Skip to main content

Full text of "Miscellaneous publication - University of Kansas, Museum of Natural History"

See other formats 


HARVARD  UNIVERSITY 


LIBRARY 


OF  THE 


Museum  of  Comparative  Zoology 


The  Library 

Museum  of  Comparative  ZoologT 

Harvard  University 


_OOLl 

'^  _^  ^,  w-  .  LIBRARY 

OCT  151971 

HARVARD 

CONTRIBUTIONS  IN  MAMmXToGY 


A   Volume  Honoring 
Professor  E.  Raymond  Hall 


EDITED  BY 

J.  Knox  Jones,  Jr. 


Museum  of  Natural  History 

The  University  of  Kansas 

1969 


University  of  Kansas  Museum  of  Natural  History,  Miscellaneous  Publications 


II 


Institutional  libraries  interested  in  publications  exchange  may  obtain  this  series  by 
addressing  the  Exchange  Librarian,  The   University   of  Kansas,   Lawrence,  Kansas    66044. 

Requests  of  individuals  are  handled  instead  by  the  Museum  of  Natural  History,  The 
University  of  Kansas,  Lawrence,  Kansas  66044.  When  individuals  request  copies  from 
the  Museum  the  amount  indicated  below  should  be  included  for  the  purpose  of  defraying 
some  of  the  costs  of  producing,  wrapping  and  mailing.  Nos.  G,  12,  17,  27,  36,  37,  and  38 
are  obtainable  only  from  the  Arctic  Institute. 

An  asterisk    (*)    indicates  that  copies  are  no  longer  available  from   the   Museum. 

*1.  The  Museum  of  Natural  History,  the  University  of  Kansas.    By  E.  R.  Hall  and  Ann  Murray.    Pp. 

1-16,  illustrated.    January  5,  1946. 
*2.  Handbook  of  amphibians  and   reptiles   of   Kansas.    By   Hobart   M.   Smith.    Pp.   1-336,   233   figures 

in  text.    September  12,  1950. 
*3.  In  memoriam,  Charles  Dean  Bunker,  1870-1948.    By  E.  Raymond  Hall.    Pp.  1-11,  1   figure  in  text. 

December  15,  1951. 
*4.  The  University  of  Kansas,   Natural  History   Reservation.    By  Henry  S.   Fitch.    Pp.   1-38,  4   plates, 

3  figures  in  text.    February  20,  1952. 
*5.  Prairie  chickens  of  Kansas.    By  Maurice  F.  Baker.    Pp.  1-68,  4  plates,  15  figures  in  text.    March 

10.  1953. 

6.  The  barren  ground  caribou  of  Keewatin.  By  Francis  Harper.  Pp.  1-163,  28  figures.  October  21, 
1955.  Copies,  paper  bound,  SI. 50  postpaid  from  the  Arctic  Institute  of  North  America,  1619 
New  Hampshire  Avenue,  N.  W.,  Washington,  D.  C.    20009. 

7.  Handbook  of  mammals  of  Kansas.  By  E.  Raymond  Hall.  Pp.  1-303,  illustrated.  December  13, 
1955.    Paper  bound   $1.50   postpaid. 

8.  Mammals  of  northern  Alaska,  on  the  arctic  slope.  By  James  W.  Bee  and  E.  Raymond  Hall. 
Pp.  1-309,  frontispiece  colored,  4  plates,  127  figures  in  text.  March  10,  1956.  Paper  bound  $1.00 
postpaid. 

9.  Handbook  of  amphibians  and  reptiles  of  Kansas.  2nd  [revised]  edition.  By  Hobart  M.  Smith. 
Pp.   1-356,  253   figures   in  text.    April   20,   1956.    Paper  bound   $1.50   postpaid. 

*10.  The  raccoon  in  Kansas.    By  Howard  J.  Stains.    Pp.  1-76,  4  plates,  14  figures  in  text.    July  6,  1956. 

*11.  The   tree   squirrels    of   Kansas.     By    Robert   L.    Packard.     Pp.    1-67,    2    plates,    10    figures    in    text. 

August  20,   1956. 
12.  The   mammals   of   Keewatin.     By   Francis   Harper.     Pp.    1-94,    6   plates.   8    figures    in    text.    1    map. 

October    26,    1956.     Copies,    paper   bound,    75    cents    postpaid    from    the    Arctic    Institute    of    North 

America,  1619  New  Hampshire  Avenue,  N.  W.,  Washington,  D.   C.    20009. 
*13.  Museum    of    Natural    History    .    .    .    University    of    Kansas.     By    Roy    R.    Moore    and    E.    R.    Hall. 

[An  unpaged,   illustrated    "flier,"   14 V^   in.   x   8I2    in.,   printed   on   both   sides,   and    folded   twice.] 

June  1,  1957. 
*14.  Vernacular  names  for  North  American  mammals  north  of  Mexico.    By  E.  Raymond  Hall,  Sydney 

Anderson,  J.  Knox  Jones.  Jr.,  and  Robert  L.  Packard.    Pp.  1-16.    June  19,  1957. 
*15.  The  ecology  of  bobwhites   in   south-central   Kansas.    By   Thane   S.   Robinson.     Pp.    1-84,   2   plates, 

11   figures  in  text.    September  6.   1957. 
*16    Natural  history  of  the  prairie  dog  in  Kansas.    By  Ronald  E.  Smith.    Pp.   1-36,  4  plates,  9  figures 

in  text.    June  17.  1958. 

17.  Birds  of  the  Ungava  Peninsula.  By  Francis  Harper.  Pp.  1-171.  6  plates,  26  figures  in  text. 
October  15,  1958.  Copies,  paper  bound,  $2.00  postpaid  from  the  Arctic  Institute  of  North 
America,   1619  New  Hampshire  Avenue,   N.   W.,   Washington,   D.   C.    20009. 

18.  Furbearers  in  Kansas:  A  guide  to  trapping.  By  Howard  J.  Stains  and  Rollin  H.  Baker.  Pp. 
1-100,  2  plates,   13  figures  in  text.    November   19,   1958.    Paper  bound   50   cents   postpaid. 

*19.  Natural   History   Museum.    By   Roy   R.    Moore    and    E.    R.    Hall.     [An   unpaged    illustrated    "flier," 
14'2  in.  X  8V2  in.,  printed  on  both  sides,  and  folded   twice. [     May  29.   1959. 

20.  Handbook  of  gastropods  in  Kansas.  By  A.  Byron  Leonard.  Pp.  1-224,  plates  1-11,  87  figures 
in  text.    November  2,   1959.    Paper  bound   $1.00   postpaid. 

21.  Management  of  channel  catfish  in  Kansas.  By  Jackson  Davis.  Pp.  1-56,  8  figures  in  text. 
November  2,   1959.    Paper  bound   50   cents   postpaid. 

22.  Hand-list  of  the  birds  of  Kansas.  By  Richard  F.  Johnston.  Pp.  1-6  [folded  twice].  May  7,  1960. 
10  cents  postpaid. 

*23.  Directory  to  the  bird-life  of  Kansas.    By  Richard  F.  Johnston.    Pp.  1-69,  1  figure  in  text.    August 

31,  1960. 
*24.  Natural   History   Museum.    By   Roy   R.   Moore   and   E.    R.    Hall.     [An   unpaged,    illustrated    "flier," 

141,2  in.  X  S'a   in.,  printed  on  both  sides,   and   folded   twice. [     October   19,   1960. 
25.  Guide   to   the   Panorama   of   North   American   Mammals.    By   E.    Raymond   Hall,    et   al.     Pp.    1-31, 

silhouettes   in  black   and   white   of   Panorama,   life-zones,   and   taped    commentary   for   each   zone. 

December  15,  1960.    Paper  bound  50  cents  postpaid. 
*26.  Beaver  in  Kansas.    By  F.  Robert  Henderson.    Pp.   1-85,   illustrated.    December   16,   1960. 

27.  Land  and  fresh-water  mammals  of  the  Ungava  Peninsula.  By  Francis  Harper.  Pp.  1-178,  plates 
1-8,  3  figures  in  text.  August  11,  1961.  Paper  bound.  $2.00  postpaid  from  tlie  Arctic  Institute 
of   North   America,    1619   New   Hampshire   Avenue,    N.    W.,   Washington,    D.    C.     20009. 

28.  Handbook  of  unionid  mussels  in  Kansas.  By  Harold  D.  Murray  and  A.  Byron  Leonard.  Pp.  1-184, 
45  plates,  42  figures  in  text.    May  10.   1962.    Paper  bound  $100  postpaid. 

(Continued  on  inside  back  cover) 


MUS.  lOO! 


~  I  _  1  \  /-I . 


1 


«/ 


CONTRIBUTIONS  IN  MAMMALOGY. 

UNiVERsrrr. 


A   Volume  Honoring 
Professor  E.  Raymona  Hall 


EDITED  BY 

J.  Knox  Jones,  Jr. 


Museum  of  Natural  History 

The  University  of  Kansas 

1969 


University    of    Kansas 
Museum    of    Natural    History 


Miscellaneous  Publication  No.  51,  pp.  1-428,  122  figs. 
Published  July  11, 1969 


Lawrence  •  Kansas 


PRINTED  BY 

THE  UNIVERSITY  OF  KANSAS  PRINTING  SERVICE 

1969 


ContriLutions  in  Mammalogy 

EDITED  BY 

J.  Knox  Jones,  Jr. 

-                      CONTENTS 
Preface    5 

Editor's  Note  6 

Stephen  D.  Durrant,  Department  of  Environmental  Biology, 
The  University  of  Utah,  Salt  Lake  City,  Utah  84112 
Eugene  Raymond  Hall — Biography  and  Bibliography 9 

Sydney  Anderson,  Department  of  Mammalogy,  The  American 
Museum  of  Natural  History,  New  York,  New  York  10024 
Taxonomic  Status  of  the  Woodrat,  Neotoma  albigula, 
in  Southern  Chihuahua,  Mexico  25 

Terry  A.  Vaughan,  Department  of  Biological  Sciences, 

Northern  Arizona  University,  Flagstaff,  Arizona  86001 
Reproduction  and  Population  Densities  ln  a 
Montane  Small  Mammal  Fauna  51 

Donald  F.  Hoffmeister,  Museum  of  Natural  History,  University 
of  Illinois,  Urbana,  Illinois  61608 
The  Species  Problem  in  the  Thomomys  bottae — 
Thomomys  umbrinus  Complex  of  Pocket  Gophers 
in  Arizona   75 

Ticul  Alvarez,  Escuela  Nacional  de  Ciencias  Biologicas, 

Instituto  Politecnico  Nacional,  Mexico  17,  D.  F.,  Mexico 
Restos  Fosiles  de  Mamiferos  de  Tlapacoya, 
EsTADO  DE  Mexico  (  Pleistoceno-Reciente  )  93 

James  S.  Findley,  Department  of  Biology,  The  University  of 
New  Mexico,  Albuquerque,  New  Mexico   87106 
blogeography  of  southwestern  boreal  and 
Desert  Mammals  113 

J.  Knox  Jones,  Jr.,  and  Hugh  H.  Genoways,  Museum  of  Natural 
History,  The  University  of  Kansas,  Lawrence,  Kansas 
66044 

HOLOTYPES  OF  ReCENT  MaMMALS  IN  THE  MuSEUM  OF 

Natural  History,  The  University  of  Kansas  129 

(3) 


Charles  L.  Douglas,  Texas  Memorial  Museum,  The  University 
of  Texas,  Austin,  Texas   78705 
Ecology  of  Pocket  Gophers  of  Mesa  Verde,  Colorado 147 

Rollin  H.  Baker,  The  Museum,  Michigan  State  University, 
East  Lansing,  Michigan  48823 
Cotton  Rats  of  the  Sigmodon  fulviventer  Group  177 

Robert  B.  Finley,  Jr.,  Section  of  Upland  Wildlife  Ecology, 

U.S.  Fish  and  Wildlife  Service,  Denver,  Colorado  80225 
Cone  Caches  and  Middens  of  Tamiasciurus  in  the 
Rocky  Mountain  Region  233 

John  A.  White,  The  Museum,  Idaho  State  University, 
Pocatello,  Idaho  83201 
Late  Cenozoic  Bats  (Subfamily  Nyctopiiylinae ) 

FROM  THE  AnZA-BoRREGO  DeSERT  OF  CALIFORNIA  275 

Henry  W.  Setzer,  Division  of  Mammals.  U.S.  National  Museum, 
Washington,  D.C.    20560 
A  Review  of  the  African  Mice  of  the  Genus 
Desmodilliscus  Wettstein,  1916  283 

Charles  A.  Long,  Department  of  Biology,  Wisconsin  State 
University,  Stevens  Point,  Wisconsin  54481 
An  Analysis  of  Patterns  of  Variation  in  Some 
Representative  Mammalia.   Part  II.   Studies  on  the 
Nature  and  Correlation  of  Measures  of  Variation 289 

E.  Lendell  Cockrum,  Department  of  Biological  Sciences, 
The  University  of  Arizona,  Tucson,  Arizona  85721 
Migration  of  the  Guano  Bat,  Tadarida  brasiliensis 303 

Robert  J.  Russell,  Department  of  Biology,  University  of 

Missouri  at  Kansas  City,  Kansas  City,  Missouri  64110 
Intraspecific  Population  Structure  of  the  Species 
Pappogeomys  castanops 337 

Robert  L.  Packard,  Department  of  Biology,  Texas 

Technological  College,  Lubbock,  Texas   79409 

Taxonomic  Reviev^  of  the  Golden  Mouse, 

ochrotomys  nuttalli  373 

Bernardo  Villa-R.  and  Martha  Villa  Cornejo,  Instituto  dc 

Biologia,  Universidad  Nacional  Autonoma  de  Mexico, 
Mexico  20,  D.  F.,  Mexico 
Algunos  Murcielagos  del  Norte  de  Argentina  407 

(4) 


PREFACE 

This  volume  is  dedicated  to  Professor  E.  Raymond  Hall  on  the 
occasion  of  his  retirement  from  the  directorship  of  The  University 
of  Kansas  Museum  of  Natural  History  on  June  30,  1967.  The  word 
"retirement"  for  someone  of  Professor  Hall's  interests  and  energies 
has  little  meaning  and  the  months  following  this  event  have  wit- 
nessed a  continuation  of  his  active  involvement  in  local,  national, 
and  international  conservation  activities,  field  work,  scholarly  re- 
search, and  graduate  teaching. 

The  fact  that  the  Museum  of  Natural  History  continues  to 
flourish  in  an  academic  world  where  museums  are  sometimes  poorly 
understood  is  a  tribute  to  Professor  Hall's  vision  and  the  single- 
mindedness  of  his  administrative  activities  through  the  years.  More 
important  surely  is  the  fact  that  the  community  of  faculty  and 
graduate  student  scholars  inhabiting  the  Museum  is  singularly  well 
adapted  to  relate  the  field  of  vertebrate  natural  history  in  a  museum 
environment  to  the  larger  problems  and  principles  of  modern  biol- 
ogy. This  circumstance  is  due  in  large  measure  to  the  personal 
philosophy  which  guided  Professor  Hall  through  his  years  as 
director. 

The  ensuing  collection  of  articles  in  the  field  of  mammalogy 
honoring  Professor  E.  Raymond  Hall  was  assembled  under  the 
editorial  supervision  of  Professor  J.  Knox  Jones  and  contains  con- 
tributions from  many  of  those  who  received  advanced  degrees  with 
Professor  Hall's  guidance.  This  collection  of  papers,  prepared 
especially  in  Professor  Hall's  honor,  is  not  only  an  affectionate 
tribute  to  him  but  also  an  exhortation  that  he  continue  unabated 
the  extraordinary  research  career  that  began  more  than  40  years 
ago. 

As  Director  of  The  University  of  Kansas  Museum  of  Natural 
History  (1944-1967),  Professor  Hall  was  a  scholar-administrator  for 
23  years  with  various  intermittent  administrative  involvements  for  a 
much  longer  period.  Professor  Hall's  career  is  proof-positive  that 
scholarly  research,  teaching,  public  service,  and  administration  can 
form  a  highly  productive  mix  and  one  which  his  successors  may 
have  difficulty  emulating. 

Philip  S.  Humphrey 


(5) 


EDITOR'S  NOTE 

Shortly  after  E.  Raymond  Hall's  retirement  as  Director  of  the 
Museum  of  Natural  History  at  Kansas,  several  of  his  former  grad- 
uate students  met  to  discuss  means  of  commemorating  the  occasion. 
It  was  decided  that  a  "Festschrift"  of  papers  in  mammalogy,  con- 
tributed by  his  students  and  honoring  a  lifetime  of  teaching  and 
research,  would  be  eminently  appropriate.  Officials  at  The  Uni- 
versity of  Kansas,  including  Chancellor  W.  Clarke  Wescoe,  Provost 
lames  R.  Surface,  and  Dean  of  Faculties  Francis  H.  Heller,  en- 
thusiastically endorsed  the  project  and  have  provided  substantial 
financial  support. 

All  those  who  received  the  Ph.D.  degree  under  Professor  Hall's 
guidance  were  asked  to  contribute  to  the  proposed  volume,  as  were 
two  leading  Mexican  mammalogists  who  studied  at  Kansas  and 
received  the  M.A.  degree  there.  Some  persons  contacted  under- 
standably were  unable  to  participate,  principally  because  admin- 
istrative or  other  duties  long  had  deprived  them  of  the  opportunity 
for  active  research  in  mammalogy.  Sexenteen,  however,  agreed  to 
do  so,  and  were  instructed  to  select  from  their  own  current  research 
a  contribution  that  they  felt  would  be  appropriate  for  the  "Fest- 
schrift." In  this  way,  it  was  hoped  that  the  collected  papers  would 
reflect,  indirectly,  one  aspect  of  Professor  Hall's  impact  on  the 
discipline  of  mammalogy. 

Editing  of  the  papers  that  comprise  "Contributions  in  Mam- 
malogy" was  held  to  a  minimum  consistent  with  the  established 
style  of  the  publications  of  the  Museum  of  Natural  History.  The 
present  volume  is  number  51  of  the  Miscellaneous  Publications  of 
the  museum,  which  is  particularly  pertinent  in  that  E.  Raymond 
Hall  was  instrumental  in  establishing  this  series  as  well  as  the 
familiar  "University  of  Kansas  Publications,  Museum  of  Natural 
History." 

The  outstanding  workmanship  and  cooperation  of  persons  at 
The  University  of  Kansas  Printing  Service  in  seeing  this  volume 
through  to  completion  is  gratefully  acknowledged,  as  is  the  sub- 
stantial editorial  assistance  provided  by  several  of  my  graduate 
students,  especially  Elmer  C.  Birney,  Hugh  H.  Genoways,  Carleton 
J.  Phillips,  James  D.  Smith,  and  Ronald  W.  Turner. 

/.  Knox  Jones,  Jr. 

(6) 


CONTRIBUTIONS  IN  MAMMALOGY 


(7) 


(8) 


EUGENE  RAYMOND  HALL— BIOGRAPHY 
AND  BIBLIOGRAPHY 

BY 

Stephen  D.  Durrant 

Professor  E.  Raymond  Hall,  honored  by  his  students  and  to 
w  honi  they  dedicate  this  \'olume,  was  born  in  Imes,  Kansas,  on  May 
11,  1902,  to  Wilbur  Downs  and  Susan  Effie  (Donovan)  Hall;  he 
married  Mary  Frances  Harkey  on  August  9,  1924;  three  sons,  Wil- 
liam Joel,  Hubert  Handel,  and  Benjamin  Downs  were  born  to  them. 

His  grammar  school  education  was  obtained  in  Kansas,  as  was 
his  high  school  with  the  exception  of  the  third  year,  which  was  taken 
at  Yakima,  Washington.  He  was  awarded  the  A.B.  degree  from 
The  University  of  Kansas  in  1924.  His  M.A.  and  Ph.D.  degrees 
were  awarded  by  the  University  of  California  at  Berkeley  in  1925 
and  1928,  respectively  ( see  "Who's  Who  in  America" ) . 

He  has  held  80  appointments  and  positions  of  local,  state,  univer- 
sity, national,  and  international  scope.  Certainly  those  of  greatest 
significance  to  him  and  to  his  contemporaries  in  his  chosen  field  are 
Associate  Professor  of  Vertebrate  Zoology  (1937-1944),  Curator  of 
Mammals  (1927-1944)  and  Acting  Director  of  the  Museum  of 
Vertebrate  Zoology  (1938-1944)  at  the  University  of  California, 
Berkeley,  and  Professor  of  Zoology  (1944-1958),  Chairman,  Depart- 
ment of  Zoology  (1944-1961),  Director,  Museum  of  Natural  History 
(1944-1967),  and  Summerfield  Distinguished  Professor  (1958- 
present)  at  The  University  of  Kansas. 

Under  his  guidance  and  through  his  efforts,  the  collection  of 
mammals  and  the  output  of  published  information  forged  ahead  at 
the  Museum  of  Vertebrate  Zoology,  University  of  California.  This 
was  perhaps  the  most  prolific  period  (1928-1944)  of  this  museum, 
both  in  the  acquisition  of  specimens  of  mammals  and  in  the  pub- 
lished works  upon  them. 

He  returned  to  his  Alma  Mater,  The  University  of  Kansas,  in 
1944,  as  Director  of  the  Museum  of  Natural  History  and  Chairman 
of  the  Department  of  Zoology,  which,  in  many  ways,  was  perhaps 
the  beginning  of  his  finest  hour.  This  museum,  in  24  short  years,  has 
matured  and  grown  in  stature  and  recognition  in  all  phases  until  at 
present  it  is  one  of  the  truly  great  institutions  of  our  land.  The 
expansion  of  the  physical  plant,  the  remarkable  acquisitions  in  kinds 
and  numbers  of  specimens,  the  gathering  of  outstanding  scholars  to 

(9) 


10  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

conduct  its  acti\'ities,  the  number  and  quality  of  its  graduate  stu- 
dents, the  hundreds  of  pages  of  printed  results,  and  the  educational 
opportunities  provided  for  the  entire  citizenry  bear  direct  testimony 
of  Professor  Hall's  industry,  devotion,  and  dedication. 

He  is  a  member  of  25  scientific  societies  and  has  held  important 
positions  in  several,  being  a  Fellow  in  the  American  Society  for  the 
Advancement  of  Science  and  Past  President  and  Honorary  Member 
of  the  American  Society  of  Mammalogists.  Moreover,  he  has 
traveled  widely  in  North  America,  Central  America,  and  Europe. 

While  these  accomplishments  are  remarkable  and  depict  great 
energy  and  purpose,  they  are  far  from  his  greatest,  which  are  his 
contributions  through  his  students,  especially  his  graduates,  and 
by  the  printed  word.  The  latter,  as  of  1968,  consisted  of  309  titles 
invohing  some  5400  pages.  The  outstanding  are  "Mammals  of 
Nevada,"  "The  Weasels  of  North  America,"  and  the  two  volume 
"Mammals  of  North  America"  with  K.  R.  Kelson  (see  attached 
Bibliography).  It  is  impossible  to  totally  assess  his  impact  upon  the 
field  of  taxonomic  mammalogy  through  his  students,  but  the  fruits  of 
his  personal  efforts  are  well  known.  He  has  named  and  described 
nine  new  genera,  a  new  subgenus,  23  new  species,  and  138  new 
subspecies  of  both  fossil  and  Recent  mammals.  Four  kinds,  Lutravus 
halli  Furlong,  Verognathoides  halli  Wood,  Microtus  longicaudus 
haUi  Ellerman,  and  Taxidea  taxus  halli  Schantz,  have  been  named 
in  his  honor,  and  another  is  named  in  this  \'olume. 

Investigations  have  failed  to  disclose  another  person  of  our  time 
who  sits  so  high  in  the  saddle  or  who  has  cast  a  wider  or  longer 
shadow  of  accomplishment  over  this  field  of  endeavor.  In  the  field 
of  taxonomic  mammalogy,  he  and  his  intellectual  sons  and  grandsons 
are  in  the  forefront  in  positions  and  prestige.  Anywhere  where  taxo- 
nomic mammalogy  is  pursured  his  influence  is  felt  by  his  writings, 
his  students,  or  his  students'  students  and  their  total  productivity. 

Penetrating  studies  of  his  disposition  and  capacities  enable  one 
to  understand  and  evaluate  the  personal  characteristics  that  have 
aided  him  to  mount  such  a  high  pinnacle  of  achievement.  Doctor 
Hall  has  an  extremely  intense  dedication  to  his  work,  a  nearly  fierce 
belief  in  the  efficacy  of  the  problem  at  hand  and  his  position  with 
reference  to  it,  and  unlimited  determination  and  physical  capacity 
to  see  projects  established  and  carried  to  completion. 

In  many  respects  he  is  a  stormy  petrel  or  at  least  a  highly  con- 
troversial figure.  Once  committed  to  a  course  of  action,  his  dogged 
persistence,  his  unrelenting  attack,  his  singleness  of  purpose,  and 


DuRRANT — Hall  Biography  and  Bibliography  11 

his  characteristic  shouldering  aside  of  opposition  usually  have  led 
him  to  the  attainment  of  his  goals.  If  thwarted  in  one  approach,  he 
stubbornly  launches  a  new  attack  from  a  new,  different  stance.  To 
friend,  he  is  a  great  source  of  support  and  encouragement;  his 
opponents  soon  learn  they  have  a  man  to  be  measured  and  an 
adversary  worthy  of  their  steel. 

To  him,  the  written  word,  whether  in  manuscript,  galley,  page 
proof,  or  printed,  possesses  a  certain  sanctity.  His  students  well 
remember  some  of  his  idiosyncrasies  in  his  efforts  to  arrive  at  excel- 
lence.  Comments  like:  "You  cannot  encounter  a  mouse";  "To  begin 
a  sentence  with  'however'  is  poor";  "Be  careful  of  the  use  of  'due  to,' 
especially  with  an  adverbial  modifier";  "Use  first  usages  in  the 
dictionary";  "Be  careful  of  the  use  of  the  word  'type.' "  Whether 
they  agreed  or  not,  all  his  students  were  impressed  by  the  fact  that 
when  their  manuscripts  were  returned,  they  were  the  best  he  was 
capable  of  making  them.  In  some  ways,  he  had  a  sixth  sense  of 
timing  to  keep  a  student  totally  productively  occupied  (sometimes 
fretfully)  from  the  inception  of  his  study  to  just  a  few  moments 
before  the  deadline. 

To  all  these  accomplishments  there  are  many  as  yet  unsensed 
and  uncatalogued  to  be  brought  into  full  fruition  in  decades  to 
come.  It  has  often  been  stated  that  no  one  can  ever  know  the  total 
effects  of  a  stone  thrown  into  a  quiet  pool.  Likewise,  it  would  be 
impossible  to  completely  assess  the  total  impact  of  our  man  of  the 
hour  upon  the  hundreds  of  persons  with  whom  he  has  come  into 
contact  both  in  and  out  of  his  chosen  field.  Moreover,  the  situation 
is  not  to  be  likened  to  that  of  a  quiet  pool  and  the  effects  in  a 
turbulent  one  undoubtedly  would  be  tremendously  pronounced  and 
increased  both  in  scope  and  intensity. 

His  students  and  colleagues  stand  in  his  honor,  respect  his  con- 
tributions, hope  for  his  continued  productivity,  and  dedicate  this 
volume  to  him  as  a  monument  to  his  endeavors  and  accomplish- 
ments. It  might  be  said  of  him  that  he  belongs  to  that  select  few 
who  during  their  lifetime  have  seen  their  realizations  far  exceed 
theii-  fondest  expectations. 

Bibliography  of  E.  Raymond  Hall,  1921-1968 

1921 

1.  Golden- winged  warbler  in  Kansas.    Auk,  38:607,  December  16. 

192.3 

2.  Occurrence  of  the  hoary  bat  at  Lawrence,  Kansas.    Join-.  Manim.,  4:192- 
193,  August  10. 


12  Misc.  PuBL.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

3.  Winter  visitors  at  Lawrence,  Kansas.    Auk,  40:701-702,  October  10. 

1925 

4.  Pelicans  versus  fishes  in  Pyramid  Lake.    Condor,  27:147-160,  July  15. 

1926 

5.  A  new  subspecies  of  the  California  spotted  skunk  ( Spilogale  phenax 
Merriam).    Jour.  Mamm.,  7:53-56,  February  15. 

6.  Changes  during  growth  in  the  skull  of  the  rodent  Otospermophilus  gram- 
murus  beecheyi.  Univ.  California  Publ.  ZooL,  21:355-404,  43  figs., 
March  9. 

7.  Notes  on  water  birds  nesting  at  Pyramid  Lake,  Nevada.  Condor,  28: 
87-91,  March  15. 

8.  A  new  marten  from  the  Pleistocene  cave  deposits  of  California.  Jour. 
Mamm.,  7:127-130,  1  pi..  May  13. 

9.  The  abdominal  skin  gland  of  Martes.    Jour.  Mamm.,  7:227-229,  August  9. 

10.  Economic  value  of  Mexican  free-tailed  bat.  California  Fish  and  Game, 
12:135-137,  September  1. 

11.  Golden  eagle  extermination  in  England.  California  Fish  and  Game,  12: 
139,  September  1. 

12.  [Review  of]  New  teachers  bulletin  on  fish  and  game  laws.  California  Fish 
and  Game,  12:143,  September  1. 

13.  Barbed  wire  fence  causes  death  of  deer.  California  Fish  and  Game,  12: 
151-152,  fig.  29,  September  1. 

14.  Forage  habits  of  pocket  gopher.  California  Fish  and  Game,  12:152,  Sep- 
tember 1. 

15.  Arkansas  still  has  no  protection  on  fish.  California  Fish  and  Game,  12:153, 
September  1. 

16.  New  method  of  predatory  mauunal  control.  California  Fish  and  Game, 
12:154,  September  1. 

17.  Oil  pollution.    California  Fish  and  Game,  12:154,  September  1. 

18.  Systematic  notes  on  the  subspecies  of  Bassariscus  astutus  with  description 
of  one  new  form  from  California.  Univ.  California  Publ.  Zool.,  30:39-50, 
pis.  2-3,  September  8. 

19.  The  muscular  anatomv  of  three  mustelid  mammals.  Mephitis,  Spilogale 
and  Martes.    Univ.  California  Publ.  Zool.,  30:7-38,  5  figs.,  September  14. 

1927 

20.  An  outbreak  of  house  mice  in  Kern  County,  California.  Univ.  California 
Publ.  Zool.,  30:189-203,  February  21. 

21.  Species  of  the  mammalian  subfauiily  Bassariscinae.  Univ.  California  Publ. 
Geol.  Sci.,  16:435-448,  pi.  64,  2  figs.,  March  17. 

22.  Notes  on  the  birds  of  Douglas  County,  Kansas.  Wilson  Bull.,  39:91-105, 
June  (with  J.  Linsdale). 

23.  The  muscular  anatomy  of  the  American  badger  (Taxidea  taxus).  Univ. 
California  Publ.  Zool.,  30:205-219,  2  figs.,  July  28. 

24.  The  deer  of  California.  California  Fish  and  Game,  13:233-259,  figs.  49-62, 
November  5. 

25.  A  commensal  relation  of  the  California  (luail  with  the  California  ground 
squirrel.    Condor,  29:271,  November  15. 

26.  The  barn  owl  in  its  relation  to  the  rodent  population  at  Berkeley,  Cali- 
fornia.   Condor,  29:274-275,  November  15. 

27.  A  new  weasel  from  Louisiana.  Proc.  Biol.  Soc.  Washington,  40:193-194, 
December  2. 

1928 

28.  A  correction  [for  The  deer  of  California].  California  Fish  and  Game,  14: 
51-52,  January  30. 


DuRRANT — Hall  Biography  and  Bibliography  13 

29.  A  new  race  of  black  bear  from  Vancoiner  Island,  British  Columbia,  with 
remarks  on  other  Northwest  Coast  forms  of  Euractos.  Uni\'.  California 
Piibl.  Zool.,  30:231-242,  March  2. 

30  Records  of  supernumerary  teeth  in  bears.  Univ.  California  Publ.  Zool., 
30:243-250,  pis.,  14-15,  1  fig.,  March  2. 

31.  Distribution  and  speciation  in  American  weasels  (subgenus  Mustela).  Pro- 
gramme, Ph.D.  degree,  Univ.  California  Grad.  Div.,  4  pp.,  May  8. 

32.  Notes  on  the  life  history  of  the  sage-brush  meadow  mouse  (Lagurus). 
Jour.  Mamm.,  9:201-204,  August  9. 

33.  Note  on  the  life  history  of  the  woodland  deer  mouse.  Jour.  Mamm.,  9: 
255-256,  August  9. 

34.  Weasels  wanted.    Missouri  Game  and  Fish  News,  4(12):  17,  December. 

1929 

35.  On  the  question  of,  "Is  it  right  to  protect  the  female  of  the  species  at  the 
cost  of  the  male?"    Canadian  Field-Nat.,  43:59,  March. 

36  A  second  new  genus  of  hedgehog  from  the  Pliocene  of  Nevada.  Univ. 
California  Publ.  Geol.  Sci.,  18:227-231,  6  figs.,  March  19. 

37.  Nhrmmals  collected  by  Charles  D.  Brower  at  Point  Barrow,  Alaska.  Univ. 
California  Publ.  Zool.,  30:419-425,  March  19. 

38.  Notes  on  the  life  history  of  the  kangaroo  mouse  (Microdipodops).  Jour. 
Mamm.,  10:298-305,  pi.  22,  November  11  (with  J.  Linsdale). 

39.  California  mink  in  marine  habitat.  Jour.  Mamm.,  10:351-352,  November 
11. 

40.  A  "den"  of  rattlesnakes  in  eastern  Nevada.  Bull.  Antivenin  Inst.  Anier., 
3:79-80,  fig.  5,  November. 

1930 

41.  A  bassarisk  and  a  new  mustelid  from  the  later  Tertiary  of  California.  Jour. 
Mamm.,  11:23-26,  2  figs.,  February  11. 

42.  Three  new  genera  of  Mustelidae  from  the  later  Tertiary  of  North  America. 
Jour.  Mamm.,  11:146-155,  pis.  7-8,  May  9. 

43.  Statement  at  "Hearing  before  the  Committee  on  Agriculture  House  of 
Representatives  seventy-first  Congress  second  session  on  H.  R.  9599  by 
Mr.  Leavitt,  a  bill  to  authorize  the  Secretary  of  Agriculture  to  carry  out 
his  ten-year  cooperative  program  for  the  eradication,  suppression,  or 
bringing  under  control  of  predatory  and  other  wild  animals  injurious  to 
agriculture,  horticulture,  forestry,  animal  husbandry,  wild  game,  and  other 
interests,  and  for  the  suppression  of  rabies  and  tularemia  in  predatory  or 
other  wild  animals  for  other  pmposes."    Pp.  57-65,  May. 

44.  Three  new  pocket  gophers  from  Utah  and  Nevada.  Univ.  California  Publ. 
Zool.,  32:443-447,  July  8. 

45.  Predatory  mammal  destruction.    Jour.  Mamm.,  11:362-369,  August  9. 

46.  Rodents  and  lagomorphs  from  the  later  Terriary  of  Fish  Lake  Valley, 
Nevada.  Univ.  California  Publ.  Geol.  Sci.,  19:295-312,  pi.  37,  29  figs., 
November  25. 

47.  Rodents  and  lagomorphs  from  the  Barstow  beds  of  southern  California. 
Univ.  California  Publ.  Geol.  Sci.,  19:313-318,  7  figs.,  November  25. 

48.  A  new  genus  of  bat  from  the  later  Tertiary  of  Nevada.  Univ.  California 
Publ.  Geol.  Sci.,  19:319-320,  pi.  38,  November  25. 

1931 

49.  On  the  occurrence  of  certain  mammals  in  extreme  southwestern  Wash- 
ington.  Murrelet,  12:22,  January. 

50.  The  poisoner  again  [in  two  parts].  Outdoor  Life,  47(4):26-27,  82-84, 
March,  and  47(5):28-29,  62-63,  April. 


14  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

51.  Critical  comments  on  mammals  from  Utah,  with  descriptions  of  new  forms 
from  Utah,  Nevada  and  Washington.  Univ.  California  Piibl.  Zool.,  37: 
1-13,  April  10. 

52.  Description  of  a  new  mustelid  from  the  later  Tertiary  of  Oregon,  with 
assignment  of  Parictis  primaevus  to  the  Canidae.  lour.  Mamm.,  12:156- 
158,  pi.  5,  May  14. 

53.  Tree  climbing  Callospermophiliis.    Murrelet,  12:54-55,  May. 

54.  A  new  subspecies  of  Peromyscus  from  San  Jose  Island,  Lower  California, 
Mexico.    Proc.  Biol.  Soc.  Washington,  44:87-88,  June  29. 

55.  The  coyote  and  his  control.  California  Fish  and  Game,  17:283-290,  figs. 
88-89,  July. 

56.  A  skull  of  Nothocyon  from  the  John  Day  Oligocene.  Univ.  Kansas  Sci. 
Bull.,  19:283-286,  pi.  29,  for  July,  1930,  but  printed  about  August,  1931 
(with  H.  T.  Martin). 

19.32 

57.  Editorial  [no  title].    Condor,  34:52,  January  15. 

58.  New  pocket  gophers  from  Nevada.  Univ.  California  Publ.  Zool.,  38:325- 
333,  February  27. 

59.  A  new  pocket  gopher  from  Lower  California,  Mexico.  Proc.  Biol.  Soc. 
Washington,  45:67-70,  April  2. 

60.  A  new  black-tailed  jack-rabbit  from  Idaho.  Proc.  Biol.  Soc.  Washington, 
45:71-72,  April  2  (with  W.  B.  Whitlow). 

61.  Three  new  pocket  gophers  from  New  Mexico  and  Arizona.  Proc.  Biol. 
Soc.  Washington,  45:95-98,  June  21. 

62.  A  new  shrew  of  the  Sorex  merriami  group  from  Arizona.  Jour.  Mamm., 
13:259-262,  pi.  13,  August  9. 

63.  A  new  weasel  from  Panama.  Proc.  Biol.  Soc.  Washington,  45:139-140, 
September  9. 

64.  New  mammals  from  St.  Lawrence  Island,  Bering  Sea,  Alaska.  Univ. 
California  Publ.  Zool.,  38:391-404,  pis.  5-6,  1  fig.,  September  17  (with 
R.  M.  Gilmore). 

65.  A  new  pocket  gopher  from  New  Mexico.  Univ.  California  Publ.  Zool., 
38:411-413,  September  20. 

66.  Remarks  on  the  affinities  of  the  mammalian  fauna  of  Vancouver  Island, 
British  Columbia,  with  descriptions  of  new  subspecies.  Univ.  California 
Publ.  Zool.,  38:415-423,  November  8. 

1933 

67.  The  Asiatic  genus  Eomellivora  in  the  Pliocene  of  California.  Jour.  Mamm., 
14:63-65,  pi.  4,  February  14  (with  C.  Stock). 

68    A  new  race  of  pocket  gopher  found  in  Oregon  and  Washington.    Proc. 

Biol.  Soc.  Washington,  46:41-44,  March  24  (with  R.  T.  Orr). 
69.  Sorex  leucogenys  in  Arizona.   Jour.  Mamm.,  14:153-154,  May  15. 
70    Mammals  of  the  Pocatello  region  of  southeastern  Idaho.    Univ.  California 

Publ.  Zool.,  40:235-277,  3  figs.,  September  30  (with  W.  B.  Whitlow). 

71.  Arrangement  of  the  obturator  muscles  with  notes  on  the  other  muscles  of 
the  thigh,  in  the  dwarf  wapiti  (Cervus  nannodes  Merriam).  Jour.  Manun., 
14:358-361,  2  figs.,  November  13  (with  A.  H.  Miller). 

72.  Dermestid  beetles  as  an  aid  in  cleaning  bones.  Jour.  Mamm.,  14:372-374, 
November  13  (with  W.  C.  Russell). 

1934 

73.  Notes  on  Arizona  Rodents.  Proc.  Biol.  Soc.  Washington,  47:51-56,  Feb- 
ruary 9  (with  W.  B.  Davis). 

74  A  new  race  of  chipmunk  from  the  Great  Basin  in  western  United  States. 
Univ.  California  Publ.  Zool.,  40:321-326,  1  Bg.,  February  12  (with  D.  M. 
Hatfield). 

75  Certain  osteological  features  of  Eudernia  maculatum  (J.  A.  Allen).  Jour. 
Mamm.,  15:68-70,  8  figs.,  February  15. 


DuRRANT — Hall  Biography  and  Bibliography  15 

76.  The  coyote  and  his  control.    Outdoor  Life,  73(4):30-32,  March. 

77.  Marmota  caHgata  broweri,  a  new  marmot  from  northern  Alaska.  Canadian 
Field-Nat.,  48:57-59,  6  figs.,  April  (with  R.  M.  Gilmore). 

78.  Sorex  melanogenys  Hall,  a  synonym  of  Sorex  vagrans  monticola  Merriam. 
Jour.  Mamm.,  15:155,  May  15. 

79.  A  new  pika  (mammalian  genus  Ochotona)  from  central  Nevada.  Proc. 
Biol.  Soc.  Washington,  47:103-106,  June  13. 

80.  Statement  of  Dr.  E.  Raymond  Hall  on  grazing  of  sheep  on  the  public 
domain  and  in  the  National  Forests.  Pp.  176-177,  in  Grazing  Sheep  in 
National  Forests,  Hearing  before  the  Special  Committee  on  Conservation 
of  Wild  Life  Resources,  United  States  Senate,  Seventy-third  Congress, 
second  session. 

81.  Two  new  rodents  of  the  genera  Glaucomys  and  Zapus  from  Utah.  Occas. 
Papers,  Mus.  Zool.,  Univ.  Michigan,  296:1-6,  November  2. 

82.  Mammals  collected  by  T.  T.  and  E.  B.  McCabe  in  the  Bowron  Lake 
region  of  British  Columbia.  Univ.  California  Publ.  Zool.,  40:363-386,  1 
fig.,  November  5. 

1935 

83.  Geographic  distribution  of  pocket  gophers  (genus  Thomomys)  in  Nevada. 
Univ.  California  Publ.  Zool.,  40:387-402,  1  fig.,  March  13  (with  W.  B. 
Davis). 

84.  A  new  mustelid  genus  from  the  Pliocene  of  California.  Jour.  Mamm., 
16:137-138,  3  figs..  May  15. 

85.  Occurrence  of  the  spotted  bat  at  Reno,  Nevada.  Jour.  Mamm.,  16:148, 
May  15. 

86.  A  new  weasel  from  Peru.  Proc.  Biol.  Soc.  Washington,  48:143-146, 
August  22. 

87.  Nevadan  races  of  the  Microtus  montanus  group  of  meadow  mice.  Univ. 
California  Publ.  Zool.,  40:417-428,  1  fig.,  October  25. 

1936 

88.  A  new  meadow  mouse  from  Bowen  Island,  British  Columbia.  Murrelet, 
17:15-16,  March  7. 

89.  Identity  of  the  Bowron  Lake  moose  of  British  Columbia.  Murrelet,  17:17, 
March  7. 

90.  A  new  pocket  gopher  from  New  Mexico.  Jour.  Washington  Acad.  Sci., 
26:296-298,  July  15. 

91.  Ranges  and  relationships  of  certain  mammals  in  southwestern  Utah.  Proc. 
Utah  Acad.  Sci.,  Arts  and  Letters,  13:211-213,  September  15  (with  C.  C. 
Presnall). 

92.  Mustelid  mammals  from  the  Pleistocene  of  North  America  with  systematic 
notes  on  some  Recent  members  of  the  genera  Mustela,  Taxidea  and 
Mephitis.  Publ.  Carnegie  Inst.  Washington,  473:41-119,  pis.  1-5,  6  figs., 
November  20. 

1937 

93.  Mustela  cicognanii,  the  short-tailed  weasel,  incorrectly  ascribed  to  Ohio. 
Amer.  Midland  Nat.,  18:304,  March. 

94.  Lafayette  sunshine.    Lafayette  Sun,  1  p.  (with  M.  F.  Hall). 

95.  [Review  of]  October  farm,  from  the  Concord  Journals  of  William  Brew- 
ster.   Jour.  Mamm.,  18:245,  May  14. 

96.  A  new  kangaroo  mouse  (Microdipodops)  of  Utah  and  Nevada.  Jour. 
Mamm.,  18:357-359,  August  14  (with  S.  D.  Durrant). 

97.  Deleterious  eftects  of  preservatives  on  study  specimens  of  mammals.  Jour. 
Mamm.,  18:359-360,  August  14. 

1938 

98.  Fur  and  the  public  domain.  The  trapper  and  sportsman,  1(1):12-14, 
January. 


16  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

99.   A  new  pocket  gopher  from   Nevada.    Proc.    Biol.   Soc.   Washington,   51: 
15-16,  February  18  (with  F.  E.  Durham). 

100.  A  new  weasel  from  Bolivia  and  Peru.  Proc.  Biol.  Soc.  Washington,  51: 
67-68,  March  18. 

101.  Gestation  period  in  the  long-tailed  weasel.  Jour.  Mamm.,  19:249-250, 
May  14. 

102.  Notes  on  the  spotted  skunks  (genus  Spilogale),  with  accounts  of  new  sub- 
species from  Me.xico  and  Costa  Rica.  Ann.  Mag.  Nat.  Hist.,  ser.  2,  1:510- 
515,  May. 

103.  Notes  on  the  meadow  mice  Microtus  niontanus  and  M.  nanus  with  de- 
scription of  a  new  subspecies  from  Colorado.  Proc.  Biol.  Soc.  Washington, 
51:131-134,  August  23. 

104.  Mammals  from  Millard  County,  Utah.  Proc.  Utah  Acad.  Sci.,  Arts  and 
Letters,  15:121-122,  June  (with  D.  H.  Johnson). 

105.  Variation  among  insular  mammals  of  Georgia  Strait,  British  Columbia. 
Amer.  Nat.,  72:453-463,  2  figs.,  September  10. 

106.  Mammals  from  Touchwood  Hills  Saskatchewan.  Canadian  Field-Nat., 
52:108-109,  October. 

107.  A  new  pika  from  southeastern  Idaho  with  notes  on  nearby  subspecies. 
Univ.  California  Publ.  Zool.,  42:335-340,  1  fig.,  October  12  (with  H.  L. 
Bowlus). 

108.  Inyo  screech  owl  at  Fallon,  Nevada.   Condor,  40:259,  November  15. 

109.  Broad-tailed  hummingbird  attracted  to  food  of  red-naped  sapsucker.  Con- 
dor, 40:264,  November  15. 

19.39 

110.  The  spotted  bat  in  Kern  County,  California.  Jour.  Mamm.,  20:103,  Feb- 
ruary 14. 

111.  Deux  sous-especes  nouvelles  du  rongeur  Dipodomijs  ordii  de  I'ouest  des 
Etats-Unis  D'Amerique.  Mammalia.  3:10-16,  1  pi.,  March  (with  S.  D. 
Durrant ) . 

112.  Three  new  pocket  gophers  (mammalian  genus  Thomomys)  from  Wash- 
ington.   Murrelet,  20:3-5,  April  30  (with  W.  W.  Dalquest). 

113.  Extension  of  the  known  geographic  range  of  the  striated  chipmunk  (Ta- 
mias  straitus).    Amer.  Midland  Nat.,  21:766,  May. 

114.  On  the  characters  of  the  pocket  gopher  Thomomys  talpoides  couchi  Gold- 
man.   Murrelet,  20:38-39,  August  (with  W.  W.  Dalquest). 

115.  Joseph  Grinnell — obituary.  Murrelet,  20:46-47,  1  photograph,  August. 

116.  A  new  subspecies  of  beaver  from  Colorado.    Jour.   Mamm.,  20:358-362, 

1  fig.,  August  14  (with  E.  R.  Warren). 

117.  The  grizzly  bear  of  California.  California  Fish  and  Game,  5:237-244,  2 
pis.,  1  fig.,  September. 

118.  Remarks  on  the  primitive  structure  of  Mustek  stolzmanni  with  a  list  of 
the  South  American  species  and  subspecies  of  the  genus  Mustcla.  Physis 
(Revista  de  la  Sociedad  Argentina  de  Ciencias  Naturales),   16:159-178, 

2  maps,  1  pi. 

119.  [Review  of]  Revision  of  the  North  American  ground  squirrels  with  a 
classification  of  the  North  American  Sciuridae.  Saugetierk.,  13:184-188, 
September  1. 

120.  Joseph  Grinnell  [obituary  notice].  Jour.  Wildlife  Mgt.,  3:366-368,  October. 

121.  Una  nueva  especie  de  comadreje  de  Bolivia  y  del  Peru.  Bol.  Mus.  Hist. 
Nat.,  Javier  Prado,  3:95-97,  October  [translation  of  contribution  no.  100]. 

122.  Geographic  races  of  the  kangaroo  rat,  Dipodomys  microps.  Occas.  Papers 
Mus.  Zool..  Louisiana  State  Univ.,  4:47-62,  3  figs.,  November  10  (with 
F.  H.  Dale). 

123.  Joseph  Grinnell  (1877  to  1839) — biographical  notes.  Jour.  Mamm.,  20: 
409-417,  November  14. 

124.  Utah  jumping  mouse  recorded  from  Idaho.    Murrelet,  20:71,  December. 


DuRRANT — Hall  Biography  and  Bibliography  17 

1940 

125.  An  ancient  nesting  site  of  the  white  pelican  in  Nevada.  Condor,  42:87-88, 
1  fig.,  January  19. 

126.  A  new  race  of  beaver  from  Oregon.  Jour.  Mamm.,  21:87-89,  Fel)ruary  14 
(with  S.  G.  Jewett). 

127.  Gifts  of  specimens  to  the  California  Museum  of  Vertebrate  Zoology,  June  1, 
1936,  to  June  30,  1939.    Univ.  California  Press,  pp.  1-21,  March  1. 

128.  Superninnerary  and  missing  teeth  in  wild  mammals  of  the  orders  Insecti- 
vora  and  Carnivora,  with  some  notes  on  disease.  Jour.  Dental  Res.  19:103- 
119,  pis.  1-12,  April. 

129.  The  Pinyon  mouse  (Peromyscus  truei)  in  Nevada,  with  description  of  a 
new  subspecies.  Univ.  CaHfornia  Publ.  Zool.,  42:401-405,  1  fig.,  April  30 
(with  D.  F.  Hofl meister ) . 

130.  Pribilof  fur  seal  on  California  Coast.   California  Fish  and  Game,  26:76-77. 

131.  Transplantation  of  the  Douglas  Ground  Squirrel.  California  Fish  and 
Game,  26:77. 

132.  A  curious  mutation  in  a  coyote  from  Kern  County,  California.  California 
Fish  and  Game,  26:393-395,  2  figs.,  December. 

133.  Capture  of  a  coati  (Nasiia  narica)  in  San  Diego  County,  California.  Cali- 
fornia Fish  and  Game,  26:395,  December. 

134.  A  new  chipmunk  of  the  Eutamias  amoenus  group  from  Nevada.  Proc. 
Biol.  Soc.  Washington,  53:155-156,  December  19  (with  D.  H.  Johnson). 

135.  A  new  race  of  Belding  ground  squirrel  from  Nevada.  Murrelet,  21:59-61, 
1  fig.,  December  20. 

1941 

136.  Freak  antlers  of  mule  deer.  CaHfornia  Fish  and  Game,  27:37-39,  2  figs., 
March  27. 

137.  Deer  has  no  gall  bladder.   Pacific  Rural  Press,  131:289,  April. 

138.  Two  new  kangaroo  mice  from  Utah.  Murrelet,  22:5-7,  April  30  (with 
S.  D.  Durrant). 

139.  New  heteromyid  rodents  from  Nevada.  Proc.  Biol.  Soc.  Washington,  54: 
55-62,  May  20. 

140.  Three  new  mammals  ( Microtus  and  Ochotona)  from  Utah.  Great  Basin 
Nat,  2:105-108,  July  20  (with  L.  C.  Hayward). 

141.  [Re\iew  of]  Distribution  and  variation  in  the  native  sheep  of  North 
America.   Jour.  Mamm.,  22:332,  August  14. 

142.  [Review  of]  Bibliography  of  fossil  vertebrates.  Jour.  Mamm.,  22:333, 
August  14. 

143.  Revision  of  the  rodent  genus  Microdipodops.  Field.  Mus.  Nat.  Hist.,  Zool. 
Ser.,  27:233-277,  8  figs.,  December  8. 

1942 

144.  The  type  specimen  of  Aplodontia  rufa  californica  (Peters).  Murrelet, 
22:4.5-51,  January  20. 

145.  Geographic  variation  in  the  canyon  mouse,  Peromyscus  crinitus.  Jovu. 
Mamm.,  23:51-65,  1  fig.,  February  14  (with  D.  F.  Hoff meister ) . 

146.  [Re\'iew  of]  Mammalia  [being  part  65  of  the  zoology  of  the  Faroes].  Jour. 
Mamm.,  23:100-101,  February  14. 

