tantilla  52 

planiceps  37,  45,  48 
Teiidae  30 
Teleostei  82,  221 
Telescopus  200-202 

rf/zara  200 

semiannulatus  200 


346 


PROCEEDINGS  OF  THE  CALIFORNIA  ACADEMY  OF  SCIENCES,  Vol.  46 


Testudinata  22 
Tetramerium 

nervosum  79 

oaxacanum  8  1 

tenuissimum  79 
Texella  122 
Thamnophis 

digueti  48-49 

elegans  37-38,  44,  48,  198 

elegans  hueyi  37 

hammondii2\,  38,  44,  48-49 

hammondii  complex  48 

marcianus  63 

sirtalis  198 

spp.  37 
Thermarces  221,  226,  240 

cerberus  226 
Titiotus  131 

137,  150 

croatii  137,  148-150 

symphyandra  150 

ra  140,  299-305,  314,  317-318,  320,  323,  325 

acuminata  320 

albertieae  304 

brenesii  302 

caliginosa  299,  304,  317,  322 

calycularis  303 

caudata  320 

co«tofa304,  317-318,  322 

crassifolia  302,  304,  317-318,  322 

dodsonorum  320 

fragrantissima  299,  305,  308,  318-320,  325 

hexandra  299,  303-305,  313,  317,  320-322,  325 

parasitica  300 

parvifolia  304,  322-323 

p/tt/m  302,  320 

storkii  303 

suaveolens  299,  305,  313,  320,  323-325 
Tragulidae  195 
Tragulus  javanicus  195 
Triaenonychidae  134 
Trimorphodon  52,  202 

biscutatus  45,  49,  202 

biscutatus  lyrophanes  38 

biscutatus  vandenburghi  38 
Trionyx  spiniferus  63 


62 

wotata  28,  48^49 
notata  notata  28 


Urosaurus  52 

graciosus  44,  48 

graciosus  graciosus  28 

fcAte/ai  28,  48 

microscutatus  28,  45,  48 

nigricaudus  48 

ornatus  63 

Urosaurus  complex  48 
t/so/z/a  131 
[7/a52 

antigua  48 

nolascensis  48 

palmeri  48 

squamata  48 

stansburiana  45,  48 

stansburiana  elegans  28 

stellata  48 
[/to  complex  48 


Valdesia  300 
Varanus  salvator  195 
Vespertilionidae  194 
Viperidae  195 
Vipiridae  38 

Washingtonia  filifera  12 

Xantusia  60 

gigilis  48 

henshawi  45-46,  48,  59,  62 

henshawi  bolsonae  49 

henshawi  henshawi  28 

spp.  47 

v/g/fo  44,  49 

v/£//w  arizonae  49 

v/g/7/5  v/gz7/5  29,  49 

wgiV/s  wigginsi  29 
Xantusiidae  28 


schidigera  12,  27 
spp.  12 

12-13 


Zoarcidae  82,  221,  235 
Zwma  131