147.  Joseph  Grinnell,  1877  to  1939.  Boone  and  Crockett  Club,  officers,  by- 
laws, treasurer's  report  and  list  of  members  for  the  years  1940-1941,  pp. 
32-33,  July  1,  1941  [unsigned]. 

148.  A  new  race  of  wood  rat  (Neotoma  lepida).  Univ.  California  Publ.,  Zool., 
46:369-370,  July  3. 

149.  Gestation  period  in  the  fisher  with  recommendations  for  the  animal's  pro- 
tection in  California.  California  Fish  and  Came,  28:143-147,  1  fig., 
August. 


18  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

150.  Fur  bearers  and  tlie  war.  Trans.  7th  N.  Amer.  Wildlife  Conf.,  pp.  472-475, 
and  discussion,  pp.  474-480,  November. 

194.3 

151.  New  genus  of  American  Pliocene  badger,  widi  remarks  on  the  relation- 
ships of  badgers  of  the  northern  hemisphere.  Abstr.  Soc.  Vert.  Paleo., 
pp.  1841-1842. 

152.  [Review  of]  The  Ohio  Recent  mammal  collection  in  the  Cleveland  Museum 
of  Natural  History.    Jour.  Mamm.,  24:105,  February  20. 

153.  [Review  of]  Furred  animals  of  Australia.  Jour.  Mamm.,  24:105-106,  Feb- 
ruary 20. 

154.  Cranial  characters  of  a  dog-coyote  hybrid.  Amer.  Midland  Nat.,  29:371- 
374,  2  figs.,  March. 

155.  Intergradation  \'ersus  hybridization  in  ground  squirrels  of  the  western 
United  States.  Amer.  Midland  Nat.,  29:375-378,  1  fig.,  March. 

156.  Criteria  for  vertebrate  subspecies,  species  and  genera:  the  mammals.  Ann. 
New  York  Acad.  Sci.,  44:141-144,  June  8. 

157.  U.S.  textbooks  for  students  in  Latin  American  universities.  Science,  98: 
15-16,  July  2. 

158.  [Review  of]  Joseph  Grinnell's  philosophy  of  nature.  Audubon  Mag.,  45: 
252-253,  August. 

159.  Oscar  Perry  Silhman — obituary  notice.    Jour.  Mamm.,  24:420,  August  17. 

1944 

160.  Pelicans  of  the  past.    Nature  Mag.,  37:156,  162,  March. 

161.  Four  new  ermines  from  the  islands  of  southeastern  Alaska.  Proc.  Biol. 
Soc.  Washington,  57:35-42,  June  28. 

162.  A  new  genus  of  American  Pliocene  badger,  with  remarks  on  the  relation- 
ships of  badgers  of  the  Northern  Hemisphere.  Publ.  Carnegie  Inst.  Wash- 
ington, 551:9-23,  2  pis.,  2  figs.,  July  18. 

163.  Classification  of  the  ermines  of  eastern  Siberia.  Proc.  California  Acad. 
Sci.,  23:555-560,  1  fig.,  August  22. 

164.  Speciation  in  the  American  genus  Mustek.    Anat.,  Rec,  89:5.50,  August. 

165.  [Review  of]  The  armadillo:  its  relation  to  agriculture  and  game.  Jour. 
Wildlife  Mgt.,  8:342-343,  October. 

1945 

166.  Four  new  ermines  from  the  Pacific  Northwest.  Jour.  Mamm.,  26:78-85, 
February  27. 

167.  Chase  Littlejohn,  1854  to  1943:  observations  by  Littlejohn  on  hunting  sea 
otters.   Jour.  Mamm.,  26:89-91,  February. 

168.  [Review  of]  The  mammals  of  Chile.   Jour.  Mamm.,  26:97-98,  Febmary  27. 

169.  Dental  caries  in  bears.    Trans.  Kansas  Acad.  Sci.,  48:79-8.5,  4  pis. 

170.  Some  mammals  of  Ozark  County,  Missouri.  Jour.  Mamm.,  26:142-145, 
July  13  (with  A.  S.  Leopold). 

171.  A  revised  classification  of  the  American  ermines  with  description  of  a  new 
subspecies  from  the  western  Great  Lakes  region.  Jour.  Mamm.,  26:175- 
182,  1  fig.,  July  13. 

1946 

172.  The  Museum  of  Natural  History,  The  University  of  Kansas.  Misc.  Publ., 
Mus.  Nat.  Hist.,  Univ.  Kansas,  1:1-16,  illustrated,  January  5. 

173.  Mammals  of  Nevada.  Univ.  California  Press,  Berkeley,  .\i  +  710  pp. 
frontispiece,  11  pis.,  485  numbered  figs,  and  54  unnum]:)ered  figs.,  2  charts, 
July  1. 

174.  [Review  of]  The  principles  of  classification  and  a  classification  of  mam- 
mals.   Jour.  Mamm.,  27:287-288,  August  14. 


DuRRANT — Hall  Biography  and  Bibliography  19 

175.  Zoological  subspecies  of  man  at  the  peace  table.  Jour.  Mamm.,  27:358- 
364,  2  figs.,  November  25. 

1947 

176.  [Re\ie\v  of]  Atlas  des  mammiferes  de  France.  Jour.  Mamm.,  28:69,  Feb- 
ruary 17. 

177.  [Re\ie\v  of]  The  California  ground  squirrel.    Ecology,  28:211,  April. 

178.  [Re\ie\v  of]  Catalogue  of  Canadian  Recent  mammals.  Jour.  Mamm.,  28: 
304,  August  19. 

179.  Subspeciation  in  pocket  gophers  of  Kansas.  Univ.  Kansas  Publ.,  Mus. 
Nat.  Hist.,  1:217-236,  2  figs.,  November  29  (with  B.  Villa-R.). 

180  A  new  bat  (genus  Myotis)  from  Mexico.  Univ.  Kansas  Publ.,  Mus.  Nat. 
Hist.,  1:237-244,  6  figs.,  December  10  (with  W.  W.  Dalquest). 

181.  Tadarida  femorosacca  (Merriam)  in  Tamaulipas,  Mexico.  Univ.  Kansas 
Publ,  Mus.  Nat.  Hist.,  1:245-248,  1  fig.,  December  10  (with  W.  W. 
Dalquest ) . 

182.  Obituary  [Ralph  Ellis  (1908-1945)].  Proc.  Linnean  Soc.  London,  159: 
158-159,  December  30. 

183.  Geographic  range  of  the  hairy-legged  vampire  in  eastern  Mexico.  Trans. 
Kansas  Acad.  Sci.,  vol.  50:315-317,  December  30  (with  W.  W.  Dalquest). 

1948 

184.  Pipistrellus  cinnamomeus  rediscovered.  Jour.  Mamm.,  29:180,  May  14 
(with  W.  W.  Dalquest). 

185.  A  new  pocket  gopher  (Thomomys)  and  a  new  spiny  pocket  mouse 
(Liomys)  from  Michoacan,  Mexico.  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
1:249-256,  6  figs.,  July  26  (with  B.  Villa-R.). 

186.  Two  new  meadow  mice  from  Michoacan,  Mexico.  Univ.  Kansas  Publ., 
Mus.  Nat.  Hist.,  1:423-427,  6  figs.,  December  24. 

1949 

187.  [Review  of]  Ecology  of  the  California  groimd  squirrel  on  grazing  lands. 
Ecology,  30:112,  January. 

188.  Paul  Rode:    1901-1948.    Science,  110:51,  July  8. 

189.  Paul  Rode:   1901-1948.    Jour.  Mamm.,  30:341-342,  August  17. 

190.  A  new  subspecies  of  the  cotton  rat,  Sigmodon  hispidus,  from  Michoacan, 
Me.xico.    Proc.  Biol.  Soc.  Washington,  62:149-150,  3  figs.,  August  23. 

191.  A  new  subspecies  of  funnel-eared  bat  (Natalus  mexicanus)  from  eastern 
Mexico.  Proc.  Biol.  Soc.  Washington,  62:153-154,  August  23  (with  W.  W. 
Dalquest ) . 

192.  A  new  harvest  mouse  from  Michoacan,  Mexico.  Proc.  Biol.  Soc.  Wasliing- 
ton,  62:163-164,  August  23. 

193.  Un  nuevo  raton  de  campo  genero  Reithrodontomys  de  Michoacan,  Mex- 
ico. Anuarie  por  1947  de  la  Comision  Impulsora  y  Coordinadora  de  la 
Investigacion  Cientifica,  pp.   173-175,  September   18   (with  B.   Villa-R.). 

194.  Five  bats  new  to  the  known  fauna  of  Mexico.  Jour.  Mamm.,  30:424-427, 
November  14  (with  W.  W.  Dalquest). 

195.  Observaciones  acerca  de  la  preparacion  de  trabajos  cientificos.  Univ.  Nac. 
Mexico,  Publ.  Inst.  Biol.,  44:1-10,  December  3. 

196.  An  annotated  list  of  the  mammals  of  Michoacan,  Mexico.  Univ.  Kansas 
Publ.,  Mus.  Nat.  Hist.,  1:431-472,  2  pis.,  1  fig.,  December  27  (with  B. 
Villa-R.). 

1950 

197.  Geographic  range  of  the  hooded  skunk.  Mephitis  macroura,  with  de- 
scription of  a  new  subspecies  from  Mexico.  Univ.  Kansas  Publ.,  Mus.  Nat. 
Hist.,  1:575-580,  1  fig.,  January  20  (with  W.  W.  Dalquest). 

198.  Pipistrellus  cinnamomeus  Miller  1902  referred  to  the  genus  Myotis.  Univ. 
Kansas  Publ.,  Mus.  Nat.  Hist.,  1:581-590,  5  figs.,  January  20  (with  W.  W. 
Dalquest ) . 


20  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

199.  A  synopsis  of  the  American  bats  of  the  genus  Pipistrelkis.  Univ.  Kansas 
Publ.,  Mus.  Nat.  Hist.,  1:591-602,  1  fig.,  January  20  (with  W.  W.  Dal- 
quest). 

200.  Speciation  in  American  weasels  (genus  Mustek).  Proc.  13th  Internat. 
Cong.  Zool.,  pp.  404-405. 

201.  Lista  anotada  de  los  mammiferos  de  Michoacan,  Mexico.  An.  Inst.  Biol., 
21:159-214,  5  figs.,  September  28  (with  B.  Villa-R.). 

202    State  administration  of  wildhfe,  a  natural  resource.    Trans.  Kansas  Acad. 

Sci.,  53:295-301,  October  3. 
203.  Kansas  wildlife  can  be  saved.    Southwest  Farmer,  Wichita  Beacon,  p.  11, 

1  fig.,  November  5. 

1951 

204  Two  new  pocket  gophers  from  Wyoming  and  Colorado.  Univ.  Kansas 
Pul)l.,  Mus.  Nat.  Hist.,  5:25-32,  February  28   (with  G.   H.   Montague). 

205.  A  northern  record  for  Centronycteris  maxmiliaiii  centralis  with  a  key  to  the 
skulls  of  the  North  American  emballonurine  bats.  An.  Inst.  Biol.,  21:431- 
433,  March  6  (with  W.  W.  Dalquest  and  W.  G.  Frum). 

206.  The  gross  anatomy  of  the  tongues  and  stomachs  of  eight  New  World  bats. 
Trans.  Kansas  Acad.  Sci.,  54:64-72,  34  figs.,  March  17   (with  H.  Park). 

207.  An  instance  of  coyote-dog  hybridization.  Trans.  Kansas  Acad.  Sci.,  54: 
73-77,  4  figs.,  March  17  (with  J.  W.  Bee). 

208.  Mammals  collected  by  Dr.  Curt  von  Wedel  from  the  barrier  beach  of 
Tamauhpas,  Mexico.  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.,  5:33-47,  1  fig., 
October  1. 

209.  Comments  on  the  taxonomy  and  geographic  distribution  of  some  North 
American  rabbits.  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.,  5:49-58,  October  1 
(with  K.  R.  Kelson). 

210.  A  new  subspecies  of  Microtus  montanus  from  Montana  and  comments  on 
Microtus  canicaudus  Miller.  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.,  5:73-79, 
October  1  (with  K.  R.  Kelson). 

211.  A  new  pocket  gopher  from  eastern  Colorado.  Univ.  Kansas  Publ.,  Mus. 
Nat.  Hist.,  5:81-85,  October  1. 

212.  An  hypothesis  to  account  for  the  winter  whitening  of  Arctic  mammals. 
Anat.  Rec.  3:  no  page  number,  November. 

213  A  synopsis  of  the  North  American  Lagomorpha.  Univ.  Kansas  Publ.,  Mus. 
Nat.  Hist.,  5:119-202,  68  figs.,  December  15. 

214  A  new  pocket  gopher  (genus  Thomomys)  from  Wyoming  and  Colorado. 
Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.,  5:219-222,  December  15. 

215.  A  new  name  for  the  Mexican  red  bat.  Univ.  Kansas  Publ.,  Mus.  Nat. 
Hist.,  5:223-226,  December  15. 

216.  In  memoriam.  Charles  Dean  Bunker  1870-1948.  Misc.  Publ.  Mus.  Nat. 
Hist.,  Univ.  Kansas,  3:1-11,  1  fig.,  December  15. 

217.  American  weasels.  Univ.  Kansas  Publ,  Mus.  Nat.  Hist.,  4:1-466,  41  pis., 
31  figs.,  December  27. 

1952 

218  Taxonomic  notes  on  Mexican  bats  of  the  genus  Rhogeessa.  Univ.  Kansas 
Publ.,  Mus.  Nat.  Hist.,  5:227-232,  April  10. 

219  Comments  on  the  taxonomy  and  geographic  distribution  of  North  Ameri- 
can microtines.  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.,  5:293-312,  Novem- 
ber 17  (with  E.  L.  Cockrum). 

220  The  subspecific  status  of  t\vo  Central  American  slodis.  Univ.  Kansas  Publ., 
Mus.  Nat.  Hist,  5:313-317,  November  21  (with  K.  R.  Kelson). 

221.  Comments  on  the  taxonomy  and  geographic  distribution  of  some  North 
American  marsupials,  insectivores  and  carnivores.  Univ.  Kansas  Publ., 
Mus.  Nat.  Hist.,  5:319-341,  December  5  (with  K.  R.  Kelson). 


DuRRANT — Hall  Biography  and  Bibliography  21 

222.  Comments  on  the  taxonomy  and  geographic  distribution  of  some  North 
American  rodents.  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.,  5:343-371,  De- 
cember 15  (with  K.  R.  Kelson). 

1953 

223.  A  synopsis  of  the  North  American  microtine  rodents.  Univ.  Kansas  Publ, 
Mus.  Nat.  Hist.,  5:373-498,  149  figs.,  January  15  (with  E.  L.  Cockrum). 

224.  A  westward  extension  of  known  geographic  range  for  the  timber  rattle- 
snake in  southern  Kansas.    Trans.  Kansas  Acad.  Sci.,  56:89,  March  21. 

225.  Seventeen  species  of  bats  recorded  from  Barro  Colorado  Island,  Panama 
Canal  Zone.  Univ.  Kansas  Publ,  Mus.  Nat.  Hist.,  5:641-646,  December  1 
(with  W.  B.  Jackson). 

1954 

226.  Occurrence  of  the  harbor  porpoise  at  Point  Barrow,  Alaska.  Jour.  Mamni., 
35:122-123,  February  10  (with  J.  W.  Bee). 

227.  A  new  subspecies  of  pocket  mouse  from  Kansas.  Univ.  Kansas  Publ.,  Mus. 
Nat.  Hist.,  7:587-590,  November  15. 

1955 

228.  On  the  legitimacy  of  scientific  authorship.    Science,  121:40-41,  January  7. 

229.  Nuevos  murcielagos  para  la  fauna  Mexicana.  Acta  Zool.  Mexicana,  1(3): 
1-2,  September  10. 

230.  A  new  subspecies  of  wood  rat  from  Nayarit,  Mexico,  with  new  name- 
combinations  for  the  Neotoma  mexicana  group.  Jour.  Washington  Acad. 
Sci.,  45:328-332,  1  fig.,  October  31. 

231.  Handbook  of  mammals  of  Kansas.  Misc.  Publ.  Mus.  Nat.  Hist.,  Univ. 
Kansas,  7:1-303,  illustrated,  December  13. 

1956 

232.  What  goes  on  under  your  farm?  Capper's  Farmer,  67(3):37,  129-131,  1 
colored  two-page  illustration,  February  13. 

233.  Harry  Harris  [obituary  notice].    Burrough  Club  Bull.,  7(4):3-4,  February. 

234.  Future  fishing  and  hunting  in  Kansas.  Kansas  Sportsman,  2(ll):4-5, 
10-11,  February. 

235.  Mammals  of  northern  Alaska.  Misc.  Publ.  Mus.  Nat.  Hist.,  Univ.  Kansas, 
8:1-309,  1  colored  frontispiece,  4  pis.,  127  figs.,  March  10  (with  J.  W. 
Bee). 

236.  Animals  that  keep  house  below  ground  give  nature  a  helping  hand. 
Topeka  Daily  Capital,  p.  13A,  1  fig.,  April  8  [essentially  a  reprint  of 
no.  232  above,  but  without  colored  illustration]. 

237.  Remarks  on  mammalian  ecology  and  germ  warfare.  Pp.  101-102,  in 
Symposium  on  ecology  of  disease  transmission  in  native  mammals,  6  + 
122  pp.,  July  6. 

238.  The  biological  relationships  lietween  American  weasels  (genus  Mustela) 
and  nematodes  of  the  genus  Skrjahingylus  Petrov,  1927  (Nematoda: 
Metastrongylidae),  the  causitive  organisms  of  certain  lesions  in  weasel 
skulls.  Revista  Iberica  de  Parasitologia,  Granada  (Espafia).  Tomo 
Extraordinario,  pp.  531-576,  8  figs.,  August  (with  E.  C.  Dougherty). 

239.  Arctic  to  tropics  in  America.  Proc.  XIV  Internat.  Cong.  Zool.,  Copenhagen, 
p.  125,  December  31. 

240.  Speciation  in  American  microtine  rodents.  Proc.  XIV  Internat.  Cong. 
Zool.,  Copenhagen,  pp.  160-162,  December  31. 

241.  Systematics  of  American  Lagomorpha.  Proc.  XIV  Internat.  Cong.,  Zool., 
Copenhagen,  pp.  521-522,  December  31. 

1957 

242.  Kansas  Natural  Resources  Conference.    Spec.  Rept.,  82:15-19,  April  1. 

243.  Museum  of  Natural  History.  An  unpaged,  illustrated  "flier,"  Mus.  Nat. 
Hist.,  Univ.  Kansas,  June  1  (with  R.  R.  Moore). 


22  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

244.  Vernacular  names  for  North  American  mammals  north  of  Mexico.  Misc. 
Publ.  Mus.  Nat.  Hist.,  Univ.  Kansas,  14:1-16,  June  19  (with  S.  Anderson, 
J.  K.  Jones,  Jr.,  and  R.  L.  Packard). 

245.  [Functions  of  a  museum  of  natinal  history].  Pp.  108-110,  in  Cultural 
Leadership  in  the  Great  Plains — a  report  of  the  Great  Plains  Conference 
on  Higher  Education  held  at  Univ.  Oklahoma,  Norman,  P.  G.  Ruggiers 
(ed. ),  Univ.  Oklahoma  Press,  viii  -|-  191  pp.,  about  November  1. 

1958 

246.  Teachers,  conser\ation  and  native  resources.  Kansas  Sportsman,  4(9): 
4-5,  13,  January  28. 

1959 

247.  Introduction  [Geographic  distri]:)ution  of  contemporary  organisms].  Publ. 
Amer.  Assoc.  Adv.  Sci.,  51:371-373,  January  16. 

248.  Publications,  University  of  Kansas  Museum  of  Natural  History.  Pp.  1-4, 
February  7. 

249.  [Review  of]  A  guide  to  Saskatchewan  mammals.    Blue  Jay,  17:44,  March. 

250.  The  mammals  of  North  America.  Ronald  Press  Co.,  New  York,  l:x.xx  -j- 
546  +  79,  and  2:ix  +  547-1083  +  7.9),  1231  illustrations,  March  31  (with 
K.  R.  Kelson). 

251.  Natural  History  Museum,  University  of  Kansas.  Misc.  Publ.  Mus.  Nat. 
Hist.,  Univ.  Kansas,  19:unpaged,  illustrated  "flier,"  May  29  (with  R.  R. 
Moore ) . 

252.  Additional  evidence  indicating  that  Sorex  hijdrodromns  Dobson  is  a  mem- 
ber of  the  Sorex  aicticus  group  of  shrews.  Pp.  263-265,  lu  Faima  of  the 
Aleutian  Islands  and  Alaska  Peninsula,  bv  O.  J.  Murie,  N.  Amer.  Fauna, 
61:xiv  +  1-364,  November  4. 

1960 

253.  Conspecificity  of  two  pocket  mice,  Perognathus  goldmani  and  P.  artus. 
Univ.  Kansas  Publ,  Mus.  Nat.  Hist.,  9:513-518,  1  map,  January  14  (with 
M.  B.  Ogilvie). 

254.  Small  carnivores  from  San  Josecito  Cave  (Pleistocene),  Nuevo  Leon, 
Mexico.   Univ.  Kansas  Publ,  Mus.  Nat.  Hist.,  9:531-538,  1  fig.,  January  14. 

255.  Elm  spray  peril:  child  safety  [letter  to  the  editor].  Lawrence  Daily  Journal- 
World,  p.  4,  January  14. 

256.  [Review  of]  A  revision  of  American  bats  of  the  genera  Euderma  and 
Plecotus.    Jour.  Mamm.,  41:144-145,  February  20. 

257.  Publications,  University  of  Kansas  Museum  of  Natural  History.  Pp.  1-4, 
May  2  (with  E.  Hudson). 

258.  The  red  fig-eating  bat  Stenoderma  rufiim  Desmarest  found  alive  in  the 
West  Indies.    Mammalia,  24:65-67,  2  figs..  May  27  (with  J.  W.  Bee). 

259.  A  new  subspecies  of  pocket  gopher  (Thomomys  imibrinus)  from  Sinaloa, 
Mexico,  with  comments  on  T.  u.  sinaloae  and  T.  u.  evexus.  Proc.  Biol. 
Soc.  Washington,  73:35-38,  August  10  (witli  C.  A.  Long). 

260.  Zoological  subspecies  of  man.  Mankind  Quart.,  1:113-119,  2  figs.,  October 
31. 

261.  Oryzomys  couesi  only  subspecifically  different  from  the  marsh  rice  rat, 
Oryzomys  palustris.    Southwestern  Nat.,  5:171-173,  November  1. 

262.  ".  .  .  Datna  as  the  generic  name  of  North  American  deer."  Jour.  Mamm., 
41:537-538,  November  11. 

263.  ".  .  .  use  of  the  generic  name  Si)cnnophihis  F.  Cuvier,  1825,  for  Mus  citel- 
his  Linnaeus  and  for  other  ground  squirrels  of  the  same  genus."  Jour. 
Mamm.,  41:538-539,  November  11   (with  others). 

264.  Guide  to  the  panorama  of  North  American  mammals.  Misc.  Publ.  Mus. 
Nat.  Hist.,  Univ.  Kansas,  25:1-31,  illustiated,  December  15. 


DuRRANT — Hall  Biography  and  Bibliography  23 

1961 

265.  Degrees  and  titles.    Science,  133:1630-1631,  May  19. 

266.  Bisoi^  bison  in  Nevada.   Jour.  Mamm.,  42:279-280,  1  fig..  May  20. 

267.  Rare  liooks  in  scientific  investigation.  Books  and  Libraries  at  the  Univer- 
sity of  Kansas,  26:7-8,  May  22. 

268.  A  new  species  of  mouse  (Peromyscus)  from  northwestern  Veracruz,  Mex- 
ico. Proc.  Biol.  Soc.  Washington,  74:203-205,  August  11  (with  T. 
Alvarez ) . 

269.  Prairie  National  Park.  Trans.  Kansas  Acad.  Sci.,  64:265-266,  November  1. 

270.  Dutch  elm  disease  and  DDT  spray.  Kansas  State  Biol.  Surv.,  Service  Leaf- 
let, no.  961-1-12,  1  p.,  December  15. 

271.  A  new  subspecies  of  pocket  gopher  (Heterogeomys)  from  northern  Vera- 
cniz.  An.  Escuela  Nac.  Ciencias  Biol.,  10:121-122,  December  20  (with 
T.  Alvarez). 

272.  A  new  subspecies  of  black  myotis  (bat)  from  eastern  Mexico.  Univ.  Kan- 
sas Publ.,  Mus.  Nat.  Hist.,  14:69-72,  1  fig.,  December  29  (with  T. 
Alvarez ) . 

273.  North  American  yellow  bats,  "Dasypterus,"  and  a  fist  of  the  named  lands 
of  the  genus  Lasiurus  Gray.  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.,  14: 
73-98,  4  figs.,  December  29  (with  J.  K.  Jones,  Jr.). 

1962 

274.  The  Prairie  National  Park.  Natl.  Parks  Mag.,  36(173):4-8,  3  photographs, 
2  figs.,  January  26. 

275.  Weasel.    Encyclopedia  Americana.    1962  ed.,  28:532,  March  15. 

276.  A  new  doglike  carnivore,  genus  Cynarctus,  from  the  Claredonian,  Pliocene, 
of  Texas.  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.,  14:135-138,  2  figs.,  April 
30  (with  W.  W.  Dalquest). 

277.  A  new  bat  (Myotis)  from  Mexico.  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
14:161-164,  1  fig..  May  21. 

278.  Collecting  and  preparing  studv  specimens  of  vertebrates.  Misc.  Publ. 
Mus.  Nat.  Hist.,  Univ.  Kansas,  30:1-46,  ,34  figs..  May  21. 

279.  [Review  of]  International  Code  of  Zoological  Nomenclature.  Jour.  Mamm., 
43:284-286,  May  29. 

280.  Natural  History  Museum,  University  of  Kansas.  Misc.  Publ.  Mus.  Nat. 
Hist.,  Univ.  Kansas,  31:  unpaged,  illustrated  "flier,"  June  1  (with  R.  R. 
Moore ) . 

281.  Two  methods  for  classifying  Recent  mammals  as  species  or  as  subspecies. 
Symposium  Theriologicum,  Czechoslovak.  Acad.  Sci.,  Praha,  pp.  147-152, 
June  30. 

282.  Preserving  the  prairie.   Sierra  Club  Bull.,  47(5):8,  2  photographs,  July  12. 

1963 

283.  ["Comments  and  recommendations"  on  Predator  Control].  Defenders  of 
Wildlife,  Winter  Bull.,  pp.  4-5,  Feliruary  2. 

284.  Document  22/3  ".  .  .  six  (6)  changes  .  .  .  [to]  be  made  in  the  Code  at  the 
XVI  International  Congress."   Bull.  Zool.  Nomenclature,  20:78,  March  18. 

285.  The  mammals  of  Veracruz.  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist,  14:165- 
362,  2  figs..  May  20  (with  W.  W.  Dalquest). 

286.  Statement  at  "Hearing  before  the  Committee  on  Finance,  United  States 
Senate,  eighty-eighth  Congress,  first  session,  on  H.  R.  1839  .  .  .  tariff  act 
...  to  provide  for  the  free  importation  of  wild  animals  and  wild  birds. 
P.  35,  July  23. 

287.  Introduction  of  exotic  species  of  mannnals.  Proc.  XVI  Internat.  Cong. 
Zool.,  1:267,  August  20. 


24  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

288.  Introduction  of  exotic  species  of  mammals.  Quart.  Bull.  Defenders  of 
Wildlife,  pp.  6-7,  October. 

289.  Opinions  [on  wildlife  management].  Lawrence  Daily  Journal-World,  pp. 
1,  14,  3  illustrations,  December  16. 

290.  Introduction  of  exotic  species  of  mammals.  Trans.  Kansas  Acad.  Sci., 
66:516-518,  December  17. 

1964 

291.  Natural  History  Museum,  University  of  Kansas.  Misc.  Publ.  Mus.  Nat. 
Hist.,  Univ.  Kansas,  35:  unpaged,  illustrated  "flier,"  March  2  (with  R.  R. 
Moore ) . 

1965 

292.  Some  support  for  predators.    Pacific  Discovery,  18:31,  1  fig.,  January. 

293.  Government  control  programs  and  the  red  fox  population.  The  Chronicle 
of  the  Horse,  28(47 ):  16-18,  July  16. 

294.  Remarks.  .  .  .  XVI  Internat.  Cong.,  Zool.,  Scientific  use  of  natural  areas 
symposium.  Field  research  projects,  Natural  Area  Studies  2,  pp.  8-11, 
October  15. 

295.  Names  of  species  of  North  American  mammals  north  of  Mexico.  Misc. 
Publ.  Mus.  Nat.  Hist,  Univ.  Kansas,  43:1-16,  December  10. 

1966 

296.  What's  wrong  with  South  Dakota.  Defenders  of  Wildlife  News,  40(5)  :1, 
January  12. 

297.  Annual  Report,  The  State  Biological  Survey  of  Kansas,  July  1,  1964  to 
June  30,  1965.  Pp.  1-15,  5  figs.,  January  20  (with  F.  B.  Cross  of  pp.  1-10; 
pp.  11-15  by  F.  B.  Cross  and  B.  A.  Simco). 

298.  A  plan  for  natural  areas  in  Kansas.  Trans.  Kansas  Acad.  Sci.,  \ol.  69,  pp. 
1-10,  June  15.    (As  one  of  many  authors;  a  committee  effort.) 

299.  [Statement  on  predatory  mammals]  at  "Hearings  before  the  Subcommittee 
on  Fisheries  and  Wildlife  Conservation  of  the  Committee  on  Merchant 
Marine  and  Fisheries,  House  of  Representatixes,  eighty-ninth  Congress, 
second  session  (serial  no.  89-22).    Pp.  89-127,  244-245,  June  20. 

300.  The  endangered  black-footed  ferret.  Defenders  of  Wildlife  News,  41:109, 
1  fig.,  June  24. 

301.  [Statement  on  predatory  mammals,  Subcommittee  on  Fisheries  and  Wild- 
life Conservation,  of  the  Committee  on  Merchant  Marine  and  Fisheries, 
House  of  Representatives].  Defenders  of  Wildlife  News,  41:152-156, 
June  24. 

302.  The  pocket  gopher  in  Kansas.  Misc.  Publ.  Mus.  Nat.  Hist.,  Univ.  Kan- 
sas, 44:1-32,  8  figs.,  October  26  (with  J.  F.  Downhower). 

303.  [Review  of]  The  wolves  of  Isle  Royal.  Natl.  Parks  Mag.,  40:23,  November 
10. 

304.  Carnivores,  sheep  and  public  lands.  Trans.  31st  N.  Amer.  Wildlife  and 
Natural  Resources  Conference,  pp.  239-245,  November  23. 

305.  [Review  of|  The  elk  of  North  America.  Defenders  of  Wildlife  News, 
41:385,  December  14. 

1967 

306.  Two  new  species  of  bats.  Genus  Myotis,  from  Pleistocene  Deposit  in  Texas. 
Amer.  Midland  Nat.,  78:531-534,  October  25  (with  J.  R.  Choate). 

1968 

307.  [Review  of]  The  bats  of  West  Africa.  Jour.  Mamm.,  49:350-351,  May  20. 

308.  [Review  of]  A  place  in  the  sun.  Defenders  of  Wildlife  News,  43:246, 
June  18. 

309.  A  new  subspecies  of  red  fig-eating  bat  from  Puerto  Rico.  Life  Sci.  Occas. 
Papers,  Royal  Ontario  Mus.,  11:1-5,  2  figs.,  July  22  (with  J.  R.  Tamsitt). 


TAXONOMIC  STATUS  OF  THE  WOODRAT, 

NEOTOMA  ALBIGULA,  IN  SOUTHERN 

CHIHUAHUA,  MEXICO 

BY 

Sydney  Anderson 

One  of  the  earliest  specimens  of  mammals  known  to  have  been 
preserved  from  the  state  of  Chihuahua,  Mexico,  was  a  woodrat, 
Neotoma.  It  was  obtained  by  Lt.  D.  N.  Couch  at  Santa  Rosaha, 
now  known  as  Ciudad  Camargo.  The  rat  was  reported  by  Baird 
(1859:44),  who  referred  it  to  Neotoma  micwptis,  which  he  had 
described  in  1855  on  the  basis  of  a  specimen  from  Charco  Escondido 
in  Tamaulipas.  The  Chihuahuan  specimen  consists  only  of  a 
mounted  skin. 

When  Goldman  (1910)  revised  the  genus  Neotoma  he  assigned 
Couch's  specimen  (USNM  561),  along  with  50  other  Chihuahuan 
specimens,  to  Neotoma  alhigula  albigula  Hartley,  1894.  Goldman 
recognized  the  species  Neotoma  micropus  and  mapped  its  range 
from  Colorado  and  Kansas  in  the  north  to  San  Luis  Potosi  in  the 
south  and  from  the  Gulf  of  Mexico  west  to  the  Rio  Grande  Valley 
in  New  Mexico  and  western  Texas.  He  did  not  record  it  from 
Chihuahua.  By  1910,  Neotoma  albigula  and  N.  micropus  each  in- 
cluded several  subspecies.  The  range  of  N.  albigula  lay  mostly  to 
the  west  of  that  of  N.  micropus  but  the  two  overlapped  in  parts  of 
New  Mexico,  Texas,  and  Coahuila.  These  distributions  were  sum- 
marized by  Hall  and  Kelson  (1959).  Baker  (1956)  studied  Neo- 
toma in  Coahuila  and  there  recognized  both  species;  however,  he 
did  not  examine  the  14  specimens  from  four  localities  from  which 
Goldman  had  reported  both  rats. 

I  recently  wrote  a  key  to  include  in  a  faunal  report  on  Chihua- 
huan mammals.  To  my  surprise  specimens  from  southeastern  Chi- 
huahua, presumably  of  the  species  Neotoma  albigula,  were  identi- 
fied as  Neotoma  micropus  when  the  key  was  used  by  non-mam- 
malogists. 

The  faunal  report  has  already  been  delayed  by  the  search  for 
answers  to  other  taxonomic  problems  and  I  do  not  wish  to  digress 
again  at  length,  but  I  will  summarize  the  problem  and  give  some 
conclusions.  The  subject  would  well  repay  additional  study  using 
morphological,  distributional,  and  ecological  evidence.    Karyologi- 

(25) 


26  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

cal,  behavioral,  serological,  and  other  types  of  study  in  the  growing 
methodological  arsenal  of  taxonomy  might  also  be  productively 
employed. 

All  specimens  from  Chihuahua  currently  assigned  to  the  species 
N.  alhigiila,  except  four  specimens  of  Neotoma  olbigiila  melanura 
from  the  southwestern  part  of  the  state,  are  from  the  plateau  east 
of  the  Sierra  Madre.  These  rats  may,  on  broad  ecological  grounds, 
be  supposed  to  have  come  from  a  single  population  or  at  least  a 
more  or  less  continuously  distributed  and  freely  interbreeding  series 
of  local  populations  in  similar  habitats.  Previous  workers  have 
assigned  most  of  these  rats  to  a  single  subspecies.  There  is  no  major 
ecological  barrier  in  this  area  and  the  gaps  that  do  occur  in  the 
known  distribution  are  in  places  where  little  collecting  has  been 
done.  However,  contrary  to  what  might  reasonably  be  supposed, 
the  rats  on  the  plateau  are  not  uniform  in  their  morphological 
characteristics.  After  the  discovery  of  micropm-\ike  characteristics 
in  two  specimens  from  southeastern  Chihuahua,  I  decided  to  look 
more  critically  at  the  other  rats  from  that  area.  I  found  that  many 
characteristics  were  involved,  not  just  the  one  or  two  used  in  the 
key.  The  rats  southeast  of  the  Rio  Conchos  are  morphologically 
intermediate  as  a  population  between  typical  N.  a.  albigida  and 
N.  micropus  and  they  are,  in  some  ways  mentioned  below,  slightly 
closer  to  micropus. 

Among  the  previously  reported  features  that  distinguish  IV. 
albigiila  from  N.  micropus  in  Chihuahua  and  adjacent  states  are: 
(1)  narrower  mesopterygoid  fossa  (about  3.2  mm.  or  less  rather 
than  4  or  more),  not  broadly  excavated  near  posterior  plane  of 
molars  (Goldman,  1910:16);  (2)  relatively  larger  bullae;  (3)  color 
not  pale  slaty  gray  but  usually  darker  and  with  a  yellowish,  brown- 
ish, or  buffy  hue;  (4)  smaller  overall  size,  especially  hind  feet  and 
diameter  of  tail  (Bailey,  1932:171);  (5)  maxillo-vomerine  notch 
present  because  medial  plate  of  vomer  not  posteriorly  extended 
(Finley,  1958:290);  and  (6)  baculum  having  slender  shaft  and  less 
massive  base  (Burt  and  Barkalow,  1942:290).  This  is  not  a  com- 
plete list  and  is  not  documented  with  all  relevant  citations,  but  it 
does  present  the  major  differences.  The  meaning  of  these  terms  and 
the  taxonomic  problems  posed  will  be  clarified  below. 

In  New  Mexico,  Goldman  (1910)  recorded  N.  albigula  from  11 
of  the  17  localities  where  N.  micropus  was  recorded  and  he  wrote 
nothing  suggesting  any  difficulty  of  identification.  In  southeastern 
Colorado,  N.  albigula  warreni  and  N.  micropus  canescens  are  sym- 


Anderson — Taxonomy  of  Chihuahuan  Woodrats  27 

patric,  and  some  problems  of  identification  were  discovered  there 
by  Finley  (1958),  who  conckided  that  certain  specimens  were 
probably  hybrids.  Baily  (1932:171)  noted  that  N.  micropus  "is 
easily  distinguished"  from  N.  alhigula  in  New  Mexico,  and  James 
S.  Findley  informs  me  that  he  has  had  no  difficulty  distinguishing 
the  two  species  in  his  years  of  work  in  that  state. 

In  northern  Chihuahua  three  specimens  with  skins  and  skulls 
and  two  partial  skulls  from  owl  pellets  seem  clearly  referrable  to 
N.  micropus.  Four  localities  are  represented  and  from  three  of 
these  localities  specimens  of  N.  alhigula  were  also  obtained. 

Prior  to  learning  of  the  problem  discussed  here,  I  examined  and 
identified  233  Neotoma  alhigula  from  the  plateau  of  Chihuahua  in 
various  museums.  Most  of  this  material  has  not  been  restudied  and 
the  specimens  are  not  here  listed.  The  specimens  were  probably 
correctly  distinguished  from  two  other  species  of  Neotoma  in 
Chihuahua.  These  are  N.  mexicana,  which  inhabits  the  Sierra 
Madre  (where  N.  alhigula  is  absent)  as  well  as  adjacent  parts  of 
the  range  of  N.  alhigula,  and  N.  goldmani,  which  is  sympatric  with 
N.  alhigula  in  southeastern  Chihuahua.  The  only  other  species  of 
Neotoma  in  Chihuahua  is  N.  micropus,  discussed  below.  All  Chi- 
huahuan records  of  N.  alhigula  and  N.  micropus  are  mapped  in 
Figure  1.  Representative  Chihuahuan  specimens  and  some  from 
other  areas  were  selected  for  study  or  restudy.  These  are  listed 
below.  In  addition  to  these,  a  larger  number  of  other  specimens 
were  reexamined  cursorily  in  connection  with  specific  questions  as 
they  arose  at  The  University  of  Kansas  (KU),  The  American 
Museum  of  Natural  History  (AMNH),  and  the  United  States 
National  Museum  (USNM),  including  the  U.S.  Biological  Survey 
collection. 

Lists  of  Specimens 

Not  all  specimens  examined  are  here  listed,  but  all  those  from  Chihuahua 
will  be  listed  in  a  later  publication.  Three  selected  groups  of  specimens  are 
listed  below. 

Localities  of  the  specimens  plotted  in  Figure  4  and  used  in  the  factor 
analysis  are  as  follows. — Neotoma  alhigula  alhi^uhi  (all  Chihuahuan,  all  KU); 
Ojo  Palomo  Viejo,  73911,  73915,  73916;  Vado  de  Fusiles,  79693,  79696,  79699, 
79703,  79705;  3.5  mi.  N  and  1  mi.  W  San  Francisco,  69991;  1  mi.  E  Samala- 
yuca,  74366;  4  mi.  NW  Chihuahua  City,  69994;  20  mi.  N  Cuahtemoc,  83369. 
Neotoma  alhigula  durangae  (all  except  the  first  are  Chihuahuan  and  at  KU); 
San  Gabriel  (Durango),  AMNH  21185  (the  holotype  of  duran<iae);  5  mi.  NE 
Las  Cruces,  82799;  Sierra  Almagre,  9  mi.  S  Jaco,  56848;  12  mi.  S  Jaco,  56838, 
56840,  56843,  56844;  15  mi.  S  and  3  mi.  E  Jaco,  56847;  5  mi.  E  Parral,  41043, 


28 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


108 


105 


108 


105 


Fig.  1.  Map  of  Cliihuahua  and  vicinity  showing  the  Cliihuahuan  localities  of 
known  occurrence  of  Neotoma  alhigula  (dots)  and  N.  micwpui  (open  circles). 
At  three  localities  both  species  are  known.  Specific  localities  outside  of 
Chihuahua  are  not  plotted.  A  heavy  line  encircles  an  area,  largely  in  die 
Sierra  Madre  Occidental,  from  which  N.  albigiila  is  absent.  Subspecies  are: 
N.  a.  mehmuni  (A),  N.  a.  albigiila  (B),  N.  a.  durangae  (C),  and  N.  a.  rohusta 
(  D ) .  The  first  and  the  last  of  these  sulxspecies  are  not  direct  subjects  of  this 
report.   The  range  of  N.  micioptis  is  shaded.   The  shaded  area  is  also  inhabited 

by  N.  albigula. 


41045,  41046;  Escalon,  83373,  83374.  Neotoma  miciopus  (all  AMNH);  Silver 
City,  Grant  Co.,  New  Mexico,  127129,  127130;  San  Antonio,  Bexar  Co.,  Texas, 
8668,    167865,    167867;   20   mi.    S   Pecos,   Reeves   Co.,   Texas,   56834,   56835; 


Anderson — Taxonomy  of  Chihuahuan  Woodrats  29 

Burnhani  Ranch,  Brewster  Co.,  Texas,  136589;  2  mi.  W  Alpine,  Brewster  Co., 
Texas;  136583;  Fort  Clark,  Kinney  Co.,  Texas,  12697,  12699. 

Specimens  numbered  1  through  22  in  Figure  5  are:  1,  AMNH  2746  (San 
Fernando  de  Preso,  Tamaulipas);  2,  USNM  35551  (Monument  15,  Boundary 
Line,  Chihuahua);  3-8,  USNM  117170,  117171,  117021,  117018,  117019,  and 
117017  (Monclova,  Coahuila);  9,  USNM  117168  (Saltillo,  Coahuila);  10, 
USNM  58072  (Jimulco,  Coahuila);  11,  KU  54812  (2  mi.  N  and  6  mi.  E 
Camargo,  Chihuahua);  12,  KU  83371  (5  km.  S  Jimenez,  Chihuahua);  13,  KU 
86067  ( 1  mi.  E  Julimes,  Chihuahua);  14-15,  AMNH  68755  and  68572  (Soledad 
and  Alvarez,  respectively,  San  Luis  Potosi);  16,  KU  89876  (3  mi.  NE  El  Fuerte, 
Sinaloa);  17,  KU  82809  (San  Francisco  de  Borja,  Chihuahua);  18-21,  KU 
73892,  73886,  73872,  and  73876  (2  mi.  S  and  5  mi.  W  San  Francisco,  Chi- 
huahua); 22,  KU  81487   (8  mi.  NE  Laguna,  Chihuahua). 

Specimens  for  which  measurements  of  bacula  are  plotted  in  Figure  6  are 
as  follows. — Neotoma  albigula  alhigula  (all  Chihuahuan,  all  KU);  Vado  de 
Fusiles,  79696;  2  mi.  S  and  5  mi.  W  San  Francisco,  73878,  73881,  73892,  73895; 
1  mi.  E  Samalayuca,  74366;  11  mi.  NNW  San  Buenaventura,  64269,  64271; 
9  mi.  WSW  San  Buenaventura,  79709.  Neotoma  albigula  durangae  (all 
Chihuahuan,  all  KU);  1  mi.  E  Julimes,  86067;  Sierra  Almagre,  12  mi.  S  Jaco, 
56843,  56845;  2  mi.  W  Jimenez,  85397;  5  mi.  E  Parral,  41046;  Escalon,  83373. 
Neotoma  micropus  (all  KU);  1  mi.  W  Santa  Fe  Airport,  Santa  Fe  Co.,  New 
Mexico,  52353;  24  mi.  E  Carlsbad,  Eddy  Co.,  New  Mexico,  100630;  La  Gloria, 
Nuevo  Leon,  49594. 

Specimens  used  to  derive  the  statistics  in  Table  2  are:  Neotoma  albigula 
albigula,  12  specimens  (all  KU )  from  within  six  miles  of  the  headquarters  of 
the  Rancho  San  Francisco  in  northwestern  Chihuahua;  Neotoma  albigula 
durangae,  seven  specimens  ( all  KU )  from  the  Sierra  Almagre  in  southeastern 
Chihuahua;  and  Neotoma  micropus,  a  composite  series  including  those  listed 
above  from  at  or  near  Pecos,  Burnham  Ranch,  Alpine,  Ft.  Clark,  and  San 
Antonio,  in  Texas,  Silver  City  in  New  Mexico,  and  Monument  15  in  Chihuahua. 

Methods 

By  means  of  the  craniometric  instrument  described  by  Anderson  (1968),  a 
nimiber  of  measurements  of  skulls  (Fig.  2)  and  bacula  (Fig.  3)  of  Neotoma 
were  taken.  Cranial  measurements  were  recorded  to  the  nearest  twentietli  of 
a  millimeter.  No  greater  accuracy  should  be  inferred  from  the  fact  that  decimal 
numbers  were  used.  For  example,  the  citation  of  a  measurement  of  6.35  means 
only  that  the  value  was  nearer  to  6.35  than  to  6.30  or  6.40. 

Positions  or  coordinates  of  36  end  points  shown  in  Figure  2  were  determined 
for  each  skull.  These  positions  were  recorded  in  six  sets:  (1)  from  posterior  to 
anterior  in  \entral  \iew;  (2)  from  midsagittal  plane  laterally  in  ventral  view; 
(3)  from  posterior  to  anterior  in  dorsal  \iew;  (4)  from  midsagittal  plane 
laterally  in  dorsal  vie\\';  (5)  from  plane  of  bullae  and  incisor  tips  dorsally; 
and  (6)  across  the  zygomatic  plate.  The  last  set  consists  of  only  two  points, 
which  are  on  a  line  oblique  to  all  of  the  planes  used  in  the  other  five  sets. 

When  a  coordinate  was  defined  by  two  bilateral  points  and  the  points  did 
not  lie  on  a  single  Hue  perpendicular  to  the  sagittal  reference  plane  tlie  recorded 
coordinate  is  that  of  a  line  midway  between  the  two  different  lines.  One  of  the 
first  things  concluded  when  skulls  are  observed  imder  a  grid  is  that  no  skull  is 


30 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Fig.  2.  Skull  of  Neototua  aJbiguIa  duiangae  (AMNH  188729)  showing  meas- 
urements described  in  the  text.  Coordinates  of  the  numbered  end  points  were 
recorded  and  the  measurements  designated  by  letters  were  calculated.  Meas- 
urements taken  laterally  from  the  midsagittal  plane  were  doubled.  The  crani- 
ometer  used  was  described  earher  (Anderson,  1968). 

exactly  symmetrical  bilaterally.  The  average  discrepancy  varies  betsveen  dif- 
ferent bilateral  points.  Studies  of  liilateral  \'ariation  as  such  are  of  interest  (see, 
for  example,  Van  Valen,  1962)  Init  the  matter  was  not  pursued  farther  here. 
When  a  bilateral  end  point  was  absent  on  one  side,  the  point  on  the  other  side 
was  used  alone.  This  is  an  advantage  of  the  present  system  of  measurement 
for  in  some  cases  complete  sets  of  measurements  are  obtainable  from  rather 
badly  damaged  skulls. 

Some  of  the  more  critical  sources  of  variability  in  measurements,  but  cer- 
tainly not  the  only  sources,  are  (1)  indistinctness  of  end  points  in  some  cases, 
(2)  differences  in  the  alignment  of  the  skull  prior  to  beginning  a  set  of 
measurements,  (3)  damage  that  occurred  while  the  animal  was  aHve  and 
which  altered  growth  in  some  way,  and  (4)  distortion  or  damage  resulting 
from  treatment  or  preparation  of  the  skull.  The  grosser  examples  of  such 
\ariation  are  recognizable  to  the  careful  observer  and  can  be  excluded  or 
sometimes  compensated  for,  but  the  lesser  examples  are  not  recognizable. 
For  example,  the  bullae  are  not  fused  with  the  other  cranial  elements  and 
excessive  or  even  moderate  maceration  or  pressure  in  cleaning  or  drying  may 
slightly  displace  them,  or  the  incisor  teeth  may  loosen  enough  to  be  moved 
in  or  out. 


Anderson — Taxonomy  of  Chihuahuan  Woodrats  31 

The  reference  points  or  end  points  from  which  measurements  were  derived 
by  sul:)traction  are  listed  below  along  with  pertinent  comments  on  some. 
(1)  The  posterior  point  of  occipital  condyle.  (2)  The  posterior  point  of 
t}'mpanic  bulla.  ( 3 )  The  anterior  point  of  anterolateral  bullar  process.  This 
process  seems  to  have  a  greater  coefficient  of  variability  than  the  overall  size 
of  the  bulla  itself,  which  the  reference  point  was  originally  selected  to  help 
measure.  ( 4 )  The  posterior  point  of  zygomatic  aperture.  There  is  considerable 
bilateral  \  ariation  in  the  anteroposterior  location  of  this  point.  ( 5 )  The  anterior 
margin  of  internal  narial  opening.  (6)  The  posterior  point  of  alveolar  margin 
of  last  molar.  ( 7 )  The  anterior  point  of  alveolar  margin  of  first  molar.  ( 8 )  The 
posterior  point  of  incisive  foramen.  ( 9 )  The  anterior  point  of  zygomatic  aper- 
ture. (10)  The  anterior  point  of  incisive  foramen.  (11)  The  anterior  point  of 
incisor  at  its  alveolus.  ( 12 )  The  anterior  point  of  the  nasal.  This  point  was 
not  used  further  in  this  study  but  is  mentioned  here  because  it  is  part  of  the 
set  of  measurements  used  elsewhere.  (13)  The  midsagittal  plane,  determined 
visually  as  the  average  of  a  number  of  discernible  midpoints.  (14)  The  lateral 
point  of  mesopterygoid  fossa.  The  point  is  indistinct  when  the  slope  of  the  bone 
is  gradual  but  it  may  be  quite  distinct  in  other  cases.  ( 15)  The  medialmost  point 
on  aUeolar  margin  of  any  molar,  usually  the  first  but  sometimes  the  second. 
(16)  The  medial  point  on  constriction  of  postdental  shelf.  (17)  The  lateral 
point  on  posterior  rostral  bulge.  ( 18)  The  lateral  point  on  crown  of  first  molar. 
Wear  of  teeth  alters  this  point.  As  wear  proceeds  beyond  the  widest  point  of 
the  tooth  the  tooth  becomes  narrower.  (19)  The  lateral  point  of  paroccipital 
process.  (20)  The  lateral  point  on  zygomatic  arch  in  a  plane  perpendicular 
to  longitudinal  reference  axis  and  passing  through  posterior  point  of  occlusal 
surface  of  first  molar.  The  longitudinal  reference  axis  is  defined  as  the  line  of 
intersection  of  the  midsagittal  plane  and  a  plane  through  the  end  points  of  a 
standard  condylobasilar  measurement.  These  are  a  posterior  point  of  the 
condyle  and  the  posterior  alveolar  margin  of  the  incisor.  The  axis  is  not  shown 
in  Figure  2.  (21)  The  lateralmost  point  on  zygomatic  arch,  usually  near 
frontojugal  junction.  (22)  The  posterior  point  of  supraoccipital  bulge.  (23)  A 
point  half  way  between  the  anterior  points  of  the  two  anteriormost  processes 
on  interparietal  margin.  This  reduces  the  influence  of  a  single  extremely  long 
process,  either  lateral  or  medial.  (24)  The  posterior  point  of  premaxillary 
process.  (25)  The  posterior  point  of  nasal.  (26)  The  posterior  point  on 
margin  of  anterior  zygomatic  notch.  (27)  The  anterior  point  of  nasal.  (28) 
The  midsagittal  plane.  (29)  The  medial  point  on  margin  of  interorbital  con- 
striction. (30)  The  lateral  point  of  bulge,  if  present,  of  the  braincase.  In  most 
Neotoma  the  posterior  root  of  the  zygoma  arises  laterally  and  obscures  the  point. 
The  reading  is  then  made  to  a  visually  approximated  point  just  dorsal  to  the 
posterior  flange  of  the  zygomatic  process  and  excluding  the  flange.  (31)  A 
plane  through  tips  of  incisors  and  ventral  points  of  bullae.  This  plane  is  not 
parallel  to  the  longitudinal  reference  axis  mentioned  under  the  twentieth 
reference  point  above  (see  Fig.  2  where  this  plane  but  not  the  reference 
axis  is  shown).  (32)  The  dorsalmost  visible  point  on  posterior  margin  of 
incisor.  (33)  The  dorsalmost  visible  point  of  bulla.  (34)  A  plane  tangential 
to  dorsal  profile  of  skull  and  parallel  to  plane  of  reference,  namely  nimiber  31. 
(35)  The  posterior  margin  of  zygomatic  plate  when  skull  is  oriented  obliquely 
so  that  plate  is  perpendicular  to  axis  of  view.  (36)  The  anterior  margin  of 
zygomatic  plate  as  viewed  for  35. 


32 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Fig.  3.  Baculuni  showing  measurements  descril:)ed  in  text.  The  specimen  is 
KU  41046,  a  relati\ely  old  indi\idual  from  5  mi.  E  of  Parral,  Chihuahua,  of 
the  subspecies  Neotoma  albigula  durangac.  The  length  of  this  specimen  is  7.65 
The  dorsal  view  is  at  left,  the  view  from  the  left  side  is  shown  at  right. 


mm. 


Measurements,  which  were  derived  by  subtracting  the  coordinates  of 
various  of  the  above-listed  end  points  ( and  doubling  measurements  lateral  from 
the  midsagittal  plane),  are  as  follows  (letters  shown  in  Fig.  2):  (A)  length  of 
I)ulla;  (B)  aheolar  length  of  molar  toothrow;  (C)  length  of  palate;  (D)  length 
of  zygomatic  aperture;  (E)  length  of  incisive  foramina;  (F)  condyloincisive 
length;  (G)  breadth  of  mesopterygoid  fossa;  (H)  breadth  of  palate;  (I)  post- 
dental  breadth;  (J)  rostral  breadth;  (K)  breadth  of  first  upper  molar;  (L)  ex- 
occipital  breadth;  (M)  anterior  zygomatic  breadth;  (N)  posterior  zygomatic 
breadth;  (O)  occipitoparietal  length;  (P)  naso-premaxillary  difference;  (Q)  oc- 
cipitonasal  length;  (R)  nasal  length;  (S)  rostral  length;  (T)  interorbital 
breadth;  (U)  breadth  of  braincase;  (V)  exposed  length  of  incisor;  (W)  depth 
of  bullae;  (X)  depth  of  skull;  (Y)  breadth  of  zygomatic  plate. 

Measurements  of  bacula  were  read  on  an  eyepiece  scale  with  100  units 
imder  a  binocular  microscope.  With  the  optical  system  used,  100  units  equalled 
8.3  mm.  If  end  points  \\'ere  read  correctly  to  the  nearest  unit  of  the  scale,  the 
measurements  are  within  0.0415  mm.  of  the  actual  \-alues.  However,  because 
of  the  difficulty  of  aligning  a  slippery  object  floating  in  glycerin  and  the  reduc- 
tion in  resolution  resulting  from  the  passage  of  light  from  the  baculum  through 
cleared  tissues  and  glycerin,  measurements  are  not  that  accurate.  However, 
more  than  90  per  cent  of  the  measurements,  other  than  those  of  dimension  B 
described  below,  probably  are  correct  to  the  nearest  tenth  of  a  millimeter. 


Anderson — Taxonomy  of  Chihuahuan  Woodrats  33 

The  measurements  of  bacula  are:  (A)  total  length;  (B)  distance  from 
proximal  margin  to  point  of  greatest  width;  (C)  greatest  width;  (D)  breadth 
at  midpoint;  (E)  depth  of  base,  or  dorsoventral  distance  bet^veen  two  parallel 
planes  tangential  to  the  bone  as  shown  in  Figure  3;  (F)  depth  of  dorsal 
cur\  ature.  The  position  on  an  anteroposterior  axis  of  the  place  of  greatest  width 
is  in  some  cases  more  indefinite  than  any  of  the  other  reference  points  or 
planes  and  the  second  measurement  (B)  is  consequendy  less  accurate  than 
the  other  measurements. 

Comparisons 

Samples  of  N.  albigtila  alhigula  from  northern  Chihuahua  were 
selected.  Other  samples,  to  which  the  name  N.  albigiila  chirangae 
will  be  applied  here,  are  from  southeastern  Chihuahua  and  repre- 
sent the  population  of  questionable  taxonomic  status.  A  sample 
representing  N.  micropus  was  drawn  from  a  large  region  extending 
from  central  New  Mexico  to  San  Antonio,  Texas.  This  sample  there- 
fore represents  a  much  larger  geographic  area  than  the  other 
samples. 

Variation  within  the  Chihuahuan  part  of  the  range  of  N.  a.  albig- 
tila and  that  of  N.  a.  chirangae  was  assessed  by  comparing  means  of 
samples  from  different  local  areas.  Two  samples  of  N.  a.  albigula 
were  compared  and  tliree  samples  of  N.  a.  chirangae  were  compared. 

Means  of  25  cranial  and  five  external  measurements  of  four  males 
and  eight  females  from  the  vicinity  of  San  Francisco  in  northwestern 
Chihuahua  were  not  found  to  be  significantly  difi^erent  (P<0.05, 
Student's  /-test),  therefore  the  sexes  were  combined  for  later  calcu- 
lations. A  series  of  12  from  the  \dcinity  of  San  Francisco  was  com- 
pared with  a  series  of  eight  from  near  Ojo  Palomo  Viejo  in  the  above 
30  measurements.  Three  measurements  were  found  to  be  signifi- 
cantly different:  breadth  of  braincase  and  breadth  of  mesoptery- 
goid  fossa  were  greater  and  depth  of  the  bulla  was  less  in  the  former 
series. 

A  comparison  of  the  means  for  25  cranial  measurements  and  five 
external  measurements  between  two  series  of  N.  a.  chirangae,  one 
from  the  vicinity  of  Parral  and  one  from  Escalon,  and  between  the 
latter  series  and  a  series  from  the  Sierra  Almagre  in  extreme  eastern 
Chihuahua,  revealed  only  two  statistically  significant  differences. 
The  alveolar  length  of  the  maxillary  toothrow  in  the  small  series  of 
three  from  Escalon  was  larger  (P<0.05,  Student's  f-test)  than  in 
the  series  of  seven  from  the  Sierra  Almagre,  and  the  tail  was  longer 
in  the  series  from  the  Sierra  Almagre  than  in  the  series  from  Escalon. 
The  latter  difference  may  be  the  result  of  measurements  by  different 
collectors  rather  than  a  real  difference  in  the  rats. 


34 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


< 

uj       4 
cc 

CD 


o 
(J 

>- 
cc 

UJ 

I- 

Q. 
O 
C/1 


3  — 


(D 


(ED 

Ov 


N.  a.  durangae 


N.  0-  albigula 


N.  micropus 


19  20  21  22  23  24 

ANTERIOR   ZYGOMATIC   BREADTH 


25 


Fig.   4.    Graph  of  two  cranial  measurements  of  individuals   of  three   taxa  of 

Neotoma  as  labelled. 


When  series  of  N.  a.  albigula  (]2  from  near  San  Francisco)  and 
N.  a.  durangae  (seven  from  the  Sierra  Ahnagre)  were  compared, 
using  means  of  30  measurements,  the  differences  were  significant 
(P<0.05,  Student's  f-test)  in  11  cranial  measurements  and  in  the 
length  of  the  ear.  The  means  of  the  former  sample  were  less  than 
those  of  the  latter  in  all  of  these  measurements  except  the  alveolar 
length  of  the  maxillary  toothrow. 

The  sample  oi  N.  a.  albigula  from  the  Rancho  San  Francisco  and 
the  sample  of  N.  a.  durangae  from  the  Sierra  Almagre  were  then 
compared  with  a  sample  of  N.  micropus,  again  using  the  95  per  cent 
confidence  level  in  the  Student's  t-test  to  compare  means.  Two  or 
three  significant  differences  among  30  measurements  had  been 
detected  in  the  comparisons  of  samples  within  each  of  the  two  sub- 
species of  N.  albigula,  and  11  significant  differences  were  present 
between  samples  of  the  two  subspecies.  Further  comparisons  then 
revealed  eight  significant  difi^erences  between  N.  micropus  and 
N.  a.  durangae  and  15  differences  between  N.  micropus  and  N.  a. 
albigula. 

In  terms  of  these  data  N.  a.  durangae  is  nearer  to  N.  micropus 
than  to  N.  a.  albigula,  and  the  possibility  is  raised  that  durangae 
may  have  been  assigned  to  the  wrong  species.  Other  possibilities 
must  also  be  considered.  Could  durangae  be  a  population  of  inter- 
grades  and  there  be  only  one  species  rather  than  two?    Could  the 


Anderson — Taxonomy  of  Chihuahuan  Woodrats  35 

large  numbers  of  significant  differences  among  the  30  measurements 
reflect  nothing  more  than  a  general  diflerence  in  size?  Could  the 
30  measurements  be  largely  irrelevant  because  diagnostic  features 
were  not  adequately  represented?  Testing  of  these  hypotheses  and 
others  requires  either  new  evidence  or  new  analysis  of  existing  evi- 
dence, or  both. 

To  further  compare  Chihuahuan  samples  of  N.  alhigiila,  the 
method  of  analysis  described  by  Lidicker  (  f962:164)  was  employed. 
The  same  samples  mentioned  above  from  the  vicinity  of  five  locali- 
ties in  Chihuahua  were  used — San  Francisco  (A)  and  Ojo  Palomo 
Viejo  (B),  both  in  northern  Chihuahua,  and  Sierra  Almagre  (C), 
Parral  (D),  and  Escalon  (E)  in  the  southern  part  of  the  state. 
Cranial  dimensions  and  external  measurements  were  used,  and  the 
"minimum  significant  difference"  was  twice  the  sum  of  the  calcu- 
lated standard  errors  of  the  means  for  the  two  samples,  rather  than 
the  estimate  used  by  Lidicker.  Only  measurements  significantly 
different  at  the  95  per  cent  level  as  shown  by  the  Student's  f-test 
were  considered.  Color  was  not  used.  Sample  A  was  compared 
with  B,  C,  and  D;  B  with  C;  C  with  D  and  E;  and  D  with  E.  The 
"Total  Differentiation"  and  "Index  of  Differentiation"  taken  together 
as  graphically  shown  by  Lidicker  (op.  c/f.:165)  were  low  on  his 
scale.  The  rank  on  his  nine  unit  scale  was  in  the  lower  two  units  in 
comparisons  between  the  two  samples  \\'ithin  northern  Chihuahua 
and  among  the  three  samples  \\'ithin  southern  Chihuahua.  The  rank 
of  the  three  comparisons  between  northern  and  southern  samples 
was  in  the  third  unit  of  Lidicker's  scale. 

The  relatively  greater  difference,  already  noted  above,  between 
N.  a.  aJbigida  (the  northern  samples)  and  N.  a.  diirangae  (the 
southern  samples)  than  between  samples  within  either  subspecies 
is  again  shown.  Of  the  29  measurements  ( a  thirtieth,  length  of  ear, 
was  not  used  here)  in  which  differences  might  have  occurred,  only 
14  were  significantly  different  in  at  least  one  pair  of  samples.  As 
pre\'iously  noted,  only  two  dimensions  differed  between  samples  of 
N.  a.  diimngae,  and  three  between  the  two  samples  of  N.  a.  aJbigida. 
Thirteen  measurements  were  involved  in  the  25  differences  found 
in  inter-subspecies  comparisons  of  samples,  which  indicates  some 
concordance. 

Some  of  the  30  measurements  taken  are  relevant  to  tliree  of  the 
six  reportedly  diagnostic  characters.  Size  in  general  influences  most 
measurements.  Two  measurements  of  the  bullae  are  included,  and 
the  breadth  of  the  mesopterygoid  fossa  is  included. 


36  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

I  examined  single  measurements  or  pairs  of  measurements  and 
found  that  N.  aJbigida  albigido  could  be  distinguished  from  N. 
micropus  with  considerable  success.  Then  I  attempted  to  decide 
which  of  these  two  taxa  was  phenetically  nearest  to  N.  a.  durangae. 
This  was  less  successful  because  dumngae  was  inter)nediate  between 
N.  a.  alhigida  and  iV.  micropus  in  all  characters  individually  exam- 
ined. I  drew  two-dimensional  graphs  of  the  most  diagnostic  meas- 
urements. The  intermediacy  of  N.  a.  durangae  and  its  partial  over- 
lap with  both  N.  a.  alhigida  and  IV.  micropus  was  always  evident 
(see  Fig.  4,  for  example).   Further  analysis  seemed  advisable. 

Factor  Analysis 

Certain  methods  of  multivariate  analysis  are  useful  in  attempting 
to  understand  situations  such  as  the  present  one  in  which  a  number 
of  more  or  less  correlated  variables  are  involved. 

Factor  analysis  is  one  such  method,  in  which  a  large  number  of 
correlated  variables  are  reduced  to  a  smaller  number  of  uncorre- 
lated  variables.  The  correlated  variables  are  the  measurements. 
The  uncorrelated  variables  are  the  "factors."  These  are  abstract 
hypothetical  components,  which  then  need  interpretation  in  bio- 
logical terms. 

For  those  who  want  to  see  some  other  recent  biological  uses  of 
factor  analysis,  the  following  are  of  interest:  Gould  (1967)  on  the 
evolution  of  pelycosaurs;  Wallace  and  Bader  (1967)  on  dentitions 
of  Mus\  Brown  et  at.  (1965)  on  human  cranial  proportions;  and 
Lawrence  and  Bossert  ( 1967 )  on  species  of  canids 

A  set  of  data  was  subjected  to  factor  analysis  by  D.  Vincent 
Manson  using  his  Multivariate  Statistical  (MUST)  program  (Man- 
son,  1967).  Computation  required  three  minutes  on  an  IBM  7094 
computer.  The  data  matrix  consisted  of  25  cranial  measurements 
of  each  of  36  individuals — 12  N.  a.  alhigula  from  several  locahties  in 
northwestern  Chihuahua,  13  N.  a.  durangae  from  several  localities 
in  southeastern  Chihuahua  and  Durango  (one  specimen  only,  the 
holotype  of  N.  a.  durangae),  and  UN.  micropus  from  several 
localities  in  New  Mexico  and  Texas. 

The  first  factor,  which  may  be  termed  the  "rat  factor,"  was 
dominant  in  all  specimens.  It  shows  a  clear  general  homogeneity 
in  the  measurements  of  the  mixed  sample  of  skulls  but  is  of  no  use 
in  discriminating  among  the  subgroups.  The  second  and  third  fac- 
tors, however,  are  useful  for  this  purpose,  and,  in  fact,  provide 
(Fig.  5)    a  better  separation  of  the  three   subsamples  than  any 


Anderson — Taxonomy  of  Chihuahuan  Woodrats 


37 


CNJ 

cr 
o 


.03 


.02 


.01 


o 

<     .00 


-.01 


-.02 


\ N.  micropus 


•^ 


20O 


FACTOR   3 

Fig.  5.  Graph  showing  relationships  of  individuals  of  Neotoma  as  represented 
by  hvo  factors  for  each  specimen.  Unnumbered  symbols  represent  specimens 
used  in  the  original  factor  analysis.  Numbered  symbols  are  of  additional 
specimens  for  comparison  as  described  in  te.xt.  Encircling  lines  are  drawn  by 
eye  to  make  it  easier  to  see  the  distributions  of  the  original  specimens  of  the 
three  taxa  labeled  on  the  graph.  Symbols  identify  all  except  three  specimens. 
These  three  are   14  and  15  (Neotoma  albif^tila  leucodon),  and   16  (Neotoma 

albigula  melanura). 

simpler  subset  of  data  such  as  the  graph  in  Figure  4.  The  inter- 
mediacy  of  N.  a.  dtirangae  and  its  overlap  with  N.  a.  albigula  and 
N.  micropus  were  previously  surmised  and  these  remain  evident. 
The  factor  analvsis  thus  extended  and  refined  the  earlier  results. 
The  interpretation  of  factors  2  and  3  is  facilitated  on  examination  of 
the  matrix  of  factor  components  ( Table  1  shows  the  composition  of 
the  factors  in  terms  of  proportional  contribution  of  each  of  the  25 
cranial  measurements).  For  example,  the  measurements  with 
greater  absolute  component  values  for  factor  2  are  those  that  will 
best  serve  to  distinguish  IV.  micropus  from  N.  a.  albigula,  and  they 
thus  provide  a  comparison  with  diagnostic  measurements  previously 


38 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Table  1. — Factor  components  for  first  three  factors  and  25  measure- 
ments (listed  by  letter  as  in  text). 


Factor  1 

Factor  2 

Factor  3 

Measurement 

components 

components 

components 

A 

0.08722 

0.01964 

0.04011 

B 

0.10493 

0.17303 

0.09230 

C 

0.09293 

—0.34159 

0.00964 

D 

0.15819 

0.08160 

0.24123 

E 

0.11445 

0.20428 

—0.02983 

F 

0.51629 

0.06707 

0.13103 

G 

0.03879 

0.35224 

0.19642 

H 

0.03628 

0.07745 

0.05867 

I 

0.08381 

0.04383 

0.00168 

J 

0.09223 

0.12933 

0.06402 

K 

0.03251 

0.04122 

0.00121 

L 

0.14823 

—0.02236 

0.44917 

M 

0.26318 

0.47117 

0.12678 

N 

0.28885 

0.38384 

0.21413 

O 

0.08443 

0.02250 

0.28622 

P 

0.03001 

0.14832 

0.28972 

Q 

0.54472 

0.16662 

—0.15326 

R 

0.19234 

0.05768 

0.48463 

S 

0.17175 

0.02826 

0.14695 

T 

0.07060 

0.12556 

0.16935 

U 

0.21438 

0.40223 

0.33691 

V 

0.06039 

0.16584 

0.06192 

W 

0.06960 

0.13902 

0.06171 

X 

0.20122 

—0.03247 

0.07969 

Y 

0.05319 

0.00249 

0.04169 

ascertained  by  other  methods.  Diagnostically,  the  best  five  measure- 
ments for  factor  2  (factor  components  given  in  Table  1)  are 
anterior  zygomatic  breadth,  mesopterygoid  breadth,  posterior  2ygo- 
matic  breadth,  breadth  of  the  braincase,  and  length  of  the  palate. 
Proportional  differences  involving  these  measurements  are  implied, 
because  absolute  size  as  such  was  eliminated  in  the  analysis  by 
normalizing  the  row  (or  specimen)  vectors.  There  are,  of  course, 
mean  differences  in  absolute  size  between  the  three  subsamples. 
These  differences  were  readily  evaluated  by  comparing  the  sums  of 
squares  for  the  normalized  data  matri.x  in  the  computer  output  or 
are  evident  in  the  data  of  Table  2.  IV.  microptis  is  largest  and  N.  a. 
durongae  is  intermediate.  There  is  some  oxerlap  between  N.  mi- 
cropus  and  N.  a.  alhi^uh.  Some  of  this  overlap  results  from  the 
inclusion  of  rats  of  different  ages  in  the  samples.  All  were  "adults" 
but  differences  in  size,  toothwear,  and  amount  of  fusion  between 
cranial  elements  all  suggest  that  were  larger  series  available  the 
range  of  inferred  ages  or  relatixe  ages  in  the  selected  sample  of 
"adults"  could  be  justifiably  reduced  further. 

Factor  3  suggests  that  IV.  a.  diirangae  tends  to  have  some  char- 
acteristics in  which  it  differs  from  both  N.  a.  alhigula  and  N.  mi- 


Anderson — Taxonomy  of  Chihuahuan  Woodrats  39 

cropus,  but  the  tendency  is  weak.  The  subsample  of  N.  a.  durangae 
has  higher  factor  coefficients  on  the  a\'erage  for  factor  3  than  do 
N.  micropus  or  N.  a.  albigula.  The  most  important  measurements 
(those  with  larger  factor  components  as  shown  in  Table  1)  con- 
tributing to  factor  3  are  length  of  nasal  bones,  exoccipital  breadth, 
and  breadth  of  braincase,  which  as  akeady  noted  also  contributes 
importantly  to  factor  2.  The  way  in  which  these  three  measure- 
ments contribute  to  distinguishing  N.  a.  durangae  from  IV.  a.  albig- 
ida  and  N.  micropus  is  not  visualized  simply.  The  measurements 
cannot  be  used  singly  or  in  pairs  to  distinguish  individuals  of  N.  a. 
durangae  from  those  of  the  other  two  kinds.  The  tendency  for 
uniqueness  in  N.  a.  durangae  among  the  three  kinds  is  much  less 
than  the  tendency  for  intermediacy  shown  in  many  individual 
measurements  and  in  factor  2. 

The  two  measurements  of  the  auditory  bullae  have  low  factor 
components  for  all  three  factors.  The  alleged  usefulness  of  bullar 
size  is  therefore  not  verified. 

Factor  analysis  in  this  case  contributes  the  following:  it  verifies 
the  usefulness  of  most  of  the  previously  suggested  diagnostic  char- 
acters; it  discounts  the  usefulness  of  bullar  size;  it  indicates  that  the 
intermediacy  of  N.  a.  durangae  involves  much  more  than  size  alone. 
The  measurements  that  are  shown  to  be  most  useful  by  factor  analy- 
sis are,  in  general,  those  regularly  used  by  systematic  mammalogists, 
and  the  measurements  not  often  used  are  shown  to  be  less  im- 
portant for  diagnostic  purposes.  This  is  not  an  argument  for  con- 
ventionality or  against  a  careful  search  for  useful  new  characters, 
but  it  does  indicate  that  seemingly  "subjective"  methods  are  some- 
times "objectively"  verifiable. 

A  new  specimen  may  be  compared  with  the  results  of  the  present 
factor  analysis,  as  follows:  record  the  25  measurements  described; 
obtain  the  sum  of  the  squares  of  these  measurements  and  extract  its 
square  root;  take  the  first  measurement,  di\'ide  it  by  the  square  root 
noted,  and  multiply  the  quotient  by  the  appropriate  factor  com- 
ponent measurement  (for  example  the  first  value  in  the  column  for 
factor  2  in  Table  1);  do  the  same  for  each  measurement  in  turn  and 
sum  the  25  products.  This  sum  is  the  value  to  be  plotted  (for  factor 
2)  as  in  Figure  5.  The  process  is  repeated  using  the  components  for 
another  factor  (for  example  factor  3  if  comparison  is  to  be  made 
with  factor  2  as  in  Fig.  5 ) .  The  process  is  not  overly  cumbersome 
if  a  desk  calculator  is  available.  A  program  has  been  written  for  the 


40  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Programma  101  Computer  (Olivetti-Underwood)  and  will  be  sent 
upon  request. 

The  information  summarized  to  this  point  answers  certain  ques- 
tions about  structural  resemblances  but  does  not  answer  some  other 
interesting  questions.  For  example,  are  one  or  two  species  involved? 
Three  areas  where  we  should  look  for  answers  are  as  follows.  Firstly, 
specimens  from  localities  in  Chihuahua  between  the  localities  al- 
ready represented  should  be  studied  to  learn  whether  intergradation 
occurs  between  N.  a.  albigtda  and  N.  a.  durangae,  and  whether  the 
zone  of  such  intergradation,  if  present,  is  narrow  or  broad  ( discussed 
below ) .  Intergradation  is  only  one  possibility  ( or  hypothesis )  to  be 
tested.  I  interpret  the  available  evidence  to  suggest  intergradation 
between  N.  a.  aJbigula  and  N.  a.  durangoe  in  Chihuahua.  Five 
specimens  especially  relevant  to  this  question  are  numbers  11,  12, 
13,  17,  and  22  in  Figure  5.  The  first  three  of  these  are  from  south  of 
the  Rio  Conchos  and  are  assigned  to  N.  a.  dtirangae.  The  last  two 
are  from  north  of  the  Rio  Conchos  and  are  assigned  to  N.  a.  albigula, 
but  these  specimens  are  not  from  localities  near  the  river.  No  data 
are  at  hand  for  specimens  from  north  of  the  river  and  nearer  to  it 
than  the  two  mentioned  above  ( and  one  other  included  in  the  orig- 
inal sample  of  N.  a.  albigula).  Study  of  series  from  the  two  banks 
of  the  ri\'er,  if  they  were  a\'ailable,  would  be  interesting. 

Secondly,  other  characters  could  be  studied.  Color  and  bacular 
characters  have  been  studied  and  ^^'ill  be  discussed  below. 

Thirdly,  specimens  from  other  parts  of  the  ranges  of  N.  albigula 
and  N.  micropus  should  be  studied.  An  especially  relevant  area  lies 
east  of  Chihuahua.  Special  attention  should  be  given  to  what  hap- 
pens distributionally,  ecologically,  and  morphologically  in  eastern 
Coahuila  where  the  two  species  are  said  to  meet.  My  deliberately 
limited  studies  in  that  direction  are  as  follows. 

I  examined  specimens  at  The  University  of  Kansas  from  Coa- 
huila assigned  by  Baker  (1956:281)  to  N.  a.  albigula.  They  may 
better  be  referred  to  what  I  am  here  calling  N.  a.  durangae.  Al- 
though I  have  not  studied  the  three  specimens  from  Durango 
reported  by  Baker  and  Greer  (1962:126)  as  N.  a.  albigula,  I  judge 
on  geographic  grounds  that  they  also  should  be  referred  to  N.  a. 
durangae. 

Goldman  (1910)  referred  specimens  from  Saltillo,  Coahuila,  to 
N.  albigula  leucodon  (USNM  117166,  117167)  and  N.  iiucropus 
micropus  ( USNM  117168) .  The  first  two  are  younger  than  the  third 
but  none  is  in  juvenile  pelage.   Pelages  of  all  are  similar  in  color. 


Anderson — Taxonomy  of  Chihuahuan  Woodrats 


41 


Table  2. — Measurements  in  millimeters  of  samples  of  three  taxa  of 
Neoioma.    Measurements  are  explained  in  tfxt;  specimens  in  each  sam- 
ple  ARE    ALSO    LISTED   IN    TEXT.     MeAN,   STANDARD   DEVIATION,    MINIMUM,    MAXI- 
MUM,  AND   SAMPLE   SIZE   ARE   GIVEN. 


N.  albi^ula 
albigula 


N.  albigula 
durangac 


N.  micro inis 


A  (length  of  bulla)  6.65+0.38 

6.15-7.2.5 
11=12 

B   (molor  toothrow)   8.50±0.29 

7.95-8.8 
n=12 

C  (length  of  palate)  7.34±0.25 

6.8-7.85 
n=12 

D  (zygomatic  aperture)  12.60±0.36 

11.95-13.05 
n=12 

E  (incisive  foramen)  9.03±0.46 

8.1-9.5 
n=12 

F  ( condyloincisi\'e  length)  40.68±0.79 

39.4-41.7 
n=12 

G  ( mesopterygoid  breadth)  2.75±0.29 

2.2-3.2 
n=12 

H  (breadth  of  palate)  2.78±0.28 

2.2-3.1 
n=12 

1  (postdental  breadth)  6.58±0.18 

6.2-6.8 
n=12 

1  ( rostral  breadth )  7.24±0.28 

6.9-7.9 
n=12 

K  (breadth  of  Ml)  2.61±0.17 

2.25-2.9 
n=12 

L  (exoccipital  breadth)   11. 64 ±0.63 

10.4-12.8 
n=ll 


M  (anterior  zygomatic  breadth) 


N  (posterior  zygomatic  breadth) 


O   (occipitoparietal  length) 


P  ( nasopremaxillary  difference)  .— 


20.42+0.63 

19.2-21.2 

n=12 

22.48±0.85 

20.4-23.8 

n=12 

6.71  ±0.47 

5.65-7.3 

n=12 

2.32±0.37 

1.8-2.9 

n=12 


7.20±0.46 

6.6-8.0 

n=7 

8.14±0.39 

7.7-8.75 
n=7 

7.81  ±0.65 

7.15-9.1 

n=7 

12.62±0.94 

11.5-14.0 

n=7 

9.16±0.84 

7.6-J0.3 

n=7 

41.96±1.36 

40.3-44.35 

n=7 

3.30±0.26 
2.9-3.6 

n=7 

3.06±0.46 
2.4-3.7 

n=7 

6.60±0.22 

6.2-6.8 

n=7 

7.81±0.39 

7.5-8.2 
n=7 

2.60±0.25 

2.3-2.8 

n=7 

12.54±0.92 
11.4-14.2 

n=7 

21.41±1.30 
19.9-21.7 

n=7 

23.34±0.95 

22.1-24.7 

n=7 

6.93±0.81 

6.1-8.3 

n=7 

2.89±0.30 

2.5-3.3 

n=7 


7.25±0.32 

6.7-7.85 

n=12 

8.76±0.59 

7.7-9.3 

n=12 

7.20±0.65 

6.3-S.l 

n=12 

13.53±0.76 
12.1-14.6 

n=12 

10.05±0.60 

9.1-11.0 

n=12 

43.32±2.30 

.38.6-45.65 

n=12 

3. 99  ±0.46 

3.4-4.8 

n=12 

3.22±0.32 

2.8-3.6 

n=12 

7.33±0.39 

6.9-8.2 

n=12 

7.54±0.65 

6.6-8.5 

n=12 

2.69±0.17 

2.45-3.05 

n=12 

12.40±0.73 

11.4-13.9 

n=12 

22.82±1.41 

20.2-24.9 

n=12 

24.88±1.44 

21.8-27.1 

n=12 

7.28±0.80 

5.85-8.3 

n=12 

2.14±0.49 

1.4-2.9 

n=12 


42  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Table  2. — Continued. 

N.  albigiila  N.  albigtila 

albigula                      durangae  N.  micropiis 

Q  (occinitonasal  length)  43.32±1.19       43.88+1.72  45.80±2.31 

41.6-45.0           41.55-46.65  41.1-47.15 

n=12                 n=7  n=12 

R  (length  of  nasals)  15.02+0.67       15.31+0.69  16.29+1.13 

13.8-16.05          14.35-16.45  14.6-18.2 

ni=12                  11=7  n=12 

S   (rostral  length)   13.37+0.57       13.80+0.81  14.51  +  1.06 

12.3-14.05          12.85-15.0  12.6-16.3 

n=12                 n=7  n=12 

T  (interorbital  breadth)  5.76±0.19         6.06+0.39  5.95+0.46 

5.4-6.0               5.5-6.7  5.3-7.0 

n=12                 n=:7  n=12 

U  (breadth  of  braincase)  17.76+0.47       17.86+0.61  17.60+0.63 

17.1-18.6           17.2-18.7  16.7-19.1 

n=12                  n=7  n=12 

V  (incisive  projection)  4.39+0.44         5.15+0.34  5.26+0.56 

3.3-5.1               4.8-5.75  4.7-5.95 

n=12                  n=7  n=12 

W  (depth  of  bulla)  5.28+0.26         5.96+0.56  5.62+0.24 

4.9-5.6               5.4-7.0  5.3-6.1 

n=12                  n=6  n=12 

X  (depth  of  sbill)  15.55+0.73       16.37+0.67  16.89±0.79 

13.85-16.75       15.7-17.25  15.2-18.25 

n=12                  n=6  n=12 

Y  (breadth  of  zygomatic  plate)  ....    4.19+0.20         4.32+0.51  4.53+0.32 

3.9-4.6                3.9-5.25  4.2-4.95 

n=12                  n=7  n=12 

Total  length  ..-    320.5+12.0       331.4+7.2  332.7±12.8 

305-340             321-341  315-348 

n=10                 n=5  n=7 

Length  of  tail 142.8+9.0         150.0+3.1  139.0+8.8 

131-160             145-153  130-153 

n=10                 n=5  n=7 

Length  of  hind  foot 32.8+1.9           33.4+1.9  35.0+2.7 

30-35                 30-36  30-39 

n=9                    n=7  n=7 

Weight               176.6+18.2       197.7+30.1  245+52.4 

155-209             153-245  191.4-296 

n=10                 n=7  n=3 

all  having  a  yellowish  hue.  No.  117168  has  a  darker  dorsal  tail  stripe 
than  the  others,  and  no.  117167  has  larger  hind  feet  (36  mm.  as 
opposed  to  33).  I  judge  that  a  single  species  is  represented.  The 
specimens  seem  more  al])i(i,tila-\ike  than  micropus-\ike.  Eight  speci- 
mens from  Monclova,  Coahuila,  include  one  (USNM  117018)  re- 
ferred by  Goldman  to  N.  alhiii^ula  albigula  and  seven  (USNM 
117017,  117019-117021,  117169-117171)  referred  by  him  to  N.  mi- 
cropiis canescens.  Two  of  the  seven  are  young  and  their  skulls  are 
damaged.    Measurements  were  taken  of  the  other  six  from  Mon- 


Anderson — Taxonomy  of  Chihuahuan  Woodrats  43 

clova  for  comparison  \\'ith  the  results  of  the  factor  analysis  already 
done.  On  the  basis  of  yellow  hue  of  pelage,  smaller  size,  and  nar- 
rower mesopterygoid  fossa  I  am  inclined  to  place  117017  and  117019 
\\ith  117018.  The  condition  of  the  vomer  posteriorly  does  not  help 
to  separate  specimens  in  the  series  from  Monclova.  I  also  measured 
a  specimen  from  Jimulco,  Coahuila  (USNM  58072,  a  large  male 
referred  to  N.  olbigula  albigula  by  Goldman),  and  one  from  the 
northern  boundary  of  Chihuahua  (USNM  35551,  a  N.  micropus  of 
moderate  size).  I  have  not  examined  the  three  specimens  from 
Jaral,  Coahuila,  the  only  other  Coahuilan  locality  of  sympatry  re- 
ferred to  by  Goldman  for  the  species  N.  dhigula  (two  specimens) 
and  A^  micropus  (one  specimen). 

The  coordinates  for  factors  2  and  3  were  calculated  for  addi- 
tional specimens  measured  and  are  plotted  in  Figure  5,  for  com- 
parison with  the  series  used  in  the  original  analysis.  The  additional 
specimen  of  N.  micropus  from  northern  Chihuahua  falls  with  the 
other  N.  micropus  but  in  the  area  of  overlap  with  N.  a.  durangae. 
The  rat  from  Saltillo  and  the  one  from  Jimulco  fall  with  N.  a. 
durangae.  The  two  rats  from  Monclova,  re-identified  by  me  as 
N.  a.  durangae,  fall  with  other  members  of  that  subspecies.  Two  of 
the  three  rats  from  Monclova  identified  by  both  Goldman  and  me  as 
N.  micropus  fall  within  the  area  of  overlap  of  N.  micropus  and  N.  a. 
durangae.  However,  the  third  falls  \\ath  N.  a.  durangae  and  outside 
the  area  on  the  graph  of  N.  micropus.  The  uncertainties  evident  in 
( 1 )  my  own  attempts  to  identify  these  rats  from  the  alleged  locality 
of  sympatry,  ( 2 )  the  differences  in  assignment  of  some  individuals 
by  Goldman,  by  me  somewhat  subjectively,  and  by  the  factor  analy- 
sis of  cranial  measurements  alone,  and  ( 3 )  the  lack  of  clear  bimodal 
clustering  in  the  Monclova  sample  all  suggest,  although  they  do  not 
prove,  that  only  one  taxon  is  present  in  the  sample  from  Monclova, 
Coahuila. 

In  short,  to  summarize  the  Coahuilan  situation,  more  work  is 
needed.  There  is  no  locality  of  certain  sympatry  between  two  taxa, 
although  the  small  sample  from  Monclova  is  suggestive.  As  my 
study  has  proceeded,  my  confidence  in  the  existence  of  a  species 
boundary  between  N.  albigula  and  N.  micropus  as  now  known,  or 
at  least  in  my  ability  to  define  the  boundary,  has  decreased.  If  they 
behave  as  species  in  eastern  Coahuila,  the  difference  between  the 
two  species  is  certainly  less  than  between  these  two  species  where 
they  are  sympatric  in  northern  Chihuahua  and  adjacent  regions  of 
New  Mexico  and  west  Texas. 


44  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Discriminant  Analysis 

Cranial  measurements  of  the  original  36  specimens  used  in  the 
factor  analysis  and  of  22  additional  specimens  ( the  same  plotted  in 
Fig.  5  and  listed  in  text)  were  examined  by  Discriminant  Analysis 
(five  minutes  on  IBM  7094  computer).  The  three  original  groups 
were  used  to  derive  the  discriminant  functions,  except  that  one 
specimen  of  N.  a.  durangae  was  placed  with  the  N.  micropus  by 
mistake.  The  three  groups  were  well  separated,  including  the  three 
specimens  of  N.  a.  durangae  that  were  not  separated  by  the  factor 
analysis  as  shown  in  Figure  5.  The  mistakenly-placed  specimen  of 
N.  a.  durangae  appeared  with  the  cluster  of  N.  micropus  in  the 
analysis,  but  at  the  side  of  the  cluster  nearest  that  of  durangae.  The 
factor  analysis  uses  all  measurements  and  incidentally  results  in  a 
scatter  in  which  groups  may  be  separated  or  partly  separated.  In 
discriminant  analysis  only  the  differences  in  measurements  between 
specified  groups  are  used.  This  enhances  the  separation  of  groups. 
Plotting  of  the  additional  22  specimens  in  terms  of  discriminant 
functions  places  them  in  scattered  positions  about  or  between  the 
three  discriminant  groups,  but  their  general  relative  positions  are 
consistent  with  the  relationships  otherwise  found. 

The  discriminant  analysis  tends  to  separate  N.  micropus  more 
clearly  from  N.  a.  albigula  and  N.  a.  durangae  than  did  the  factor 
analysis.  Both  analyses  tend  to  place  additional  specimens  of  N. 
micropus  (nos.  1-4  in  Fig.  5)  with  others  of  that  species,  and  to 
place  additional  specimens  of  N.  alhigula  (nos.  13,  15,  and  16  in 
Fig.  5),  which  are  of  subspecies  other  than  those  already  studied, 
with  the  specimens  of  that  species  ah'eady  studied. 

This  technique  is  still  being  developed  and  evaluated,  and,  there- 
fore, more  detailed  comments  on  these  results  are  not  included. 

Color 

Color  is  notoriously  more  difficult  to  describe  or  report  on  than 
are  measurements  to  which  numerical  values  easily  can  be  assigned. 
I  have  mentioned  that  N.  micropus  canescens,  the  widely  distributed 
subspecies  that  meets  N.  albigula  in  the  Chihuahuan  region,  differs 
from  N.  albigula  albigula  in  grayness,  paleness,  and  relative  lack  of 
yellowish  or  buffy  hues  in  dorsal  pelage.  However,  there  is  consid- 
erable variation  in  color  \\'ithin  the  wide  range  of  N.  a.  albigula  even 
after  its  reduction  by  the  present  reassignment  of  specimens  south 
of  the  Rio  Conchos  of  Chihuahua  to  N.  a.  durangae.  There  are 
specimens  of  N.  a.  durangae  that  are  not  distinguishable  by  color 


Anderson — Taxonomy  of  Chiiiuahuan  Woodrats  45 

from  many  N.  a.  albigula  from  Chihuahua.  However,  on  the  average 
N.  a.  durangae  is  grayer  and  paler  than  N.  a.  albigida.  The  palest 
specimens  examined  of  N.  a.  durangae  are  from  Escalon. 

The  single  specimen  of  N.  a.  durangae  that  falls  in  the  midst  of 
the  N.  a.  alhigtda  in  Figure  5  is  from  Escalon.  It  therefore  resembles 
N.  a.  alhigtda  cranially  and  resembles  N.  micropus  in  color.  When 
only  one  or  two  cranial  characters  and  color,  which  is  known  to  be 
subject  to  strong  local  natural  selection,  were  found  to  be  intermedi- 
ate in  N.  a.  durangae,  it  seemed  reasonable  to  suggest  that  selection 
might  have  produced  a  convergence  in  these  few  characteristics. 
However,  evidence  for  intermediacy  has  accumulated,  and  a  large 
number  of  characters  are  involved.  Some  of  these  are  not  struc- 
turally related,  are  poorly  correlated  (when  individuals  in  a  single 
series  are  compared),  and  are  not  obviously  influenced  by  local 
selection.  Convergence  as  a  result  of  selection  is  less  reasonable 
than  the  alternative  interpretation  that  these  animals  have  the  com- 
plex suite  of  characters  because  of  common  ancestry.  The  interest- 
ing question  now  becomes  whether  genetic  exchange  is  occurring 
anywhere  in  nature  between  N.  micropus  and  N.  albigida,  and,  if  so, 
what  form  the  exchange  takes.  Does  free  inter  gradation  occur,  or 
occasional  hybridization,  or  do  different  events  occur  at  different 
places? 

Baculum 

The  structure  of  the  male  phallus,  and  of  the  baculum  therein  in 
particular,  has  been  studied  for  Neotoma  albigida  and  Neotoma 
micropus  by  Burt  and  Barkalow  (1942)  and  by  Hooper  (1960). 
These  authors  suggested  that  N.  albigula  differs  from  N.  micropus 
in  (1)  slenderness  of  shaft,  (2)  shallower  U-shape  of  base,  (3) 
larger  distal  knob,  (4)  smaller  glans,  (5)  terminal  hood  relatively 
larger  and  not  cleft  distally,  as  opposed  to  slightly  cleft,  (6)  bone 
longer  and  its  spine  shorter  relative  to  both  glans  and  foot  lengths. 
The  first  three  differences  were  based  on  study  of  15  albigida  and 
four  micropus  by  Burt  and  Barkalow.  In  their  table  of  measure- 
ments the  range  for  micropus  is  within  that  of  albigula  for  length 
and  base  measurements  (both  dorsoventral  and  lateral)  but  near 
the  middle  of  the  shaft  there  is  little  overlap  in  dorsoventral  di- 
ameter and  none  in  lateral  diameter  (0.51  to  0.73  for  albigula,  and 
0.80  to  0.91  for  micropus).  Hooper  examined  three  specimens  each 
of  albigula  and  micropus  and  one  of  Neotoma  floridana  and  noted 
that  the  three  species  are  so  similar  as  to  suggest  that  they  are  con- 


46 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


LU 

h- 
< 
> 

O 

Ll 
O 


Q. 
LlI 
Q 


mm. 

1            1 

1        1        1 

1      i_  ■  1 

/5^^ 

<^       1    - 

0.6 

-            //-^ 

O    / 

7    / 

~ 

fo 

o/(3 

0.4 



1° 

O    y 

-  W- 

1      1 

1      1      1 

(o    o)        - 
1      1      1 

0.4  06  0.8  10  "1^- 

WIDTH  OF  BACULUM  AT  MlD-POINT 

Fig.  6.  Graph  of  two  measurements  of  the  bacula  of  individuals  of  tliree  taxa 
of  Neotoma.  Large  open  circles  are  N.  alhigula  albigitla;  dots  are  N.  a. 
diirangae;  small  open  circles  are  N.  micropus.  Encircling  lines  are  drawn 
freehand  simply  to  aid  perception  of  the  groups.  As  in  Figures  4  and  5, 
durangae  is  intermediate  to  the  other  two  taxa. 

Specific  or  are  sibling  species.  Differences  4  through  6  as  Hsted 
above  were  said  to  be  shght  (Hooper,  1960:5)  and  I  have  not 
attempted  to  study  them  in  my  material. 

Series  of  bacula  were  selected  for  study  and  comparison  as  fol- 
lows: N.  albigtda  alhigula  (22),  N.  albigitla  durangae  (16),  and 
N.  micropus  (13).  Eighteen  were  measured,  and  their  values  for 
two  measurements  are  plotted  in  Figure  6.  These  are  the  two  meas- 
urements of  the  six  taken  that  best  distinguish  N.  a.  alhigula  from 
N.  micropus  and  which  therefore  are  most  revealing  as  to  the  posi- 
tion of  N.  a.  durangae.  As  in  the  cranial  characters,  in  color,  and 
in  size  of  entire  animal,  bacular  characters  show  intermediacy  for 
N.  a.  durangae,  and  are  not  clearly  nearer  either  N.  a.  alhigula  or 
N.  micropus. 

Structure  of  Vomer 

The  degree  of  development  and  form  of  the  vomer  differ  in 
different  taxa  of  Neotoma.  Finley  (1958)  described  the  differences 
between  the  species  N.  alhigula  (as  represented  by  the  subspecies 
N.  a.  icarreni)  and  N.  micropus  (represented  by  N.  m.  canescens), 
and  he  used  these  characters  in  evaluating  evidence  for  species 
hybridization  that  he  found  in  southeastern  Colorado.  Figure  7 
illustrates  the  stages  of  development.  The  major  variable  is  the 
degree  of  enlargement  posteriorly  of  the  medial  vomerine  plate 
( shown  stippled ) .  In  the  least  developed  stage  ( as  in  7B ) ,  a  distinct 


Anderson — Taxonomy  of  Chihuahuan  Woodrats 


47 


Fig.  7.  Drawings  of  vomer  and  sinrounding  elements  in  skulls  of  Neotoma. 
A — dissection  from  the  right  to  a  sagittal  plane  near  the  middle  of  skull  but 
leaving  the  medial  partitions  of  ethmoid,  cartilage,  and  vomer  intact.  Labelled 
elements  are:  be,  braincase;  ca,  cartilage;  et,  ethmoid;  fr,  frontal;  ma,  maxillary; 
?ia,  nasal;  pin,  premaxillary;  ps,  presphenoid;  and  vo,  vomer.  Cut  elements  are 
shown  liy  diagonal  lines,  and  the  vomerine  partition  is  stippled  for  ease  of 
comparison  with  B,  C,  D,  and  E.  These  are  diagrams  showing  four  states  of 
vomerine  dexelopment  arranged  in  morphological  sequence  from  least  developed 
to  most  developed.  States  in  the  range  of  B  and  C  occur  in  N.  albigula  alhigiila, 
those  in  the  range  of  C  and  D  occur  in  N.  albigula  durangae,  and  those  in  the 
range  of  D  and  E  occur  in  N.  micropus:  An  oblique  view  looking  dorsally  and 
slightly  posteriorly  into  the  incisive  foramina  is  shown  in  F  and  H,  which 
correspond  in  state  to  B  and  D,  respectively.    The  oblique  orientation  of  skull 

in  F  and  H  is  shown  by  G. 


48  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

rounded  opening  is  apparent  in  the  medial  plane  as  seen  in  the 
incisi\'e  foramen  (as  in  7F).  When  the  plate  is  larger,  conditions 
such  as  in  7C,  7D,  and  7E  are  evident.  The  legend  for  Figure  7 
provides  other  details. 

A  condition  in  Neoto7na  pakitina  that  exceeds  that  of  Figure  7E 
was  kindly  pointed  out  to  me  by  Professor  E.  R.  Hall,  who  is  study- 
ing the  taxonomic  status  of  this  species.  In  some  of  the  specimens 
that  he  has  assembled  at  The  University  of  Kansas  the  posterior 
projection  of  the  vomer  is  so  great  that  it  protrudes  visibly  behind 
the  posterior  palatal  margin  (to  the  position  of  the  letter  E  in 
Fig.  7E). 

Certain  modifications  in  adjacent  structures  are  positively  cor- 
related with  degree  of  vomerine  enlargement:  (1)  shortness  of 
medial  maxillary  spine;  (2)  shortness  or  loss  of  articulation  of  this 
spine  with  premaxillary;  (3)  amount  of  contact  of  spine  with  vomer, 
at  least  in  stages  such  as  7C  and  7D,  although  in  some  specimens  a 
slit  may  separate  these  elements  (as  in  7E);  (4)  shortening  of 
medial  slit  between  the  presphenoid  and  medial  part  of  vomer  and 
bordered  laterally  by  posterolateral  wings  of  vomer  ( the  opening  is 
reduced  in  a  form  such  as  7E  to  a  posteriorly  directed  median 
opening);  and  (5)  narrowing  of  maxillo- vomerine  gap  posteroven- 
tral  to  median  part  of  vomer  from  a  broadly  rounded  aperture  to 
a  narrow  slit. 

In  this  complex  of  related  characters,  as  in  all  other  diagnostic 
characters  studied,  N.  a.  chirangae  is  intermediate  between  iV. 
albigula  albigula  and  N.  micropus. 

Discussion 

I  judge  it  unwise  to  synonymize  N.  aUngula  and  N.  micropus  on 
the  basis  of  available  information,  although  the  contrary  conclusion 
is  based  on  tenuous  evidence.  The  basic  questions  about  species 
limits  cannot  be  resolved  without  additional  information  from  the 
field.  Nomenclatorially,  I  think  the  conservative  course  is  to  use  the 
names  as  presently  arranged  until  better  evidence  is  available.  A 
change  at  this  time  would  add  nothing  to  our  knowledge  and  would 
not  be  especially  useful.  There  is,  furthermore,  a  possibility  that 
not  only  is  N.  albigula  synonymous  with  N.  micropus.  but  that  they 
are  both  conspecific  and  hence  synonymous  at  the  species  level  with 
N.  floridami.  If  this  be  so,  the  name  of  the  species  would  become 
N.  fioridana,  for  that  is  the  oldest  name  among  these  three.  The 
relationships  of  N.  micropus  and  N.  fioridana  are  currently  being 


Anderson — Taxonomy  of  Chihuahuan  Woodrats  49 

studied  by  Elmer  C.  Birney  at  The  University  of  Kansas  and  perhaps 
by  other  students  elsewhere.  In  any  case,  N.  fioridana,  N.  micropus, 
and  N.  olbigula,  as  suggested  by  previous  authors,  are  certainly 
closely  related. 

The  following  are  hypotheses  for  later  consideration.  N.  alhi<i,ula 
dunin<iae  may  be  similar  to  a  stock  that  was  ancestral  to  both  IV. 
cilbigula  and  N.  micropus.  There  may  or  there  may  not  be  present 
day  interbreeding  between  the  two  species  in  the  region  of  eastern 
Coahuila.  The  original  contact,  geographically  broad  though  it  has 
become,  between  the  two  species  in  northern  Chihuahua,  southern 
New  Mexico,  and  western  Texas  may  have  been  both  secondary  to 
species  divergence  and  later  than  the  contact  in  northeastern  Mex- 
ico. The  possible  hybridization  of  N.  alhigula  and  N.  micropus  in 
southeastern  Colorado  reported  by  Finley  (1958)  needs  to  be  in- 
vestigated in  greater  detail. 

I  have  not  studied  Mexican  specimens  from  south  of  Chihuahua 
and  Coahuila  sufficiently  to  offer  any  conclusion  about  the  southern 
limits  of  N.  alhigula  durangae,  or  about  the  relationships  of  the 
three  most  southern  subspecies,  N.  alhigula  leucodon,  N.  a.  zaca- 
tecae,  and  N.  a.  suhsolamis. 

The  discoveries  here  reported  indicate  that  fascinating  scientific 
problems  remain  to  be  discovered  and  solved,  even  in  the  system- 
atics  of  common  and  reasonably  well  known  groups.  The  present 
results  of  study  of  Neotoma  in  northern  Mexico  add  to  our  knowl- 
edge of  variation  and  relationships  and  also  direct  attention  to 
further  problems  warranting  study,  to  the  places  where  they  may 
best  be  approached,  and  to  some  methods  likely  to  be  useful. 

Acknowledgements 

I  am  especially  grateful  to  Professor  E.  Raymond  Hall  for  his  continued 
support  of  my  studies  of  Chihuahuan  mammals  and  for  his  recent  willingness  to 
discuss  problems  of  taxonomy  in  the  genus  Neotoma,  including  his  own  current 
work  with  Neotoma  palatina.  My  work  at  The  University  of  Kansas  was  also 
greatly  aided  by  Dr.  J.  Knox  Jones,  Jr.,  and  his  curatorial  assistants  among  the 
graduate  students.  Various  persons  at  the  U.S.  National  Museum,  including 
especially  Dr.  Ronald  Pine  on  my  most  recent  visit,  were  helpful.  My  colleagues 
at  The  American  Museum  of  Natural  History,  Drs.  Richard  G.  Van  Gelder,  Karl 
F.  Koopman,  and  Guy  G.  Musser,  provided  valued  comments  both  in  the  course 
of  my  work  and  upon  the  completion  of  the  manuscript  itself.  Our  Scientific 
Assistant,  Miss  Elizabeth  Fryatt  helped  in  various  ways,  including  the  computa- 
tions done  on  the  Programma  101  Computer.  Dr.  D.  Vincent  Manson  pro- 
grammed and  ran  the  Factor  Analysis  and  Discriminant  Analysis  and  aided  in 
my  statistical  interpretation  of  the  results.    He  should  not,  however,  be  held 


50  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

responsible  for  tlie  biological  interpretations.  The  preliminary  art  work  and 
plotting  of  graphs  were  done  by  me  and  then  efficiently  completed  by  graphic 
artists  under  the  direction  of  Mr.  Joseph  Sedacca  of  The  American  Museum  of 
Natural  History. 

Literature  Cited 

Anderson,  S. 

1968.     A  new  craniometer  and  suggestions  for  craniometry.    Jour.  Mamm., 
49:221-228,  1  fig. 
Bailey,  V. 

1932.     Mammals  of  New  Mexico.   N.  Amer.  Fauna,  53:1-412,  22  pis.,  58 
figs. 
Baird,  S.  F. 

1859.     Mammals  of  the  Boundary.    In  report  on  U.S.  and  Mexican  Boun- 
dary Survey,  under  W.  H.  Emory,  2,  pt.  2:1-62,  27  pis. 
Baker,  R.  H. 

1956.     Mammals    of   Coahuila,    Mexico.     Univ.    Kansas    Publ.,    Mus.    Nat. 
Hist,  9:125-335,  76  figs. 
Baker,  R.  H.,  and  J.  K.  Greer 

1962.     Mammals  of  the  Mexican  State  of  Durango.    Publ.  Mus.  Michigan 
State  Univ.,  Biol.  Series,  2:29-154,  4  pis.,  6  figs. 
Brown,  T.,  M.  J.  Barrett,  and  J.  N.  Darboch 

1965.     Factor  analysis  in  cephalometric  research.    Growth,  29:97-107. 
Burt,  W.  H.,  and  F.  S.  Barkalow,  Jr. 

1942.  A  comparative  study  of  the  bacula  of  wood  rats  (subfamily  Neo- 
tominae).   Jour.  Mamm.,  23:287-297,  3  figs. 

Finley,  R.  B.,  Jr. 

1958.  The  wood  rats  of  Colorado:  distribution  and  ecology.  Univ.  Kansas 
Publ.,  Mus.  Nat.  Hist.,  10:213-552,  34  pis.,  8  figs. 

Goldman,  E.  A. 

1910.     Revision  of  the  wood  rats  of  the  genus  Neotoma.    N.  Amer.  Fauna, 
31:1-124,  8  pis.,  14  figs. 
Gould,  S.  J. 

1967.  Evolutionary  patterns  in  pelycosaurian  reptiles:  a  factor-analytic 
study.   Evolution,  21:385-401,  7  figs. 

Hall,  E.  R.,  and  K.  R.  Kelson 

1959.  The  mammals  of  North  America.  The  Ronald  Press  Go.,  New  York, 
2:viii  +  547-1083  +  79,  illustrated. 

Hooper,  E.  T. 

1960.  The  glans  penis  in  Neotoma  (Rodentia)  and  aUied  genera.  Occas. 
Papers  Mus.  Zool.,  Univ.  Michigan,  618:1-20,  11  pis. 

Lawrence,  B.,  and  W.  H.  Bossert 

1967.  Multiple  character  analysis  of  Canis  lupus,  latrans,  and  familiaris, 
with  a  discussion  of  the  relationships  of  Canis  niger.  Amer.  Zoologist, 
7:223-232,  3  figs. 

LiDiCKER,  W.  Z.,  Jr. 

1962.  The  nature  of  subspecies  boundaries  in  a  desert  rodent  and  its  im- 
plications for  subspecies  taxonomy.    Syst.  Zool.,  11:160-171,  3  figs. 

Manson,  D.  V. 

1967.     Factor  analysis  of  petrochemical  data.    Pp.  251-258,  in  Basalts:    the 
Poldervaart  treatise  on  rocks  of  basaltic  composition  (H.   H.   Hess 
and  A.  Poldervaart,  eds.).    Interscience,  New  York. 
Van  Valen,  L. 

1962.     A  study  of  fluctuating  asymmetry.   Evolution,  16:125-142. 

Wallace,  J.  T.,  and  R.  S.  Bader 

1967.  Factor  analysis  in  morphometric  traits  of  the  house  mouse.  Syst. 
Zool.,  16:144-148. 


REPRODUCTION  AND  POPULATION  DENSITIES 
IN  A  MONTANE  SMALL  MAMMAL  FAUNA 


BY 


Terry  A.  Vaughan 

Small  mammals  occupying  subalpine  environments  are  adapted 
in  \arioiis  ways  to  the  severe  annual  climatic  cycle  and  to  the  short 
growing  season.  In  the  present  study  area  in  Colorado  snow  covers 
the  ground  and  little  plant  growth  occurs  for  approximately  seven 
months  of  the  year,  from  November  through  May.  During  the  five- 
month  snow-free  period  the  plants  undergo  their  brief  growth  and 
flowering.  Reproduction  in  the  small  mammals  is  limited  almost 
completely  to  this  period,  and  population  levels  of  small  mammals 
in  a  given  year  are  partly  determined  by  reproductive  success  and 
late  summer  populations  of  the  pre\aous  summer.  Although  faunal 
studies  have  yielded  information  on  the  geographic  distributions 
and  habitat  preferences  of  montane  mammals  (see  Grinnell  and 
Storer,  1924;  Grinnell  et  al,  1930;  Bailey,  1932;  Hall,  1946;  Warren, 
1942),  and  some  information  is  available  on  their  reproduction  (see 
appropriate  accounts  beyond),  population  fluctuations  and  repro- 
ductive patterns  of  montane  small  mammals  remain  poorly  known. 
This  report  contributes  information  to  fill  this  gap  in  our  knowledge. 
Because  of  their  abundance  in  the  study  area,  this  report  considers 
primarily  the  following  species:  Sorex  vagrans,  vagrant  shrew; 
Eutamios  minimus,  least  chipmunk;  Thomomys  talpoides,  northern 
pocket  gopher;  Peromyscus  manictilatus,  deer  mouse;  Microtus  mon- 
tamis,  montane  vole. 

Acknowledgments 

For  assistance  with  laboratory  and  field  work  I  am  indebted  especially  to: 
Reldon  Beck,  Robert  Casady,  Daniel  Hemphill,  Lloyd  Reed,  Donnie  Sparks,  and 
Wendy  Weil.  I  profited  from  critical  discussions  of  the  study  with  R.  M.  Hansen 
and  J.  J.  Norris.  This  research  was  supported  by  National  Science  Foundation 
Grant  GB-3686. 

Study  Area 

The  study  area  was  three  miles  southwest  of  Rabbit  Ears  Pass,  in  Grand 
County,  Colorado,  at  an  elevation  of  9900  feet.  The  census  quadrats  were  in 
rolling,  semi-open  subalpine  "parks"  tliat  were  nearly  devoid  of  trees  except  for 
scattered  open  stands  of  spruce,  Ficea  engelmanii,  and  fir,  Abies  lasiocarpa. 
As  indicated  liy  estimates  of  plant  composition,  the  vegetation  was  composed 
of  18  per  cent  grasses,  seven  per  cent  sedges,  and  75  per  cent  forbs  (Table  1). 

(51) 


52  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Table  1. — Relative  freqtjencies  of  the  common  plants  in  eight  half- 
acre   QUADRATS   IN   NORTHERN   COLORADO. 

Species  Relative  frequency 

Grasses  and  sedges 

Stipa  lettermani  14.4 

Bromu.s  ciliata  1.2 

B ramus  polyanthus  1.2 

Stipa  Columbiana  .8 

Carcx  sp.   7.4 

Foibs 

CoUomia  linearis  ■. 13.4 

Achillea  lanulosa  11.6 

Viola  mtttalli   8.6 

Taraxicum  officinale  6.6 

Polygonum  douglasii  4.1 

Lupinus  argenteus  3.8 

Agoseris  glauca  3.6 

Senecio  crassultis  2.6 

Solidago  inissouriensis  1.6 

All  other  plants  19.1 


Scattered  patches  of  gooseberry  (Rihes  montigenum)  occurred  locally,  and 
lupine  (Lupinus  argenteus)  was  often  conspicuous  late  in  summer.  Logs  and 
stimips  were  present  in  most  quadrats  and  were  centers  of  chipmunk  activity. 
In  the  years  from  1964  through  1967,  the  area  was  never  free  of  snow  until 
June  5,  and  most  of  the  plants  flowered  in  August.  By  early  September  of 
each  year  many  forbs  had  been  killed  by  frost.  The  earliest  snows  came  in 
September,  but  a  continuous  snow  cover  usually  did  not  develop  until  Novem- 
ber. Maximum  snow  depth  was  usually  attained  in  March,  and  varied  between 
6  and  11  feet. 

The  assemblage  of  mammals  in  tlie  study  area  was  typical  of  subalpine  areas 
in  northern  Colorado.  Table  2  lists  the  mammals  observed  or  trapped  in  the 
study  area. 

Methods 

An  electric  fence  similar  to  tliat  developed  by  Pequegnat  and  Thompson 
(1949)  was  used  to  study  population  densities.  Aroimd  each  half-acre  quadrat 
a  fine-mesh  chicken  wire  fence  18  inches  high  was  erected,  and  a  copper  wire 
was  stretched  one  inch  inside  tlie  fence  and  about  three-fourths  of  an  inch  above 
the  ground.  The  copper  wire  was  connected  to  a  Model  T  Ford  spark  coil 
powered  by  a  six  volt  automobile  battery.  Any  small  mammal  ( less  than  about 
500  grams)  that  contacted  the  wire  was  killed,  and  from  the  position  of  the 
animal  it  was  clear  whether  it  was  entering  or  leaving  the  quadrat.  Within 
each  quadrat  a  grid  of  169  snap  traps  was  set;  the  traps  were  at  intervals  of 
10.5  feet.  Macabee  gopher  traps  were  set  where  there  were  signs  of  pocket 
gopher  activity.  The  quadrats  were  operated  for  at  least  four  days,  and  for 
tlie  first  24  hours  the  quadrats  were  checked  every  four  hours.  Because  small 
mammals  were  unable  to  enter  or  leave  the  quadrat  on  the  surface  of  the 
ground  without  being  killed,  and  because  even  the  mammals  not  attracted  to 


Vaughan — Montane  Small  Mammal  FauxVa 


53 


Table  2. — Mammals  observed  or  trapped  in  the  study  area  in  northern 

Colorado. 

Scientific  name  Vernacular  name 

Sorex  cinereus Masked  shrew 

Sorex  vaprans Vagrant  shrew 

Sorcx  pahisfris  Water  shrew 

Microsoiex  hoiji  Pygmy  shrew 

Eutamias  iiiinimus  _. Least  chipmunk 

Marmota  flaviveutris  Yellow-bellied  marmot 

SpcnnopJiihis  lateralis Golden-mantled  ground  squirrel 

TlH>moiiiys  talpoides  Northern  pocket  gopher 

PcioDiysctis  maniculatus  .._  Deer  mouse 

Clcthrionomijs  gapperi  Red-backed  vole 

Plienacomijs  intermedins Heather  vole 

Microtns  7nontanns  Montane  vole 

Microtns  longicaudus Long-tailed  vole 

Zapns  princeps  Meadow  jumping  mouse 

Cants  latrans Coyote 

Vulpes  vnlpes Red  fox 

Martes  americana  Marten 

Mustcla  crminea  Ermine 

Mustela  frenata  Long-tailed  weasel 

Cervns  canadensis American  elk 

Odocoiletis  hemionus Mule  deer 


the  traps,  such  as  shrews,  seemed  eventually  to  contact  the  copper  wire,  the 
quadrats  seemed  at  the  least  to  provide  reliable  indices  to  the  abundance  of 
small  mammals. 

Pocket  gophers  were  common  in  the  stiidy  area,  and  because  these  rodents 
are  fossorial  it  would  be  expected  that  they  would  regularly  pass  beneath  the 
fence,  making  accurate  estimates  of  their  density  impossible.  It  seemed,  how- 
ever, that  pocket  gophers  were  as  xidnerable  to  the  wire  as  were  other  rodents. 
When  the  quadrats  were  established  a  strip  about  six  inches  wide  beneatli  the 
wire  was  completely  cleared  of  vegetation,  and  usually  about  one  inch  of  soil 
was  remo\ed  with  the  \egetation.  This  cleared  strip  offered  a  strong  attraction 
to  pocket  gophers;  repeatedly  they  tunneled  to  the  surface  in  the  strip  and 
were  killed  by  the  wire.  The  direction  from  which  they  came  could  be  deter- 
mined by  tracing  their  burrows.  Due  to  the  attraction  offered  by  the  cleared 
strip,  a  good  indication  was  obtained  of  pocket  gophers  moving  in  or  out  of 
the  quadrat,  and  the  estimates  of  their  population  densities  made  within  the 
quadrats  were,  in  my  judgment,  reasonably  accurate. 

A  series  of  quadrats  was  established  four  times  in  1965,  four  times  in  1966, 
and  three  times  in  1967.  Each  series  consisted  of  four  half-acre  quadrats.  In 
each  of  tlie  first  two  summers,  therefore,  a  total  of  eight  acres  was  sampled,  and 
in  the  third  summer  six  acres  were  studied.  Because  of  adverse  weather  condi- 
tions and  late  snowmelts  in  two  summers,  only  in  1966  were  quadrats  operated 
in  June.  In  the  other  two  years  they  were  operated  in  early  July.  The  first  set 
of  quadrats  each  summer  was  established  as  soon  after  snowmelt  as  possible. 
At  this  time  most  of  the  plants  were  just  beginning  their  annual  growth  and 
the  soil  at  many  sites  was  still  saturated  with  water.  The  last  set  of  quadrats 
each  summer  was  usually  operated  in  early  September,  after  the  first  late 
summer  frosts,  at  a  time  when  the  annual  growtli  of  most  plants  had  been 
completed.    An  attempt  was  made  to  situate  all  quadrats  in  the  same  plant 


54  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

community.  Each  set  of  quadrats  was  established  at  a  different  place  in  a 
given  summer,  but  for  a  given  set  of  quadrats  the  same  sites  were  used  each 
summer. 

A  total  of  1639  small  mammals  \\'ere  taken  during  this  study.  Because  the 
quadrats  did  not  always  yield  adequate  samples  for  reliable  information  on 
reproduction,  additional  trapping  was  done  outside  the  quadrats  and  at  a 
considerable  distance  from  them.  Animals  were  frozen  in  the  field  soon  after 
captrne  and  were  dissected  later  in  the  laboratory. 

The  flora  was  studied  by  sampling  the  frequency  of  the  plants  in  the 
census  quadrats.  One  hundred  randomly  placed  plots,  each  six  inches  square, 
were  sampled  each  summer  in  each  quadrat. 

Population  Densities 

In  each  year  of  the  study  the  population  density  of  small  mam- 
mals was  low  immediately  after  snowmelt  and  relatively  high  in  late 
summer.  Although  this  is  probably  the  basic  pattern  of  change  that 
occurs  in  most  summers,  the  relati\'e  abundance  of  the  species  and 
the  monthly  changes  in  population  densities  of  each  species  seem- 
ingly change  sharply  from  year  to  year. 

Summer  of  1965 

Snowmelt  was  late  in  this  summer  due  to  an  unusually  deep 
winter  snowpack  and  a  cold  spring.  The  first  set  of  quadrats  was 
not  in  operation  until  July  8,  when  the  soil  was  still  saturated  with 
water  and  scattered  snow  drifts  persisted  in  shaded  situations;  in 
most  places  the  summer  growth  of  vegetation  was  barely  underway. 
An  occasional  dead  M.  montamis  or  T.  tolpoides  was  found  in  places 
where  water  had  coxered  the  surface  of  the  ground  for  several  weeks 
during  snowmelt. 

The  density  of  small  mammals  was  low  at  this  time  (Figs.  1-2); 
the  census  quadrats  indicated  a  density  of  17.5  per  acre.  The  four 
most  abundant  species  and  their  densities  per  acre  were:  P.  manicu- 
latiis,  7.5;  E.  minimus,  4.0;  M.  montanus,  3.0;  T.  falpoides,  2.5.  Be- 
cause of  the  difficulty  with  which  pocket  gophers  are  trapped  imme- 
diately after  snowmelt,  the  figure  for  T.  tolpoides  is  probably  lower 
than  was  the  actual  density.  Although  the  pattern  of  changes  in 
density  differed  between  species,  by  early  September  the  density  of 
small  mammals  had  risen  roughly  three-fold  to  51.0  per  acre  (Fig. 
3).  Densities  were:  T.  talpoides,  14;  E.  minimus,  14;  P.  manicidatus, 
6.5;  S.  vogrons,  6.0.  The  figures  for  E.  minimus  seem  unreasonably 
high  and  should  be  considered  as  an  index  to  abundance  rather 
than  an  expression  of  actual  density.  On  two  occasions  chipmunks 
were  observed  to  leap  onto  the  chicken  wire  fence  enclosing  a  quad- 


Vaughan — Montane  Small  Mammal  Fauna 


55 


10 
8 

• 

1965 

17.5/acre 

6 

4 

■ 

■ 

2 

u 
< 

10 

UJ  8 

6 

• 

1966 

8.0/acre 

^4 
-2 

■ 

■ 

LU 

Q   0 

10 

1967 

34.0/acre 

8 

• 

6 

4 
2 

. 

r^ 

1 

Tt        Em 


Pm       Mm 
SPECIES 


Sv        Sc 


Fig.  1.  Early  summer  densities  of  six  species  as  determined  in  four  half-acre 
quadrats  each  summer.  Quadrats  were  operated  soon  after  snowmelt  in  June 
or  early  July.  The  following  symbols  are  used:  Tt,  Thomomijs  talpoides; 
Em,  Eutamias  minimus;  Pm,  Peromyscus  maniciilatiis;  Mm,  Microtus  montanus; 

Sv,  Sorex  vagrans;  Sc,  S.  cinereus. 

rat  and  scramble  over  the  fence  without  contacting  the  copper 
wire;  the  large  number  of  chipmunks  taken  in  the  quadrats  may 
have  resulted  in  part  from  this  type  of  entry.  Usually,  however, 
chipmunks  attempted  to  enter  or  leave  the  quadrat  by  crawling 
beneath  or  through  the  fence;  these  animals  were  killed  almost 
instantly  when  they  contacted  the  copper  wire.  Additional  indica- 
tion of  the  abundance  of  chipmunks  in  the  study  area  in  early  Sep- 
tember of  1965  is  indicated  by  the  fact  that  the  total  number  of 


56 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


19 


15 


10 


19 


^  15 


< 


10 


z 

UJ 

a 


15 


10 


5  • 


Peromyscus 
Eutamias 
Thomomys 
Microtus 


1967 


QUADRATS 

Fig.  2.  Changes  in  densities  of  some  small  mammals  in  summer.  Quadrats  were 
operated  as  follows:  series  1,  soon  after  snowmelt,  in  late  June  or  early  July; 
series  2,  in  mid-  or  late  July;  series  3,  in  August;  series  4,  in  late  August,  or 

early  September. 

chipmunks  taken  during  operation  of  the  fourth  series  of  quadrats, 
including  individuals  killed  trying  to  enter  the  quadrats,  was  51. 

The  ground  remained  mostly  free  of  snow  until  mid-November. 
Trapping  from  November  4  to  7  revealed  that  both  T.  talpoides  and 
P.  maniculatus  were  common,  but  no  M.  montonus  were  taken. 
Chipmunks  were  apparently  in  hibernation,  for  none  were  seen  or 
trapped. 

Summer  of  1966 

Snowmelt  was  fairly  early  in  this  summer  and  much  of  the  study 
area  was  free  of  snow  by  about  May  25.  The  first  set  of  quadrats 
was  in  operation  on  June  15  and  yielded  8.0  small  mammals  per 
acre  (T.  talpoides,  5.0;  P.  maniculatus,  2.0;  E.  minimus,  1.0).  The 
last  series  of  quadrats  was  established  on  September  2.  The  density 


Vaughan — Montane  Small  Mammal  Fauna 


57 


15 

. 

e 

10 

- 

1965 

51.0/acr 

5 

1 

15 

■ 

. 

LU 

■ 

1966 

< 

33.5/acre 

TY    PER 

CO 

1            1 

z 

lU 

^15 

■ 

1967 

k                                                1 

10 

■ 

32.5/acrt 

5 

Tt 

Em 

Pm 

Mm 
SP 

Sv 
ECIE 

s 

Sc 

Fig.  3.    Late  summer  densities  of  six  species  as  determined  in  four  half-acre 

quadrats    each    summer.     Quadrats    were    operated    in    late    August    or    early 

September,  after  the  first  late-summer  frosts.    Symbols  for  species  are  the  same 

as  in  Fig.  1,  but  Zp,  for  Zapiis  princeps,  is  also  used. 

of  small  mammals  was  four  times  as  great  at  this  time  (33.5  per 
acre)  as  it  was  in  June  (Figs.  1-3).  The  most  abundant  species  and 
their  densities  were  T.  talpoides  (13.5),  P.  manicuhius  (10.5),  and 
E.  minimus  (8.0). 

Summer  of  1967 
Snowmelt  persisted  through  the  first  half  of  June  and  occasional 
snow  fell  through  mid-June.  Although  the  first  set  of  quadrats  were 
not  studied  in  1967,  it  was  clear  that  the  pattern  of  change  in  the 
density  of  small  mammals  that  typified  the  previous  two  summers 
was  at  least  partly  reversed  in  1967  (Figs.  1-3).  The  set  of  quadrats 
started  on  July  1  indicated  a  fairly  dense  population  of  34  small 


58  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

mammals  per  acre  (P.  manicukitus,  12.5;  E.  minimus,  11.5;  T.  tal- 
poides,  5.0;  M.  montamis,  5.0).  In  1967,  rather  than  a  steady  rise  in 
density  through  late  summer,  as  occurred  in  the  other  years,  there 
was  a  decline  by  late  August  to  32.5  per  acre.  This  was  due  largely 
to  drops  from  July  to  late  August  in  the  densities  of  both  P.  manicii- 
latits  (from  12.5  to  6.0)  and  E.  minimus  (from  11.5  to  5.5).  Popu- 
lations were  lower  in  late  summer  of  1967  than  they  were  at  com- 
parable times  in  either  of  the  other  years. 

Population  Dynamics 

Sharp  fluctuations  in  small  mammal  populations  were  seemingly 
the  rule  throughout  the  study.  As  shown  in  Figures  1  to  3,  the 
relative  densities  and  the  absolute  densities  of  the  different  species 
shifted  markedly  from  summer  to  summer  as  well  as  from  month 
to  month  within  a  summer.  Early  in  the  summer  of  1965,  for 
example,  the  most  abundant  species  was  P.  manicuJatus;  T.  talpoides 
was  fourth  in  terms  of  abundance.  This  relationship  was  altered 
two  months  later,  when  T.  talpoides  was  the  most  abundant  mam- 
mal. Whereas  the  density  of  Thomomys  rose  steadily  during  the 
summer,  that  of  P.  maniculatus  was  lower  early  in  September  than 
it  was  just  after  snowmclt.  The  summer  of  1965  was  unusually  cool 
and  rainy;  perhaps  this  weather  resulted  in  poor  survival  of  young 
Peromyscus.  In  any  case,  the  summer  of  1966  presented  a  different 
picture.  In  early  summer  T.  talpoides  was  the  most  abundant 
species  with  P.  maniculatus  second.  Although  the  rate  of  increase 
was  different  in  each  species  (Fig.  2),  this  order  was  maintained 
through  early  September.  As  a  further  difference  between  these 
two  summers,  M.  montanus  was  taken  commonly  and  consistently 
throughout  the  summer  of  1965,  but  only  one  individual  was  taken 
in  the  quadrats  in  1966.  Densities  of  small  mammals  were  generally 
low  in  1967,  but  M.  montanus  was  commoner  late  in  that  summer 
than  at  any  other  time  in  the  study.  The  density  of  the  entire  com- 
munity of  small  mammals  in  late  summer  also  fluctuated,  being 
highest  in  1965  (51  per  acre),  and  roughly  the  same  in  1966  and 
1967  (33.5  per  acre  and  32.5  per  acre,  respectively). 

The  period  of  dispersal  of  young  of  the  year  was  signalled  by 
the  occasional  capture  at  the  quadrats,  usually  as  the  animals  tried 
to  enter,  of  species  that  did  not  regularly  occur  there  and  that  prob- 
ably found  optimal  conditions  in  nearby  moist  situations.  Species 
of  this  type  were  Microsorex  hoyi,  Phenacomys  intermedins,  and 


Vaughan — Montane  Small  Mammal  Fauna  59 

Zapiis  princeps.  Sorex  cinereus,  a  species  that  was  taken  occasion- 
ally throughout  the  summer,  was  taken  much  more  frequently  in 
late  summer.  In  nearby  areas  in  Wyoming  this  species  is  known  to 
favor  moist  habitats  ( Brown,  1967 ) .  Probably  largely  as  a  result  of 
the  dispersal  of  young  animals,  the  greatest  diversity  of  species  was 
usually  encountered  in  late  summer. 

In  the  study  area  the  population  levels  of  small  mammals  were 
influenced  strongly  by  the  survival  of  young.  In  southern  Colorado, 
Hansen  (1962)  found  that  high  survival  of  young  was  associated 
with  high  densities  of  pocket  gophers  and  low  survival  of  young  was 
associated  with  a  declining  population.  Age  ratios  at  the  end  of  the 
breeding  season,  then,  can  be  a  partial  basis  for  predictions  of  future 
densities.  Limited  data  on  age  ratios  of  pocket  gophers  in  the  study 
area  are  relevant  to  a  consideration  of  population  fluctuations. 
Density  of  pocket  gophers  was  high  in  late  summer  of  1965  ( 14  per 
acre)  and  survival  of  young,  as  indicated  by  the  fact  that  87  per 
cent  of  the  38  animals  taken  were  young,  was  high.  The  density 
was  about  the  same  in  late  summer  of  1966,  but  survival  of  young 
was  probably  low  (50  per  cent  of  22  animals  were  young).  Pocket 
gopher  density  declined  in  1967  to  but  six  per  acre  in  late  summer. 
This  drop  was  perhaps  due  both  to  low  survival  of  young  the  previ- 
ous summer  and  to  continued  low  survi\'al  in  the  summer  of  1967. 
At  that  time,  48  per  cent  ( 10  of  21 )  of  the  animals  taken  were  young. 

Apparently  there  were  changes  in  the  density  of  the  long-tailed 
weasel  (Mustela  frenata)  in  the  period  of  this  study.  Evidence  for 
these  changes  consists  of  sight  records,  made  during  roughly  40 
days  spent  in  the  field  each  summer,  and  on  the  number  of  weasels 
killed  at  the  quadrats.  (A  total  of  approximately  2400  linear  feet  of 
electric  fence  was  in  operation  each  time  a  series  of  quadrats  was 
studied. )  No  weasels  were  taken  at  the  quadrats  in  1965,  and  only 
two  weasels  were  seen.  In  1966  six  weasels  were  killed  by  the  elec- 
tric fences  and  weasels  were  observed  regularly;  in  1967  four 
weasels  were  taken  by  the  electric  fences  and  many  were  seen. 
Although  the  data  give  only  a  general  idea  of  abundance,  weasels 
were  clearly  far  more  abundant  in  1966  and  1967  than  they  were 
in  1965. 

Reproductive  Cycles 

The  reproductive  cycles  of  only  those  species  taken  with  regu- 
larity in  the  quadrats  are  discussed.  Reproductive  data  are  pre- 
sented in  Tables  3-4  and  in  Figs.  4-8. 


60  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Table  3. — Litter  sizes,  as  indicated  by  numbers  of  fetuses  and  placental 

SCARS,    IN    SEVEN    SPECIES.      ThE    MEAN    NUMBER    IS    GIVEN,     FOLLOWED    BY    THE 
standard  ERROR,  THE  SIZE  OF  THE  SAMPLE    (  IN  PARENTHESES),  AND  THE   RANGE. 

Species  Fetuses  Placental  scars 

Sorex  vagrans  5.6  ±:  .43  (7),  4-7 

Eutamias  minimus  5.7  ±  .15  (52),  3-8         5.7  ±  .15  (97),  2-10 

Thomomy.s  ialpoides  4.0  ±  .53  ( 12),  2-7         5.0  ±  .16  (66),  2-8 

Peiomyscus   maniculatus  5.6  ±  .12  (111),  2-9       5.9  ±  .31  (34),  2-10 

Clethrionomijs  gapped  6.1  ±  .50  (10),  4-8 

Phenacomijs  intermedins  4.8  ±  .75  (4),  3-6  5.3  ±  .52  (11),  2-9 

Microtus  montanus  5.8  ±  .25  (46),  2-10 

Sorex  cinereus 

Tliis  species  was  uncommon  in  the  quadrats,  but  was  more  abundant  in 
nearby  moist  areas.  Of  the  19  indixiduals  taken  in  the  quadrats,  four  (21  per 
cent)  were  adults.  Breeding  apparently  extends  at  least  into  September,  for 
a  pregnant  yearling  female  with  six  fetuses  was  taken  on  September  3,  and 
males  with  spermatozoa  in  their  testes  were  taken  on  September  2.  Young 
animals  were  recorded  as  early  as  July  14. 

Sorex  vagrans 

Because  of  the  small  sample  of  this  shrew  (N=73)  only  a  general  picture 
of  its  reproductive  cycle  was  gained.  Clothier  (1955)  recorded  pregnant 
S.  vagrans  from  April  1  until  August  8  in  Montana.  The  seven  pregnant 
shrews  from  the  present  study  area  were  taken  from  June  15  to  August  15. 
There  was  no  evidence  that  breeding  females  (yearlings)  had  more  than  one 
litter  in  their  second  siuumer,  and  no  young  of  either  sex  were  recorded  in 
breeding  condition  in  their  first  summer.  Fertile  males  (all  yearlings)  were  taken 
from  June  through  early  September.  Whereas  but  11  per  cent  of  the  31  females 
taken  in  August  and  September  were  yearlings,  56  per  cent  of  the  18  males 
from  this  period  were  yearlings.  Doubtless  few  females  surxive  their  second 
sununer.  The  cessation  of  breeding  in  the  late  summer,  therefore,  may  be 
due  primarily  to  the  loss  of  yearling  females  from  the  population.  The  mean 
number  of  fetuses  for  seven  females  was  5.6  (Table  3).  Clothier  determined 
a  mean  litter  size  of  6.4,  based  on  33  pregnant  females  from  Montana. 

Eutamias  minimus 

This  species  hibernates  from  October  or  November,  depending  on  snow 
conditions,  to  the  time  of  snowmelt  in  spring.  In  each  year  of  the  study,  chip- 
munks were  seen  as  soon  as  any  snowfree  ground  appeared.  Individuals  were 
first  noted  in  the  study  area  as  follows:  June  13,  1965;  May  15,  1966; 
June  7,  1967. 

Seemingly  not  all  female  chipmunks  breed  as  yearlings.  Of  93  females 
taken  in  June  and  July  of  1966,  for  example,  17  (18  per  cent)  had  threadlike 
uteri  and  gave  no  morphological  indication  of  approaching  estrous.  These 
animals  (hereinafter  termed  non-breeding  females)  averaged  appreciably 
lighter  in  weight  than  did  l^eeding  females.  Twenty-six  breeding  females  in 
a  sample  from  June,  1966,  averaged  45.5  grams  (range  37.2  to  58.7),  whereas 
nine  non-breeding  females  averaged  38.2  (33.6  to  42.1).  Perhaps  females  that 
do  not  attain  a  certain  critical  weight  in  their  first  summer  do  not  undergo 


Vaughan — Montane  Small  Mammal  Fauna 


61 


100 

80 
60 
40 


5     20 
< 


o 
a. 


estrous 

pregnant 

post-partum 


20 
40 
60 
80 
lOOh 


o 

Z 


MONTHS 


Fig.  4.    Summary  of  the  reproductive  cycle  of  Eutamias  minimus  based  on 
samples  of  177  males  and  247  females. 


estrous  in  their  second  summer.  If  this  he  true,  large  numljers  of  non-breeding 
females  would  be  expected  in  a  year  following  a  summer  made  unusually 
short  by  a  late  snowmelt  and  by  early  autumn  snows.  Reproduction  in  such 
a  summer  would  be  unusually  low  and  the  population  density  would  be 
strongly  affected. 

The  female  reproductive  organs  undergo  development  in  preparation  for 
breeding  before  the  animals  emerge  from  hibernation,  because  females  are  in 
estrous  soon,  within  roughly  a  week,  after  emergence  from  hibernation.  A 
sample  of  18  females  was  taken  at  the  height  of  snowmelt  (May  18,  1966)  at 
sites  probably  free  of  snow  for  one  to  two  weeks.  Four  of  these  animals  had 
embryos,  nine  had  swollen,  flaccid,  highly  vascular  uteri  and  were  judged  to 
be  in  or  near  estrous,  and  five  were  non-breeding  females.  In  some  cases 
pregnant  females  were  taken  on  soil  saturated  with  water  and  adjacent  to 
snowbanks  up  to  fi\e  feet  deep.  Much  chasing  of  one  animal  by  another 
occurred  during  snowmelt,  an  activity  that  I  assumed  to  be  associated  with 
breeding.  Virtually  all  except  the  non-l^reeding  females  were  pregnant  by 
mid-June.  Of  31  reproductively  active  females  taken  at  this  time,  one  had 
placental  scars  and  the  rest  were  pregnant.  For  these  females  the  mean  size 
of  the  fetuses  was  12  mm.  (crown-rump  length,  full-term  fetuses  measured 
about  25  mm.).  Parturition  was  largely  completed  by  mid-July,  for  of  59 
reproductive  females  taken  in  July  only  four  had  fetuses;  the  rest  had  placen- 
tal scars  and  enlarged  mammae  (see  Fig.  4).  The  mean  number  of  fetuses 
was  5.7;  the  mean  number  of  placental  scars  also  was  5.7  (Table  3).  Litter 
size  as  indicated  by  this  type  of  data  did  not  differ  appreciably  from  year 
to  year. 


62 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


100 


90 
80 
70  h 


^  60 
°50 


O   40 

I— 
2  30 

LU 

u 

OC 

m    20 

O- 

10 

0 


MONTHS 


Fig.  5.  Percentages  of  young  in  the  total  monthly  samples  (representing  1965, 
1966  and  1967)  of  four  species.  Sample  sizes  are  as  follows:  Eutamias  minimus, 
424;  Thomomys  talpoides,  353;  Peromi/scus  maniculatus,  480;  Microtiis  mon- 

tanus,  195. 

Young  chipmunks  apparently  stay  in  the  nest  roughly  30  days  and  begin 
appearing  above  ground  in  early  August.  From  this  time  onward,  they  form 
a  progressi\'ely  larger  proportion  of  the  population.  As  shown  in  Figure  5, 
54  per  cent  of  the  chipmunks  taken  in  August  ( N  =  1 62 )  \\'ere  young; 
for  September  the  corresponding  figure  was  70  per  cent  (N=:61).  There  was 
no  indication  that  any  females  enter  estrous  in  their  first  summer. 

Records  from  1966  illustrate  best  the  reproductive  cycle  of  the  male. 
Snowmelt  was  at  its  height  on  May  18,  when  roughly  half  of  the  surface  of  the 
ground  was  still  imder  snow.  Of  a  sample  of  30  males  taken  on  this  date,  80 
per  cent  were  in  lireeding  condition.  In  these  animals,  die  testes  and  the  caudal 
epididymides  contained  spermatozoa  and  the  seminal  vesicles  were  enlarged 
and  tiugid.  The  non-breeding  males  taken  at  this  time  were  not  individuals 
that  had  yet  to  reach  breeding  condition  and  would  breed  later,  but  rather 
were  males  that  would  not  reach  breeding  condition  during  the  summer.  Each 
monthly  sample  contained  such  reproductively  inactive  yearling  males.  Prob- 
ably all  males  that  would  have  bred  in  1966  were  in  breeding  condition  at 
the  time  of  snowmelt.  The  reproductive  organs  of  males  doubtless  enlarge 
before  the  animals  emerge  from  hibernation,  for  the  testes  and  seminal  vesicles 
averaged  largest  in  recently-emerged  males;  both  testes  and  seminal  vesicles 
regressed  rapidly  in  size  from  mid-July  until  August  (Table  4).  Spermatogene- 
sis was  occurring  in  all  of  the  reproductive  males  taken  in  May  and  mid-June, 
whereas  of  the  12  reproductixe  males  taken  in  July  only  four  had  spermatozoa 
in  the  testes.  A  continued  reduction  in  spermatogenesis  in  August  probably 
occurs,  but  I  lack  sufficient  data  to  illustrate  this  trend. 


Vaughan — Montane  Small  Mammal  Fauna 


63 


Thoniomys  talpoides 

This  species  is  active  throughout  the  winter  beneath  the  snowpack.  In 
February,  1966,  burrows  were  found  extending  at  least  one  foot  above  the 
ground  into  the  snowpack,  which  was  roughly  50  inches  deep.  This  species  is 
a  sharply  seasonal  breeder,  and  its  reproductive  cycle  in  my  study  area  was 
similar  to  the  cycles  described  for  this  species  elsewhere  in  Colorado  by  Hansen 
(1960)  and  Vaughan  (1967). 

In  summer  young  of  this  species  ( animals  that  have  not  been  through  a 
breeding  season)  can  be  distinguished  easily  from  adults  (animals  that  are  in 
breeding  condition  or  have  been  through  a  reproductive  cycle).  In  addition  to 
differences  in  total  weights,  adult  females  ha\'e  pubic  gaps  whereas  young 
do  not  (Hisaw,  1924;  Miller,  1946;  Hansen,  1960).  Adult  males  in  breeding 
condition  have  much  larger  bacula  and  testes  than  do  young.  In  adult  males 
with  regressed  testes,  these  organs  are  flaccid,  reddish,  and  often  partially 
wrinkled,  whereas  those  of  young  males  are  tingid,  yellowish  or  whitish,  and 
have  a  smooth  surface. 

Females  had  a  single  litter  a  year  and  bred  early,  in  May  or  June.  More 
than  half  ( 69  per  cent )  of  the  June-taken  females  already  had  borne  young, 
and  only  10  per  cent  had  not  entered  estrous  (Fig.  6).  In  August  only  seven 
per  cent  of  the  adult  females  (N=31)  were  pregnant;  the  remainder  had 
placental   scars.     No   evidence   was   found   of   females   breeding   in   their   first 


MONTHS 


Fig.  6.    Summary  of  the  reproductive  cycle  of  Thomomys  talpoides  based  on 
samples  of  191  males  and  162  females. 


64  Misc.  Publ.  51,  Unw.  Kansas  Mus.  Nat.  Hist. 

Table  4. — Lengths  of  the  testes  and  seminal  vesicles  in  the  summer 

MONTHS    IN    FOUR    SPECIES.      ThE    MEAN    LENGTH    IS    GIVEN,     FOLLOWED    BY    THE 
standard  error   and  THE   SIZE   OF  THE   SAMPLE    ( IN   PARENTHESES). 

Species  Month  Testes  Seminal  vesicles 

Eutamias  minimus  May  10.1  ±.38  (30)  6.81  ±  .32(30) 

June  8.9  ±.26  (64)  6.2    ±  .19(61) 

July  6.5  ±.26  (49)  4.2    ±  .24(47) 

August  5.2  ±  .20  (28)  3.6    ±  .22  (28) 

September       6.3  ±  .38  (6) 

Thomomtjs  talpoides           May  19.0  ±  .53  (4)  14.8  ±  .07(4) 

June  16.8  ±  .33  (33)  13.8  ±  .75(33) 

July  14.5  ±  .45  (21)  10.8  ±  .70(21) 

August  11.3  ±.42  (31)  8.8  ±  ..57(31) 
September       9.7  ±  .95  (2) 

Pcromiiscus  maniculahis     May  8.4  ±  .19  (8)  8.4  ±  .28(8) 

June  8.7  ±.10  (98)  9.9  ±  .17(99) 

July  9.2  ±  .18  (36)  10.8  ±  .36  (36) 

August  8.9  ±.39  (16)  10.0  ±  .52(16) 

September  5.1  ±  .11  (4) 

Microtus  montaniis  June  10.8  ±  .17  (7)  13.4  ±     .43(7) 

July  11.0  ±.46  (3)  10.5  ±1.4     (3) 

August  9.9  ±  .40  (27)  12.6  ±     .60  (27) 

September  11.8  ±  .67  (2)  12.4  ±     .14(2) 

summer  or   of  females   bearing   two   litters   per   year.     The   mean   number  of 
fetuses  and  placental  scars  per  female  was  5.0  and  4.0,  respectively  (Table  3). 

The  reproductive  organs  of  males  undergo  marked  seasonal  changes  in  size 
and  probably  reach  maximum  dexelopment  at  about  the  time  of  snowmelt 
(late  May  or  early  June),  after  which  the  testes  and  seminal  vesicles  of  adults 
become  progressively  smaller  through  the  rest  of  the  summer  (Table  4).  In 
August  only  30  per  cent  of  the  adult  males  examined  (N=30)  were  fertile. 
Young  males  do  not  become  fertile  in  their  first  summer. 

Young  pocket  gophers  were  first  recorded  in  June,  and  formed  an  ever 
increasing  part  of  tlie  population  during  the  rest  of  the  summer  (Figs.  5-6). 
Seventy  six  per  cent  of  the  pocket  gophers  taken  in  September  (N=58)  were 
young. 

Peromyscus  maniculatus 

Litde  information  bearing  on  the  winter  activity  of  this  species  is  available 
from  the  study  area.  Several  specimens  have  been  taken  in  the  winter  in 
partially  snow-free  areas,  and  during  snowmelt  in  the  spring  some  of  the 
runways  and  nests  beneath  log  piles  and  matted  vegetation  were  probably 
those  used  by  Peromyscus  in  the  winter.  Such  evidence  suggests  that  in  the 
study  area  this  species  remained  active  at  least  part  of  the  winter  beneath  the 
deep  snowpack.  Tracks,  judged  to  be  those  of  Peromyscus,  were  occasionally 
seen  on  the  surface  of  the  snow  in  midwinter,  but  this  animal  probably  was 
active  mostly  in  the  depth  hoar  beneath  the  snowpack. 

Females  breed  soon  after  snowmelt.  About  33  per  cent  of  the  females  in 
a  small  sample  taken  during  snowmelt  in  May  of  1966  were  pregnant  (Fig.  7). 
Because  no  young  animals  were  encountered  at  this  time  it  seems  unlikely  that 
breeding  began  before  snowmelt.  The  percentage  of  pregnant  females  in 
monthly  samples  declined  from  a  high  of  89  per  cent  in  June  to  38  per  cent  in 


Vaughan — Montane  Small  Mammal  Fauna 


65 


100 


80 


60 


LU40 
Q- 

<20 


°   0 

z 

LU 

u20 

UJ 
Q. 

40 

60 

80 

100 


FEMALES 

MALES 

Y/4f, 

post  -  par tum 
pregnant 

J) 

o 

•E 

0) 

o 
> 

-o 

■  0 

J^^^^^j 

B 

W/ 

1^ 

_aj 

,  0 

E 

Li- 

> 

-D 
0 

i 

0) 

a 

01 

c 
o 
Z 

c 
o 

Z 

, 

. 

M  J  J  A  S  N  M  J  J 

MONTHS 


Fig.  7.    Summary  of  the  reproductive  cycle  of  Peromtjscus  maniculatus  l:)ased 
on  samples  of  260  males  and  220  females. 


August  (Fig.  7),  and  no  pregnant  females  were  taken  after  August.  Some 
yearling  females  that  bred  in  May  or  early  June  had  a  second  litter  before 
September  as  females  that  had  olniously  suckled  young  and  were  carrying 
fetuses  were  noted  regidarly.  In  addition,  some,  but  seemingly  not  all,  young 
females  bred  during  their  first  summer;  a  few  small,  partially  gray-pelaged 
females  with  fetuses  were  taken  in  August  of  1965  and  1966.  The  late  summer 
decline  in  the  percentage  of  pregnant  females  in  the  population  seems  due  to 
a  cessation  of  breeding  by  yearling  or  older  females  (and  to  their  death)  and 
to  a  low  percentage  of  young  females  that  breed  in  their  first  summer.  The 
reproductive  burden  for  the  population,  then,  seems  to  be  borne  primarily  by 
the  overwintering  yearling  females. 

Litter  sizes  were  large  in  the  study  area.  The  mean  number  of  fetuses  and 
placental  scars  was  5.6  and  5.9,  respecti\'ely  (Table  3).  These  averages  did 
not  differ  significantly  from  year  to  year.  Large  litters  were  judged  by  Spencer 
and  Steinhoff  ( 1968 )  to  be  typical  of  popidations  of  P.  maniculatus  occupying 
areas  with  short  growing  seasons,  where  the  animals  must  make  the  most  of  a 
short  period  of  food  abundance.  Mean  litter  size  in  the  present  study  was 
considerably  larger  than  the  4.6  determined  by  Jameson  (1953)  for  P. 
maniculatus  occurring  at  elevations  between  3500  and  5000  feet  on  the  western 
slope  of  the  Sierra  Nevada  Mountains  in  California,  where  the  breeding  season 
in  one  year  of  his  study  extended  from  April  through  No\'ember. 

Considering  both  sexes,  the  age  structure  of  the  population  of  P.  manicu- 
latus in  the  present  study  area  imderwent  striking  and  similar  changes  each 
summer  of  the  study.  June  samples  contained  almost  entirely  yearling  ( or 
older)    breeding   animals.     Reflecting   the   sudden   onset   of   breeding  in   early 


66  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

summer,  young  appeared  suddenly  and  in  large  numbers  in  July,  when  they 
averaged  40.5  per  cent  of  the  female  population  and  61.0  per  cent  of  the  male 
population.  Young  animals  formed  a  progressively  higher  percentage  of  the 
population  through  the  rest  of  the  summer  and  autumn  (Figs.  5  and  7).  A  small 
sample  (N=21)  from  November  consisted  entirely  of  young  of  the  year, 
indicating  that  individuals  of  either  sex  rarely  sur\'i\ed  a  second  autumn.  The 
breeding  population  of  May  and  June,  therefore,  probably  consisted  entirely 
of  animals  born  the  previous  summer. 

All  adult  males  taken  in  May  and  nearly  all  (97  per  cent)  taken  in  June 
were  fertile,  but  the  percentage  of  fertile  males  in  the  population  became 
progressively  lower  from  July  (65  per  cent)  a:  d  August  (61  per  cent)  to 
September,  when  none  was  taken.  The  testes  and  seminal  vesicles  of  yearling 
males  do  not  regress  appreciably  in  size  through  the  summer  (Table  4).  Such 
males  probably  remain  continuously  in  breeding  condition  through  the  summer, 
but  by  September  the  reproductive  organs  of  the  few  surviving  yearlings  are 
regressed  and  the  animals  are  not  fertile.  Young  males  born  early  in  the 
summer  begin  producing  sperm  when  the  animals  weight  but  15  grams,  less 
than  75  per  cent  of  the  weight  of  fully  grown  males,  whereas  young  that  are 
born  later  in  the  summer  do  not  become  fertile  until  the  following  summer. 
None  of  the  yoimg  individuals  taken  in  September  and  November,  even  those 
weighing  more  than  15  grams,  was  fertile. 

Clethrionomys  gapperi 
This  was  one  of  the  least  common  rodents  on  the  study  area.  Breeding 
extended  at  least  from  May  through  November,  because  juvenile  mice  were 
taken  at  irregular  intervals  from  May  through  late  December.  Ten  pregnant 
individuals  were  captured  in  July  and  August,  and  the  mean  number  of 
fetuses  was  6.1  (Table  3).  Adult  females  have  more  than  one  litter  per 
summer,  and  young  animals  breed  in  their  first  summer.  Several  small  August- 
taken  females  with  nearly  unworn  teeth  carried  fetuses  or  had  placental  scars 
and  were  judged  to  have  been  born  earlier  in  the  summer.  In  several  small 
males  of  roughly  the  same  age  as  these  females,  spermatozoa  were  abundant 
in  the  testes  and  the  caudal  epididymides. 

Phenacomys  intermedius 

This  rodent  was  never  common  in  the  (luadrats  and  most  individuals  were 
taken  in  late  summer  or  in  autumn.  The  occurrence  of  fertile  males  from  the 
time  of  snowmelt  in  May  in  one  year  (1966)  until  late  August,  and  the 
occurrence  of  pregnant  females  from  June  through  September,  indicates  that 
breeding  takes  place  through  most  of  the  sununer.  Females  are  polyestrous  in 
Colorado  according  to  Warren  (1942).  An  August-taken  female  weighing 
only  16  grams  (roughly  half  the  weight  of  a  fully  grown  female),  and  having 
nearly  unworn  teeth,  was  pregnant,  and  males  of  similar  size  from  mid- 
summer had  sperm  in  their  caudal  epididymides.  This  evidence  indicates  that 
young  animals  born  early  in  the  summer  breed  in  their  first  summer.  Young 
animals  born  later  in  the  year  seemingly  do  not  reach  breeding  condition  their 
first  autunm,  however,  for  none  of  four  November-taken  males,  weighing  from 
17  to  23  grams,  was  fertile,  and  three  young  females  (18  to  29  grams)  taken  at 
the  same  time  had  transparent,  threadlike  uteri.  The  mean  number  of  embryos 
and  placental  scars  was  4.8  and  5.3,  respectively  (Table  3). 


Vaughan — Montane  Small  Mammal  Fauna 


67 


Miciotus  montanus 

Tliis  species  is  not  known  to  hilx'inate.  In  the  stndy  area  abundant  evi- 
dence, in  the  form  of  runways  and  nests  made  in  winter  and  uncoxered  dining 
snowmelt,  indicated  that  montane  \'o!es  were  active  through  the  winter  beneath 
the  snow.  Because  no  specimens  were  taken  in  the  winter,  and  because  few 
were  caught  until  July,  it  is  not  known  if  l:)reeding  occurred  in  the  winter. 

Compared  to  other  rodents  of  the  area,  these  \oles  have  an  unusually  long 
breeding  period.  The  females  taken  earliest  in  the  year  were  caught  during 
snowmelt  on  May  17,  1966;  these  two  individuals  had  placental  scars,  had 
recently  suckled  young,  and  were  taken  at  a  site  that  had  been  free  of  snow 
for  but  a  few  days.  These  animals  iniqnestionably  had  their  litters  beneath  the 
snowpack.  Breeding  continued  from  snowmelt  at  least  through  August,  when 
54  per  cent  of  the  females  taken  carried  embryos  (Fig.  8).  Hall  (1946)  found 
that  in  Nevada  this  species  breeds  through  November,  and  this  may  well  be 
true  for  the  present  study  area.  Young  of  the  year  were  distinguished  from 
yearlings  primarily  on  the  basis  of  weight  ( indi\'iduals  of  more  than  3.5  grams 
were  classed  as  yearlings),  and  it  seemed  that  most  of  the  pregnant  individuals 


100 


80 


60 


^40 

CL 

<20 


Z    0 


20 


> 


3 

o 


Q. 

<u 

Of 


FEMALES 


VTTA 

/jy  /y  post-partum 


pregnant 


t  ft  Yd 


MALES 


u 

> 

Z^o 

u 

a. 

3 

u 

o 

60 

'    Q. 

4J 

c 

80 

.   O 

Z 

100 


c 
o 

z 


-I  J  A  J  J  A  5 

MONTHS 

Fig.  8.    Summary  of  the  reproductive  cycle  of  Micwtus  iiiouiaiuis  based  on 
samples  of  111  males  and  84  females. 


68  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

taken  in  August  were  animals  born  earlier  in  the  summer.  Small  females 
weighing  but  17  grams  had  embryos,  and  some  indi\iduals  weighing  less  than 
20  grams  had  placental  scars.  Of  33  pregnant  females  taken  in  August,  only 
five  (15  per  cent)  were  judged  to  be  yearlings.  Seemingly,  tlien,  as  yearling 
females  that  bred  early  in  the  siunmer  are  increasingly  outnumbered  by  young 
females,  a  progressively  greater  share  of  the  reproductive  burden  shifts  to 
young  females.  The  shaip  rise  observed  in  August  in  two  years  in  the  popula- 
tion density  of  M.  montanus  indicates  that  the  young  breeding  females  that 
form  a  major  part  of  the  relatively  high  August  population  probably  contribute 
considerably  more  yovmg  to  the  population  than  do  the  few  yearling  females 
that  survive  die  winter  and  participate  in  early  summer  breeding.  It  would  be 
expected,  therefore,  that  the  size  of  the  population  in  the  autimin  would  be 
determined  primarily  by  the  survival  rates  of  early  summer  litters. 

The  mean  number  of  fetuses  for  46  pregnant  females  was  5.8,  and  did  not 
differ  significantly  between  summers.  This  is  similar  to  the  litter  size  recorded 
by  Hall  (1946)  in  Nevada. 

Most  of  the  males  trapped  each  summer  from  June  through  August  were 
fertile  (Fig.  8).  Males  began  producing  sperm  whe  i  they  were  roughly  20 
to  25  grams  in  weight,  and  only  three  of  the  37  males  (eight  per  cent)  that 
weighed  more  than  25  grams  were  not  fertile.  In  the  mondis  from  which 
specimens  are  available,  no  important  changes  in  the  mean  sizes  of  the  seminal 
vesicles  and  testes  of  males  heavier  than  25  grams  were  noted  (Table  4), 
indicating  that  males  stay  in  breeding  condition  through  the  summer.  Young 
males  far  outnumber  yearling  males  in  July  and  August,  and  probably  most  of 
the  breeding  from  mid  July  or  August  until  winter  is  done  by  young  males. 
Young  males  first  appeared  in  the  population  in  June  and  formed  an  increasing 
segment  of  the  population  through  the  remainder  of  the  summer  (Fig.  5).  Of 
75  males  taken  in  August,  65  (88  per  cent)  were  judged  to  be  young. 

Discussion 

The  yearly  cycle  of  breeding  and  of  changes  in  the  densities  of 
the  small  mammals  of  the  study  area  has  been  influenced  by  several 
demanding  environmental  features:  (1)  a  short  annual  growing 
season,  averaging  about  thee  months,  and  a  snow-free  period  of 
about  five  months;  (2)  low  temperatures  tlirough  most  of  the  year 
and  a  frost-free  period  of  only  about  60  days;  (3)  annual  "catas- 
trophes," first  in  the  form  of  snowmelt,  attended  typically  by  flood- 
ing of  much  of  the  surface  of  the  ground,  and  second,  in  the  form 
of  periodic  "open"  autumns,  \\'hen  temperatures  approach  zero  but 
no  snow  cover  that  protects  small  mammals  from  the  cold  has 
developed.  Some  characteristic  aspects  of  the  life  cycles  of  sub- 
alpine  small  mammals  probably  haxe  developed  in  response  to 
restrictions  imposed  by  these  environmental  features. 

The  reproducti\  e  cycles  of  the  small  mammals  considered  here 
seem  adapted  to  the  short  growing  season.  Adaptations  include 
large  litters  and  few  litters  per  year.    In  addition,  reproductive 


Vaughan — Montane  Small  Mammal  Fauna  69 

organs  enlarge  and  mature  while  snow  still  covers  the  ground,  and 
breeding  occurs  during  or  immediately  after  snowmelt.  Of  the  five 
most  common  species,  three — T.  talpoides,  E.  minituus  and  S. 
vagram — ha\'e  but  one  litter  per  year.  Another,  P.  maniculaius,  is 
polyestrous,  but  breeding  is  confined  to  the  summer  months  and 
ceases  in  September.  Only  M.  niontanus  has  a  long  breeding  period; 
this  season  coincides  roughly  with  the  snow-free  part  of  the  year. 
Even  this  species,  however,  has  large  litters  relative  to  those  of  other 
voles  in  less  boreal  areas.  Corthum  (1967)  found  that  in  Indiana 
M.  ochrogaster  and  M.  pennsylvanicus  had  litter  sizes  of  3.9  and  4.5, 
respectively,  and  that  breeding  occurred  throughout  the  year.  In 
the  present  study  area,  by  contrast,  M.  monfanus  had  a  litter  size 
of  5.8  and  roughly  a  fixe-month  breeding  period,  giving  this  species 
a  substantially  lower  reproductive  potential  than  that  of  either 
species  studied  by  Corthum.  The  presence  of  a  faii'ly  high  per- 
centage of  non-reproductive  yearling  E.  minimus  in  the  population 
is  noteworthy.  Perhaps  the  time  from  birth  to  first  hibernation  is 
so  short  at  high  elevations  that  some  individuals  born  unusually  late 
do  not  reach  a  critical  stage  of  development  ( size? )  before  hiberna- 
tion and  therefore  do  not  become  sexually  active  in  the  usual 
breeding  season  during  and  immediately  after  snowmelt.  All  of 
the  small  mammals  in  the  present  study  area,  despite  large  litter 
sizes,  ha\'e  fairly  low  reproductive  potentials  because  few  litters  are 
produced  annually.  It  may  be,  however,  that  the  partial  protection 
against  predation  afforded  by  a  continuous,  deep  snow  cover  for 
seven  months  of  the  year  compensates  for  a  low  reproductiv^e  rate. 
The  timing,  length  and  severity  of  snowmelt  probably  has  a 
strong  influence  on  small  mammal  populations.  Most  of  these  mam- 
mals are  forced  to  abandon  low-lying  or  poorly-drained  sites  during 
the  height  of  snowmelt,  and  local,  temporary  shifts  to  drier  sites 
occur  (see  Ingles,  1949;  Hansen,  1962).  Local  concentrations  of 
chipmunks  were  observed  during  the  snowmelt  periods  of  1965 
and  1966.  From  June  14  to  18  of  1965,  for  example,  continuous  deep 
snow  covered  roughly  75  per  cent  of  the  surface  of  the  ground. 
Strips  of  bare  but  saturated  soil  occured  along  the  southern  borders 
of  stands  of  conifers  and  on  certain  south-  or  west-facing  slopes. 
Chipmunks  had  emerged  from  hibernation  in  these  areas  and  were 
concentrated  around  logs,  stumps,  and  rock  piles,  sites  where  the 
animals  could  mostly  avoid  the  saturated  soil  and  running  water. 
Although  these  shifts  occur  when  population  densities  are  lowest, 
'f  the  occupancy  of  these  refuges  is  maintained  locally  into  the  time 


70  Misc.  Publ.  51,  Unw.  Kansas  Mus.  Nat.  Hist. 

of  natality  by  a  prolonged  snowmelt  or  by  an  unusually  heavy  run- 
off, the  resultant  crowding  and  heightening  of  interspecific  competi- 
tion would  be  expected  to  decrease  markedly  the  sur\  ival  of  young. 

In  the  early  summer  of  1965  dead  montane  voles  and  pocket 
gophers  were  found  occasionally  at  low-lying  sites;  Hansen  (1962) 
also  found  dead  pocket  gophers  during  snowmelt  in  southern  Colo- 
rado. This  limited  exidence,  supported  by  our  lack  of  success  in 
trapping  small  mammals  in  areas  saturated  with  water  from  snow- 
melt, suggests  that  animals  that  are  not  able  to  move  to  dry  refuges 
during  snowmelt  are  drowned  or  die  of  exposure.  Jenkins  (1948) 
thought  that  periodic  flooding  of  meadows  in  the  Sierra  Nevadas  of 
California  affected  population  levels  of  M.  montanus.  One  of  the 
difficulties  small  mammals  face  at  such  times  \\'as  made  apparent 
when  a  tent  that  I  pitched  on  frozen  ground  in  the  morning  on 
May  15,  1966,  was  awash  with  flowing  water  when  I  returned  in 
late  afternoon  of  the  same  day.  At  this  time  of  the  year  J.apus 
princeps  and  Sorex  pohistris  were  taken  occasionally  on  open  slopes 
where  they  were  never  recorded  later  in  the  summer. 

Another  transition  period,  that  from  summer  to  winter,  may  be  a 
time  of  stress  for  small  mammals.  As  pointed  out  by  Formozov 
(1946)  andPruitt  (1957,  1960)  certain  small  mammals  (for  example, 
shrews  and  \  oles )  retreat  beneath  the  snow  when  it  reaches  a  depth 
of  six  inches  or  more.  These  animals  spend  the  winter  in  a  moist 
subnivean  en\  ironment  where  the  temperature  remains  fairly  con- 
stant at  close  to  freezing  and  where  they  are  insulated  by  snow  from 
the  violent  climatic  fluctuations  frequently  occurring  above  them. 
If  a  snowpack  does  not  develop  before  extreme  cold  descends,  how- 
ever, small  mammals  (such  as  M.  montanus)  that  seek  plant  material 
for  food  on  the  surface  of  the  ground  are  forced  to  forage  for  frozen, 
low  quality  food  without  protection  from  extreme  cold  that  imposes 
considerable  metabolic  demands.  Appreciable  mortality  may  occur 
at  such  times.  Such  snowless  conditions  persisted  into  early  No\'em- 
ber  of  1965,  and  at  this  time  no  M.  montanus  were  taken  on  the 
surface  of  the  ground  despite  intensixe  trapping.  At  this  time  the 
soil  was  frozen  to  a  depth  of  several  inches,  and  temperatures  of 
-4°F  were  recorded  on  two  nights.  The  unusually  low  populations 
of  M.  montanus  in  the  summer  of  1966  may  have  been  due,  in  part, 
to  this  late  de\  elopment  of  a  snow  cover  in  the  autiunn  of  1965. 
When  snow  does  not  co\'er  the  ground,  voles  may  avoid  difficult 
conditions  on  the  surface  by  confining  much  of  their  activity  to 
beneath  the  ground.   In  early  summer  of  1966  the  few  M.  montanus 


Vaughan — Montane  Small  Mammal  Fauna  71 

taken  were  trapped  in  abandoned  burrows  of  pocket  gophers;  no 
voles  were  taken  at  this  time  on  the  surface  of  the  ground. 

Finally,  what  are  the  most  important  factors  controlling  fluctua- 
tions of  population  densities  of  small  mammals  in  the  study  area? 
Natality  seems  fairly  constant  for  each  species  from  year  to  year. 
A  consideration  of  sources  of  mortality,  consequently,  seems  most 
germane.  Under  some  conditions  abundance  or  quality  (or  both) 
of  food  are  thought  to  influence  mortality  ( Keith  et  al.,  1959;  Schultz, 
1964).  Food  is  seemingly  not  a  limiting  factor  in  the  study  area, 
however,  for  I  considered  the  vegetation  and  the  food  habits  of  the 
mammals  ( unpublished  data )  and  found  ample  food  to  be  available 
each  year.  In  my  judgment,  the  most  important  factors  controlling 
populations  are  the  length,  timing,  and  severity  of  snowmelt  in  the 
spring,  the  time  at  which  a  continuous  snowpack  first  develops  in 
the  autumn,  and  the  intensity  of  predation.  Only  predation  needs 
further  comment. 

The  most  abundant  and  seemingly  most  important  mammalian 
predator  of  small  mammals  in  the  study  area  seemed  to  be  the  long- 
tailed  weasel.  Elsewhere  in  Colorado  I  have  observed  weasels 
chasing  chipmunks  and  have  found  Microtiis  in  weasel  stomachs, 
and  weasels  have  frequently  been  taken  in  pocket  gopher  burrows  in 
Colorado  (Vaughan,  1961).  It  seems  reasonable,  therefore,  that 
weasels  in  the  study  area  were  pre\'ing  mostly  on  terrestrial  rodents. 
As  mentioned  earlier,  weasel  density  fluctuated  in  the  period  of  this 
study,  but  was  high  during  the  summer  of  1967,  a  time  when  density 
of  small  mammals  was  low.  As  pointed  out  by  Pearson  ( 1966 ) ,  the 
most  important  predator-prey  interactions,  in  terms  of  eff^ects  on 
rodent  populations,  are  those  that  occur  when  rodent  density  is  low. 
It  may  be  that  in  montane  situations  the  patterns  of  rodent  cycles 
are  modified  by  periodically  intense  predation  when  weasels  are 
abundant  and  populations  of  preferred  prey,  such  as  pocket  gophers, 
are  low. 

Summary 

A  series  of  electric  fences,  each  fence  enclosing  a  half-acre  quad- 
rat, was  used  to  study  population  densities  of  small  mammals  in  a 
subalpine  area  in  northern  Colorado,  and  information  on  reproduc- 
ti\e  cycles,  based  on  dissections  of  1639  animals,  was  obtained  for 
the  eight  most  common  species. 

The  climate  of  the  study  area  is  sharply  seasonal,  with  snow  on 
the  ground  for  about  seven  months  of  the  year.  Reproduction  and 
population  cycles  of  small  mammals  seem  dominated  by  this  de- 


72  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

manding  climatic  pattern;  in  most  species  the  period  of  natality  is 
short  relatixe  to  the  long  period  of  mortality  and  population  decline. 
Following  are  some  basic  features  of  the  life  cycles  of  the  small 
mammals  of  the  study  area: 

1.  Population  densities  fluctuated  widely,  but  an  annual  cycle 
involving  low  populations  immediately  after  snowmelt  in  early  sum- 
mer and  high  populations  in  late  summer  seemed  characteristic. 

2.  Yearly  and  month-to-month  changes  in  the  relative  and  abso- 
lute densities  of  the  different  species  in  the  summer  were  the  rule. 
Population  densities  of  Microtus  montanus  seemed  most  variable. 

3.  Because  nearly  100  per  cent  mortality  occurred  each  summer 
among  yearling  animals  of  three  species  (Peromyscus  manictdohis, 
Sorex  vagrans,  and  M.  montanus),  and  young  of  the  year  comprised 
most  of  the  autumn  populations  of  the  other  two  of  the  five  most 
common  species  (Thomomijs  tolpoides  and  Eutamios  minimus), 
survival  rates  of  young  animals  was  a  primary  factor  influencing 
population  levels. 

4.  Large  litters  are  characteristic  of  all  species,  and  most  species 
have  short  seasons  of  natality.  Of  the  five  most  common  species, 
three  (T.  tolpoides,  E.  minimus,  and  S.  vagrans)  have  one  litter  a 
year  and  two  (P.  maniculatus  and  M.  montanus)  are  polyestrous. 
In  E.  tninimus  a  substantial  segment  of  the  yearling  population  each 
year  does  not  breed,  a  unique  situation  perhaps  associated  with 
short  growing  seasons. 

5.  The  striking  fluctuations  in  the  densities  of  small  mammals 
seem  most  strongly  influenced  by  the  duration,  severity,  and  timing 
of  snowmelt,  by  the  time  at  which  the  ground  is  first  continuously 
covered  by  snow  in  the  autumn,  and  by  predation,  particularly  when, 
as  occurred  in  1967,  high  populations  of  weasels  and  low  popula- 
tions of  rodents  occur  concurrently. 

Literature  Cited 

Bailey,  V. 

1932.     Mammals  of  New  Mexico.    N.  Amer.  Fauna,  53:1-412,  22  pis.,  58 
figs. 
Browx,  L.  N. 

1967.  Ecological  distribution  of  six  species  of  shrews  and  comparison  of 
sampling  mediods  in  die  central  Rocky  Mountains.  Tour.  Manim., 
48:617-623,  1  fig. 

Clothier,  R.  R. 

1955.  Contribution  to  the  life  historv  of  Sorex  vagrans  in  Montana.  Jour. 
Mamm.,  36:214-221. 

CoRTHUM,  K.  W.,  Jr. 

1967.  Reproduction  and  duration  of  placental  scars  in  the  prairie  ^•ole  and 
the  eastern  vole.    Jour.  Mamm.,  48:287-292. 


Vaughan — Montane  Small  Mammal  Fauna  73 

FoRMOZov,  A.  N. 

1946.  The  covering  of  snow  as  an  integral  factor  of  the  environment  and 
its  importance  in  the  ecology  of  mammals  and  liirds.  Material  for 
Fauna  and  Flora  of  USSR,  New  Series  Zoology,  5:1-141,  26  figs. 
(Russian  with  11  unnumbered  siunmary  pages  in  French). 

GrIXXELL,  J.,  J.    DiXOX,   AXD    T.    M.    LiXSDALE 

1930.  Vertebrate  natural  history  of  a  section  of  northern  California  through 
the  Lassen  Peak  region.  Univ.  California  Publ.  Zool.,  35:v  +  1-594, 
181  figs. 

Grinnell,  J.,  AND  T.  I.  Storer 

1924.  Animal  life  in  the  Yosemite.  .  .  .  Univ.  California  Press,  Berkeley, 
xviii  +  752  pp.,  60  pis.,  65  figs. 

Hall,  E.  R. 

1946.  Mammals  of  Nevada.  Univ.  California  Press,  Berkeley,  xi  +  710  pp., 
frontispiece,  11  pis.,  485  figs. 

Haxsex,  R.  M. 

1960.     Age  and  reproductive  characteristics  of  mountain  pocket  gophers  in 

Colorado.   Jour.  Mamm.,  41:323-335,  August  15. 
1962.     Movements  and  survi\al  of  Thomomtjs  talpoides  in  a  mima-mound 

habitat.   Ecology,  43:151-154,  3  figs. 

Hisaw,  F.  L. 

1924.  The  absorption  of  the  puliic  symphysis  of  the  pocket  gopher,  Geomys 
bursarius   (Shaw).    Amer.  Nat.,  58:93-96. 

IXGLES,  L.  G. 

1949.  Ground  water  and  snow  as  factors  affecting  the  seasonal  distribution 
of  pocket  gophers,  Thomomijs  monticola.  Jour.  Mamm.,  30:343-350, 
2  pis.,  1  fig. 

Jameson,  E.  W. 

1953.  Reproduction  of  deer  mice  (Peromiisctts  maniculatus  and  P.  hoylei) 
in  the  Sierra  Nevada,  California.    Jour.  Mamm.,  34:44-58,  4  figs. 

Jexkins,  H.  O. 

1948.  A  population  study  of  the  meadow  mice  (Microiiis)  in  three  Sierra 
Nevada  meadows.  Proc.  California  Acad.  Sci.,  ser.  4,  26:43-67, 
11  figs. 

Keith,  1.  O.,  R.  M.  Hansen,  and  A.  L.  Ward 

1959.  Effect  of  2,4-D  on  abundance  and  foods  of  pocket  gophers.  Jour. 
Wildlife  Mgt.,  23:137-145,  2  figs. 

Miller,  M.  A. 

1946.     Reproductive  rates  and  cycles  in  the  pocket  gopher.    Jour.  Mamm., 

27:335-358,  6  figs. 

Pearson,  O.  P. 

1966.  The  prey  of  carni\'ores  during  one  cycle  of  mouse  abundance.  Jour. 
Anim.  Ecol.,  35:217-233,  9  figs. 

Pequegnat,  W.  E.,  and  D.  H.  Thompson 

1949.  An  electric  fence  for  studving  rodent  populations.  Jour.  Entomol. 
Zool.,  41  (3):  1-37,  6  pis.,  9  figs. 

Pruitt,  W.  O.,  Jr. 

1957.  Obserx  ations  on  the  microclimates  of  some  taiga  mammals.  Arctic, 
10:131-138. 

1960.  Animals  in  the  snow.    Scientific  Amer.,  203(6)  :61-68. 

SCHULTZ,  A.  M. 

1964.  The  nutrient-recovery  hypothesis  for  arctic  microtine  cycles.  II.  Eco- 
systemic  variables  in  relation  to  arctic  microtine  cycles.  Pp.  57-68 
in  Grazing  in  terrestrial  and  marine  environments  ( D.  J.  Crisp,  ed.). 
British  Ecol.  Soc,  Symp.  no.  4,  Blackwell,  Oxford,  England. 

Spexcer,  a.  W.,  and  H.  W.  Steixhoff 

1968.  An  explanation  of  geographic  variation  in  litter  size.  Jour.  Mamm., 
49:281-286,  1  fig. 


74  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Vaughan,  T.  a. 

1961.     Vertebrates  inhabiting  pocket   gopher  burrows   in  Colorado.    Jour. 

Mamni.,  42:171-174. 
1967.     Two  parapatric  species  of  pocket  gophers.    Evolution,  21:148-158, 

5  figs. 

Warrex,  E.  R. 

1942.     The   mammals   of   Colorado,    their   habits   and    distrilxition.     Uniw 
Oklahoma  Press,  Norman,  xviii  +  330  pp.,  frontispiece,  50  pis. 


THE  SPECIES  PROBLEM  IN  THE  THOMOiMYS 

BOTTAE— THOMOMYS  UMBRINUS  COMPLEX  OF 

POCKET  GOPHERS  IN  ARIZONA 

BY 

Donald  F.  Hoffmeister 

In  southwestern  United  States  and  northern  Mexico,  there  are 
two  common,  small  pocket  gophers,  Thomomijs  hottae  and  Tho- 
momijs  umbrinus.  These  are  allopatric  nearly  everywhere  (Fig.  1). 
Where  the  two  occur  together  or  adjacent,  there  has  always  been 
considerable  doubt  as  to  which  "species"  any  population  should  be 
referred.  This  has  been  true  in  southwestern  Texas  ( Davis  Moun- 
tains), southern  Arizona,  eastern  Coahuila,  and  eastern  Sinaloa. 
The  two  "species"  have  been  regarded  as  conspccific  on  some  occa- 
sions and  as  distinct  species  on  other  occasions.  In  1959,  Hall  and 
Kelson  regarded  the  tw^o  as  conspecific  basing  this  on  the  remarks 
of  Hoffmeister  and  Goodpaster  (1954:95). 

Gophers  in  the  Huachuca  Mountains,  Arizona 
Hoffmeister  and  Goodpaster,  working  in  southern  Arizona 
(1954),  regarded  Thomomijs  hottae  and  T.  umbrinus  as  conspecific 
for  a  variety  of  reasons.  Mearns  (1897:719)  described  a  pocket 
gopher  "from  the  aspen  and  spruce  zone  at  the  summit  of  the 
Huachuca  Mountains"  that  had  the  characters  of  Thomomys  um- 
brinus. Our  collecting  on  the  summit  of  these  mountains  in  a  clear- 
ing in  the  aspens  and  fir  revealed  an  isolated  population  of  gophers 
that  had  the  features  of  T.  bottae,  and  since  two  characters  sup- 
posedly valuable  in  distinguishing  bottae  and  umbrinus  were  vari- 
able— namely  color  and  number  of  pectoral  mammae — we  con- 
cluded that  in  the  Huachuca  Mountains,  Arizona,  it  is  "advisable  to 
refer  all  material  to  one  species,  for  which  the  earliest  name  is  T. 
bottae  [but]  by  this  we  do  not  mean  to  imply  that  T.  umbrinus  is 
necessarily  a  synonym  of  T.  J)ottae"  (Hoffmeister  and  Goodpaster, 
1954:95).  Further  collecting  in  the  Huachuca  Mountains  indicated 
that  at  intermediate  elevations,  between  the  summit  and  the  bahada, 
a  distinct  kind  of  gopher  exists  and  these  are  referable  to  Thomomys 
un\hrinus.  Re-examination  of  the  87  gophers  that  Hoffmeister  and 
Goodpaster  studied  in  the  Huachucas  indicates  that  seven  are  refer- 
able to  T.  umbrinus,  and  these  seven  are  from  the  oak-belt  on  the 

(75) 


76 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


I  BOTTAE 

i   ; 


Fig.  1.  Distribution  of  Thomomys  hotiac  and  Thomomijs  umhiimis  in  western 
North  America.  Specimens  from  parts  of  the  range  marked  with  a  question  mark 
have  been  regarded  as  more  bottacAike  than  umbrmus-]ike  by  some  authors. 


Mountains.  The  type  of  Mearns'  gopher  indeed  is  umbriniis-\ike. 
We  doubt  that  it  came  from  the  summit  of  the  higher  peaks  of  the 
Huachucas.  In  any  event,  it  is  clear  that  in  the  Huachuca  Moun- 
tains, Arizona,  there  are  two  kinds  of  pocket  gophers — T.  bottae  and 
T.  unihrinus. 


HOFFMEISTER ThOMOMYS  IN  ARIZONA 


/  / 


Fig.   2.     Range   of   Thomomys   bottae   and    Thomomijs   iimbrimis   in   Arizona. 
Sycamore   Canyon,    Patagonia    Mountains,   the   area   of   study,   is   immediately 

above  the  tip  of  the  arron'. 


Gophers  in  the  Patagonia  Mountains,  Arizona 

In  Arizona,  Thomoimjs  hottae  and  Thomomys  umhrinus  are  allo- 
patric  in  three  counties,  Cochise,  Santa  Cruz,  and  Pima  (Fig.  2), 
and  are  to  be  found  within  a  half-mile  or  less  of  each  other  at  more 
than  10  localities.   One  of  the  most  interesting  of  these  is  Sycamore 


78 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Canyon  on  the  west  side  of  the  Patagonia  Mountains,  Santa  Cruz 
County.  In  the  higher  parts  of  the  Patagonias,  timbrinus-]ike  go- 
phers are  found;  in  the  Santa  Cruz  River  Valley,  to  the  west  of  these 
mountains,  the  gophers  are  l)ottae-\i\:e.  T.  bottae  is  found  up  the 
Canyon  as  far  as  4400  feet  elevation;  T.  umhriniis  down  the  canyon 
to  4100  feet  elevation  (Fig.  3).  In  some  places  along  this  canyon, 
the  two  kinds  of  gophers  are  to  be  found  "together"  and  in  other 
places  seem  to  be  intermediate,  as  if  intergrading  or  hybridizing. 
In  southern  Arizona  where  the  two  kinds  occur  so  close  together, 
there  is  no  character  displacement  and  it  is  difficult  to  select  charac- 
ters useful  in  differentiating  the  two  kinds. 

Previous  Methods  for  Distinguishing  the  Two  Kinds 

Bailey  (1915),  in  his  revision  of  the  pocket  gophers  of  the  genus 
Thomomijs,  recognized  a  Thomomys  umhrimis  group  but  did  not 
formally  characterize  it.  However,  features  of  one  pair  of  pectoral 
mammae,  dichromatic  color,  and  short  skull  were  mentioned.  Nelson 
and  Goldman  (1934)  more  clearly  defined  T.  umbriniis,  pointing 
out  that  the  species  is  "normally  recognizable  by  the  diflering  num- 


Ifl 

f 

1       f       9 

3 

4  5b'     °-  «  ;■- 

] 

2 

.  .  .-..  ■♦■: 

.•.♦.1^ '  ''''■'■' 

> 

• 

,  «.  , 

■.■■:-: 

5700 
5300 
4300 
4500 
4100 
3700 


Fic;.  3.  Collecting  sites  in  Sycamore  Canyon.  Elevation  in  feet  is  indicated  in 
the  upper  transect.  The  dotted  line  indicates  the  road  up  the  Canyon.  The 
scale  is  for  one  mile.  The  Santa  Cruz  River  Valley  is  to  the  left,  \\'est  of 
Arizona  Highway  89.    The  crest  of  the  Patagonia  Mountains  it  to  the  light. 


HOFFMEISTER ThOMOMYS  IN   ARIZONA 


79 


bcr  of  pectoral  mammae  (one  pair  in  umhrinus,  two  pairs  in  hottae) 
and  in  the  summation  of  cranial  details,  none  of  which  is  very 
trenchant"  (p.  105).  Blair  (1939)  characterized  T.  umhrinus  by 
"small,  rounded  skull,  small  size,  and  weak  forefeet."  Davis  (1946: 
266)  in  characterizing  T.  umhrinus  employed  the  variation  in  "the 
margin  of  the  anterior  base  of  the  zygoma  where  it  meets  the 
frontal."  This  character  is  shown  in  Figs.  8-9.  Goldman  (1947:6) 
characterized  the  TJwmomys  umhrinus  group  by  certain  color  and 
cranial  features,  but  not  by  the  number  of  pectoral  mammae.  Baker 
(1953),  in  studying  the  pocket  gophers  in  Coahuila,  used  the  char- 
acter of  the  maxillo-frontal  suture,  the  position  of  the  lacrimal,  and 
the  procumbency  of  the  incisors  in  separating  umhrinus  and  hottae. 
Anderson  (1966:196)  noted  that  17  characters  are  seemingly  useful 
in  distinguishing  the  two  species  in  Chihuahua,  but  specifically 
mentioned  only  seven  of  the  17  characters,  and  these  only  in  a  casual 
way.  Dunnigan  (1967:142-144)  pointed  out  several  characteristics 
useful  in  separating  Sinaloan  specimens  of  the  two  species. 


in 

—  11-1 

o 

1       f       9 

3 

4    5-6-J       i^     . 

2 

.  -  -♦.■ 

,»  .»,1^'-V  ■■•-•• 

:' 

t  ' , ,  ■,■■■■ 

■.■,■-■.•, 

-  ■:-; 

5700 
5300 
4900 
4500 
4100 
3700 


bottae 

intermediote 

umbrlnus 

37 


0 

0 
dorsal     coloration 


6  7  2  0  0 
10  2  1  1 
0     9   6  7       5 


1 

i 
24 


Fig.  4.  Distribution  of  gophers  with  hottae  color  (1  and  2  of  Fig.  6),  inter- 
mediate, or  r(/)i/jn'jiH.s-color  (5  and  6  of  Fig.  6).  The  37  gophers  from  below 
Sycamore  Canyon  (lower  left  corner)  all  are  hottae-]ike  in  color;  two  of  the 
wnbiimis  (lower  right  corner)  are  intermediate  in  color.  On  the  map  at 
locality  6,  for  example,  two  specimens  were  like  bottae  in  color,  two  were 
intermediate,  and  six  resembled  umhrinus. 


80 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


The  characters  employed  by  these  various  workers  are  not  useful 
individually  in  separating  the  two  kinds  in  Arizona  nor  is  a  com- 
bination of  their  characters  highly  reliable  either.  Coloration,  in- 
cluding the  presence  or  absence  of  a  broad,  dark  dorsal  stripe,  is 
variable,  although  often  useful.  Most  specimens  regarded  as  T. 
umhrinus  have  one  pair  of  pectoral  mammae  and  most  T.  hottae 
have  two  pair,  although  there  is  variation  in  this  character  as  well. 
Among  approximately  200  specimens  of  T.  hottae  from  Arizona,  on 
which  field  examination  for  number  of  mammae  was  made,  five 
have  only  one  pair,  not  two.  Nelson  and  Goldman  (1934:117)  I 
recorded  at  least  one,  possibly  three,  other  T.  hottae  with  a  single 
pair  of  pectoral  mammae.  Variation  in  the  shape  of  the  maxillo-  j 
frontal  suture  and  in  the  position  of  the  lacrimal  at  one  locality  in 
Arizona  is  summarized  in  Figs.  8-9.  Some  of  the  variation  in  cranial 
and  external  measurements  in  the  population  in  Sycamore  Canyon, 
Arizona,  is  shown  in  Table  1.  The  variation  from  throughout  the 
range  of  T.  hottae  or  T.  umhrinus  would  show  the  overlap  in  l 
measurements  to  be  even  greater.  ' 


<n 

f 

,        «        9 

1 

4     5    6-'        °       « 

1 

2 

.■.-.■♦.■: 

.■**-"'^''  ' 

• 

■...,•-,■:  :■:■:■■:■■ 

:  :;;; 

5700 
5300 
4900 
4500 
4100 
3700 


bottae 
umbrinus 

23 

b     » 


3    12  0      0      0 


1    II  4  5      4       I 


pectoral    mammae 


N 

1      s 

0^ ^ 

12     U 


Fig.  5.  Distribution  of  gophers  with  two  pair  of  pectoral  mammae  (hottae)  or 
one  pair  (umhrinus).  Of  the  females  from  the  Santa  Cruz  River  Valley,  23  had 
the  mammae-count  of  hottae,  none  the  count  of  umhrinus  (see  arrow  at  lower 
left).  All  12  in  the  Patagonia  Moimtains  had  the  count  of  umhrinus.  Counts 
for  localities  1  to  3  are  not  available;  for  localities  7  to  11,  the  counts  are  all 
iimbrinus-hke.    Note  the  variation  at  localities  4,  5,  and  6. 


HOFFMEISTER ThOMOMYS  IN  ARIZONA  81 

Special  Problem  in  Sycamore  Canyon,  Patagonia  Mountalns 
At  the  mouth  of  Sycamore  Canyon,  the  dorsal  coloration  of 
specimens  is  typically  like  that  of  T.  hottae  (1  and  2  in  Fig.  6). 
Part  way  up  Sycamore  Canyon,  some  have  the  color  of  T.  hottae, 
some  of  T.  umhrimis,  and  some  are  clearly  intermediate.  This  is 
true  at  localities  4,  5,  6,  7,  and  8,  as  shown  in  Fig.  4.  All  23  speci- 
mens of  Thomomys  examined  for  mammae  from  below  Sycamore 
Canyon  had  the  number  typical  of  T.  hottae.  All  12  specimens 
examined  from  the  Patagonia  Mountains  had  the  number  typical  of 
T.  umhrimis.  Proceeding  up  the  Canyon,  at  localities  4,  5,  and  6, 
some  specimens  have  one  pair,  others  two  pair  of  pectoral  mammae 
(Fig.  5).  Field  work  in  1968  produced  an  additional  two  females 
from  locality  5,  not  listed  on  the  map  (Fig.  5),  one  of  which  had 
two  pair  of  pectoral  mammae,  the  other  one  paii-.  The  coloration 
of  the  dorsum  did  not  always  correlate  with  the  number  of  mammae, 
nor  did  the  size  of  the  baculum  or  the  skull.  Within  Sycamore 
Canyon,  some  pocket  gophers  are  typical  T.  hottae  or  typical  T. 
umhrimis,  but  some  have  characteristics  of  both,  especially  at  inter- 
mediate elevations,  as  between  localities  3  to  8. 

The  ecological  differences  of  the  two  species,  if  any,  must  be 
subtle.  At  the  lower  end  of  the  Canyon  T.  hottae  lives  in  the  loose, 
mostly  rock-free  soils  without  heavy  stands  of  oak.  T.  umhrimis  at 
the  upper  end  of  the  Canyon  lives  in  rocky,  shallow  soil  and  in 
subterranean  runways  that  by  necessity  are  of  such  small  diameter 
that  it  is  nearly  impossible  to  place  a  trap  within  the  burrow.  Often 
they  are  within  stands  of  oak.  There  is  no  clear  break  between  these 
two  ecological  situations  at  intermediate  localities  within  Sycamore 
Canyon.  Rocky  soil  and  friable  soil  does  occur  very  close  together, 
and  within  openings  in  oak  forests,  within  the  Canyon. 

Materials  and  Methods 

Since  Sycamore  Canyon  appeared  to  be  an  area  of  hybridization  or  sec- 
ondary intergradation,  11  collecting  stations  were  established  between  elevations 
of  4100  and  5400  feet  (Fig.  3).  In  1968,  substations  were  established  below 
locality  3,  at  0.85  mi.  (one  specimen),  0.4  mi.  (one),  and  0.2  mi.  (two). 
These  are  included  with  locality  3.  Some  stations  (4  to  7)  were  only  1/10  of 
a  mile  apart.  A  total  of  84  specimens  was  collected  between  1959  and  1968 
from  the  stations  in  Sycamore  Canyon  together  with  44  specimens  from  the 
adjacent  Santa  Cruz  River  Valley  ( listed  as  hottae  in  subsequent  discussions ) 
and  33  specimens  from  the  higher  parts  of  the  Patagonia  Mountains  ( Hsted  as 
wnhiinns).  Twenty-five  additional  specimens  from  Sycamore  Canyon  were 
borrowed  from  the  University  of  Arizona.  These  specimens  were  assigned  to 
one  of  the  11  collecting  sites,  although  in  some  cases  they  probably  do   not 


82 


Misc.  Publ.  51,  Untiv.  Kansas  Mus.  Nat.  Hist. 


Fig.  6.    Dorsal  coloration  in  "typical"  bottae  (1  and  2)  and  "typical"  nnihiinus 
(5  and  6).    Intermediate  banding  is  indicated  by  3  and  4. 

precisely  correspond.  Tliis  piobably  is  true  for  tbose  listed  as  locality  6,  which 
appear  out  of  place  in  Table  2. 

Most  females  were  checked  for  number  of  mammary  glands  before  skin- 
ning, although  some  of  the  earliest  collections  and  borrowed  specimens  lack  this 
information.  Bacula  were  cleared  in  KOH,  stained  with  alazarin  red,  and  the 
height  of  the  head  was  measured  with  an  ocular  micrometer. 

Color  was  valued  from  light  to  dark  as  one  to  six.  Specimens  without  a  dark 
dorsal  stripe  were  coded  as  one,  those  with  a  pronounced  black  stripe  as  six, 
and  intermediates  were  coded  between   (see  Fig.  6).    In  hottae-hke  animals, 


Fig.  7.  Method  of  taking  cranial  measurements.  AA',  basilar  length;  BB', 
greatest  length  of  nasal;  CC',  greatest  zygomatic  breadth;  DD',  mastoid  breadth; 
EE',  least  interorbital  breadth;  FF',  maxillary  toothrow;  BG',  length  of  rostrum; 
HH',  breadth  of  rostrum.  Features  for  aging  are  X,  basioccipital-lxisisphenoid 
suture;  Y,  supraoccipital-exoccipital  suture. 


HOFFMEISTER ThOMOMYS  IN   ARIZONA  83 

without  a  prominent  dorsal  stripe,  the  sides  have  an  ochraceoiis  color;  in 
umbrinus-hke  animals,  and  nearl>'  all  those  color-coded  4,  5,  and  6,  have  a 
somewhat  purplish  cast  o\'erlying  the  ochraceous  on  the  sides. 

Cranial  measurements  were  taken  as  indicated  in  Figure  7.  Only  animals 
judged  to  be  adult  were  used.  Specimens  were  placed  in  four  age-groups. 
In  the  oldest  group,  the  basioccipital-basisphenoid  suture  is  closed,  the  supra- 
occipital-exoccipital  suture  closed,  and  the  temporal  ridges  prominent  in  both 
sexes,  but  less  so  in  females.  In  the  next  younger  group,  the  ]>asioccipital- 
basisphenoid  suture  is  not  closed  (X  in  Fig.  7),  the  supraoccipital-exoccipital 
suture  mostly  closed  (Y  in  Fig.  7),  and  the  temporal  ridges  present  but  only 
moderately  prominent.  The  two  younger  groups  were  not  used.  Males  and 
females  were  treated  separately  because  of  the  secondary  sexual  variation 
(Table  1). 

The  position  where  the  maxillo-frontal  suture  comes  in  contact  with  the 
lacrimal  varies  (Fig.  8),  and  this  variation  is  given  a  value  of  1  for  hottaeAike, 
2  for  intermediate,  and  3  or  3.5  for  umbrinusAike.  The  shape  of  the  maxillo- 
frontal  suture,  from  concave  (Fig.  9)  to  convex,  is  given  a  similar  value  of  1, 
2,  and  3  or  3.5  from  /;o^f«c'-like  to  umbrinusAike. 

Characters  Useful  in  Distinguishing  Gophers 
IN  Southern  Arizona 

In  separating  bottacAike  and  ti>nJ)iinus-]ike  gophers  in  southern  Arizona, 
the  following  color  and  morphological  features  were  most  useful.  They  are 
probably  of  decreasing  value  in  the  order  given  except  for  the  analysis  of 
chromosomes. 

Dorsal  coloration. — T.  umbrimis  has  a  dark  band  running  from  the  tip  of 
the  nose  to  the  base  of  the  tail.  The  width  of  the  band  varies.  Usually  it  is 
broadest  in  the  middle  of  the  back  and  also  extends  completely  between  the 
eyes  and  ears.  Immediately  below  the  black  band,  the  lateral  coloration  is 
ochraceous,  usually  with  an  o\erlay  of  color,  hard  to  describe,  that  gives  an 
iridescent  or  purplish  cast.  T.  bottae  has  the  lateral  coloration  extending  over 
the  back,  with  a  slight  sprinkling  of  dark  hairs  in  the  middorsal  area.  Rarely  is 
there  an  indication  of  a  dark  band  (see  Fig.  6). 

Number  of  pectoral  mammae. — T.  umhrinus  has  only  one  pair  of  pectoral 
mammae;  T.  bottae  two  pairs.  Variation  from  this  arrangement  is  discussed 
above. 

Length  of  bactilum. — In  eastern  Santa  Cruz  County,  the  greatest  length  of 
the  baculum  in  T.  bottae  is  usually  more  than  11.0  mm.;  in  T.  umbrinus,  usually 
less  than  10.2  mm. 

External  measurements. — For  eastern  Santa  Cruz  County,  body  length  in 
adult  male  T.  bottae  is  usually  more  than  152  mm.;  in  T.  umbrinus,  less  than 
149;  in  females,  there  is  overlap,  with  T.  bottae  more  than  138.  T.  umbrinus 
less  than  143.  The  hind  foot  is  usually  longer  than  29.5  mm.  in  male  T.  bottae, 
less  than  29.2  in  male  T.  umbrinus;  more  than  27.0,  rather  than  less,  in  female 
T.  bottae.  There  is  overlap  in  length  of  tail,  but  in  male  T.  bottae  it  is  usually 
more  than  60  mm.,  in  T.  umbrinus,  less. 

Cranial  measurements. — The  most  useful  cranial  measurements  were  basilar 
length,  zygomatic  breadth,  mastoid  breadth,  and  length  of  nasals.  There  is  less 
overlap   in   these   measurements    between    males   of   the    two   species    than   in 


84 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Table  1. — Measurements  (ix  millimeters),  valltes  (position  and  shape  of 

MAXILLO-FROXTAL    SUTURe),    AND    NUMBERS    (mAMMAe)    FOR    VARIOUS    FEATURES 

OF  MALE  AND  FEMALE  Thomomys  bottac  AND  T.  umbrinus  from  the  Santa 
Cruz  River  Valley  near  Nogales  and  the  Patagonia  Mountains,  respec- 
tively.   The  mean  is  followed  by  plus  or  minus  one  standard  devlatioxN. 


Measurement 

Males 

Females 

bottae 

umbrinus 

bottae 

umhrimis 

(xN=ll) 

(N=9) 

(N=26) 

(N  =  17) 

Length  of 

body  

..163.6  ±5.57 

137.8  ±5.59 

152.7  ±7.30 

130.5  ±12.52 

Length  of 

tail    

.  70.1   ±5.17 

55.1   ±3.41 

63.8  ±4.08 

52.1  ±  4.78 

Length  of 

hind  foot  

_  31.0  ±0.77 

27.1   ±1.05 

29.7  ±1.38 

25.2  ±  0.97 

Basilar 

length  

-  35.S6±0.68 

31.04±0.91 

32. 89  ±1.06 

29.40±   1.11 

Length  of 

nasals   

..  14.39±0.63 

12.69±0.45 

12.76±0.65 

11.45±  0.62 

Zygomatic 

breadth    

-  25.87±1.00 

22.31  ±0.81 

23.50±0.66 

21.02±  0.84 

Mastoid 

breadth - 

-  21.31±0.51 

18.38±0.52 

19.87±0.64 

17.36±  0.55 

Interorbital 

breadth    

.     6.76±0.08 

6.73±0.29 

6.85±0.26 

6.69±  0.18 

Length  max. 

toothrow    

_     8.22±0.30 

7.63±0.41 

8.30±0.43 

7.62 ±  0.25 

Length  of 

rostrum     

.  17.52±0.42 

15.07±0.39 

15.72±0.63 

14.04±  0.71 

Breadth  of 

rostrum     

.     8.68±0.35 

7.91±0.32 

7.95±0.29 

7.34±  0.26 

Suture  position 

( 1-3.5 )     

.     1.18±0.40 

2.33±0.50 

1.58±0.50 

2.71±  0.47 

Suture  shape 

(1-3.5)     

.     1.18±0.40 

2.22±0.45 

1.23±0.38 

2.71  ±  0.56 

Dorsal  color 

(1-6)    

.     1.27±0.47 

4.67±1.87 

2.04±0.82 

5.41  ±  0.79 

Length  of 

baculum  

.  11.97±0.51 

9.19±0.52 

Height  of 

baculum    ( base 
Pairs  pectoral 

)     2.08±0.25 

1.90±0.19 

mammae 

2 

1 

females.  In  males  in  eastern  Santa  Cruz  County,  basilar  length  is  usually  more 
than  34.4  mm.  in  T.  bottae,  less  than  32.9  in  T.  umbrinus;  zygomatic  breadth 
more  than  23.9  mm.  rather  than  less  than  23.9;  mastoid  breadth  more  than 
20.2  mm.  rather  than  less  than  19.4;  nasals  usually  longer  than  13.4  mm.  rather 
than  shorter.  In  females  the  overlap  is  indicated  by  the  fact  that  basilar  lengtli 
is  usually  more  than  30.8  mm.  in  T.  bottae,  less  than  31.6  in  T.  umbrinus; 
zygomatic  breadth  is  more  than  22.2  mm.  in  T.  bottae,  less  than  22.7  in 
T.  urnbrinits;  mastoid  breadth  is  more  than  18.5  mm.  in  T.  bottae  rather  than 
less;  nasals  are  longer  than  11.5  mm.  in  T.  bottae,  shorter  than  12.6  in 
T.  umbrinus. 

Position  and  shape  of  the  maxillo-frontal  suture. — The  position  and  shape  of 
the  maxillo-frontal  suture  are  discussed  above  and  in  Figures  8-9.  The  position- 
\alues  given  in  these  figures  indicate  the  variability  of  each  character.    Further- 


HOFFMEISTER ThOMOMYS  IN  ARIZONA 


85 


umbrinus 
bottae 


1 

2 
3 
4 
5 
6 
7 
8 
9 

10 

11 


42 

49 

5 
1 
8 

7 

;6 

9 
9 
6 
2 
2 
2 


4 
1 

5 
1 
2 
1 
J 
0 
0 
0 
0 


; 

n 
f 

e 
r 
m 
e 
d 
I 

a 
f 
e 


5 
16 

0 

0 

2 

2 

1 

2 

2 

0 

0 

1 

0 


1 
0 
7 
2 
12 
5 
6 
3 
2 
1 
1 


0 
0 
0 
2 
I 
I 
0 
3 
0 
0 
7 


Fig.  8.  Position  of  the  maxillo-frontal  sutiire  relative  to  the  lacrimal  (see 
arrow).  In  hottacAike  gophers,  the  suture  reaches  the  lacrimal  near  the  center 
of  that  bone,  but  in  (//»/;)/;H/.s-like  gophers  the  suture  reaches  the  lacrimal  near 
the  medial  side.  The  42  specimens  of  tnnhriniis,  49  of  hotfae,  and  those  from 
each  of  the  11  localities  in  Sycamore  Canyon  (left  colimin)  are  scored  as  to 
whether  they  are  bottae-Mke,  ttmbrinttsAike,  or  intermediate. 


more,  it  is  frequently  difficult  to  ascertain  if  the  maxillo-frontal  suture  is  straight 
or  convex,  or  its  precise  position  relati\ e  to  the  lacrimal. 

Length  and  breadth  of  rostrum. — In  males  of  T.  bottae,  the  length  and 
breadth  of  the  rostrum,  respectively,  is  usually  more  than  16.3  mm.  and  8.3  mm. 
rather  than  less;  in  females,  usually  more  than   14.9  and  7.6  rather  than  less. 

CJiromosomes. — Analysis  of  the  chromosome  numl^er  and  karyotypes  in  the 
Patagonia  Mountains  has  been  made  by  Patton  and  Dingman  (1968),  and  the 
following  summary  is  from  their  report.  "Typical"  specimens  of  T.  bottae  have 
2n=:76,  no  acrocentrics  or  minute  chromosomes,  with  all  being  metacentrics, 
submetacentrics,  or  subtelocentrics.  "Typical"  specimens  of  T.  umbrinus  have 
2n=78,  62  acrocentrics  and  minute  chromosomes,  with  the  remainder  being 
biarmed.  However,  populations  of  T.  bottae  from  within  100  miles  of  the 
Patagonias  display  considerable  variation  in  the  number  of  acrocentrics — one 
population  from  less  than  50  miles  away  has  18  acrocentrics.  One  wonders  how 
great  the  variation  in  number  of  acrocentrics  might  be  throughout  the  range 
of  T.  bottae,  judging  from  the  amount  of  morphological  variation.  Only  three 
populations  of  one  subspecies  of  T.  tmd^rinus  was  studied  and  the  variation 
was  minor. 

Employment  of  characters. — Each  character  listed  above  displays  overlap 
between  the  two  species.  In  dorsal  coloration,  number  of  pectoral  mammae,  and 


86 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


bof  foe-  like 

umbrinus 

42 

5 

botfae 

48 

27 

I 

5 

4 

2 

2 

0 

3 

8 

4 

4 

7 

2 

5 

16 

0 

c 

6 

9 

1 

0 

c 
o 

5 

7 

9 

1 

U 

o 

8 

6 

0 

« 

0 

c 

0 

9 

3 

0 

E 

U) 

o 

o 

10 

2 

0 

>. 

o 

t/) 

a. 

11 

2 

0 

/ 

n 
f 
e 
r 
m 
e 
d 

rP 

=4' 

i 
a 

tf 

A 

f 

u  m 

brinus 

-like 

1 

33 

8 

13 

1 

0 

1 

0 

3 

J 

2 

2 

2 

12 

2 

5 

2 

3 

7 

5 

0 

3 

7 

0 

7 

1 

3 
0 

0 
0 
0 
1 

2 
7 
3 
0 
0 
7 
0 


Fig.  9.    Shape  of  the  maxillo-frontal  suture   (arrow).    It  is  concave  in  hottae, 
straight  or  convex  in  uuibrinus,  or  intermediate.    The  specimens  are  scored  as 

in  Fig.  8. 


length  of  baeulum,  there  is  least  overlap.  Scatter  diagrams  were  prepared  for 
nearly  all  combinations  of  characters,  including  color  as  coded,  but  there  was 
no  obvious  means  of  separating  T.  hottae  and  T.  umbrinus  in  the  area  of  this 
study  by  use  of  these  diagrams.  The  use  of  all  characters  in  combination  is 
necessary  if  the  two  are  to  be  distinguished. 

Using  a  computer  program  BIMD  05  as  modified  by  Charles  Thaeler 
(1968),  employing  discriminant  coefficients,  16  characters  were  used  in  the 
analysis  for  males  (including  length  and  height  of  head  of  baeulum)  and  15 
for  females.  The  discriminant  function  coefficient  was  determined  for  each 
character  and  these  summed  for  each  specimen.  The  values  for  male  T.  hottae 
range  from  5.822  to  6.450  and  for  male  T.  timhiinus  from  4.102  to  5.018. 
The  values  for  the  male  pocket  gophers  from  Sycamore  Canyon  localities  1  to  11 
(Fig.  10)  range  from  3.514  to  6.581.  Some  are  well  within  the  range  of  one  or 
the  other  species  and  some  fall  between  the  two.  The  \alues  for  female  T. 
hottae  range  from  4.452  to  5.444  and  for  female  T.  umhrinus  from  2.064  to 
2.870.  The  values  for  the  female  pocket  gophers  from  localities  1  through  11 
range  from  1.614  to  5.373.  The  values  for  the  males  and  females  are  corrected 
to  correspond  and  both  are  plotted  on  one  graph  (Fig.  10).  T^^'enty-nine  speci- 
mens in  Sycamore  Cainon  tall  between  the  range  of  the  samples  of  the  t\vo 
species  and  this  suggests  that  these  specimens  are  hybrids  or  intergrades  (see 
Fig.  10  and  Table  2). 


HOFFMEISTER ThOMOMYS  IN  ARIZONA 


87 


Independently  of  the  al)o\e  procedure,  another  approach  was  undertaken. 
The  relationships  of  the  specimens  were  determined  by  taxonomic  distance 
coefficients  ( Sokal  and  Sneath,  1963:147,  300).  The  same  characters  were 
employed  as  in  the  analysis  liy  the  discriminant  function  coefficients.  In  using 
these  taxonomic  distance  coefficients,  the  lesser  the  distance  between  two,  the 
greater  their  phenetic  relationship.  Each  specimen  was  judged  to  be  a  T.  hottae, 
aberrant  bottae,  hybrid,  aberrant  umhriniis,  or  T.  itmbrinus.  The  results  of  this 
analysis,  listed  b\-  localit>',  are  shown  in  Table  2.    Note  that  the  results  from 


_     „.4-- 


ry-'-^ 


n 

1_J  L        j_ , 


r 

"■'1 

:1 1"  " 

1  )»^  , 

I 

umbrinus 


-hybrid 


Fig.  10.  Distribution  of  pocket  gophers  from  Sycamore  Canyon,  Patagonia 
Mts.,  as  determined  by  discriminant  function  coefficients.  Light  diagonal  lines 
are  T.  hottae.  dots  are  T.  tnnhrimis.  and  hea\y  diagonal  lines  represent  various 
localities   within    the   Canyon.     Each    square    represents    one   individual.     The 

hybrids  are  indicated. 

this  analysis  do  not  differ  greatly  from  those  of  the  analysis  by  the  discriminant 
function  coefficients  or  the  next  analysis  made.  It  should  be  noted  that  the 
sample  sizes  may  differ  in  the  different  analyses. 

Still  another  procedure  was  employed  to  segregate  T.  bottae  or  T.  iimhnmis. 
A  lower  or  upper  limit  of  the  range  of  variation  was  set  at  two  standard  devia- 
tions. For  males,  11  characters  were  used — color,  length  of  baculum,  three 
external,  four  cranial,  and  position  and  shape  of  the  maxillo-frontal  suture. 
For  females,  10  characters  were  used,  with  the  one  for  length  of  baculum 
omitted.  Each  character  for  each  specimen  was  scored,  as  bottae  if  it  fell  within 
two  standard  de\iations  of  the  known  sample  of  T.  bottae,  or  T.  unibrintis,  or 
intermediate.  The  result  was  that  a  male  specimen  might  be  scored  as  10/11 
hottae  (=10  characters  bottae-\\ke)  and  1/11  intermediate,  and  regarded  as 
T.  hottae;  or  4/11  bottae,  5/11  umbrinus,  2/11  intermediate,  and  regarded  as  a 
hvbrid.  The  results  of  this  analysis  are  given  in  Table  2.  Although  the  con- 
cordance of  this  system  with  the  others  is  not  so  great  as  between  the  first  two, 
and  one  \\'ould  not  expect  it  to  be,  there  still  is  considerable  similarity. 


Discussion 

A  large  number  of  the  specimens  from  Sycamore  Canyon  are 
intermediate  between  T.  hottae  and  T.  umbrinus  (Table  2).  These 
are  from  "intermediate"  localities  also.  It  seems  apparent  that  there 
is  hybridization  or  intergradation  between  the  two  species.  From 
localities  1  through  11  within  the  Canyon,  29  specimens  are  "inter- 
mediate" as  determined  by  the  discriminant  function  analysis. 
Using  the  taxonomic  distance  coefficients,  21  specimens  are  "inter- 


88  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Table  2. — Placement  of  pocket  gophers  from  the  11  localities  in  Syca- 
more Canyon  (nos.  1  through  11)  as  to  one  or  the  other  species  or  as 
"intermediate"  using  three  different  methods  of  analysis. 

T.  bottae         123456789        10       II    T.umbrinus 


Discriminant  Function  Analysis 


bottae 
hybrid 
umhrinus 

44 

5 

2 
2 

6 
9 

2 

8 
1 

5 
10 

3*- 

4       1 
13       9       7     12 

Ta.xonomic 

Distance  Coefficient 

bottae 

"aberrant' 

hybrid 

"al^errant' 

umhrinus 

37 

5 

2 
2 

2 
2 
9 

2 

1 

5 
3 
1 

4 

2 

11 

1* 

3* 
1 

2       3               3 
13     10       7       8 

Two 

Sta 

ndard  Deviations 

bottae 

"alierrant' 

hyl^id 

"aberrant' 

umhrinus 

36 

1 

5 

1 

1 

2 

6 

1 
5 
1 
3 

6 

1 
2 

4 
1 

12 

1* 
2* 
4 

1               1 
13     10       7     10 

1 

32 


26 


26 


"  These  may  not  be  from  locality  6,  and  perhaps  should  be  identified  with  a  lower  locality. 

mediate"  and  an  additional  22  are  regarded  as  aberrant  (possibly 
hybrids,  or  possibly  referable  to  a  species ) .  Using  the  2  SD  method, 
21  specimens  are  regarded  as  "intermediate"  and  an  additional  nine 
as  aberrant. 

In  parts  of  the  Patagonia  Mountains,  Thomomijs  bottae  and 
T.  uml)rinus\  which  are  usually  allopatric,  ha\'e  not  evolved  effective 
genetic  isolation.  Where  the  two  groups  of  populations  make  sec- 
ondary contact,  either  sympatrically  or  in  allopatric  zones,  intro- 
gressive  hybridization  occurs.  In  Sycamore  Canyon,  at  locality  1 
and  below,  all  specimens  are  T.  bottae;  at  locality  8  and  above,  all 
are  T.  umbrinus.  From  localities  2  through  7,  a  distance  of  less  than 
three  miles,  there  are  to  be  found  T.  bottae,  hybrids,  and  T.  um- 
brinus. Most  of  the  hybridization  occurs  between  locality  6  and  a 
place  one-half  mile  below  locality  3,  a  distance  of  about  one  and  a 
half  miles.  This  is  approximately  the  same  area  where  Patton  and 
Dingman  ( 1968 )  found  four  hybrids  in  Sycamore  Canyon.  The 
zone  of  hybridization  seems  well  marked  but  narrow. 

On  the  basis  of  karyology,  nine  specimens  from  Sycamore  Can- 
yon are  hybrids  or  backcross  hybrids.  These  are  from  a  narrow  zone 
also — localities  2  through  7.  Four  of  these  hybrids  have  been  re- 
ported upon  and  discussed  by  Patton  and  Dingman  ( 1968 ) .  One  is 
regarded  as  a  Fi  hybrid,  three  as  backcrosses.  Five  specimens  that 
were  karyotyped  by  M.  R.  Lee  and  E.  Zimmerman  indicate  that 


HOFFMEISTER ThOMOMYS  IN  ARIZONA  89 

three  are  Fi  hybrids  and  two  are  probably  backcrosses.  Of  these 
five,  three  are  from  locaHty  3  and  two  from  two  tenths  of  a  mile 
below  locality  3. 

A  basic  consideration  is  whether,  in  southern  Arizona,  T.  hottae 
and  T.  timbrinus  are  distinct  species.  Samples  of  gophers  taken  from 
the  Santa  Cruz  River  Valley  and  from  the  upper  parts  of  Sycamore 
Canyon,  a  distance  of  only  eight  miles,  would  leave  little  doubt  but 
that  the  samples  were  from  distinct  species  if  collections  were  not 
available  from  the  inter\'ening  area.  But  such  questions  arise  as 
these:  (1)  Are  all  gophers  that  live  on  poor,  rocky  soil  of  small  size 
(umhrimis-like)?  Often  they  are  small,  but  do  not  have  the  other 
characteristics  of  umbrinus.  (2)  Do  small-sized  gophers  frequently 
have  but  one  pair  of  pectoral  mammae  (iimbrinus-like)?  There  is 
no  evidence  that  such  is  the  case.  ( 3 )  Do  gophers  that  live  at  higher 
elevations  ha\'e  a  darker  dorsal  coloration?  Frequently  at  higher 
elevations  the  soils  are  darker  and  so  are  the  gophers,  but  the  dark 
color  is  not  necessarily  restricted  to  a  middorsal  band  as  in  um- 
brinus. T.  bottoe  and  T.  umbrinus  are  judged  on  chromosomal 
distinctiveness  as  two  species  in  Arizona  (Patton  and  Dingman, 
1968 ) .  Yet,  in  view  of  the  survey  of  the  variation  in  the  karyology 
of  only  a  few  T.  bottae  in  Arizona,  are  the  differences  great  enough 
to  regard  these  as  distinct  species?  Probably,  but  further  evidence 
will  be  needed. 

With  all  of  the  evidence  here  marshalled,  it  still  is  a  matter  of 
interpretation  as  to  whether  the  populations  of  bottae  and  umbrinus 
in  southern  Arizona  are  subspecificalh'  or  specifically  distinct.  I 
would  interpret  the  evidence  as  indicating  that  they  are  distinct 
species,  with  introgressive  or  allopatric  hybridization  occurring  in 
a  narrow  zone  in  one  place  in  southern  Arizona,  and  possibly  in 
several  other  areas  where  the  range  of  the  two  species  come  to- 
gether, and  especially  where  it  has  been  difficult  to  assign  specimens 
to  one  or  the  other  species.  The  narrowness  of  the  zone  of  hybrid- 
ization, in  spite  of  the  extensive  interchange  of  genetical  material 
within  this  zone,  would  suggest  specific  difFerentiation.  Admittedly, 
the  suggested  presence  of  backcross  hybrids,  based  on  karyology, 
and  the  seeming  vigor  of  these  and  other  hybrids,  may  argue  against 
specific  status. 

Does  this  hybridization  that  occurs  at  one  place,  and  perhaps 
several  places,  represent  the  incipient  formation  of  two  species  or 
rather  the  infrequent  intermingling  of  genetical  material  of  two 
established  species?  Although  difficult  to  decide,  I  would  guess  the 


90  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

latter.  The  allopatric  distribution  of  umbrinus-\ikc  gophers  and 
hottae-\ike  gophers  over  large  areas  ( Fig.  1 )  suggests  that  these  two 
evolved  independently.  In  at  least  one  place  where  the  two  species 
come  together,  allopatric  hybridization  occurs  with  nearly  complete 
genetic  interchange  and  backcrossing,  yet  the  parental  populations 
maintain  their  identities. 

The  situation  in  Sycamore  Canyon,  Arizona,  seemingly  fits  the 
definition  of  Remington's  suture-zone  hybridization,  judging  from 
the  summary  of  Yang  and  Selander  (1968:139).  There  is  extensive 
hybridization  in  a  narrow  zone  between  morphologically  dissimilar, 
allopatric  populations.  There  is  backcrossing  with  no  reproductive 
isolation  or  hybrid  inferiority.  Despite  interbreeding  and  back- 
crossing,  introgression  of  genes  into  the  parental  population  is 
limited.  Where  the  species  Tlwmojmjs  hottae  and  T.  umbrinus  live 
sympatrically,  be  it  Sycamore  Canyon  or  the  few  other  places 
throughout  their  ranges,  even  though  there  may  be  only  partial 
ecological  and  reproductive  isolation  between  the  two,  the  area  of 
contact  is  so  small  that  swamping  does  not  occur. 

Summary 

In  places  in  southeastern  Arizona,  Thomomijs  hottae-\\Ve  go- 
phers and  Thomomys  wnbrinus-like  gophers  ocur  together  or  within 
a  few  yards  of  each  other.  Characters  useful  in  distinguishing  these 
two  species  elsewhere  will  not  serve  here  for  some  animals  have 
some  characters  of  both  species.  Three  analyses  of  the  population 
in  Sycamore  Canyon,  west  side  of  the  Patagonia  Mountains,  were 
made  using  discriminant  function  analysis,  taxonomic  distance  co- 
efficients, and  two  standard  de\'iations  of  each  side  of  the  mean. 
Characters  employed  were  dorsal  coloration,  number  of  pectoral 
mammae,  length  of  baculum,  various  external  and  cranial  measure- 
ments, and  position  and  shape  of  the  ma.xillo-frontal  suture  relative 
to  the  lacrimal.  These  analyses  indicate  that  somewhere  between  21 
and  29  specimens,  out  of  109,  are  intermediates  or  hybrids.  Within 
a  3  mile  area  of  contact,  at  intermediate  elevations  within  Sycamore 
Canyon,  extensive  hybridization,  not  intergradation,  occurs  between 
the  two  species,  T.  bottoe  and  T.  umbrinus. 

Acknowledgments 

Numerous  persons  liiue  assisted  and  advised  with  the  computer  analysis  of 
these  data.  I  should  especially  like  to  thank  Dr.  Charles  Thaeler,  New  Mexico 
State  University,  and  Dr.  Da\  id  Eades  and  Dr.  Richard  Selander,  University  of 
Illinois.    John  Lynch  measured  most  of  the  skulls;  Harry  Henriksen  prepared 


HOFFMEISTER ThOMOMYS  IN   ARIZONA  91 

the  illustrations.  Dr.  James  Patton,  then  at  the  University  of  Arizona,  loaned 
the  material  he  reported  upon.  Earl  Zimmerman  and  Dr.  Raymond  Lee 
prepared  chromosome  smears  and  photomicrographs.  Without  the  skill  and 
perseverance  in  collecting  by  Woodrow  and  Lois  C.oodpaster,  this  study  would 
not  ha\e  been  possible.  The  National  Science  Foundation  supported  much 
of  this  research  on  grant  GB-1432. 

Literature  Cited 

Anderson,  S. 

1966.  Taxonomy  of  gophers,  especially  TJwiuomijs  in  Chihuahua,  Me.xico. 
Syst.  Zool.,  15:189-198,  6  figs. 

Bailey,  V. 

1915.  Revision  of  the  pocket  gophers  of  the  genus  Thomomys.  N.  Amer. 
Fauna,  39:1-136,  8  pis.,  10  figs. 

Bakeh,  R.  H. 

1953.  The  pocket  gophers  (genus  Thomomys)  of  Coahuila,  Mexico.  Univ. 
Kansas  Publ,  Mus.  Nat.  Hist.,  5:499-514,  1  fig. 

Blair,  W.  F. 

1939.  New  mammals  from  Texas  and  Oklahoma,  with  remarks  on  the 
status  of  Thomomys  texemis  Bailey.  Occas.  Papers  Mus.  Zool.,  Univ. 
Michigan,  403:1-7. 

Davis,  W.  D.,  and  H.  K.  Buechner 

1946.  Pocket  gophers  (Thomomys)  of  the  Davis  Mountains,  Texas.  Jour. 
Mamm.,  27:265-271,  2  figs. 

Dunnigan,  p.  B. 

1967.  Pocket  gophers  of  the  genus  Thomomys  of  the  Mexican  state  of 
Sinaloa.    Radford  Rev.,  21:139-168,  4  figs. 

Goldman,  E.  A. 

1947.  The  pocket  gophers  (genus  Thomomys)  of  x\rizona.  N.  Amer. 
Fauna,  59:1-39,  2  figs. 

Hall,  E.  R.,  and  K.  R.  Kelson 

1959.  The  mammals  of  North  America.  Ronald  Press,  New  York,  l:xxx 
-f  546  +  7.9,  illustrated. 

HOFFMEISTER,  D.   F.,  AND  W.  W.   GOODPASTER 

1954.  The  mammals  of  the  Huachuca  Mountains,  southeastern  Arizona. 
Ilfinois  Biol.  Monog.,  24  :v  +  152,  27  figs. 

Mearns,  E.  a. 

1897.  Descriptions  of  six  new  mammals  from  North  America.  Proc.  U.S. 
Nat.  Mus.,  19:719-724. 

Nelson,  E.  W.,  and  E.  A.  Goldman 

1934.  Pocket  gophers  of  the  genus  Thomomys  of  Mexican  mainland  and 
bordering  territory.   Jour.  Mamm.,  15:105-124. 

Patton,  J.  L.,  and  R.  E.  Dingman 

1968.  Chromosome  studies  of  pocket  gc^^hers,  genus  Thomomys.  I.  The 
specific  status  of  TJiomomi/s  umbrinus  (Richardson)  in  Arizona. 
Jour.  Mamm.,  49:1-13,  8  figs. 

SoKAL,  R.  R.,  AND  P.  H.  A.  Sneath 

1963.  Principles  of  numerical  taxonomy.  W.  H.  Freeman  and  Co.,  San 
Francisco  and  London,  xvi  -|-  359  pp.,  illustrated. 

Thaller,  C.  S.,  Jr. 

1968.  An  analysis  of  three  hybrid  populations  of  pocket  gophers  ( genus 
Thomomys,).    Evolution,  22:543-555,  3  figs. 

Yang,  S.  Y.,  and  R.  K.  Selandfr 

1968.  Hybridization  in  the  grackle  Qiiisccdus  qiiiscula  in  Louisiana.  Syst. 
Zool.,  17:107-143,  19  figs. 


RESTOS  FOSILES  DE  MAMIFEROS  DE  TLAPACOYA, 
ESTADO  DE  MEXICO  (PLEISTOCENO-RECIENTE) 

FOR 

TicuL  Alvarez 

Las  excavaciones  realizadas  por  el  personal  del  ]3epartamento 
de  Prehistoria,  Institute  Nacional  de  Antropologia  e  Historia,  en  el 
cerro  de  Tlapacoya,  Mexico,  durante  los  afios  de  1966  y  1967,  han 
proporcionado  una  gran  cantidad  de  huesos  de  vertebrados,  siendo 
por  ello  hasta  el  momento,  una  de  las  mas  iniportantes  localidades 
para  el  conocimiento  de  la  Fauna  Pleistocenica  de  la  parte  central 
de  Mexico,  conocida  como  Valle  de  Mexico.  Estas  excavaciones  son 
tambien  muy  valiosas  porque  por  primera  vez  se  tienen  de  esta 
region  de  Mexico,  fechas  de  C^\  que  nos  dan  una  idea  de  cuando 
dichas  faunas  existieron  y  haran  posible  futuras  correlaciones,  tanto 
con  descubrimientos  del  pasado,  como  con  los  descubrimientos 
venideros. 

La  fauna  estudiada  y  que  dio  origen  a  este  trabajo  se  limita  a 
los  mamiferos  encontrados  en  las  excavaciones  del  aiio  de  1966, 
realizadas  por  los  Arqueologos  Elizabeth  y  Mickael  Goodlife, 
quienes  laboraron  para  el  Departamento  de  Prehistoria.  Ademas 
de  los  fosiles  de  mamiferos,  se  rescataron  gran  cantidad  de  huesos 
de  aves,  que  son  objeto  de  minucioso  estudio  por  parte  del  Dr. 
Allan  R.  Phillips  del  Institute  de  Biologia  de  la  Universidad  Na- 
cional Autonoma  de  Mexico,  y  del  Dr.  Pierce  Brodkorb  de  la 
Universidad  de  Florida. 

Las  excavaciones  se  realizaron  en  la  falda  sureste  del  cerro  de 
Tlapacoya,  a  1.5  kilometres  al  sur  del  pueblo  del  mismo  nombre 
(113°  02'  40"  long.  E  y  19°  17'  49"  lat.  N),  estado  de  Mexico  y 
que  a  su  vez  se  encuentra  a  26  kilometres  al  Sureste  del  centre  de 
la  Ciudad  de  Mexico  (Fig.  1).  Los  hallazgos  se  llevaron  acabo  en 
dos  calas  e  trincheras  denominadas  respectivamente  alfa  y  beta, 
situadas  a  37  metres  una  de  la  etra. 

La  cala  alfa  tenia  30  metres  de  extension  y  anchura  entre  dos  y 
tres  metres.  La  capa  principal  donde  se  encontraron  los  huesos 
fue  denominada  con  los  numeres  XLI  a  XLII.  La  trinchera  beta 
niidio  53  metres  de  large  y  de  dos  a  cuatro  metres  de  anche.  La 
capa  en  la  que  se  encontro  mayor  numere  de  restes  oseos  fue  de 
cenizas  volcanicas,  que  se  encentraba  en  contacte  con  la  roca 
madre.    La  capa  de  ceniza  se  denemino  con  el  numere  XLVL 

(93) 


94 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


9  ?^ 


S8»  SO' 


19°  30 


I  9« 


' 

\  _^ 

-'"   "^^ 

r'  ' ' 

^\     HIDALGO         \ 

MEXICO 

,- 

,'                                            / 
___•               x      , / 

/ 

©Toluca 

(  MexicoP 

'v                 TLAXCALA 

Ayotia            '-_^^^'\  TIaxcala 
°  Tlapacoya  \                ^^^- 

\ 
\ 

^ 
1 

• 

Distrifo        j 
Federal\ 

\                                \ 

\PUEBLA    V        "^--^ 

;                    Pueblal^ 

) 

MORELOS 
1 



/    1 

I9">  30' 


19° 


99° 


9e°30> 


9e»55' 


I9°I8' 


Kilometre 


IS'IS' 


98°  55' 


Fig.  1.    Mapa  del  area  y  sitio  exacto  de  las  excavaciones. 

La  estratificacion  do  las  trincheras  es  muy  complicada,  habien- 
dose  dividido  hasta  en  54  diferentes  estratos  (Goodlife  y  Goodlife, 
1969);  sin  embargo,  solo  dos  son  importantes  desde  el  punto  de 
vista  paleontologico,  porque  en  ellos  se  encontraron  la  mayor  parte 
de  los  restos  oseos  (Fig.  2). 


Alvarez — Pleistocene  Fossils  from  Mexico 


95 


El  mas  mock'ino  se  encontro  en  la  trinchera  alfa  y  se  supone 
haya  sido  un  hogar  del  honibre  prehistorico  por  la  gran  acumula- 
cion  de  huesos,  principalmente  de  mamiferos  grandes,  asi  como  por 
haberse  enconlrado  un  area  linipia,  con  carbon  y  grandes  piedras 
alrededor  (Goodlife  y  Goodlife,  1966).  Los  restos  de  carbon  de 
este  lugar  ban  sido  fechados  por  C^^  en  24,000  ±  4000  aiios  antes 
del  presente  (Haynes,  1967). 

El  otro  estrato  en  que  se  encontraron  gran  numero  de  huesos, 
es  ceniza  xolcanica  en  contacto  con  la  roca  del  cerro.  No  existe 
fecha  de  material  de  esta  capa,  pero  suponemos  que  sea  alrededor 
de  los  30,000  anos,  ya  que  la  muestra  de  una  capa  superior  a  la  de 
cenizas  data  de  hace  24,200  +  500. 


Madera  carbonizada 
22,200+  2.600 


Huesos  de  mami'feros  en 
limo  amorlllo 
(Copas  XLI-XLII  <  ) 


Madera  descompuesta 
24,200+  500 


Superficie 
artificial 


Ceniza  volcdnica  de  grano 

mediono  ,  amarillo  claro  y  gris 


Roca  volca'nfca   de  grano   fino 

con  fenocristales   negros 

Aglomerado  volcanico  grueso  de 

cenizas,  cantos   y  huesos 
(Capos  XLVI   /3  ) 

Fig.  2.   Estiatigrafia  generalizada  de  los  calas  alfa  y  beta  de  Tlapacoya  (66-1), 
basada  en  la  de  Haynes  (1967),  con  modificaciones. 


Los  restos  de  esta  capa  estan  constituidos  principalmente  por 
aves  y  pequeiios  mamiferos,  aunque  tambien  existen  algunos  restos 
de  animales  de  mayor  tamaiio. 


96  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Los  ejemplares  que  forman  la  base  de  este  estudio  se  encuentran 
catalogados  en  el  Laboratorio  de  Paleozoologia,  del  Departamento 
de  Prehistoria,  Institiito  Nacional  de  Antropologia  e  Historia  (DP). 
Dos  mandibulas  de  la  nueva  especie  de  Odocoileus  aqui  descrita, 
se  encuentran  en  la  coleccion  de  Paleontologia  de  Vertebrados  de 
la  Smithsonian  Institution.  Si  no  se  especifica  de  otra  manera,  todas 
las  medidas  estan  en  milimetros. 

Agradezco  al  Prof.  Jose  Luis  Lorenzo,  Jefe  del  Departamento 
de  Prehistoria  su  sugerenica  y  ayuda  para  realizar  este  trabajo.  A  la 
arqueologa  Lorena  Mirambell,  encargada  del  proyecto  de  Tlapa- 
coya,  por  su  valiosa  cooperacion  para  aclarar  diferentes  problemas 
estratigraficos  y  de  otra  indole.  A  los  arqueologos  Elizabeth  y 
Mickael  Goodlife  por  ser  ellos  quienes  realizaron  las  excavaciones 
y  pusieron  todo  su  empeiio  por  rescatar  los  materiales  oseos  en  el 
mejor  estado  posible. 

En  especial  agradezco  al  Dr.  Clayton  Ray  de  la  Smithsonian 
Institution  por  sus  valiosas  sugerencias  respecto  a  la  identificacion 
del  material  estudiado.  Al  Dr.  J.  R.  Macdonald  del  Museo  de  Los 
Angeles  County  por  el  prestamo  de  varias  piezas  de  "Sangamona" 
de  San  Josecito,  Nuevo  Leon. 

Lista  de  Especies 

Chiroptfra 
Phyllostomatidae 

Mormoops  megalophylla  (Peters,   1864) 
Material— Humero  (928  DP). 

El  hueso  fosil  encontrado  en  Tlapacoya,  no  difiere  en  nada  de  los  usados 
para  su  comparacion,  por  lo  ciial  se  identifica  conio  Mormoops  megalophylla, 
murcielago  muy  abundante  en  las  zonas  bajas  de  Mexico,  pero  que  no  se  ha 
registrado  del  Valle  de  Mexico  en  epocas  lecientes. 

La  capa  de  donde  proviene  tiene  antiguedad  mayor  a  los  24,000  anos  antes 
del  presente  y  se  encuentra  justamente  arriba  de  las  capas  formadas  por 
cenizas  de  la  erupcion  volcanica. 

Lagomorpha 
Leporidae 

Sylvilagus  cunicularius  (Waterhouse,   1848) 

Material— CaMneo  (929  DP). 

El  calcaneo  se  ha  identificado  como  perteneciente  a  la  especie  S.  ctmicu- 
larius,  con  base  en  su  tamaiio,  que  es  realmente  mayor  que  el  de  S.  ftoridamis 


Alvarez — Pleistocene  Fossils  from  Mexico  9/ 

o  que  Romerolagus,  los  otros  dos  Leporidae  que  existen  en  el  valle  de  Mexico. 
S.  cunicitlarius  es  todavia  un  conejo  muy  aliundante  en  todo  el  sur  del  altiplano 
mexicano,  extendiendo  su  distribucion  hasta  las  costas  de  Sinaloa.  Medidas: 
longitud  del  hueso  26.4;  ancho  maximo  del  mismo,  11.3. 

Como  fosil  se  ha  registrado  de  Veracruz  ( Dalquest,  1961).  El  calcaneo  de 
Tlapacoya  tiene  antiguedad  aproximada  de  12,000  anos  ye  se  encontro,  for- 
mando  parte  de  el  un  conjunto  de  huesos  de  un  hogar,  en  la  cala  alfa. 

Sylvilagus  floridanus  (J.  A.  Allen,  1890) 
Material. — Fragmento  de  mandiiiula  con  pm2-m2   (930  DP). 

Al  contrario  del  calcaneo  anterior,  este  material  se  identifico  como  S. 
floridanus  por  se  menor  tamaiio.  Al  compararlo  con  mandibulas  de  floridanus 
no  encontramos  diferencias  palpables  en  cuento  a  forma  y  tamano,  no  siendo 
asi  con  S.  cunicularius  que  si  bien  la  forma  no  difiere,  el  tamaiio  de  cunicularius 
es  notablemente  mas  grande. 

S.  floridanus  es  un  animal  que  ha  sobrevivido  a  la  presion  demograiica  del 
Valle  de  Mexico  y  todavia  es  posible  encontrarlo  con  cierta  facilidad. 

La  mandibula  de  S.  floridanus  fue  encontrada  en  la  capa  de  ceniza  de  la 
erupcion  volcanica,  que  se  calcula  de  una  antiguedad  mayor  a  24,000  aiios. 

RODENTIA 

Geomyidae 

I 
Pappogeomys  sp. 

Material. — Fragmento  de  mandibula,  con  i-m2,  sin  procesos;  ibid.,  sin 
incisi\'o;  fragmento  maxilar  y  premaxilar  con  incisivo   (936-938  DP). 

Debido  a  lo  fragmentado  y  escaso  del  material,  nos  ha  sido  imposible  poder 
determinar  la  especie  de  los  restos  de  tuzas  encontrados  en  Tlapacoya,  con 
antiguedad  mayor  a  los  24,000  anos;  sin  embargo,  si  creemos  que  el  genero 
este  bien  determinado  ya  que  los  incisivos  superiores  presentan  el  surco  medio 
tan  caracteristico  de  este  genero. 

A  juzgar  por  el  tamano,  los  fosiles  de  Tlapacoya  no  pertenecen  a  ninguna 
de  las  especies  que  actualmente  viven  en  el  Valle  de  Mexico,  de  los  cuales 
Tappogeomys  merriami  es  la  mas  grande  y  la  que  habita  actualmente  en  el 
area  de  donde  se  rescataran  los  fosiles,  la  otra  especie  P.  tylorhinus  es  mas 
pequena  y  habita  al  noroeste  del  Valle  de  Me.xico.  En  la  Figura  3  se  compara 
el  ancho  de  los  incisivos  inferiores  y  la  longitud  del  premolar  del  fosil,  con 
ejemplares  de  P.  merriami  y  P.  tylorhinus  viendose  que  el  fosil  queda  por 
abajo,  de  las  medidas  de  los  adultos  de  merriami  y  tylorhinus  y  solo  es  igual 
que  algunos  ejemplares  muy  jovenes  de  las  especies  citadas. 

Cualitativamente,  el  caracter  que  mayor  significado  tiene  y  por  el  cual  no 
hemos  querido  asignar  los  ejemplares  de  Tlapacoya  a  una  de  las  especies 
vivientes,  principalmente  de  las  del  Valle  de  Mexico,  es  que  el  surco  medio 
de  la  cara  frontal  del  incisivo  esta  mas  afuera  de  la  mitad  de  la  cara  del  incisivo, 
de  tal  manera  que  la  porcion  interna  (1.4)  es  el  127.2  por  ciento  de  la  porcion 
externa  ( 1.1 )   (Fig.  4A).   En  20  ejemplares  de  P.  merriami  tomadas  al  azar  (en 


98 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


cuanto  a  edad  y  sexo)  la  media  es  96.3  por  ciento,  con  111.1  a  80.0  de  varia- 
cion  y  en  18  ejeniplares  de  P.  tylorhimis  es  de  97.6  (102.0-85.2)   por  ciento. 

Otro  caracter  se  presenta  en  el  premolar  inferior,  en  el  cual  el  angulo 
entrante  externo  que  separa  a  los  dos  lobulos  que  forman  el  diente,  presenta 
una  proyeccion  interna  y  hacia  adelante  (Fig.  4B).  Esta  proyeccion  no  la 
hemos  encontrado  tan  bien  marcada  en  ninguno  de  los  50  ejemplares  de 
P.  merriami  y  P.  tylorliinus  que  hemos  examinado,  aunque  si  se  encuentra  en 
muchos  de  ellos  una  indicacion  de  dicha  proyeccion.  Sin  embargo,  considera- 
mos  que  la  variacion  en  este  respecto  puede  ser  mayor,  por  lo  que  su  signifi- 
cado  taxonomico  se  desvanece. 

Merriam  (1895)  coloca  dentro  de  los  generos  que  poseen  el  surco  medio 
del  incisivo  en  la  parte  externa  a  Geomys  y  Zygogeomys,  dos  generos  que 
presentan  dos  surcos  en  vez  de  uno  como  sucede  entre  otros  Pappogeomys. 
El  autor  antes  citado  no  menciona  nada  al  respecto  de  la  variacion  de  los 
angulos  en  el  premolar  inferior. 


mm 

ANCHO 

DEL 

INCISIVO 

5  0 

_ 

INFERIOR 

• 

• 

• 

8 

" 

• 
•• 

• 

• 

• 

»» 

• 

• 

• 

■  6 

- 

•• 

• 

.4 

• 

• 

• 

• 

•  • 

• 

*• 

- 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

.2 

- 

• 

• 

• 

t 

• 

• 

40 
.8 

- 

• 

• 
• 

♦ 

• 

t 

• 

t 

t 

• 

.6 
.4 

- 

• 

* 

• 

♦ 

t 

• 

u 

• 

t. 

* 

• 

• 

2 

30 

♦ 

♦ 

t 

* 

♦ 

* 

♦ 

• 

.8 

.6 

- 

fosil 

♦ 

• 

♦ 

• 
• 

.4' 

•* 

2.2 

' 

_L 

+ 

1 

1     1 

1 

1 

1 

1 

1 

3.2 


6  .8  4.0  .2  4 

LONGITUD  DEL  PREMOLAR    INFERIOR 


5.0   mm 


Fig.  3.    Grafica  de  la  relacion  entre  el  ancho  del  incisivo  inferior  y  la  longitud 

del  premolar  inferior  en  Pappogeomys.     Los   puntos   representan   medidas   de 

P.  merriami  y  cruces  de  P.  tylorhinus. 


Considerando  que  el  material  fosil  de  Tlapacoya  no  coincidio  en  elgunas 
caracteristicas  con  las  especies  del  Valle  de  Mexico,  preferimos  asignar  este 
material  solo  al  genero  Pappogeomys,  esperando  que  e.xista  mas  material,  tanto 
fosil  como  viviente,  para  poder  determinar  si  se  trata  de  una  especie  no 
descrita  o  bien,  solo  un  extremo  de  la  variacion  de  alguna  de  las  especies 
actuales. 


Alvarez — Pleistocene  Fossils  from  Mexico 


99 


Cricetidae 

Peroniyscus  maldonatloi  Alvarez,   1967 

Material. — Fragmento  de  mandihula  derecha,  sin  procesos  ni  horde  inferior 
de  la  rania,  ni  la  mitad  posterior  de  m3  (1153  DP). 

Tomando  en  consideracion  el  taniano  de  la  mandihula,  asi  coino  los  pocos 
caracteres  que  se  puden  apreciar  en  la  mandilnila  procedente  de  Tlapacoya, 
esta  es  muy  semejante  tipo  de  Peromijscus  maldonadoi  espeeie  descrita  del 
Pleistoceno  de  Teqnesquinahua,  Mexico. 

En  la  estructura  y  tamano  de  los  molares  encontramos  tanihien  mucha 
similitud.  La  serie  de  molares  inferiores  mide  en  el  tipo  de  P.  maldonadoi  5.8, 
igual  que  en  el  ejemplar  de  Tlapacoya.  Aunque  la  estructura  de  los  molares 
en  el  tipo  de  P.  maldonadoi  no  se  puede  apreciar  en  detalle,  dehido  al  desgaste 
de  los  mismos,  si  es  posihle  apreciar  que  existe  una  similitud  en  la  forma 
general  entre  este  ejemplar  y  la  del  de  Tlapacoya  que  es  mas  joven;  asi,  el 
segundo  molar  es  relati\amente  grande  y  cuadrado  en  amhos  ejemplares.  La 
linica  diferencia  aparente  es  la  ausencia  de  ectostilidio  en  el  tipo  de  P.  ?»a/- 


ont. 


936  DP 


937  DP 


938  DP 


1  mm 


1  mm 


Fig.   4.    A,  corte   transxersal   del  incisivo   superior  del   Pappogeomtjs  fosil;   B, 
premolares  y  molares  del  Pappogeomtjs  fosil;  C,  primero  y  segundo  molares  de 

Peromijscus  maldonadoi. 


100  Misc.  Publ.  51,  Unw.  Kansas  Mus.  Nat.  Hist. 

donadoi  y  que  si  esta  bien  desarrollado  en  el  ejemplar  de  Tlapacoya,  especial- 
mente  en  el  primer  molar,  ya  que  en  el  segundo  dicha  estruchira  no  esta  bien 
maicada  (Fig.  4C).  Sin  embargo,  esta  diferencia  no  es  de  tomarse  mucho  en 
cuenta  ya  que  segun  Hooper  (1957)  la  presencia  de  stilos  y  lofos  es  muy 
variable  en  las  especies  de  Peromtjscus. 

Debido  a  la  semejanza  de  estructura  de  la  mandibula  procedente  de 
Tlapacoya,  con  el  tipo  de  P.  rnaklonadoi,  asi  como  que  ambas  proceden  de 
localidades  muy  cercanas  y  de  depositos  Pleistocenicos,  consideramos  que  sean 
de  la  misma  especie.  Tlapacoya  se  encuentra  a  42  kilometres  al  sureste  de 
Tequesquinahua,  Mexico  y  el  estrato  en  que  se  encuentra  la  mandibula  de 
Tlapacoya  fue  fechado  en  mas  de  24,000  aiios;  aunque  del  tipo  de  P.  mal- 
donadoi  no  tenemos  la  edad,  suponemos  que  proviene  del  Pleistocene  por  el 
tipo  de  fauna  con  que  se  encontro  asociado  (Alvarez,  1967). 

Para  completar  mas  la  diagnosis  de  la  especie  P.  rnaldonadoi,  describiremos 
a  continuacion  la  estructura  de  los  molares  del  ejemplar  de  Tlapacoya,  siguiendo 
la  terminologia  dada  por  Hooper  (1957).  El  primer  molar  presenta  proto- 
conidio  dividido  en  la  parte  anterior  y  con  el  doblez  menor  bien  marcado; 
el  mesostilidio  unido  por  medio  del  mesolofidio  a  la  cara  anterior  del  entoco- 
nidio;  cingulo  posterior  sin  llegar  al  borde  interno  del  molar;  con  ectostilidio, 
pero  sin  ectolofidio. 

El  segundo  molar  presenta  el  cingulo  anterior  extendido  hasta  el  borde 
externo  del  molar,  ectostilidio  poco  desarrollado,  sin  ectolofidio;  mesolofidio 
unido  al  ectoconidio  en  su  parte  anterior,  sin  mesolofidio;  cingulo  posterior 
extendido  hasta  el  borde  lingual  del  molar  y  continuandose  en  cierta  forma 
con  el  borde  del  ectoconidio. 

La  estructura  de  los  molares,  asi  como  el  tamano  de  los  mismos  es  muy 
parecido  a  P.  zarhinchus,  del  cual  difiere  en  que  el  primer  pliegue  primario 
tiene  direccion  oblicua  al  eje  longitudinal  del  molar,  ademas  de  ser  mas 
profunda. 

Medidas  del  ejemplar  de  Tlapacoya:  longitud  de  la  serie  de  molares,  5.8; 
longitud  de  ml,  2.23,  ancho,  1.43;  longitud  de  m2,  1.82,  ancho,  1.50. 

Neotoma  mexicana  Baird,  1855 
Material. — Primer  molar  inferior  izquierdo  (939  DP). 

El  molar  difiere  en  algunos  detalles  de  los  10  ejemplares  usados  para  com- 
paracion,  pero  la  variacion  entre  ellos  es  muy  grande,  de  tal  manera  que 
consideramos  que  las  diferencias  entre  el  molar  de  Tlapacoya,  que  ademas  no 
son  constantes,  y  todos  los  otros  molares,  sean  mas  bien  de  caracter  individual 
que  especifico. 

Freudenberg  (1922:  104)  enlisto  restos  de  Neotoma  del  Valle  de  Mexico, 
sin  mencionar  las  especies  ni  dar  ninguna  medida  o  figura  que  nos  pudiera  dar 
una  idea  de  que  especie  se  trata.  Alvarez  (1967)  describio  N.  magnodonta 
del  Pleistoceno  del  Valle  de  Mexico,  que  difiere  de  N.  mexicana  en  varios 
caracteres  morfologicos  y  sob  re  todo  en  su  mayor  tamaiio. 

Los  restos  aqui  registrados  como  iV.  mexicana,  se  encontraron  formando 
parte  de  la  concentracion  mayor  de  huesos  de  la  cala  beta,  que  se  supone 
fueron  depositados  por  el  hombre  hace  unos  24,000  anos.  Existe  la  posibilidad 
de  que  el  molar  de  Neotoma  haya  llegado  hasta  el  hogar  por  acarreo  a  traves 
de  los  tuneles  de  tuza. 


Alvarez — Pleistocene  Fossils  from  Mexico  101 

Microtus  mexicanus  (Saussure,  1861) 

Material. — Mandibula  con  incisivo,  ml-2.  Sin  proceso  angular  y  coronoide 
(940  DP). 

La  linica  diferencia  con  las  mandihulas  que  se  usaron  para  comparacion, 
estriha  en  que  el  fosil  es  ligeramente  mayor.  Procede  del  area  del  hogar  y 
presenta  la  misma  posibilidad  que  el  molar  de  Neotoma. 

Neochoerus  pinckneyi  Hay,  1926 
Material. — Primer  molar  izquierdo  inferior  (941  DP). 

El  molar  de  carpincho  encontrado  en  Tlapacoya,  tentativamente  se  asigna  a 
AT.  pinckneyi,  especie  que  se  ha  registrado  en  America  del  Norte.  Actualmente 
se  esta  realizando  un  estudio  del  Material  de  Carpinchos  que  proviene  de 
Chapala  y  Zacualco,  Jalisco  y  en  el  que  se  ha  encontrado  una  variacion  tan 
grande  en  la  forma  y  medida  del  premolar  inferior,  que  nos  hace  pensar  que 
los  diferente^  generos  de  la  familia  Hydrochoeridae  son  variacion  de  un 
misnio  ta.xon  y  que  por  lo  tanto,  muchas  especies  deberan  ser  puestas  en 
sinonimia. 

La  corona  del  molar  de  Tlapacoya  mide  de  largo  9.8  y  de  ancho,  6.9  y 
procede  de  la  capa  mas  profunda  que  se  excavo  o  sea  la  de  cenizas  volcanicas. 

Carnivora 
Canidae 

Canis  sp. 
Material. — Un  fragmento  de  mandibula,  sin  ningun  diente  (942  DP). 

Aunque  la  falta  de  dientes  hace  mas  dificil  la  identificacion  de  este  frag- 
mento, la  consideramos  como  Canis  por  la  forma  del  borde  inferior  de  la 
mandibula;  posicion  de  la  fosa  coroidea  del  alveolo  de  m3,  caracteres  que  no 
difieren  de  las  mandibulas  de  Canis  lupus,  C.  Intrans,  C.  familiaris  con  que  fue 
comparado. 

Se  encontro  este  fragmento  en  la  capa  de  ceniza  volcanica,  que  data  de 
mas  de  24,000  anos. 

Ursidae 

Ursus  americanus  Pallas,  1780 

Material. — Craneo    fragmentado,     dos     mandibulas;     dos     femures;     tibias 

izquierda  y   derecha;   pelvis  fragmentada;   fragmentos   de   escapula   izquierda; 

calcaneo,  uno  completo  y  otro  roto;  astragalo;  ocho  vertebras;  12  falanges;  dos 

ulnas  fragmentadas  y  dos  radios  (950-976  DP). 

Todos  los  materiales  identificados  como  de  oso  se  encontraron  en  una 
misma  capa  y  muy  juntos  entre  si,  por  lo  que  suponemos  que  se  trate  de  un 
solo  individuo. 

La  comparacion  de  las  medidas  de  los  molares  del  ejemplar  de  Tlapacoya 
con  los  dados  por  Kurten  (1963:5)  para  25  ejemplares  de  Ursus  americanus 
del  Reciente,  nos  muestra  que  nuestro  ejemplar  se  aparta  ligeramente  de  la 
amplitud  dada  por  dicho  autor.  P4,  M2,  m2,  y  m3  son  menores,  en  cambio 
M2  y  p4  son  mayores  que  el  maximo  de  la  variacion  dada  por  Kurten  (op.  cit.). 


102  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Tomando  en  cuenta  las  medidas  de  Stock  (1950)  para  una  mandiliula  de 
U.  ameiicanus  de  la  ciieva  de  San  Josecito,  Nuevo  Leon,  la  mandiljiila  de 
Tlapacoya,  es  im  poco  mayor,  pero  los  dientes  son  notablemente  mas  pequenos. 
Stock  (op.  cit.)  identifica  como  U.  ameiicanus  el  oso  de  San  Josecito,  aunque 
senala  sn  parecido  en  la  denticion  con  U.  optimus,  especie  del  Pleistoceno  de 
California  caracterizado  por  sn  robiista  dentadura.  El  ejemplar  de  Tlapacoya 
creo  que  sea  tipico  U.  americanus,  aunque  posiblemente  una  subespecie  mas 
pequeiia  que  los  actualmente  conocidos. 

Los  huesos  del  oso  se  encontraron  formando  la  mayor  parte  del  grupo 
designado  como  hogar  en  la  cala  alfa  y  como  ya  se  indico,  data  de  24,000 
aiios  aproximadamente. 

Procyonidae 
Procyon  lotor  (Linnaeus,  1758) 

Material. — Mitad  anterior  de  mandibula,  desde  el  borde  posterior  de  ml,  con 
pm4  y  ml.  Fragmento  maxilar  con  ah'eolo  de  Ml  y  con  el  M2;  mitad  inferior 
del  humero;  mitad  inferior  de  tibia;  un  astragalo;  tercio  superior  de  ulna 
(943-948  DP). 

La  mandibula  pertenecio  a  un  individuo  joven,  sin  embargo,  tanto  esta 
como  los  restos  postcraneales  son  ligeramente  de  mayor  tamafio  que  los  usados 
para  comparacion.  Esta  diferencia  probablemente  se  deba  a  distinto  desarrollo 
o  sexo  del  material  comparado. 

Los  restos  de  Procijon  fueron  encontrados  en  tres  capas  dife rentes.  La  mas 
antigua  corresponde  a  las  ceni/as  \'olcanicas  de  la  cala  beta,  que  como  ya  se 
ha  indicado,  data  de  mas  de  24,000  anos;  otros  restos  se  hallaron  en  la  aglomera- 
cion  de  huesos  y  piedras  denominada  como  hogar  y  cjue  tiene  una  edad  de 
24,000  aiios  y  por  ultimo,  se  localizaron  algimos  restos  en  capas  mas  super- 
ficiales  que  las  del  hogar,  que  corresponden  a  una  aproximada  de  8000 
a  9000  anos. 

Mustelidae 
Lutra  canadensis  (Schreber,   1776) 

Material. — Fragmentos  de  un  craneo:  region  jjasal;  frontales;  pterigoides, 
maxila  izquierda  con  la  carnacia  "//;  situ"  y  alveolos  de  premolares  y  molares 
(949  DP). 

Se  comparo  con  cuatro  craneos  de  ejemplares  recientes  y  se  encontro  que 
difiere  solo  en  pequeiios  detalles,  como  presentar  el  talonoide  de  la  carnacia 
mas  angosto;  el  borde  posterior  del  mismo  es  ligeramente  concavo  en  contra 
de  convexo  en  los  actuales.  EI  taniaiio  del  carnacia  es  mas  pequeiio  que  en 
los  ejemplares  de  comparacion  que  presentan  el  mismo  desgaste  en  el  diente. 
Medidas  del  carnacia  del  fosil,  seguidas  de  las  de  dos  ejemplares  recientes: 
largo,  12.2,  14.1,  12.2;  ancho  10.0,  11.9,  10.2. 

Este  genero  no  se  habia  registrado  anteriorme  ite  en  el  Pleistoceno  mexi- 
cano,  ni  tampoco  se  conoce  del  Reciente  en  el  Valle  de  Mexico.  Los  restos  de 
Lutra  de  Tlapacoya  fueron  rescatados  en  el  mismo  ni\'el  que  los  de  Neo- 
choerus  o  sea  el  estrato  mas  profundo,  formado  por  los  depositos  de  una 
erupcion  volcanica. 


Alvarez — Pleistocene  Fossils  from  Mexico  103 

Ahctiodactyla 
Cenidae 

Odocoileus  virginianus  (Zimmermann,  1780) 

Material. — Rama  mandibular  derecha  con  dientes;  mitad  posterior  de  la 
mandibnla  derecha  desde  el  primer  premolar;  astragalo;  region  parietal  muy 
fragmentada,  con  la  base  de  las  astas;  tres  falanges  segundas;  dos  falanges 
primeras  completas  y  dos  mitades;  fragmento  de  calcaneo;  articulacion  inferior 
de  la  escapula;  fragmento  basal  de  asta  con  nna  pnnta;  dos  femures  completos; 
metatarso,  dos  mitades  superiores  de  metatarso;  atlas  (977-1000,  1151-1152 
DP). 

Todo  el  material  que  se  refiere  a  Odocoileus  virginianus  se  identifico  como 
tal,  con  base  principalmente  en  el  tamafio,  ya  que  cualitativamente  no  difiere 
del  material  asignado  a  la  especie  de  Odocoileus  aqui  descrita  y  que  es  de 
mucho  mayor  tamaiio. 

Los  restos  de  O.  virginianus  se  encontraron  en  diferentes  capas;  sin  embargo, 
la  mayor  frecuencia  esta  en  la  capa  mas  profunda,  tanto  en  la  cala  alfa,  como 
la  beta  que  pertenece  a  la  erupcion  volcanica.  Existen  restos  en  una  capa  de 
la  cala  beta,  situada  por  arriba  de  la  playa,  que  se  ha  fechado  en  22,200  aiios, 
pero  por  abajo  de  la  otra  capa  fechada  en  14,000  anos  de  antiguedad.  Por 
ultimo,  encontramos  tambien  algimos  restos  de  \enado  cola  blanca  en  el  hogar 
y  otros  pocos  en  la  capa  mas  superficial. 

Odocoileus  halli,  especie  nue\'a 

Holottjpo. — Mandibula  izquierda  y  derecha  del  mismo  ejemplar  procedente 
de  la  capa  XLII  de  la  trinchera  alfa  de  Tlapacoya,  Mexico,  y  catalogada  con 
los  numeros  1082  y  1083.  Departamento  de  Prehistoria,  Instituto  Nacional  de 
Antropologia  e  Historia.  Mandibula  derecho  completa,  sin  incisixos,  ni  tercer 
molar,  sin  proceso  coronoide;  mandibula  izquierda  con  los  premolares  y 
molares,  sin  incisivos,  le  falta  la  region  desde  el  ultimo  molar  hacia  atras 
(Fig.  5). 

Paratipos. — Ocho  mandibulas  sin  incisivos  pero  con  premolares  y  molares; 
una  rama  mandibular  sin  dientes;  cuatro  series  de  dientes  superiores  con 
fragmento  del  maxilar,  15  premolares  y  molares  sueltos;  region  posterior  del 
craneo,  con  la  base  de  las  astas;  parietales  de  ambos  lados  con  la  base  de  las 
astas;  region  parietal;  tres  fragmentos  de  asta;  numerosos  elementos  post- 
craneales   (1001-1150,  1177-1180,  1191  DP). 

Edad. — Pleistoceno-Reciente,  fecha  de  C  14  de  22,000  ±  2600  anos. 

Diagnosis. — Tamaiio  grande  para  el  genero,  longitud  alveolar  de  la  man- 
dibula de  104  a  115;  maxilar,  117.3.  Forma  de  los  premolares  y  molares  igual 
a  Odocoileus  virginianus,  crestas  de  la  cara  lingual  de  los  premolares  y  molares 
bien  desarrollada;  segunda  mitad  de  pm4  menos  de  la  mitad  de  la  anterior; 
la  mayoria  de  los  molares  con  un  estilo  entre  los  dos  lobidos  externos. 

Description  de  los  Molares 
La  descripcion  de  las  estructiuas  molares  esta  basada  en  tres  ejemplares 
que  presentan  desde  el  que  no  tiene  desgaste  alguno  hasta  el  ejemplar  viejo, 
con  los  molares  muy  desgastados  (Fig.  5). 


104 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


1147   DP 


Fig.  5.  Molares  y  mandibula  de  Odocoileus  halli.  De  arriba  hacia  abajo, 
primero  una  serie  compuesta  de  premolares  y  molares  superiores  (1147,  1102, 
1191  DP);  serie  de  dientes  superiores  con  menos  desgaste  (1117  DP);  vista 
lingual  de  los  dientes  (1117  D.P. );  vista  oclusal  de  una  serie  de  molares  in- 
feriorcs  (1119  DP);  vista  de  los  molares  inferiores  izquierdos  del  tipo  (1083 
DP)   y  de  la  mandibula  derecha  del  tipo   (1082  DP). 


I 


Alvarez — Pleistocene  Fossils  from  Mexico  105 

Molares  infcriores. — El  pm2  es  sencillo  y  consta  de  una  sola  cuspide  central 
y  dos  crestas  oblicuas  posteriores,  las  cuales  con  el  desgaste  van  despareciendo, 
el  tamano  del  diente  es  como  la  mitad  del  premolar  siguiente,  el  cual  ya 
presenta  \anas  coniplicaciones  en  sii  estructura.  La  cara  externa  esta  dividida 
en  dos  lobnlos  por  una  escotadura  poco  profunda,  el  161:)ulo  posterior  es 
aproximadamente  un  tercio  del  lobulo  anterior.  La  cara  lingual  presenta  una 
entrada  en  forma  de  U  que  separa  de  cara  oclusal  en  dos  lobulos,  el 
anterior  es  muy  pequeno  y  esta  a  su  vez  dividio  en  dos  pequeiios  lobulos  los 
cuales  con  el  desgaste  se  \"an  uniendo  hasta  verse  en  los  animales  muy  viejos 
solo  una  pequeiia  entrants.  El  lobulo  posterior,  esta  a  su  vez  dividido  por  dos 
angulos  entrantes  muy  profundos.  La  combinacion  de  estos  dos  angidos  y  el 
angulo  externo,  dan  a  la  parte  posterior  del  premolar  forma  de  E  con  sus  tres 
salientes  hacia  el  lado  lingual. 

El  pm4,  es  ya  muy  parecido  a  los  molares,  esta  constituido  por  dos  lobulos, 
el  posterior  es  mas  pequefio  que  el  anterior.  Cada  lobulo  tiene  dos  crestas 
longitudinales,  la  lingual  es  casi  recta;  en  cambio,  la  externa  es  en  forma  de 
V  o  de  U  segun  el  desgaste. 

En  el  ejemplar  con  poco  desgaste  solamente  en  la  parte  anterior  de  las 
crestas,  la  dentina  se  une;  en  cambio,  en  todos  los  otros  la  dentina  se  encuentra 
rodeada  de  esmalte;  conforme  el  desgaste  se  continua,  las  crestas  se  van 
uniendo  hasta  formar  ima  sola  placa  con  dos  islas  ovales  de  dentina,  sitiiada 
mas  o  menos  en  el  centro  de  los  lobulos.  En  la  cara  externa  entre  los  dos 
lobulos  y  en  la  base  de  la  cara,  existe  un  pequefio  esHlo. 

Los  molares  1  y  2  son  iguales  al  premolar  4,  pero  el  segundo  lobulo  tiende 
a  ser  mas  grande,  hasta  alcanzar  un  tamaiio  igual  al  primer  lobulo  en  el  tercer 
molar.  EI  ml  presenta  mayor  desgaste  que  m2.  Los  estilos  entre  los  lobulos 
estan  mejor  desarrollados  que  en  pm4. 

El  m3  presenta  tres  loliidos,  los  dos  primeros  mas  o  menos  del  mismo 
tamaiio  y  el  tercero  muy  reducido. 

La  parte  anterior  de  las  crestas  del  primer  lobulo  presentan  la  dentina 
continua,  asi  como  la  parte  posterior  interna  del  primer  lobulo  y  la  parte 
anterior  de  la  cresta  externa.  El  borde  posterior  interno  de  la  cresta  lingual 
del  segundo  lobulo  y  la  parte  anterior  de  la  cresta  lingual  del  tercer  lobulo, 
tambien  se  continua,  asi  como  las  partes,  posteriores  de  las  crestas  del  tercer 
lobulo.  Conforme  avanza  el  desgaste,  la  dentina  se  va  uniendo,  sin  embargo 
la  parte  posterior  de  la  cresta  lingual  del  lobulo  primario  no  se  une  con  los 
del  segundo  lobulo. 

En  la  cara  interna  tanto  de  los  dos  ultimos  premolares  como  de  los  tres 
molares,  hay  unas  pequeiias  crestas  que  van  de  la  parte  mas  alta  de  la  corona 
a  la  base  de  la  misma. 

Dientes  stiperiores. — Los  tres  premolares  estan  formados  por  un  solo  lobulo, 
en  cambio  los  molares  estan  formados  por  dos.  El  PM2  es  el  mas  largo  de  los 
premolares.  La  cara  labial  esta  dividida  por  un  angulo  entrante  muy  cerrado, 
situado  en  el  primer  quinto  anterior  del  diente,  despues  tiene  otra  entrante 
mas  abierta,  en  la  base  de  la  cual  no  existe  ningi'm  estilo.  La  cara  oclusal  se 
encuentra  dividida  en  la  mitad  por  una  isla  de  esmalte  que  se  extiende  longi- 
tudinalmente,  con  entrantes  y  salientes  irregulares.  La  cara  lingual  tambien 
tiene  una  entrante  muy  tenue,  situada  en  el  primer  tercio  anterior. 


106  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

El  PM3  es  mas  corto  que  el  PM2,  y  su  estnictura  mas  sencilla,  consta  de 
dos  crestas  con  dentina  enmedio,  la  laliial  es  recta  y  la  lingual  en  forma  de  U. 
La  parte  anterior  y  la  posterior  de  las  crestas,  se  continuan.  En  la  cara  interna 
de  la  cresta  lingual  existen  dos  proyecciones  hacia  la  cara  interna  de  la  cresta 
labial,  sin  llegar  a  unirse  con  ella.  En  la  cara  externa  existen  tres  hordes  bien 
marcados  que  van  de  la  parte  mas  alta  del  molar  a  la  base  de  la  corona,  en 
donde  se  juntan  formando  un  borde  horizontal  en  la  base.  Las  crestas  hajan 
de  los  extremos  anterior  y  posterior  de  la  corona  y  de  la  parte  media  de  la 
misma.    El  PM4  es  igual  al  anterior,  pero  ligeramente  mas  corto. 

Los  tres  molares  son  semejantes  entre  si  y  estan  formados  de  dos  lohulos, 
cada  uno  de  igual  magnitud  y  separados  por  ima  entrante  lingual  en  cuya  base, 
en  algunos  ejemplares  existe  un  estilo  mas  o  menos  bien  desarrollado. 

Las  crestas  lahiales  de  los  dos  lohulos  de  cada  molar  se  continuan  entre  si, 
lo  mismo  que  con  la  parte  anterior  y  posterior  de  las  crestas  lahiales,  no  asi 
los  hordes  posterior  interne  de  las  crestas  linguales  del  primer  y  segundo 
lohulo  respectivamente,  los  cuales  estan  separados  entre  si  y  de  la  cresta  labial. 

En  la  cara  externa  de  los  molares  existen  cinco  crestas  bien  marcadas  que 
hajan  de  la  parte  alta  de  la  corona  a  la  base  de  la  misma,  en  donde  se  unen. 
Las  crestas  se  encuentran  en  la  parte  anterior  del  molar,  otra  en  la  posterior, 
ima  en  donde  se  unen  los  dos  lohulos  y  dos  mas  en  la  parte  media  de  cada 
16]:)ulo.  Entre  cresta  y  cresta  existe  una  depresion  en  forma  de  V  6  de  U  con 
la  parte  mas  profunda  hacia  la  base. 

Descripcion  de  las  Astas 

De  las  astas  de  Odocoileus  halli  solo  se  recohraron  cinco  fragmentos,  de  los 
cuales  dos  estan  muy  destruidos  y  son  muy  pequenos  para  poder  sacar  alguna 
conclusion,  de  los  otros  tres,  el  mejor  de  ellos  es  el  no.  1137  D.P.,  y  forma 
parte  de  un  craneo  fragmentado  de  donde  se  recohraron  los  occipitales,  region 
basal  y  fragmentos  de  los  frontales  con  una  parte  de  asta  de  15  cm  de  longitud 
y  la  region  basal  de  otra  asta  unida  al  frontal.  El  corte  transversal  (Fig.  6) 
de  la  base  del  asta  de  mayor  tamano  es  ligeramente  o\alado,  con  el  diametro 
mayor  transversal  de  37.3  y  el  anteroposterior  de  29.5.  El  corte  del  asta  por 
arriha  de  la  primera  punta  es  mas  circular,  con  el  diametro  transversal  de 
35.84  y  el  anteroposterior  de  33.5.  Toda  la  superficie  del  asta  es  rugosa;  falta 
el  anillo  basal  por  lo  que  no  podemos  decir  que  forma  y  tamaiio  pudo  haher 
tenido.  La  superficie  anterior  del  asta  se  abre  aproximadamente  a  165°  con 
respecto  a  la  superficie  del  frontal  (Fig.  6). 

Otro  fragmento  de  asta  forma  parte  de  una  region  frontal  (1095  DP)  con 
la  base  de  ambas  astas,  las  cuales  se  elevan  en  angulo  de  mas  o  menos  170°  con 
respecto  a  la  superficie  del  frontal.  El  corte  transversal  de  la  misma  es  de  forma 
circular  en  sus  caras  anteriores  y  laterales  pero  completamente  plana  en  la  cara 
posterior  (Fig.  6).  Los  diametros  transversal  y  anteroposterior  son  respectiva- 
mente de  31.8  y  22.8. 

Por  ultimo,  existe  un  fragmento  de  asta  (1085  DP)  (lue  consta  de  unos 
18  cm  de  la  rama  principal  con  la  region  basal  de  una  punta.  El  corte  trans- 
versal de  la  rama  principal  antes  de  la  punta  ( Fig.  6 )  es  ovalado,  su  diametro 
mayor  es  de  37.3,  en  cambio  el  anteroposterior  es  de  24.0.  El  corte  despues  de 
la  punta  conserva  la  misma  forma  o\a]ada  aunque  ya  no  tan  marcada,  prin- 
cipalmente  porque  presenta  un  abnltamiento  anterointerno;  el  diametro  trans- 
\ersal  es  de  26.5  y  anteroposterior  de  18.1 


Alvarez — Pleistocene  Fossils  from  Mexico 


107 


1137  DP 


pos 


1095  D 


F'iG.  6.    Difeientes  fragmentos  de  asta  de  O.  halli  con  cortes  transversales  de 

los  misinos. 


108  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Comparacion 

De  todas  las  especies  conocidas,  tanto  fosiles  como  vivlentes  Odocoileus 
halli  difiere  por  su  mayor  tamaiio,  que  es  casi  igual  al  de  un  pequeno  Cervus. 

La  unica  especie  que  se  le  parece  en  tamano  es  Odocoileus  brachyodontus 
Oelrich,  1953,  pero  difiere  en  que  las  crestas  anterointerna  y  posterointerna  estan 
unidas  a  la  cresta  transversal,  en  cambio  en  O.  brachyodontus  dichas  crestas 
se  encuentran  separadas.  Ademas,  existe  diferencia  en  epocas,  )  a  que  O.  brachy- 
odontus es  del  Pliocene  y  O.  halli  es  del  fin  del  Pleistocene. 

Notas 
Desde  el  momento  en  que  el  material  de  Tlapacoya  fue  \isto  por  el  Dr. 
Clayton  E.  Ray  de  la  Smithsonian  Institution,  me  sugirio  la  idea  de  que  pudiera 
corresponder  nuestro  material  a  Sangamona  fugitiva,  sin  embargo  al  examinar 

CUADRO    1. MeDIDAS    MANDIBULARES    Y   de    LOS    DIEXTES    MAXILARES    DE 

Odocoileus  halli. 

Medidas  No.         No.        No.        No.        No.         No.         No. 

mandibulares  1001       1002      1082      1083      1116      1119      1177 

pm2  largo    

anclio  

pm3  largo     

anclio  

pm4  largo     17.9 

ancho  12.7 

ml     largo     18.4 

ancho  13.8 

m2     largo     20.4 

ancho  14.5 

m3     largo     27.8 

ancho 13.2 

longitud   alveolar   115.3 

minima  altura  de  la  rama  mand.  21.9 

Altura  rama  bajo  ml-2  43.6 


9.8 

10.0 

10.6 

11.8 

7.0 

6.9 

6.8 

7.0 

16.3 

15.7 

15.9 

15.6 

10.1 

10.9 

10.0 

9.8 

18.2 

17.6 

16.3 

17.8 

17.8 

16.7 

12.6 

12.7 

12.4 

12.5 

12.6 

11.3 

19.4 

16.3 

16.5 

17.1 

15.5 

17.8 

14.0 

12.9 

12.3 

12.9 

11.5 

13.2 

20.7 

20.5 

20.4 

19.7 

20.4 

21.0 

14.4 

14.0 

14.7 

13.7 

13.4 

15.1 

28.0 

27.0 

30.0 

25.9 

24.6 

13.2 

14.4 

13.7 

13.2 

13.9 

115.2 

105.2 

104.5 

115.0 

111.2 

20.2 

20.4 

18.5 

20.8 

44.5 

39.8 

39.0 

38.4 

38.1 

Medidas  de  los  No.  No.         No.         No.         No. 

dientes  ma.xilares  1060       1092      1117       1145      1180    Suellos 

PM2  largo ._-..  ......      . ._....  16.2  18.3  17.7 

ancho .....  ......      . . 12.9  15.9  15.5 

PM3  largo    16.0  ......      . 16.4  14.9  16.4 

ancho 16.8  ......  ...._      . 14.3  16.9  19.9 

PM4  largo    17.1  ....  17.2      . 15.4      . . 

ancho 17.5 ..-  18.7  ......  -.... 

Ml     largo    22.4  19.7  22.4  __....  ......      . ...... 

ancho 18.3  20.9  20.9  19.5  ......  ..-..  ...... 

M2     largo    23.6  21.3  23.7  18.9  ......      . . 

ancho 20.5  23.0  22.7  21.2  ......  ......      . 

M3     largo    23.3  21.8  23.1  20.0  .....  ......  ...... 

ancho  18.2  21.4  20.6  21.1  .....  ......  ...... 

la  descripcion  de  este  genero  y  ver  el  molde  del  tipo  que  nos  fue  enviado  por  el 
Dr.  Ray,  encontramos  que  difiere  mucho.  De  acuerdo  con  la  descripcion 
original  Sangamona  difiere  de  los  otros  \'enados  por  carecer  de  las  costillas  en 
las  caras  externas  de  los  lobulos  de  los  molares.  Dichas  costillas  estan  bien 
marcadas  en  todos  los  molares  de  Tlapacoya;  comparando  el  tipo  de  Sangamona 
con  material  de  Odocoileus  halli,  encontramos  que  efectivamente  las  costillas 


Alvarez — Pleistocene  Fossils  from  Mexico  109 

son  niiiy  pnco  marcadas  y  que  la  cara  externa  bajo  el  paracono  y  inetacono  no 
presenta  ninguna  costilla,  en  cambio  en  este  mismo  liigar  Odocoileus  halli  al 
igual  que  todas  las  otras  especies  de  Odocoileus,  presentan  una  costilla  que 
desciende  del  metacono  y  paracono. 

Sau'^amona  presenta  tambien  un  estilo  entre  los  dos  lobulos  internos,  dicho 
estilo  se  presenta  en  la  mayoria  de  los  niolares  de  Odocoileus  halli,  asi  como  en 
otras  especies  que  se  pudieron  ver,  O.  hcmionus  y  O.  virginianus  en  ejemplares 
y  O.  salinae,  O.  cascensis,  O.  hrachijodontus  en  las  figuras  publicadas  en  su 
descripcion;  otra  forma  que  nos  fue  sugerida  con  posibilidad  de  ser  igual  a 
nuestro  material  fue  Rangifer  fiicki  Schultz  y  Howard  1935.  Al  comparar  el 
material  de  Tlapacoya  con  la  figura  publicada  en  la  descripcion  original  de 
R.  fricki,  encontramos  una  semejanza  muy  estrecha  en  la  forma  de  los  dientes, 
asi  como  en  sus  dimenciones.  Las  unicas  diferencias  estriban  en  la  longitud 
de  algunos  dientes:  el  segundo  premolar  inferior  es  ligeramente  mas 
largo  en  R.  fricki;  en  cambio,  el  tercer  molar  inferior  es  mas  corto  que  en 
O.  halli.  Los  molares  superiores  en  general  son  mas  largos  y  sobre  todo  mas 
anchos  en  O.  haUi  que  en  R.  fiicki.  A  pesar  de  que  la  semejanza  en  forma  es 
muy  estrecha  entre  Odocoileus  halli  y  R.  fricki  y  que  se  asignan  a  diferentes 
generos,  preferimos  poner  la  especie  halli  dentro  de  el  genero  Odocoileus, 
porque  como  ya  se  indica,  la  estructura  de  los  molares  no  difiere  mas  que  en 
su  tamano  mayor,  pero  no  en  la  estructura  esencial  de  los  dientes,  creemos  que 
mas  bien  Rangifcr  fricki  puede  ser  un  forma  de  Odocoileus. 

Odocoileus  halli  ha  sido  nombrado  en  honor  al  Dr.  E.  Raymond  Hall,  como 
un  estimonio  de  gratitud  por  todas  las  enseiianzas  y  atenciones  que  siempre 
he  recibido  de  el. 

DiSCUSION 

Las  excavaciones  realizadas  en  el  cerro  de  Tlapacoya,  Mexico, 
durante  el  ano  de  1966,  nos  dan  por  primera  vez  dentro  del  llamado 
Valle  de  Mexico,  algunas  fechas  relacionadas  con  la  fauna  existente 
en  dicha  area.  Los  restos  de  animales  fueron  encontrados  al  realizar 
dos  calas  perpendiculares  a  la  linea  de  la  falda  del  cerro.  Dentro  de 
los  muchos  estratos  que  se  identificaron,  dos  de  ellos  proporcionaron 
la  mayoria  de  los  huesos  estudiados. 

El  estrato  mas  antiguo  en  que  se  encontraron  huesos,  esta  for- 
mado  por  cenizas  volcanicas  que  en  este  punto  se  hallan  en  contacto 
con  la  roca  basal  y  que  tienen  un  espesor  de  un  metro  aproximada- 
mente.  La  edad  de  esta  capa  no  se  determine  directamente,  pero  se 
infiere  que  sea  mayor  a  los  24,000  anos,  ya  que  los  restos  procedentes 
de  una  capa  superior  dieron  la  fecha  antes  indicada. 

La  fauna  de  este  estrato  esta  formada  por  numerosos  restos  de 
aves  acuaticas  y  de  mamiferos  como  Pappogeomijs  sp.,  Stjlvilagus 
fioridanu.s,  Neochoerus  pinckneyi,  Ltitra  canadensis,  Peromijscus 
maldonadoi,  Odocoileus  virginianus,  y  Mormoops  megalophijlla. 

La  asociacion  de  las  aves  acuaticas  con  mamiferos  como  Neo- 
choerus y  Lutra,  nos  hacen  pensar  que  hace  24,000  aiios  la  region 


110  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

de  Tlapacoya  era  miicho  mas  hiimeda  y  que  debio  de  existir  un 
lago,  en  cuyas  margenes  la  vegetacion  era  exuberante.  For  los  restos 
encontrados  en  esta  capa,  es  difieil  dicernir  el  tipo  de  clima  existente 
en  esa  epoca  con  respecto  a  la  temperatura,  ya  que  la  asociacion 
faunistica  no  es  exclusiva  de  tierras  calientes,  aunque  si  se  le  encu- 
entra  con  preferencia  en  tal  tipo  de  climas;  ademas,  la  presencia  de 
Mormoops,  genero  de  murcielago  que  muy  raramente  se  encuentra 
en  climas  frios  y  que  vive  mas  bien  en  lugares  calidos,  nos  inclina  a 
pensar  que  el  clima  de  esa  epoca  ero  mas  calido  y  humedo  que 
el  actual. 

El  otro  estrato  en  que  fueron  abundantes  los  restos  oseos,  se 
cncontro  principalmente  en  la  trinchera  alfa  y  data  de  22,000  aiios 
aproximadamente. 

Los  restos  encontrados  en  este  sitio,  fueron  principalmente 
huesos  de  mamiferos  grandes  y  formaban  una  aglomeracion  muy 
marcada  alrededor  de  unas  grandes  piedras  que  rodeaban  un  claro, 
en  donde  se  cncontro  carbon,  asi  como  restos  de  utensilios  humanos, 
ha  sido  interpretado  como  un  hogar  del  hombre  prehistorico. 

Aunque  existen  restos  de  pequeiios  mamiferos,  la  gran  mayoria 
corresponden  a  Odocoileus  halli,  especie  de  venado  muy  grande  y 
que  sin  duda  alguna  proporciono  al  hombre  buena  cantidad  de 
carne.  Tambien  se  cncontro  en  la  trinchera  beta,  pero  no  se  ha- 
llaron  restos  de  este  animal  en  los  mas  antiguos  de  24,000  afios,  lo 
que  nos  hace  suponer  que  dicho  venado  fue  un  inmigrante  reciente, 
exterminado  por  la  accion  del  hombre;  en  cambio  Odocoileus  vir- 
ginianus,  cuya  talla  es  mucho  menor  que  O.  halli,  ya  existia  en  el 
Valle  de  Mexico  desde  hace  mas  de  24,000  aiios  y  ha  logrado  resistir 
desde  entonces  el  impacto  de  la  accion  humana,  ya  que  este  venado 
es  uno  de  los  pocos  mamiferos  de  tamaiio  regular  que  todavia  se 
encuentran  en  los  bosques  de  coniferas  alrededor  del  Valle  de 
Mexico. 

Summary 

The  Departamento  de  Prehistoria  of  the  Institute  Nacional  de 
Antropologia  e  Historia  has  been  working  for  several  years  at 
Tlapacoya  in  the  Valley  of  Mexico,  looking  for  remains  of  prehistoric 
man.  In  1966,  two  trenches  were  dug  on  the  south  side  of  Cerro 
Tlapacoya.  Only  two  of  the  many  strata  found  in  the  trenches  fur- 
nished good  quantities  of  bones.  The  oldest  is  a  \'olcanic  ash,  which 
at  this  point  is  on  bed  rock.  The  exact  age  of  this  stratum  is 
unknown,  but  the  one  above  it  was  aged  by  C^^  analysis  as  24,000 
years  old. 


Alvarez — Pleistocene  Fossils  from  Mexico  111 

Bones  in  this  ash  belong  mainly  to  birds.  Among  the  mammals 
it  was  possible  to  identify  Pappogeoimjs  sp.,  Sijlvilagus  floridamis, 
Neoclioerus  pinckneyi,  Lutra  canadensis,  Peromysciis  maldonadoi, 
Odocodeus  virg,inianus,  and  Mormoops  megaloplu/lla. 

The  association  of  a(]uatic  birds  principally  ( cormorants,  herons, 
grebes,  etc. )  and  mammals  like  Lutra  and  Neoclioerus  indicates  that 
24,000  years  ago  the  region  of  Tlapacoya  was  more  humid  than  now; 
probably  the  vegetation  was  more  luxuriant  than  in  Recent  times. 
It  is  possible  also  that  the  average  temperature  was  higher  than  it 
now  is,  as  indicated  by  the  presence  of  Mormoops,  a  bat  which  is 
abundant  in  warm  regions.  On  the  other  hand,  the  rest  of  the 
mammalian  fauna  is  found  today  in  both  warm  and  cool  climates. 
Thus  the  fauna  does  not  clearly  indicate  a  particular  climate. 

The  other  stratum  with  bones  was  found  in  the  trench  alpha  and 
was  dated  as  22,000  years  old.  Bones  from  this  level  were  of  large 
mammals  and  it  is  believed  that  they  were  brought  in  by  man, 
because  they  were  grouped  around  several  big  stones  and  associated 
with  charcoal.  Most  of  the  bones  belong  to  a  big  deer,  Odocodeus 
haUi,  here  described  on  the  basis  of  its  large  size — larger  than  any 
other  known  Odocodeus.  Comparison  of  O.  halli  with  Sangamona 
reveals  that  they  differ  in  the  construction  of  the  lingual  ridges  of 
the  molars. 

Other  remains  found  associated  with  O.  liaUi  were  identified  as 
follows:  Urusus  americanus,  an  almost  complete  skeleton;  Si/Jvda- 
gus  cunicularius,  Pappogeomys  sp.,  Neotoma  mexicana,  Microtus 
mexicanus,  Canis  sp.,  Procyon  lotor  and  Odocodeus  virginianus.  All 
these  mammals  except  the  two  first  named  still  live  today  in  the 
Valley  of  Mexico. 

LiTERATURA  CiTADA 

Alvarez,  T. 

1966.  Roedores  fosiles  del  Pleistocene  de  Tequesquinahiia,  Estado  de 
Mexico,  Mexico.   Acta  Zool.  Mexicana,  8  (3):  1-16. 

Dalquest,  W.  W. 

1961.  Sylvilagiis  cunicularius  in  the  Pleistocene  of  Mexico.  Jour.  Mamm., 
42:   408-409. 

Freudenberg,  W. 

1922.  Die  Saugetierfauna  des  Pliociins  und  Pospliocans  von  Mexiko.  II 
Teil:  Mastodonten  und  Elefanten.  Geol  u.  Paleont.  Abhand.,  14: 
103-176. 

C.oodlife,  E.,  y  M.  Goodlife 

1966.     Un  sitio  Pleistocenico  en  Tlapacoya,  Estado  de  Mexico.    Bol.   Inst. 

Nac.  Antiopol.  e  Hist.,  23:  30-32. 
1969.     Excavaciones  en  el  cerio  de  Tlapacoya,  Estado  de  Mexico,  Sitio  I. 
En  prensa. 


112  Misc.  Publ.  51,  Unr  .  Kansas  Mus.  Nat.  Hist. 

Haynes,  C.  v.,  Jr. 

1967.     Muestras  de  C14,  de  Tlapacoya,  Estado  de  Mexico.    Bol.  Inst.  Nac. 
Antropol.  e  Hist.,  29:  49-52. 

Hooper,  E.  T. 

1957.     Dental  patterns  in  mice  of  the  genus  Peromijscus.    Misc.  Publ.,  Mus. 
Zool.,  Univ.  Michigan,  99:  1-59,  24  figs. 

KXJRTEN,   B. 

1963.     Fossil  bears  from  Texas.    Pearce  Sellards  Ser.,  Texas  Mem.   Mus., 
1:  3-15,  6  figs. 

Merriam,  C.  H. 

1895.     Monographic  revision  of  the  pocket  gophers.  .  .  .    N.  Amer.  Fauna, 
8:  1-258,  19  pis.,  71  figs. 

Stock,  C. 

1950.     Bears    from    the    Pleistocene    cave   of   San   Josecito,    Nuevo    Leon, 
Mexico.    Jour.  Wasliington  Acad.  Sci.,  40:  317-321,  1  fig. 


BIOGEOGRAPHY  OF  SOUTHWESTERN  BOREAL 
AND  DESERT  MAMMALS 

BY 

James  S.  Findley 

The  southwestern  part  of  the  United  States  is  a  funnel-shaped 
region,  consisting  ecologically  of  an  eastern  and  a  western  desert 
bisected  by  forested  highlands  through  which  only  one  or  two 
lowland  gaps  permit  the  interchange  of  desert  organisms.  During 
later  Pleistocene  time  the  two  deserts  have  been  sundered  and 
reunited,  and  the  biota  of  the  forested  highlands  has  alternately 
spread  and  been  fragmented  into  island-like  refugia.  It  is  my  thesis 
that  these  historical  events  have  been  the  principal  agents  in  the 
remarkable  diversification  of  mammals  in  the  Southwest.  This  con- 
clusion is  reached  primarily  on  the  basis  of  the  observation  of 
contemporary  distribution  and  variation  patterns  and  secondarily 
on  the  basis  of  deductions  from  known  facts  of  southwestern 
climatic  history. 

If  the  patterns  of  distribution  and  variation  of  mammals  ap- 
peared non-correlated,  one  might  conclude  that  a  diversity  of 
factors  had  shaped  the  patterns.  If  on  the  other  hand  certain  pat- 
terns reappear  commonly,  one  might  suppose  that  only  a  few 
factors  were  involved,  and  it  might  be  possible  to  identify  these. 
It  seems  to  me  that  the  kinds  of  patterns  displayed  by  southwestern 
mammals  are  indeed  limited  and  often  similar.  The  most  frequent 
of  these  patterns  are  described  in  the  following  section.  I  have  not 
attempted  to  be  exhaustive,  hoping  that  the  selected  examples  are 
typical  and  point  the  way  to  the  overall  picture. 

The  Boreal  Pattern 

Boreal  mammals  are  those  that  occupy  montane  coniferous  forest 
zones  where  soil  moisture  is  available  and  aridity  is  not  an  important 
limiting  factor.   The  major  features  of  the  boreal  pattern  are: 

1.  Reduction  of  number  of  kinds  of  strictly  lioreal  species  with  decreasing 
latitude,  even  given  seemingly  comparable  habitats. 

2.  Gradual  replacement  of  northern  boreal  species  with  others  at  succes- 
sively lower  latitudes. 

3.  Altitudinal  zonation  of  northern  and  southern  boreal  species  in  areas 
of  sympatry. 

4.  Increase  in  lower  altitudinal  limits  of  northern  kinds  at  successively 
more  southerly  latitudes. 

(113) 


114 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist, 


106 


Fig.  1.    Geographic  distribution  of  three  species  of  Eiitamias:    black,  E.  mini- 
mus; blotched  pattern,  E.  dorsalis;  stipple,  E.  quadrivittatus  group. 


N 


Fig.  2.    Altitudinal  relationships  of  three  species  of  Eiiiamias  on  southwestern 
mountain  ranges:    black,  E.  ininimiis;  stipple,  £.  quadrivittatus  group;  blotched 
pattern,  E.  dorsalis;  A,  Animas  Mountains;  B,  Black  Range;  W,  White  Moun- 
tains; J,  Jemez  Mountains. 


FiNDLEY — Southwestern  Mammals 


115 


Selected  examples  ol  southwestern  boreal  mammals  displaying 
\'arious  aspects  of  this  pattern  follow. 

Chipmunks  of  the  Genus  Eutomios 
Figures  1  and  2  show  geographic  and  ecologic  distribution  of 
chipmunks  in  a  part  of  the  Southwest.  Using  the  White  Mountains 
of  Arizona  as  a  focal  point  for  discussion,  it  can  be  seen  that  in  the 
higher  parts  of  this  range  the  least  chipmunk,  Eutamias  minimus, 
occurs  in  a  rich,  mesic,  mixed  coniferous  forest  along  with  the 
gray-necked  chipmunk,  E.  cinereicollis,  which  is  found  to  the  lower 
limit  of  the  ponderosa  forest  and  there  overlaps,  and  is  replaced 
by,  the  cliff  chipmunk,  E.  dorsaJis.  The  latter  may,  under  suitable 
circumstances,  occur  nearly  to  sea  level  in  the  Sonoran  desert. 
Eutamias  dorsalis  is  widespread  around  the  margins  of  the  Sonoran 
desert,  occurring  in  chaparral,  encinal,  pinyon-juniper  woodland, 
and  similar  xeric  habitats.  However  in  the  absence  of  other  chip- 
munks, this  species  may  occur  in  pine  or  mixed  coniferous  forest. 
The  gray-necked  chipmunk,  together  with  its  close  relatives  and 


Fig.  3.   Geographic  distribution  of  two  species  of  Microtiis:   black,  M.  nionianus; 

blotched  pattern,  M.  mexicanus. 


116 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


;   long 


Jackson   Hole 


_.n VtYYYt 

////////.  •.•,:.•.♦.•.•.•.•.•.•. 
pen  n    '  ' 


M 


J  emez 


I   / 


lUi 


.iiiiiiiiiiiiiiii 


Y  YYY  V\/ 


-•,•     •      •     •     •     •      •     ••      ••      ••-      •.• 

••••      ••      •      ••      •       ••      ff*       ••      •• 


•       ••••• 


•     •  '  •     •     • 


I     mon  tonus 


White 


iiiiiiiiiiiii 


nnex  iconus 


M  ogol  lo  n 


> uYjYtn 


Fig.  4.    Diagram  of  habitat  relationships  of  four  species  of  Microtus  on  three 
southwestern  mountain  ranges  and  in  Jackson  Hole,  Wyoming. 


ecological  vicariants,  E.  qiiadrivittatus,  E.  canipes,  and  £.  hulleri, 
is  widespread  in  mixed  coniferous  and  ponderosa  forest.  In  the 
absence  of  dorsalis,  the  species  qiiadrivittatus  and  canipes  occur 
down  to  the  lower  edge  of  the  woodland,  and,  in  places,  even  into 


FiNDLEY — Southwestern  Mammals  117 

the  grassland.  Eutamias  minimus,  in  the  Southwest,  is  limited  to 
some  of  the  most  mesic  ranges,  but  farther  north  occurs  over  wide 
continuous  areas,  even  into  the  sage-grassland  in  Wyoming  and 
adjacent  states.  That  the  least  chipmunk  is  well  adapted  to  boreal 
conditions  is  suggested  by  its  extensive  postglacial  spread  across 
the  Canadian  coniferous  forest. 

Voles  of  the  Genus  Microtus 

Figures  3  and  4  depict  geographic  and  ecologic  distribution  of 
several  species  of  Microtus  in  the  Southwest.  Again  using  the  White 
Mountains  as  a  focal  point,  we  find  there  the  northern  species, 
M.  montamis,  living  in  the  most  mesic  microtine  habitat — grass- 
sedge  meadows  around  streams  and  ponds.  In  drier  grasslands  at 
the  same  altitude,  and  down  to  the  lower  edge  of  ponderosa  forest 
occurs  M.  mexicanus.  The  montane  and  Mexican  voles  seem  to 
occupy  somewhat  analogous  positions  to  the  least  and  gray-necked 
chipmunks,  respectively,  but  there  is  no  microtine  analog  to  the 
cliff  chipmunk.  If  we  look  below  the  pine  forest  in  the  Southwest 
for  a  grazer  that  is  a  grass-tunnel  dweller,  we  find  various  species  of 
the  cricetine  genus  Sigmodon,  an  animal  much  more  tolerant  of  xeric 
conditions  than  any  Microtus.  Aside  from  its  drought-tolerant  quali- 
ties, Sigmodon  is  so  Microtus-\ike  in  reduction  of  countershading, 
reduction  of  appendages,  burrowing  and  tunnel-dwelling  propensi- 
ties, diurnal  activity,  grazing  habit,  and  ability  to  respond  to  grass 
growth  with  population  irruptions,  that  it  might  be  thought  of  a 
"microtoid"  cricetine.  The  species  M.  longicaudus  is  not  primarily 
grass-dwelling,  but  rather  a  species  of  the  forest  edge,  and  thus  does 
not  enter  directly  into  competition  with  the  other  two  Microtus,  nor 
is  it  closely  related  to  them,  usually  being  placed  in  another  sub- 
genus. Thus  its  distribution  in  the  Southwest  is  not  directly  related 
to  the  history  of  the  other  two  species.  On  those  ranges  where  the 
montane  vole  is  not  found,  the  Mexican  vole  occupies  the  most 
mesic  habitats  as  well  as  more  xeric  areas,  and  I  earlier  ( Findley 
and  Jones,  1962)  postulated  a  competitive  relationship  between 
the  two  species,  mexicanus  invading  those  places  in  post-pluvial 
times  where  montanus  had  become  extinct. 

Rabbits  of  the  Genus  Sylvilogus 
Geographic  and  ecologic  distribution  of  three  species  of  Sylvi- 
lagus  in  the  Southwest  is  shown  in  Figures  5  and  6.    The  higher 
elevations  of  the  White  Mountains  are  inhabited  by  S.   rmttaUii. 


118 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Fig.  5.    Geographic  distribution  of  three  species  of  SlJlviIa^tls:    black,  S.  niii- 
iallii;  blotched,  S.  floridamis;  wliite,  S.  auduhonii. 


More  southerly  and  lower  ranges  are  occupied  by  S.  floridamis, 
which  occurs  so  low  as  2000  feet  in  Arizona  ( Hoffmeister  and  Lee, 
1963).  In  the  woodlands,  grasslands,  and  deserts  at  all  lower  eleva- 
tions in  the  Southwest  occurs  S.  auduhonii.  Farther  north  S.  nuttalUi 
is  more  continuously  distributed  and  often  occurs  at  low  elevations 
in  the  absence  of  the  other  two  species.  That  S.  floridamis  is  not, 
at  first  glance,  a  convincing  analog  of  the  Eutamias  quadrivitatus 
group  and  Microtus  mexicamis  cannot  be  debated.  Sijlvilagus  flori- 
damis is  widely  distributed  in  the  eastern  United  States  and  in 
Mexico,  and  attempting  to  derive  the  species  by  pluvial  abandon- 
ment in  the  Southwest  may  seem  to  be  in  violation  of  the  precepts 
of  Occam's  razor.  However,  the  range  of  S.  floridamis  in  the  South- 
west is  highly  disjunct,  and  western  populations  are  largely  sep- 
arated from  those  to  the  east.  The  populations  of  the  region, 
including  those  of  Mexico,  are  quite  variable,  and  it  seems  to  me 
that  the  conspecificit)'  of  this  widespread  assemblage  is  far  from 
obvious.    There  is  at  least  a  possibility  that  the  montane,  forest- 


FiNDLEY — Southwestern  Mammals 


119 


SYLV  ILAGUS 


Fig.  6.    Diagram  of  altitiidinal  relationships  of  tliree  species  of  Stjlvilagus  on 
selected  mountain  ranges  in  the  Southwest. 


MYOTIS 


lucifugus 


»'..""/'■»« .V r  I  .V' ^an^vryumanensis  .•;'".'.->^S3 


vel  ifer 


Desert 


Grass  &  woodland  Boreal     forest 


Fig.  7.    Diagram  of  habitat  relationships  of  six  species  of  Mijoiis  in  the  South- 
west. 


dwelling  S.  floridamis  of  the  Southwest  and  the  Sierra  Madre  Occi- 
dental is  a  derivative  of  the  closely  related  S.  miftallii.   Serological 


120 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Fig.   8.    Distribution  of  bats  of  the  Myotis  evotis  group:     solid   lines   enclose 
range  of  M.  evotis;  blotched  pattern,  M.  keenii;  stipple,  M.  thijsanodes. 


and  karyological  studies  are  strongly  indicated  for  this  group  of 
mammals. 

Bats  of  the  Genus  Myotis 
Three  groups  of  three  species  each  may  be  delineated  among 
the  Myotis  of  the  Southwest.  ( 1 )  The  lucifugtis  group,  consisting 
of  hicifiigus  (including  occtdttis),  ytimanensis,  and  velifer,  alike  in 
having  unkeeled  calcars,  large  feet,  relatively  short  ears,  and  rela- 
tively short  broad  rostra  with  relatively  large  molars,  the  paralophs, 
metalophs,  and  hypocones  of  which  are  well-developed.  Usually 
these  bats  are  found  near  sources  of  permanent  water  of  relatively 
large  size.  (2)  The  evotis  group,  consisting  of  evotis,  keenii  (races 
apache  and  aurictihis),  and  thysanodes,  alike  in  having  long  ears, 
relatively  long,  slender  rostra,  slightly  keeled  or  non-keeled  calcars, 
a  tendency  to  have  a  fringed  trailing  edge  to  the  uropatagium,  and 
upper  molars  with  reduced  or  absent  lophs,  and  a  reduced  hypo- 
cone.    (3)  The  volans  group,  consisting  of  volans,  Jcihii  (formerly 


FiNDLEY — Southwestern  Mammals 


121 


Fig.  9.    Distrilmtion  of  bats  of  the  Myotis  hicifti^iis  group:    solid  lines  enclose 
range  of  M.  hicifiigus;  blotched  pattern,  M.    ijwnanensis;   stipple,   M.   vclifer. 


subulatus),  and  califoniicus,  alike  in  small  size,  medium  ears, 
strongly  keeled  calcar,  small  feet,  and  hicifugiis-\ike  molars.  Sum- 
mer geographic  and  ecologic  distribution  of  some  of  these  species 
in  the  Southwest  is  shown  in  Figures  7  through  9.  In  each  case, 
one  member  of  a  species  group  is  definitely  boreal,  one  is  most  com- 
mon at  middle  elevations  in  pine  forests  and  woodlands,  and  one 
is  a  desert  and  grasslands  species.  The  lucifugus  group  requrres 
special  comment  in  this  regard.  Maternity  colonies  of  M.  lucifugus 
occultus  and  M.  yumanensis  occur  at  some  rather  low  elevations 
as  at  Socorro,  New  Mexico,  and  Blythe,  California.  Here,  however, 
the  colonies  are  located  near  large  permanent  bodies  of  water,  the 
Rio  Grande  and  the  Colorado  River.  In  the  postpluvial  retreat  of 
these  water-loving  species,  it  might  be  expected  that  they  would 
persist  in  the  few  water-rich  lowland  habitats,  especially  the  more 
arid-adapted  yumanensis,  which,  of  course,  is  much  more  common 
in  south^\'estern  lowlands  than  is  lucifugus. 


122  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

The  Desert  Patterns 

Desert  mammals  are  those  living  in  lowlands  where  aridity  exerts 
a  dominant  limiting  and  molding  force  on  the  animals.  Desert  mam- 
mals in  the  region  under  consideration  exist  in  an  eastern  ( Chihua- 
huan)  center  and  a  western  (Sonoran)  center,  and  frequently  extend 
from  one  center  to  the  other  through  the  lowland  gap  ( the  Deming 
Plain)  extending  across  southwestern  New  Mexico  and  southeastern 
Arizona.   Several  patterns  seem  to  recur  among  these  mammals. 

1.  Widespread  continuous  distribution  through  both  deserts  with 
little  striking  geographic  variation.  Dipodomys  merriami  seems 
typical  of  this  pattern.  Teroiinathiis  peniciUatus  recently  studied 
by  Hoffmeister  and  Lee  (1967)  seems  to  show  several  concordant 
character  shifts,  one  of  which  coincides  with  the  Continental  Divide. 
Other  species  include  Notiosorex  crawfordi,  SijJvilogus  auduhonii, 
Lepus  californicus,  Peromyscus  eremicus,  Omjchomys  torridus,  and 
Neotoma  albigida. 

2.  As  above  but  with  one  or  more  sharp  character  breaks,  often 
in  the  area  of  the  Continental  Divide,  or  between  southern  and 
central  Mexican  Plateau  populations.  Unpublished  results  of  a 
study  by  myself  and  Gerald  L.  Traut  reveal  that  the  desert  bat 
PipistreUtis  hesperus,  while  showing  numerous  local  adaptations, 
is  di\'isible  into  a  large  eastern  and  a  small  western  population. 
Whereas  in  some  areas  size  of  this  species  is  responsive  to  climate, 
this  is  not  the  explanation  for  the  present  distribution  of  the  two 
populations.  The  two  kinds  contact,  with  a  steep  clinal  change  in 
size,  along  the  Continental  Dixide  ( Fig.  10 ) . 

The  widespread  hispid  cotton  rat,  Sigmodon  hispidus,  occurs  in 
parts  of  the  southwestern  deserts  where  a  cover  of  grass  is  available. 
While  these  animals  extend  across  the  Deming  Plain,  Mohlhenrich 
( 1961 )  thought  that  they  had  entered  this  area  recently,  and  Gen- 
naro  ( 196S )  concurred  as  a  result  of  his  finding  that  cotton  rats  in 
the  Deming  Plain  were  not  adaptively  colored.  Gennaro  thought 
that  these  animals  had  entered  the  Deming  Plain  from  both  east 
and  west.  Recently  Earl  Zirnmermann  (personal  communication) 
has  revealed  that  animals  from  southern  Ai-izona  ha\e  a  fundamen- 
tal chromosome  number  of  38,  while  those  from  Lubbock,  Texas, 
to  the  east  have  a  fundamental  number  of  52  to  54.  It  may  well  be 
that  the  two  differently  colored  hispid  cotton  rats  of  the  Deming 
Plain  also  differ  karyologically,  providing  an  important  example  of 
this  pattern. 


FiNDLEY — Southwestern  Mammals 


123 


Fig.   10.    Size  in  PipistreUus  Jicsperus  leased  on   10  size  variables.    The  darker 
the  circle  the  larger  the  bats  from  that  area.    Black  areas  are  forested  highlands. 


Fig.  11.    Eastern  limits  of  western  desert  species  described  in  text.    Black  areas 

are  forested  highlands. 


124  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

3.  Largely  confined  to  the  Sonoran  desert,  barely  extending  to 
the  Continental  Divide  area.  Selected  species  are  Choeromjcteris 
mexicana,  Perognathus  haileyi,  Macrotus  tvaterhousii,  Spermophilus 
tereticaudus,  Perognathus  longimembris,  Perognathus  amplus,  Dipo- 
domys  deserti,  and  Peromyscus  merriami.  Eastern  limits  of  selected 
species  in  this  category  are  shown  in  Figure  11. 

4.  Largely  confined  to  the  Chihuahuan  desert,  barely  extending 
into  the  Continental  Divide  area.  Selected  species  are  Spermophilus 
spilosoma,  Perognathus  merriami,  Perognathus  flavus,  Perognathus 
hispidus,  Dipodomys  spectahilis,  Dipodomys  ordii,  Onychomys 
leucogaster,  ReitJirodontomys  montanus,  Peromyscus  leucopus,  and 
Neotoma  micropus.  Western  limits  of  selected  species  from  this 
category  are  shown  in  Figure  12. 

5.  Like  3  or  4  above,  but  with  a  close  relative  in  the  other 
desert.  Determination  of  this  situation  depends  upon  sound  knowl- 
edge of  relationships,  frequently  not  available.  That  the  white-sided 
jackrabbits,  Lepus  caUotis  and  L.  alleni,  are  closer  to  each  other 
than  to  other  species  seems  accepted  by  students  of  the  group. 
Ranges  of  the  two  are  basically  Chihuahuan  and  Sonoran,  re- 
spectively. The  antelope  squirrels,  Ammospermophihis,  show  a 
similar  pattern,  more  striking  because  of  disjunction,  and  because 
the  trans-Coloradan  isolate,  A.  leucurus,  has  encircled  the  Chihua- 
huan desert  from  the  north  and  thus  closely  approaches  its  Chi- 
huahuan relative  (A.  interpres).  Perognathus  intermedins  and  P. 
nelsoni  seem  to  fit  this  pattern  although  here  the  Sonoran  isolate, 
P.  intermedins,  has  extended  well  into  the  Chihuahuan  desert, 
whereas  its  Chihuahuan  relative,  P.  nelsoni,  seems  to  have  been 
rather  sedentary.  Ranges  of  two  of  these  species  pairs  are  shown  in 
Figure  13. 

Development  of  Patterns 

That  major  ages  of  cool,  moist  climates,  recurring  during  the 
Pleistocene,  caused  depression  and  coalescence  of  montane  forests 
in  the  Southwest  seems  established  beyond  reasonable  doubt.  This 
is  the  only  easily  acceptable  explanation  for  the  existence  of  south- 
western montane  boreal  islands. 

If  more  than  one  cycle  of  pluvial-interpluvial  conditions  affected 
the  Southwest,  and  if  montane  highlands  existed  there  to  act  as 
interpluvial  refugia  for  boreal  organisms,  it  is  possible  to  hypothe- 
size a  series  of  events  that  could  have  led  to  the  present  diversifica- 
tion of  the  southwestern  boreal  and  sub-boreal  mammalian  fauna. 


FiNDLEY — Southwestern  Mammals 


125 


Fig.   12.    Western  limits  of  eastern  desert  and   grassland  species  described  in 
text.    Black  areas  are  forested  highlands. 


Fig.  13.   Ranges  of  related  pairs  of  species:    solid  lines,  Lepiis  aUeni  (west)  and 
L.   callotis    (east);   dashed   lines,   Aminospermophilus   hairi.sii    (west)    and   A. 

inierpres  (east). 


126 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


A 


A 


A 


Fig.  14.    Diagrams  showing  hypothetical  sequence  of  events  in  development  of 
zonation  of  lioreal  mammals  in  the  Southwest.    See  text  for  explanation. 


These  hypothetical  events  are  diagrammed  in  Figure  14,  where  the 
stippled  pattern  represents  the  distribution  of  a  boreal  mammalian 
taxon,  moving  northward  and  southward,  as  well  as  up  and  down 
mountain  slopes,  with  Pleistocene  climatic  fluctuations.  Figure  14A 
represents  conditions  during  a  pluvial  time  when  boreal  species  A  is 
continuously  distributed   in  the  Southwest.    In  figure   14B,   inter- 


FiNDLEY — Southwestern  Mammals  127 

plux'ial  climates  have  resulted  in  northw aid  and  upward  \\'ithdrawal 
of  species  A  with  its  isolation  on  a  mountain  range.  Figure  14C 
represents  a  later  time  during  the  interpluvial  when  the  montane 
isolate  has  differentiated,  perhaps  in  response  to  the  arid  marginally 
boreal  conditions  in  which  it  was  trapped  as  well  as  through  the 
vagaries  of  random  genetic  phenomena,  and  now  is  different  from 
its  northern  relative  and  hence  designated  B,  and  shown  with  a 
different  pattern.  Figure  14D  shows  another  pluvial  time.  Species 
A  and  B  have  both  mo\'ed  southward,  B,  by  viitue  of  its  adaptation 
to  less  boreal  conditions,  being  commonest  in  lower  and  more 
southern  areas.  Note  that  A  and  B  may  coexist  and  hence  are  con- 
sidered to  ha\e  speciated.  Figure  14E  depicts  another  interpluvial. 
Species  A  and  B  have  moved  northward  and  upward.  On  the  moun- 
tain both  are  now  isolated,  A,  still  more  borealy  adapted  than  B, 
occupying  higher  more  mesic  areas.  A  continuation  of  this  cycle 
of  exents  might  result  in  still  more  complex  stratification  of  boreal 
mammals.  Such  a  situation  would  result  only  if  the  taxon  involved 
could  adapt  to  the  somewhat  more  arid  conditions  obtained  on  small 
boreal  islands.  Many  boreal  isolates  would  become  extinct  with 
each  interplu\'ial,  leaving  no  descendants. 

If  the  postulated  series  of  extents  ever  took  place,  we  might 
expect  to  find  some  cases  of  closely  related,  probably  congeneric 
species  occupying  successive  altitudinal  zones,  or  at  least  local 
habitats  differing  in  available  moisture,  on  southwestern  mountain 
ranges. 

With  each  expansion  of  the  boreal  environment  during  pluvial 
times,  the  highlands  of  the  Deming  Plain  must  have  been  largely 
unavailable  to  desert  mammals  with  the  result  that  Sonoran  and 
Chihuahuan  segments  of  the  ranges  of  many  species  must  have  been 
separated  from  one  another.  Under  these  ciix-umstances,  divergence 
of  the  two  segments  could  ha\'e  proceeded  in  response  to  differential 
selective  pressures  and  random  phenomena,  with  the  following  pos- 
sible results.  ( 1 )  Elimination  of  one  segment  of  the  population. 
With  postpluvial  spread  of  desert  organisms  the  surviving  population 
reoccupies  the  entire  desert  area,  as  has  Dipodomys  merriami.  (2) 
The  two  segments  differentiate  but  do  not  attain  reproductive  isola- 
tion, hence  secondary  intergradation  takes  place  when  they  reestab- 
lish contact,  as  seems  to  be  the  case  with  PipistreUtis  liesperus. 
(3)  Only  one  population  survives,  but  fails  to  reoccupy  the  entire 
desert  area,  perhaps  because  it  has  become  too  dependent  on  low 
or  high  desert  conditions.    Examples  might  be  Spermophihis  tereti- 


128  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

caiuJus  and  S.  spilosoma.  (4)  The  two  populations  attain  repro- 
ducti\'e  isolation.  Perhaps  there  are  no  cases  of  this  among  mam- 
mals where  the  specific  level  of  the  populations  is  demonstrated  by 
natural  sympatry,  though  this  may  be  an  artifact  of  our  imperfect 
knowledge  of  degrees  of  relationships  between  species.  However, 
there  seems  little  doubt  that  Lepus  callotis  and  L.  alleni  are  spe- 
cifically distinct,  and  many  species  pairs  such  as  the  two  Omjchomys 
may  be  examples  of  this  situation. 

Some  patterns  suggest  that  in  immediate  postpluvial  times 
eastern  grassland  species  spread  westward  across  the  Deming  Plain 
into  the  grasslands  of  southern  Arizona.  Subsequent  progressive 
aridity  limited  these  grassland  species  to  grassland  zones  between 
desert  and  encinal  on  many  desert  mountain  ranges,  thus  resulting 
in  doughnut  shaped  relict  ranges  for  some  species  such  as  Dipo- 
domijs  ordii  in  southern  Arizona.  Of  course,  such  grassland  species 
may  have  existed  on  both  sides  of  the  Continental  Divide  during 
pluvial  times,  since  pluvial  grasslands  were  evidently  widespread  in 
southern  Arizona  and  in  Sonora. 

It  is  evident  that  repeated  dividing  and  reuniting  of  deserts  could 
provide  the  mechanism  for  the  proliferation  of  desert  species. 

Summary 
So  far  back  in  the  Pleistocene  as  the  present  physiography  has  existed  in 
the  Southwest,  we  may  imagine  an  oscillating  series  of  boreal  expansions 
coupled  with  desert  contractions,  alternating  with  desert  expansions  coupled 
with  boreal  retreats.  The  former  phases  resulted  in  desert  speciation,  the  latter 
in  boreal  speciation.  It  seems  to  me  likely  that  the  majority  of  distribution 
and  variation  patterns  in  the  Southwest  and  indeed  in  all  of  western  North 
America  will  be  found  to  be  largely  shaped  by  these  events. 

Literature  Cited 

FiXDLEY,  J.   S.,  AND  C.  J.   JONES 

1962.  Distribution  and  variation  of  voles  of  the  genus  Microtiis  in  New 
Mexico  and  adjacent  areas.  Jour.  Mamm.,  43:154-166,  5  figs., 
May  29. 

Genxaro,  a.  L. 

1968.  Color  \ariation  of  the  hispid  cotton  rat  in  New  Mexico.  Jour. 
Mamm.,  49:317-318,  May  20. 

HOFFMEISTER,   D.   F.,   AND  M.   R.   LeE 

1963.  Taxonomic  review  of  cottontails,  Sylvilagus  floridanus  and  Sylvilagus 
nuttalhi,  in  Arizona.    Amer.  Midland  Nat.,  70:138-148,  2  figs.,  July. 

1967.  Revision  of  the  pocket  mice,  Perognathus  penicillatus.  Jour.  Mamm., 
48:361-380,  5  figs.,  August  21. 

MOHLHEXRICH,   J.    S. 

1961.  Distribution  and  ecology  of  the  hispid  and  least  cotton  rats  in  New 
Mexico.   Jour.  Mamm.,  42:13-24,  3  figs.,  February  20. 


HOLOTYPES  OF  RECENT  MAMMALS  IN  THE 

MUSEUxM  OF  NATURAL  HISTORY, 

THE  UNIVERSITY  OF  KANSAS 


BY 


J.  Knox  Jones,  Jr.,  and  Hugh  H.  GExNOWays 

Various  museums  in  recent  years  have  published  Hsts  of  type 
specimens  housed'  in  their  collections.  Such  lists  provide  a  useful 
point  of  reference  for  systematists  and,  for  that  reason,  have  been 
encouraged  by  the  International  Council  of  Museums.  In  1968,  that 
organization  issued  "A  preliminary  list  of  catalogues  of  type  speci- 
mens in  zoology  and  palaeontology,"  compiled  by  A.  W.  F.  Banfield, 
and  a  revised  list  is  planned  for  1971.  This  catalogue  of  holotypes  of 
mammals  in  The  University  of  Kansas  Museum  of  Natural  History  is 
particularly  appropriate  for  the  present  volume,  because  Professor 
E.  Raymond  Hall  was  directly  responsible  for  the  descriptions  of 
20  taxa  here  listed,  and  was  instrumental  in  arranging  support  for 
field  work  that  resulted  in  the  collection  of  many  others. 

The  first  type  specimen  of  a  mammal  designated  from  the  col- 
lections at  Kansas  was  the  holotype  of  ''Reitliroclontomys  dijchei,' 
described  by  J.  A.  Allen  (1895:120).  This  specimen,  originally  KU 
(old  series)  5232,  was  renumbered  as  10127/8431  in  the  mammal 
collection  of  the  American  Museum  of  Natural  History,  where  it 
now  is  housed.  Similarly,  the  holotype  of  "Mimoii  cozwnelae," 
named  by  E.  A.  Goldman  (1914:75),  previously  was  in  the  collec- 
tion at  Kansas  (original  number,  if  any,  unknown),  but  was  pre- 
sented to  the  U.  S.  National  Museum,  where  it  now  is  deposited  as 
USNM  203191. 

Ninety-nine  holotypes  and  one  allotype  were  among  the  nearly 
120,000  specimens  of  Recent  mammals  housed  in  the  Museum  of 
Natural  History  as  of  December  31,  1968.  These  include  two  insecti- 
vores,  17  bats,  five  lagomorphs,  73  rodents,  and  two  carnivores. 
Four  of  the  holotypes — Nijcteris  vinsoni  Dalquest,  Scotophiliis  al- 
vensJebeni  Dalquest,  Eumops  perotis  renatae  Pirlot,  and  Ochotona 
princeps  howelli  Borell — were  donated  to  the  museum  subsequent 
to  the  original  description.  Additionally,  two  holotypes  of  the 
rodent  genus  Ochrotomijs  are  designated  in  this  volume. 

(129) 


130  Misc.  Publ.  51,  Unw.  Kansas  Mus.  Nat.  Hist. 

List  of  Holotypes 

In  the  following  list,  holotypes  are  arranged  under  the  name  by 
which  they  were  originally  described.  Condition  of  specimens  and 
current  nomenclatorial  status  of  taxa  are  mentioned  where  appropri- 
ate. Under  each  ordinal  name,  genera  are  listed  phylogenetically 
(those  from  North  America  after  Hall  and  Kelson,  1959),  whereas 
species  and  subspecies  are  arranged  alphabetically. 

Insectr'Ora 

Cryptotis  euryrhynchis  Genoways  and  Choate,  Proc.  Biol.  Soc.  Washington, 
80:203,  December  1,  1967. 

Holotype. — Adult  male,  skin  and  skull,  KU  107143,  from  Volcan  de  Fuego 
(also  called  Volcan  de  Colima),  9800  ft.,  Jalisco;  obtained  July  10,  1966,  by 
Percy  L.  Clifton,  original  number  11059. 

Remarks. — Braincase  of  skull  smashed  and  coronoid  process  of  left  ramus 
broken. 

Scalopus  montanus  Baker,  Univ.  Kansas  Publ.,  Mus  Nat.  Hist.,  5:19,  Feb- 
ruary 28,  1951. 

Holotype. — Adult  male,  skin,  skull,  and  body  skeleton,  KU  35668,  from 
Club  Sierra  del  Carmen,  2  mi.  N  and  6  mi.  W  Piedra  Blanca,  Coahuila;  obtained 
April  7,  1950,  by  J.  R.  Alcorn,  original  number  11093. 

Chiroptera 

Nycteris  vinsoni  Dalquest,  Jour.  Mamm.,  46:256,  May  20,  1965. 

Holotype. — Adult  female,  skin  and  skull,  KU  105221,  from  S  bank  Save 
River,  212  km.  SSW  Beira,  Mozambique;  obtained  October  8,  1963,  by 
Walter  W.  Dalquest,  original  number  18739. 

Leptonycteris  nivalis  longala  Stains,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
9:355,  January  21,  1957. 

Holotype.— Aduh  female,  skin  and  skull,  KU  33087,  from  12  mi.  S  and  2 
mi.  E  Arteaga,  7500  ft.,  Coahuila;  obtained  July  11,  1949,  by  W.  K.  Clark, 
original  number  787. 

Remarks. — Regarded  as  a  synonym  of  Leptonycteris  nivalis  (Saussure, 
1860)  by  Davis  and  Carter  (1962:194). 

Stumira  ludovici  occidentalis  Jones  and  Phillips,  Univ.  Kansas  Publ.,  Mus. 
Nat.  Hist.,  14:477,  March  2,  1964. 

Holotype.— Adult  female,  skin  and  skull,  KU  92798,  from  Plumosas,  2500 
ft.,  Sinaloa;  obtained  August  31,  1962,  by  Percy  L.  Clifton,  original  number 
2939. 

Remarks. — Canine  and  first  premolar  of  right  ramus  missing. 

Natalus  mexicanus  saturatus  Dalquest  and  Hall,  Proc.  Biol.  Soc.  Washing- 
ton, 62:153,  August  23,  1949. 

Holotype. — Adult  male,  skin  and  skull,  KU  23815,  from  3  km.  E  San 
Andreas  [=Andres]  Tuxtla,  1000  ft.,  Veracruz;  obtained  January  10,  1948,  by 
Walter  W.  Dalquest,  original  number  8621. 

Remarks. — Hole  in  the  mid-ventral  region  of  skin;  left  upper  canine 
missing.    Arranged   as   Natalus   .stramineus   saturatus   by   Goodwin    (1959:7). 


Jones  and  Genoways — Holotypes  of  Mammals  131 

Regarded   as   a   synonym   of  Nataltis  .stiaininctis   nicxicanus    Miller,    1902,   by 
Handley  (1966:770). 

Myotis  argentatus  Dalquest  and  Hall,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
1:239,  December  10,  1947. 

Holotype.— Adult  male,  skin  and  skull,  KU  19228,  from  14  km.  SW 
Coatzocoalcos,  100  ft.,  Veracruz;  obtained  February  2,  1947,  by  Walter  W. 
Dalquest,  original  number  7052. 

Myotis  elegans  Hall,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.,  14:163,  May  21, 
1962. 

HoIoUjpe.—Aduh  female,  skin  and  skull,  KU  88398,  from  12)^  mi.  N 
Tihuatlan,  300  ft.,  Veracruz;  obtained  September  24,  1961,  by  Percy  L. 
Clifton,  original  number  985. 

Remarks. — Right  upper  incisors  and  canine,  left  upper  canine  and  second 
small  premolar,  and  lower  left  first  molar  missing,  both  zygomatic  arches 
broken,  and  both  auditory  bullae  separated  from  skull. 

Myotis  evotis  auriculus  Baker  and  Stains,  Univ.  Kansas  Publ.,  Mus.  Nat. 
Hist.,  9:83,  December  10,  1955. 

Holottjpe. — Adult  female,  skin  and  skull,  KU  55110,  from  10  mi.  W  and 
2  mi.  S  Piedra,  1200  ft.,  Sierra  de  Tamaulipas,  Tamaulipas;  obtained  June  9, 
1953,  by  Gerd  H.  Heinrich,  original  number  7061. 

Remarks. — Skin  with  hole  along  ventral  slit.  Arranged  as  Myotis  keenii 
auriculus  by  Findley  (1960:18).  Arranged  as  Myotis  auriculus  auriculus  by 
Genoways  and  Jones  (1969:10). 

Myotis  nigricans  dalquesti  Hall  and  Alvarez,  Univ.  Kansas  Publ.,  Mus.  Nat. 
Hist.,  14:71,  December  29,  1961. 

Holotype. — Adult  male,  skin  and  skull,  KU  23839,  from  3  km.  E  San  Andres 
Tuxtla,  1000  ft.,  Veracruz;  obtained  January  5,  1948,  by  Walter  W.  Dalquest, 
original  number  8444. 

Remarks. — Skin  with  hole  along  right  side;  right  zygomatic  arch  broken. 

Myotis  planiceps  Baker,  Proc.  Biol.  Soc.  Washington,  68:165,  December  31, 
1955. 

Holotype. — Adult  male,  skin  and  skull,  KU  48242,  from  7  mi.  S  and  4  mi. 
E  Bella  Union,  7200  ft.,  Coahuila;  obtained  June  24,  1952,  by  Albert  A. 
Alcorn,  original  number  920. 

Remarks. — Parietals  and  major  portion  of  frontals  missing. 

Myotis  thysanodes  pahasapensis  Jones  and  Genoways,  Jour.  Mamm.,  48:231, 
May  20,  1967. 

Holotype. — Adult  male,  skin  and  skull,  KU  100704,  from  6  mi.  N  New- 
castle, 6000  ft.,  Weston  Co.,  Wyoming;  obtained  July  2,  1965,  by  Ronald  W. 
Turner,  original  number  156. 

Myotis  velifer  brevis  Vaughan,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.,  7:509, 
July  23,  1954. 

Holotype. — Adult  male,  skin  and  skull,  KU  22631,  from  Madera  Canyon, 
5000  ft.,  Santa  Rita  Mountains,  Pima  Co.,  Arizona;  obtained  March  12,  1948, 
by  J.  R.  Alcorn,  original  number  5571. 

Pipistrellus  subflavus  clarus  Baker,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
7:585,  November  15,  1954. 

Holotype. — Adult  female,  skin  and  skull,  KU  48270,  from  2  mi.  W  Jimenez, 


132  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

850  ft.,  Coahuila;  obtained  June  19,  19.52,  l>y  Rollin  H.  Baker,  original  number 
2062. 

Scotophilus  alvenslebeni  Dalquest,  Jour.  Mamm.,  46:258,  May  20,  1965. 

Holotijpe. — Adult  male,  skin  and  skull,  KU  105222,  from  S  bank  Save  River, 
212  km.  SSW  Beira,  Mozambique;  obtained  October  9,  1963,  by  Walter  W. 
Dalquest,  original  number  18767. 

Remarks. — Regarded  as  synonym  of  Scotophilus  gigas  Dobson,  1875,  by 
Dalquest  (1966:134). 

Lasiurus  intermedius  insularis  Hall  and  Jones,  Univ.  Kansas  Publ.,  Mus. 
Nat.  Hist.,  14:85,  December  29,  1961. 

Holotijpe. — Adult  female,  specimen  stored  in  spirits  with  skull  removed, 
KU  81666,  from  Cienfuegos,  Las  Villas  Prov.,  Cuba;  obtained  January  3,  1948, 
by  D.  Gonzales  Muiioz,  original  number,  if  any,  unknown. 

Remarks. — Skull  missing  left  auditory  bulla. 

Antrozous  bunkeri  Hibbard,  Jour.  Mamm.,  15:227,  August  10,  1934. 

Holotype. — Adult  female,  skin  and  skull,  KU  9302,  from  7  mi.  S  [actually 
4/2  mi.  S,  ^4  mi.  E]  Sun  City,  Barber  Co.,  Kansas;  obtained  September  2,  1933, 
by  Hobart  Smith  and  Claude  W.  Hibbard,  original  number  538  (Hibbard). 

Remarks. — Arranged  as  Aiitrozous  pallidits  hunkeri  by  Krutzsch  and 
Vaughan  1955:97,  99)  and  by  Morse  and  Class  (1960:15).  A  corrected  state- 
ment of  the  type  locality  was  published  by  Jones  et  al.  ( 1967:25). 

Molossops  greenhalli  mexicanus  Jones  and  Genoways,  Proc.  Biol.  Soc.  Wash- 
ington, 80:207,  December  1,  1967. 

Holotype.— Adult  male,  skin  and  skull,  KU  108609,  from  VA  mi.  SE  Teco- 
mate,  1500  ft.,  Jalisco;  obtained  December  7,  1966,  by  Percy  L.  Clifton, 
original  number  11998. 

Eumops  perotis  renatae  Pirlot,  Le  Naturaliste  Canadian,  92:5,  January,  1965. 

Holotype. — Adult  female,  specimen  stored  in  spirits  with  skull  removed, 
KU  115920,  from  Cumana,  Sucre,  Venezuela;  obtained  on  unknown  date  by 
R.  P.  C.  Prieto,  original  niunber,  if  any,  unknown. 

Remarks. — Skull  cracked  across  interorbital  region  and  longitudinally  along 
the  palate;  part  of  right  maxillary,  part  of  fourth  upper  premolar,  and  part  of 
left  parietal  missing. 

This  specimen,  which  was  originally  deposited  in  the  Collegio  San  Jose  in 
Cumana,  Venezuela,  was  obtained  for  the  Museum  of  Natural  History  1)y 
James  D.  Smith  from  R.  P.  C.  Prieto.  Smith  entered  the  specimen  in  his  field 
catalogue  as  number  2852.    Pirlot  (1968:90)  claimed  specific  rank  for  renatae. 

Lagomorpha 

Ochotona  princeps  howelli  Borell,  Jour.  Mamm.,  12:306,  August  24,  1931. 

Holotype. — Adult  male,  skin  and  skull,  KU  45705  (originally  number  8744 
in  collection  of  Ralph  Ellis),  from  summit  of  Smith  Mountain  (near  head  Bear 
Creek,  S  end  of  Seven  Devils  Mountains),  7500  ft.,  Adams  Co.,  Idaho;  obtained 
July  16,  1930,  by  Raymond  M.  Gilmore,  original  number  1325. 

Ochotona  princeps  obscura  Long,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.,  14: 
.538,  July  6,  1965. 

Holotype. — Subadult  male,  skin  and  skull,  KU  32918,  from  Medicine  Wheel 
Ranch,  9000  ft.,  28  mi.  E  Lo\ell,  Big  Horn  Co.,  Wyoming;  obtained  July  7, 
1949,  by  J.  W.  Twente,  original  number  232. 


Jones  and  Genoways — Holotypes  of  Mammals  133 

Remarks. — Braincase  cracked  but  intact,  tips  of  nasals  and  posterior  exten- 
sion ot  the  left  zygomatic  arch  missing. 

Sylvilagus  floridanus  nelsoni  Baker,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
7:611,  April  8,  1955. 

Holotype. — Adult  female,  skin  and  skull,  KU  57771,  from  22  mi.  S  and 
5  mi.  W  Ocampo,  5925  ft.,  Coahuila;  obtained  April  4,  1952,  by  Rollin  H. 
Baker,  original  number  2571. 

Remarks. — Posterior  portion  of  right  supraorbital  process  missing.  Regarded 
as  synonym  of  Sylvilagus  floridanus  robusttis  (Bailey,  1905)  by  Raun  (1965: 
521). 

Lepus  americanus  seclusus  Baker  and  Hankins,  Proc.  Biol.  Soc.  Washing- 
ton, 63:63,  May  25,  1950. 

Holotype. — Adult  male,  skin  and  skull,  KU  20897,  from  12  mi.  E  and  2  mi. 
N  Shell,  7900  ft..  Bighorn  Mts.,  Big  Horn  Co.,  Wyoming;  obtained  July  8,  1947, 
by  Gilbert  Winemiller,  original  number  22  of  Joao  Moojen. 

Remarks. — Baker  (1959:145)  believed  seclusus  to  be  preoccupied  by  Lepus 
timid  us  seclu.sus  Degerbpl,  1940,  and  therefore  proposed  the  replacement  name 
Lepus  americanus  setzeri  for  this  subspecies.  Long  (1965a:548:  1965b:  125- 
126)  considered  the  name  Lepus  americanus  seclusus  Baker  and  Hankins  not 
to  be  preoccupied  by  Lepus  timidus  seclusus  Degerb0l  because  the  latter  name 
pertains  to  a  "variety"  or  "forma,"  which  ha\e  no  standing  under  the  Code; 
therefore  he  regarded  Lepus  americanus  .setzeri  Baker  as  a  junior  synonym  of 
L.  a.  seclusus. 

Lepus  califomicus  curti  Hall,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist,  5:42, 
October  1,  1951. 

Holotype. — Adult  female,  skin  and  skull,  KU  35470,  from  an  island,  88  mi. 
S  and  10  mi.  W  Matamoros,  Tamaulipas;  obtained  March  19,  1950,  by  E.  R. 
Hall,  original  number  6783. 

RODENTIA 

Eutamias  minimus  silvaticus  White,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
5:261,  April  10,  1952. 

Holotype. — Adult  female,  skin  and  skull,  KU  20050,  from  3  mi.  NW  Sun- 
dance, 5900  ft..  Crook  Co.,  Wyoming;  obtained  July  4,  1947,  by  H.  W.  Setzer, 
original  mmiber  1692. 

Eutamias  umbrinus  fremonti  White,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
5:575,  December  1,  1953. 

Holotype. — Adult  male,  skin,  skull,  and  prepared  baculum,  KU  41790,  from 
31  mi.  N  Pinedale,  8025  ft.,  Sublette  Co.,  Wyoming;  obtained  July  8,  1951,  by 
Rollin  H.  Baker,  original  number  1596. 

Eutamius  umbrinus  monlanus  White,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
5:576,  December  1,  1953. 

Holotype. — Adult  male,  skin,  skull,  and  prepared  baculum,  KU  20105,  from 
)2  mi.  E  and  3  mi.  S  Ward,  9400  ft.,  Boulder  Co.,  Colorado;  obtained  August  1, 
1947,  by  E.  L.  Cockrum,  original  number  721. 

Marmota  monax  bunkeri  Black,  Jour.  Mamm.,  16:319,  November  15,  1935. 

Holotype.— Aduh  female,  skin  and  skull,  KU  3089,  from  7  mi.  SW  Law- 
rence, Douglas  Co.,  Kansas;  obtained  March  8,  1920,  by  Fred  Hastie,  original 
number,  if  any,  unknown. 


134  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Spermophilus  spilosoma  oricolus  Alvarez,  Univ.  Kansas  Publ.,  Mus.  Nat. 
Hist.,  14:123,  March  7,  1962. 

Holotijpe. — Adult  female,  skin  and  skull,  KU  55497,  from  1  mi.  E  La  Pesca, 
Tamaulipas;  obtained  March  27,  1953,  by  Gerd  H.  Heinrich,  original  number 
6933. 

Thomoniys  bottae  angustidens  Baker,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
5:508,  June  1,  1953. 

Holotijpe. — Adult  male,  skin  and  skull,  KU  48481,  from  Sierra  del  Pino, 
6  mi.  N  and  6  mi.  W  Acebuches,  Coahuila;  obtained  July  3,  1952,  by  Rollin 
H.  Baker,  original  number  2141. 

Remarks. — Hall  and  Kelson  (1959:416)  arranged  tliis  and  other  subspecies 
of  Thomomys  hottae  as  races  of  Thomomijs  umhrimis;  however,  several  recent 
authors  (Anderson,  1966:189;  Dunnigan,  1967:144;  Patton  and  Dingman, 
1968:2)  have  shown  the  t\\'o  species  to  be  distinct. 

Thomomys  bottae  caneloensis  Lange,  Proc.  Biol.  Soc.  Washington,  72:131, 
Noveml^er  4,  1959. 

Holotijpe. — Adult  male,  skin  and  skull,  KU  51788,  from  Huachuca  Moun- 
tains, west  foothills,  Canelo,  10  mi.  S  Elgin,  5100  ft.,  Santa  Cruz  Co.,  Arizona; 
obtained  November  30,  1952,  by  Gerd  H.  Heinrich,  original  number  5551. 

Allotype.— Adult  female,  skin  and  skull,  KU  51786;  obtained  November  29, 
1952,  by  Gerd  H.  Heinrich,  original  number  5549. 

Thomomys  bottae  humulis  Baker,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
5:503,  June  1,  19.53. 

Holotype.—Adu\t  male,  skin  and  skull,  KU  35746,  from  3  mi.  W  Hda. 
[Hacienda]  San  Miguel,  2200  ft.,  Coahuila;  obtained  April  4,  1950,  l^y  J.  R. 
Alcorn,  original  number  11039. 

Thomomys  bottae  retractus  Baker,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
5:507,  June  1,  1953. 

Holotijpe.— Aduh  male,  skin  and  skull,  KU  44826,  from  Fortin,  3300  ft., 
33  mi.  N  and  1  mi.  E  San  Geronimo,  Coahuila;  obtained  March  29,  1952,  by 
Rollin  H.  Baker,  original  number  1971. 

Thomomys  bottae  rubidus  Youngman,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
9:.376,  February  21,  19.58. 

Holotijpe.- Aduh  female,  skin  and  skidl,  KU  72954,  from  2  9/10  mi.  E 
Caiion  City,  5344  ft.,  Fremont  Co.,  Colorado;  obtained  March  17,  1957,  by 
Richard  S.  Miller  and  Phillip  M.  Yougman,  original  number  253  (Youngman). 

Thomomys  bottae  villai  Baker,  Uni\ .  Kansas  Publ.,  Mus.  Nat.  Hist.,  5:505, 

June  1,  1953. 

Holotype.— Adult  female,  skin  and  skull,  KU  44816,  from  7  mi.  S  and  2  mi. 
E  Boquilias,  1800  ft.,  Coahuila;  obtained  March  1,  1952,  by  Rollin  H.  Baker, 
original  number  1816. 

Thomomys  talpoides  attenuatus  Hall  and  Montague,  Univ.  Kansas  Publ., 
Mus.  Nat.  Hist.,  5:29,  February  28,  1951. 

Holotijpe.— Adult  male,  skin  and  skull,  KU  15095,  from  3)2  mi.  W  Horse 
Creek  Post  Office,  7000  ft.,  Laramie  Co.,  Wyoming;  obtained  July  16,  1945, 
by  Henry  W.  Setzer,  original  number  629. 

Thomomys  talpoides  meritus  Hall,  Univ.  Kansas  Pul)l.,  Mus.  Nat.  Hist., 
5:221,  December  15,  1951. 


Jones  and  Genoways — Holotypes  of  Mammals  135 

Holottjpe. — Adult  male,  skin  and  skull,  KU  25628,  from  8  mi.  N  and  19/2 
mi.  E  Savery,  8800  ft.,  Carbon  Co.,  Wyoming;  obtained  July  19,  1948,  by 
George  M.  Newton,  original  number  4. 

Thomomys  talpoides  rostralis  Hall  and  Montague,  Univ.  Kansas  Publ.,  Mus. 
Nat.  Hist.,  5:27,  February  28,  1951. 

Holottjpe. — Adult  female,  skin  and  skull,  KU  17096,  from  1  mi.  E  Laramie, 
7164  ft.,  Albany  Co.,  Wyoming;  obtained  August  26,  1946,  by  C.  Howard 
Westman,  original  number  320. 

Remarks. — The  date  on  which  the  holotype  was  captured  was  incorrectly 
given  as  July  16,  1945,  in  the  original  description. 

Thomomys  umbrinus  varus  Hall  and  Long,  Proc.  Biol.  Soc.  Washington, 
73:35,  August  10,  1960. 

Holotype. — Adult  male,  skin  and  sk-ull,  KU  75271,  from  1  mi.  S  El  Dorado, 
Sinaloa;  obtained  November  14,  1957,  by  William  L.  Cutter,  original  number 
1452. 

Remarks. — Left  upper  premolar  missing.  Regarded  as  a  synonym  of 
Thomorinjs  bottae  sinaloae  Merriam,  1901,  by  Dunnigan   (1967:149). 

Geomys  bursarius  industrius  Villa-R.  and  Hall,  Univ.  Kansas  Publ.,  Mus. 
Nat.  Hist.,  1:226,  November  29,  1947. 

Holotype. — Adult  male,  skin  and  skull,  KU  14083,  from  Ui  mi.  N  Fowler, 
Meade  Co.,  Kansas;  obtained  December  30,  1941,  by  H.  H.  Hildebrand,  orig- 
inal number  16. 

Remarks. — Exposed  portion  of  right  upper  premolar  missing. 

Heterogeomys  hispidus  latirostris  Hall  and  Alvarez,  An.  Escuela  Nac.  Cien. 
Biol.,  10:121,  December  20,  1961. 

Holotype. — Adult  female,  skin  and  skull,  KU  82968,  from  Hacienda  Tamia- 
hua,  Cabo  Rojo,  Veracruz;  obtained  April  2,  1960,  by  M.  R.  Lee,  original 
number  1822. 

Remarks. — Catalogue  number  of  the  holotype  was  incorrectly  listed  in  the 
original  description  as  83968.  Arranged  as  Orthogeomys  hispidus  latirostris  by 
Russell  (1968a:  531). 

Pappogeomys  alcomi  Russell,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.,  9:359, 
January  21,  1957. 

Holotype. — Adult  female,  skin  and  skull,  KU  39806,  from  4  mi.  W  Maza- 
mitla,  6600  ft.,  Jalisco;  obtained  October  18,  1950,  by  J.  R.  Alcorn,  original 
number  12835. 

Pappogeomys  bulleri  infuscus  Russell,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
16:610,  August  5,  1968. 

Holotype. — Adult  male,  skin  and  skull,  KU  33451,  from  Cerro  Tequila, 
10,000  ft.,  7  mi.  S  and  2  mi.  W  Tequila,  Jalisco;  obtained  May  13,  1949,  by 
J.  R.  Alcorn,  original  number  9186. 

Pappogeomys  bulleri  lutulentus  Russell,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
16:612,  August  5,  1968. 

Holotype. — Adult  female,  skin  and  skull,  KU  92984,  from  Sierra  de  Cuale, 
7300  ft.,  9  km.  N  El  Teosinte  ( =Desmoronado),  Jalisco;  obtained  October  28, 
1962,  by  Percy  L.  Clifton,  original  number  3236. 

Pappogeomys  castanops  elibatus  Russell,  Univ.  Kansas  Publ.,  Mus.  Nat. 
Hist,  16:672,  August  5,  1968. 


136  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Holotype.—Aduk  female,  skin  and  skull,  KU  58092,  from  12  mi.  W  San 
Antonio  de  las  Alazanas,  about  7500  ft.,  Coahuila;  obtained  January  10,  1954, 
by  Robert  W.  Diekerman,  original  number  2268. 

Pappogeomys  castanops  parviceps  Russell,  Univ.  Kansas  Publ.,  Mus.  Nat. 
Hist.,  16:673,  August  5,  1968. 

Holotype. — Adult  female,  skin  and  skull,  KU  87152,  from  18  mi.  SW  Ala- 
mogordo,  4400  ft.,  Otero  Co.,  New  Mexico;  obtained  June  30,  1961,  by  M. 
Raymond  Lee,  original  number  4067. 

Pappogeomys  castanops  perexiguus  Russell,  Univ.  Kansas  Publ.,  Mus.  Nat. 
Hist.,  16:676,  August  5,  1968. 

Holotype. — Adult  female,  skin  and  skull,  KU  55584,  from  6  mi.  E  Jaco, 
Chihuahua,  4500  ft.,  in  Coahuila;  obtained  March  18,  1953,  by  Gerd  H.  Hein- 
rich,  original  number  6262. 

Pappogeomys  castanops  pratensis  Russell,  Unix.  Kansas  Publ.,  Mus.  Nat. 
Hist.,  16:653,  August  5,  1968. 

Holotype. — Adult  female,  skin  and  skull,  KU  52051,  from  8  mi.  W  and  3  mi. 
S  Alpine,  5100  ft.,  Brewster  Co.,  Texas;  obtained  December  30,  1952,  by  Gerd 
H.  Heinrich,  original  number  5684. 

Pappogeomys  castanops  surculus  Russell,  Univ.  Kansas  Publ.,  Mus.  Nat. 
Hist.,  16:688,  August  5,  1968. 

Holotype. — Adult  female,  skin  and  skull,  KU  62470,  from  La  Zarca,  Duran- 
go;  obtained  May  29,  1954,  by  Robert  W.  Diekerman,  original  number  3361. 

Remarks. — Left  upper  premolar  missing. 

Pappogeomys  castanops  torridus  Russell,  Univ.  Kansas  Publ.,  Mus.  Nat. 
Hist.,  16:665,  August  5,  1968. 

Holotype. — Adult  female,  skin  and  skull,  KU  84461,  from  3  mi.  E  Sierra 
Blanca,  about  4000  ft.,  Hudspeth  Co.,  Texas;  obtained  August  13,  1960,  by 
M.  R.  Lee,  original  number  2659. 

Pappogeomys  tylorhinus  brevirostris  Russell,  Univ.  Kansas  Publ.,  Mus.  Nat. 
Hist.,  16:733,  August  5,  1968. 

Holotype. — Adult  female,  skin  and  skull,  KU  66151,  from  2  mi.  E  Celaya, 
5800  ft.,  Guanajuato;  obtained  January  17,  1955,  by  Robert  W.  Diekerman, 
original  number  4844. 

Remarks. — Tips  of  upper  incisors  missing. 

Cratogeomys  castanops  buUatus  Russell  and  Baker,  Univ.  Kansas  Publ., 
Mus.  Nat.  Hist.,  7:597,  March  15,  1955. 

Holotype.— Aduh  female,  skin  and  skull,  KU  48498,  from  2  mi.  S  and  B'A 
mi.  E  Nava,  810  ft.,  Coahuila;  obtained  June  16,  1952,  by  Robert  J.  Russell, 
original  number  276. 

Remarks. — Exposed  portion  of  third  upper  molar  missing  on  both  sides. 
Arranged  as  Pappogeomys  castanops  huUatus  by  Russell  (1968b:632). 

Cratogeomys  castanops  jucundus  Russell  and  Baker,  Uni\ .  Kansas  Publ., 
Mus.  Nat.  Hist.,  7:599,  March  15,  1955. 

Holotype. — Adult  female,  skin  and  skull,  KU  56603,  from  Hermanas, 
1205  ft.,  Coahuila;  obtained  December  5,  1953,  In-  Robert  \V.  Diekerman, 
original  number  2051. 

Remarks. — Arranged  as  Pappogeomys  castanops  jucundus  by  Russell 
(1968b: 648). 


Jones  and  Genoways — Holotypes  of  Mammals  137 

Cratogeoniys  castanops  soididulus  Russell  and  Baker,  Univ.  Kansas  Piihl., 
Mus.  Nat.  Hist.,  7:600,  March  15,  1955. 

Holotype. — Adult  female,  skin  and  skull,  KU  56614,  from  IJ2  mi.  NW 
Ocampo,  3300  ft.,  Coahuila;  obtained  December  16,  1953,  by  Robert  W. 
Dickerman,  original  number  2164. 

Remarks. — Arranged  as  Pappogeomys  castanops  soididulus  by  Russell 
(1968b: 658). 

Cratogeomys  castanops  ustulatus  Russell  and  Baker,  Univ.  Kansas  Publ., 
Mus.  Nat.  Hist.,  7:598,  March  15,  1955. 

Holotype. — Adult  female,  skin  and  skull,  KU  34589,  from  Don  Martin,  800 
ft.,  Coahuila;  obtained  August  19,  1949,  by  W.  Kim  Clark,  original  number 
1034. 

Rouaiks. — Arranged  as  Pappogeomys  castanops  ustulatus  by  Russell  ( 196815: 
667). 

Cratogeoniys  gymnurus  atratus  Russell,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
5:539,  October  15,  1953. 

Holotype. — Adult  female,  skin  and  skull,  KU  31880,  from  top  of  Cerro 
Viejo  de  Cuyutlan,  9700  ft.,  19  mi.  S  and  9  mi.  W  Cuadalajara,  Jalisco;  obtained 
February  17,  1949,  by  J.  R.  Alcorn,  original  number  7902. 

Remarks. — Arranged  as  Pappogeomys  tylorhinus  atratus  by  Russell  (1968b: 
731). 

Cratogeomys  gymnurus  tellus  Russell,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
5:537,  October  15,  1953. 

Ho/of(//)C.— Adult  female,  skin  and  skull,  KU  33454,  from  3  mi.  W  Tala, 
4300  ft.,  Jalisco;  obtained  June  2,  1949,  by  J.  R.  Alcorn,  original  number  9376. 

Remarks. — Arranged  as  Pappogeomys  gymnurus  tellus  by  Russell  (1968a: 
482,  571;  1968b:756). 

Cratogeoniys  zinseri  niorulus  Russell,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
5:541,  October  15,  1953. 

Holotype. — Adult  male,  skin  and  skull,  KU  36679,  from  N  end  Lago  Sayula, 
4400  ft.,  9  mi.  N  and  2  mi.  E  Atoyac,  Jalisco;  obtained  March  23,  1950,  by 
J.  R.  Alcorn,  original  number  10889. 

Remarks. — Exposed  portion  of  left  upper  incisor  missing.  Considered  a 
synonym  of  Pappogeomys  gymnurus  gy?uniirus  (Merriam,  1892)  by  Russell 
(1968b: 751). 

Cratogeomys  zinseri  zodius  Russell,  ITniv.  Kansas  Publ.,  Mus.  Nat.  Hist., 
5:540,  October  15,  1953. 

Holotype. — Adult  male,  skin  and  skull,  KU  31879,  from  13  mi.  S  and  15  mi. 
W  Guadalajara,  Jalisco;  obtained  February  6,  1949,  by  J.  R.  Alcorn,  original 
number  7747. 

Remarks. — Exposed  portion  of  left  upper  incisor  and  left  lower  third  molar 
missing.  Arranged  as  Pappogeomys  tylorhinus  zodius  by  Russell  (1968a:535; 
1968b:742). 

Perognathus  flavescens  cockrunii  Hall,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
7:589,  November  15,  1954. 

Holoty pe.—Suhadu\t  female,  skin  and  skull,  KU  13045,  from  4)2  mi.  NE 
Danville,  Harper  Co.,  Kansas;  obtained  December  1,  1939,  by  Sam  Tihen, 
original  number  99  of  J.  A.  Tihen. 

Remarks. — Both  zygomatic  arches  missing. 


138  Misc.  PuBL.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Perognathus  flavus  bunkeri  Cockrum,  Univ.  Kansas  Publ.,  Mus.  Nat    Hist 
5:205,  December  15,  1951. 

Holotype.— Adult  female,  skin  and  skull,  KU  11716,  from  Conrad  Farm, 
1  mi.  E  Coolidge,  Hamilton  Co.,  Kansas;  obtained  July  1,  1936,  by  F.  Parks 
and  C.  W.  Hibbard,  original  nimiber  894  (Hibbard). 

Remarks.— Leit  zygomatic  arch  missing,  right  ramus  broken,  and  hole  in 
right  parietal. 

Perognathus  flavus  medius  Baker,  Univ.  Kansas  Publ.,  Mus.  Nat  Hist 
7:343,  February  15,  1954. 

Holotype.— Adult  female,  skin  and  skull,  KU  48583,  from  1  mi.  S  and 
6  mi.  E  Rincon  de  Romos,  6550  ft.,  Aguascalientes;  obtained  July  14,  1952,  by 
Rollin  H.  Baker,  original  number  2215. 

Remarks. — Right  zygomatic  arch  missing. 

Perognathus  flavus  pallescens  Baker,  Univ.  Kansas  Publ.,  Mus.  Nat  Hist 
7:345,  February  15,  1954. 

Holotype.— Adult  male,  skin  and  skull,  KU  40298,  from  1  mi.  SW  San 
Pedro  de  las  Colonias,  3700  ft.,  Coahuila;  obtained  February  9,  1951,  by 
J.  R.  Alcorn,  original  number  14177. 

Perognathus  flavus  parviceps  Baker,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist 
7:344,  February  15,  1954. 

Holotype. — Adult  female,  skin  and  skull,  KU  38402,  from  4  mi.  W  and 
2  mi.  S  Guadalajara,  5100  ft.,  Jalisco;  obtained  June  15,  1950,  by  J.  R.  Alcorn, 
original  number  12020. 

Remarks. — Right  zygomatic  arch  missing. 

Dipodomys  ordii  largus  Hall,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.  5:40 
October  1,  1951. 

Holotype.— Adult  female,  skin  and  skull,  KU  27234,  from  Mustang  Island, 
14  mi.  SW  Port  Aransas,  Aransas  Co.,  Te.xas;  obtained  June  30,  1948,  by  W.  K. 
Clark,  original  number  543. 

Dipodomys  ordii  parvabullatus  Hall,  Univ.  Kansas  Publ..  Mus.  Nat.  Hist 
5:38,  October  1,  1951. 

Ho/of (//)c.— Adult  male,  skin  and  skull,  KU  35454,  from  an  island,  88  mi. 
S  and  10  mi.  W  iMatamoros,  Tamaulipas;  obtained  March  19,  1950,  by  E.  R. 
Hall  and  Curt  \on  Wedel,  original  number  6778  (Hall). 

Remarks. — Both  zygomatic  arches  missing  and  hole  in  right  parietal. 

Reithrodontomys  fulvescens  meridionalis  Anderson  and  Jones,  Univ.  Kansas 
Publ.,  Mus.  Nat.  Hist.,  9:522,  January  14,  1960. 

Holotype.— Adult  male,  skin  and  skull,  KU  71388,  from  9  mi.  NNW  Esteli, 
Esteli,  Nicaragua;  obtained  July  15,  1956,  by  J.  R.  Alcorn,  original  number 
21464. 

Reithrodontomys  gracilis  insularis  Jones,  Proc.  Biol.  Soc.  Washington, 
77:123,  June  26,  1964. 

Holotype.— Adult  male,  skin  and  skull,  KU  92262,  from  8  mi.  ENE  Ciudad 
del  Carmen,  Isla  del  Carmen,  Campeche;  obtained  July  7,  1962,  by  William 
C.  Stanley,  original  number  .373. 

Reithrodontomys  spectabilis  Jones  and  Lawlor,  Univ.  Kansas  Publ.,  Mus. 
Nat.  Hist.,  16:413,  April  13,  1965. 

Holotype.— Adult  male,  skin  and  skull,   KU  92294,  from  2}i  km.   N   San 


Jones  and  Genoways — Holotypes  of  Mammals  139 

Miguel,  Isla  Coziimel,  Quintana  Roo;  obtained  August  8,  1962,  by  Ticul 
AKarez,  original  numlier  848. 

Peromyscus  angustirostris  Hall  and  Alvarez,  Proc.  Biol.  Soc.  Washington, 
74:203,  August  11,  1961. 

H olotype. —0\d  adult  male,  skin  and  skull,  KU  83226,  from  3  km.  W 
Zacaulpan,  6000  ft.,  Veracruz;  obtained  April  12,  1960,  by  M.  Raymond  Lee, 
original  number  1886. 

Remarks. — Regarded  as  a  synonym  of  Peromtjscus  furvtts  J.  A.  Allen  and 
Chapman,  1897,  by  Mus.ser  (1964:12). 

Peromyscus  boylii  cansensis  Long,  Univ.  Kansas  Publ.,  Nhis.  Nat.  Hist., 
14:101,  December  29,  1961. 

Holotype. — Adult  male,  skin  and  skull,  KU  81830,  from  4  mi.  E  Sedan, 
Chautauqua  Co.,  Kansas;  obtained  December  30,  1959,  by  C.  A.  Long,  orig- 
inal number  456. 

Remarks. — Both  zygomatic  arches  broken.  Regarded  as  a  synonym  of  Pero- 
myscus boylii  attwateri  J.  A.  Allen,  1895,  by  Choate  et  al.  (1967:312). 

Peromyscus  difficilis  petricola  Hoffmeister  and  de  la  Torre,  Proc.  Biol.  Soc. 
Washington,  72:167,  November  4,  19.59. 

Holotype. — Adult  female,  skin  and  skull,  KU  33239,  from  12  mi.  E  San 
Antonio  de  las  Alazanas,  9000  ft.,  Coahuila;  obtained  August  2,  1949,  by 
W.  Kim  Clark,  original  number  979. 

Remarks. — Left  ramus  broken. 

Peromyscus  maniculalus  ozarkiarum  Black,  Jour.  Manim.,  16:144,  May  15, 
1935. 

Holotype. — Adult  male,  skin  and  skull,  KU  10104,  from  3  mi.  S  Winslow, 
Washington  Co.,  Arkansas;  obtained  August  .30,  1934,  by  Ruby  Black,  original 
number  853  of  J.  D.  Black. 

Remarks. — Right  zygomatic  arch  broken. 

Peromyscus  melanophrys  coahuilensis  Baker,  L^ni\.  Kansas  Publ.,  Mus.  Nat. 
Hist.,  5:2.57,  April  10,  19.52. 

Holotype. — Adult  female,  sldn  and  skull,  KU  35019,  from  7  mi.  S  and  1  mi. 
E  Gomez  Farias,  6500  ft.,  Coahuila;  obtained  November  20,  1949,  by  W.  K. 
Clark,  original  number  1293. 

Remarks. — Skin  with  slippage  on  belly  and  right  flank.  In  the  original 
description  in  the  statement  of  particulars  about  the  holotype,  the  locality  of 
capture  was  incorrectly  given  as  "Gomez  Farias,  6500  ft.,  Coahuila,"  although 
elsewhere  in  the  paper  it  is  located  correctly. 

Peromyscus  melanophrys  micropus  Baker,  Univ.  Kansas  Publ.,  Mus.  Nat. 
Hist.,  5:255,  April  10,  1952. 

Holotype. — Adult  male,  skin  and  skull,  KU  31760,  from  3  mi.  N  Guadala- 
jara, Jalisco;  obtained  January  18,  1949,  by  J.  R.  Alcorn,  original  number  7402. 

Remarks. — The  date  on  which  the  holotype  was  obtained  was  incorrectly 
given  in  the  original  description  as  January  11,  1949. 

Peromyscus  ochraventer  Baker,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist,  5:213, 
December  15,  1951. 

Holotype.— Aduh  female,  skin  and  skull,  KU  36958,  from  70  km.  [by  high- 
\\ay]  S  Ciudad  Victoria  and  6  km.  W  of  the  [Pan  American]  highway  [at  El 
Carrizo],  Tamaulipas;  obtained  January  12,  1950,  l)y  William  J.  Schaldach,  Jr., 
original  number  566. 


140  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Peromyscus    truei   erasmus   Finley,    Univ.    Kansas    Publ.,    Mus.    Nat.    Hist 
5:265,  May  23,  1952. 

Holotype.— Young  adult  female,  skin  and  skull,  KU  34417,  from  8  mi.  NE 
Durango,  6200  ft.,  Durango;  obtained  August  16,  1949,  by  J.  R.  Alcorn,  original 
number  10255. 

Reryiarks. — Right  z>'gomatic  arch  broken.  Regarded  as  a  synonym  of  Pero- 
mijscus  truei  gentilis  Osgood,  1904,  by  Baker  ( 1960:321 ). 

Baiomys  musculus  pullus  Packard,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
9:401,  December  19,  1958. 

Holotype.— Aduh  female,  skin  and  skull,  KU  71605,  from  8  mi.  S  Condega, 
Esteli,  Nicaragua;  obtained  July  15,  1956,  by  A.  A.  Alcorn,  original  number 
4218. 

Baiomys    taylori   canutus    Packard,    Univ.    Kansas   Publ.,    Mus.    Nat     Hist 
9:643,  June  16,  1960. 

Holotype.— Adult  male,  skin  and  skull,  KU  62075,  from  1  mi.  S  Pericos, 
Sinaloa;  obtained  June  14,  1954,  by  A.  A.  Alcorn,  original  number  1754. 

Remarks. — Left  ear  missing  and  some  slippage  of  pelage  on  left  shoulder. 

Baiomys  taylori  fuliginatus  Packard,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
9:645,  June  16,  1960. 

Holotijpe.— Adult  male,  skin  and  skull,  KU  36765,  from  10  mi.  E  and  2  mi. 
N  Ciudad  de  Maiz,  4000  ft.,  San  Luis  Potosi;  obtained  January  17,  1950,  by 
J.  R.  Alcorn,  original  number  10400. 

Sigmodon  hispidus  solus  Hall,  Uni\.  Kansas  Publ.,  Mus.  Nat.  Hist.,  5:42, 
October  1,  1951. 

Holotype.— Adult  male,  skin  and  skull,  KU  35468,  from  an  island,  88  mi.  S 
and  10  mi.  W  Matamoros,  Tamaulipas;  obtained  March  22,  1950,  by  E.  R. 
Hall  and  Curt  \()n  Wedel,  original  number  6806  (Hall). 

Remarks. — Hole  in  left  orbit  and  left  zygomatic  arch  cracked  in  region  of 
infraorbital  foramen. 

Neotoma  albigula  subsolana  Alvarez,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
14:141,  April  30,  1962. 

Holotype.— Adult  male,  skin  and  skull,  KU  56950,  from  Miquihuana,  6400 
ft.,  Tamaulipas;  obtained  July  20,  1953,  by  Gerd  H.  Heinrich,  original  number 
7553B. 

Neotoma  angustapalata  Baker,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.,  5:217, 
December  15,  1951. 

Holotype. — Subadult  male,  skin,  skull,  and  prepared  baculum,  KU  36976, 
from  70  km.  [by  highway]  S  Ciudad  Victoria  and  6  km.  W  of  the  [Pan  Ameri- 
can] highway  [at  El  Carrizo],  Tamaulipas;  obtained  January  14,  1950,  by 
William  J.  Schaldach,  Jr.,  original  number  578. 

Neotoma  mexicana  eremita  Hall,  Jour.  Washington  Acad.  Sci.,  45:328, 
October  31,  1955. 

Holotype. — Adult  female,  skin,  skull,  and  body  skeleton,  KU  64532,  from 
1  mi.  S  San  Francisco,  50  ft.,  Nayarit;  obtained  January  27,  1955,  by  J.  R. 
Alcorn,  original  number  17830. 

Neotoma  mexicana  scopulorum  Finley,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
5:529,  August  15,  1953. 


Jones  and  Genoways — Holotypes  of  Mammals  141 

Hohttjpe.—0]d  adult  male,  skin  and  skull,  KU  37137,  from  37°  47'  N, 
103°  28'  W,  3  mi.  WV  Higbee,  4300  ft.,  Otero  Co.,  Colorado;  ol)tained  May  16, 
1950,  by  R.  B.  Finley,  Jr.,  original  number  500516-1. 

Remarks. — Left  upper  first  molar  appears  to  ha\e  been  lost  in  life. 

Nelsonia  neotomodon  cliftoni  Cenoways  and  Jones,  Proc.  Biol.  Soc.  Wash- 
ington, 81:97,  April  30,  1968. 

Holotii])c.—\i\\\\i  female,  skin  , and  skull,  KU  1094.37,  from  2;^  mi.  ENE 
Jazmin,  6800  ft.,  Jalisco;  obtained  October  20,  1966,  by  Percy  L.  Clifton, 
original  number  11706. 

Microtus  montanus  codiensis  Anderson,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
7:497,  July  23,  1954. 

Holotype. — Adult  female,  skin  and  skull,  KU  27578,  from  3  1/5  mi.  E  and 
3/5  mi.  S  Cody,  5020  ft..  Park  Co.,  Wyoming;  obtained  August  11,  1948,  by 
James  W.  Bee,  original  number  18-8-11-48. 

Microtus  montanus  pratincolus  Hall  and  Kelson,  Uni\.  Kansas  Publ.,  Mus. 
Nat.  Hist.,  5:75,  October  1,  1951. 

Holotype. — Adult  female,  skin  and  skull,  KU  34004,  from  6  mi.  E  Hamil- 
ton, .3700  ft.,  Ravalli  Co.,  Montana;  obtained  August  14,  1949,  by  John  A. 
White,  original  number  477. 

Remarks. — Subspecilic  name  emended  to  pratincola  by  Hall  and  Cockrum 
(1953:417). 

Microtus  montanus  zygomaticus  Anderson,  Univ.  Kansas  Publ.,  Mus.  Nat. 
Hist.,  7:500,  July  23,  1954. 

Holoiypc. — Adult  male,  skin  and  skull,  KU  32761,  from  Medicine  Wheel 
Ranch,  9000  ft.,  28  mi.  E  Lo\ell,  Big  Horn  Co.,  Wyoming;  obtained  July  8, 
1949,  by  R.  Freiburg,  original  number  105. 

Microtus  ochrogaster  taylori  Hibbard  and  Rinker,  Univ.  Kansas  Sci.  Bull., 
29:256,  October  15,  1943. 

Holotype. — Adult  female,  skin  and  skull,  KU  14126,  from  U2  mi.  N  Fowler, 
Meade  Co.,  Kansas;  obtained  June  17,  1942,  by  George  C.  Rinker,  original 
number  1195  of  C.  W.  Hibbard. 

Remarks. — Hole  in  left  auditory  bulla. 

Microtus  pennsylvanicus  alcomi  Baker,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist, 
5:105,  November  28,  1951. 

Holotype. — Adult  female,  skin  and  skull,  KU  21552,  from  6  mi.  SW  Kluane, 
2550  ft.,  Yukon  Territory,  Canada;  obtained  August  24,  1947,  by  J.  R.  Alcorn, 
original  number  5240. 

Microtus  pennsylvanicus  finitus  Anderson,  Univ.  Kansas  Publ.,  Mus.  Nat. 
Hist,  9:96,  May  10,  1956. 

Holotype. — Adult  female,  skin  and  skull,  KU  50204,  from  5  mi.  N  and  2  mi. 
W  Parks,  Dundy  Co.,  Nebraska;  obtained  August  16,  1952,  by  J.  Knox  Jones,  Jr., 
original  number  906. 

Remarks. — Right  zygomatic  arch  broken. 

Microtus  pennsylvanicus  pullatus  Anderson,  Univ.  Kansas  Publ.,  Mus.  Nat. 
Hist.,  9:97,  May  10,  1956. 

Holotype. — Adult  male,  skin  and  skull,  KU  37873,  from  12  mi.  N  and  2  mi. 
E  Sage,  6100  ft.,  Lincoln  Co.,  Wyoming;  obtained  July  19,  1950,  by  Rollin  H. 
Baker,  original  number  1343. 


142  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Microtus  pennsylvanicus  tananaensis  Baker,  Univ.  Kansas  Publ.,  Mus.  Nat. 
Hist.,  5:107,  November  28,  1951. 

Holotype. — Adult  female,  skin  and  skull,  KU  21509,  from  Yerrick  Creek, 
21  mi.  W  and  4  mi.  N  Tok  Junction,  Alaska;  obtained  July  20,  1947,  by  J.  R. 
Alcorn,  original  number  5023. 

Remarks. — Left  auditory  bulla  broken. 

Microtus  pennsylvanicus  uligocola  Anderson,  Univ.  Kansas  Publ.,  Mus.  Nat. 
Hist.,  9:94,  May  10,  1956. 

Holotype. — Adult  male,  skin  and  skull,  KU  26898,  from  6  mi.  W  and  32  mi. 
S  Loveland,  5200  ft.,  Larimer  Co.,  Colorado;  obtained  July  26,  1948,  by 
James  O.  Lounquist,  original  number  349. 

Synaptomys  cooperi  paludis  Hibbard  and  Rinker,  Univ.  Kansas  Sci.  Bull., 
28:26,  May  15,  1942. 

Holotype. — Adult  male,  skin  and  skull,  KU  13713,  from  Meade  County 
State  Park,  14  mi.  SW  Meade,  Meade  Co.,  Kansas;  obtained  July  12,  1941,  by 
Claude  W.  Hibbard,  original  number  528. 

Synaptomys  cooperi  relictus  Jones,  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist., 
9:387,  May  12,  1958. 

Holotype. — Adult  female,  skin  and  skull,  KU  51617,  from  Rock  Creek  Fish 
Hatchery,  5  mi.  N  and  2  mi.  W  Parks,  Dundy  Co.,  Nebraska;  obtained  Novem- 
ber 1,  1952,  by  J.  Knox  Jones,  Jr.,  original  number  995. 

Zapus  hudsonius  pallidas  Cockrum  and  Baker,  Proc.  Biol.  Soc.  Washington, 
63:1,  April  26,  1950. 

Holotype. — Adult  male,  skin,  skull,  and  body  skeleton,  KU  22953,  from  NW 
corner  sec.  4,  T.  12  S,  R.  20  E,  6)2  mi.  N  and  1%,  mi.  E  Lawrence,  Douglas  Co., 
Kansas;  obtained  May  4,  1948,  by  E.  Lendell  Cockrum  and  Rollin  H.  Baker, 
original  number  916  (Cockrum). 

Carnivora 

Taxidea  taxus  kansensis  Schanz,  Jour.  Mamm.,  31:346,  August  21,  1950. 

Holotype. — Adult  female,  tanned  (cased)  skin  and  skull,  KU  21989,  from 
4  mi.  SE  McLouth,  Leavenworth  Co.,  Kansas;  obtained  November  30,  1947,  by 
Henry  Murr  and  prepared  by  James  O.  Lounquist,  original  number  46. 

Remark.^. — Bullet  hole  entering  skull  through  right  parietal  and  leaving 
through  right  alisphenoid,  but  braincase  intact.  The  taxonomic  status  of  this 
nominal  subspecies  is  uncertain  because  the  species  is  in  need  of  systematic 
review. 

Mephitis  macroura  eximius  Hall  and  Dalquest,  Univ.  Kansas  Publ.,  Mus. 
Nat.  Hist.,  1:579,  January  20,  1950. 

Holotype.— Adult  female,  skin  and  skull,  KU  19272,  from  15  km.  W  Piedras 
Negras,  300  ft.,  Veracruz;  obtained  January  13,  1947,  by  J.  Mazza  and  Walter 
W.  Dalquest,  original  number  7017  (Dalquest). 

Geographic  Origin  of  Type  Specimens 

The  countries,  states,  and  territories  whence  originated  the  holotypes  in  the 
foregoing  list  are  here  arranged  alphabetically.  Names  of  species  and  sulispecies 
are  listed  alphabetically,  in  the  combination  used  in  the  original  description, 
under  the  place  names. 


Jones  and  Genoways — Holotypes  of  Mammals  143 

Canada 

Yukon  Territory 

Microtus  pennsylvanicus  alcorni  Baker 
Cuba 

Lasiuiu.s  intermedins  iusularis  Hall  and  Jones 
Mexico 

Aguascalientes 

Peio^nathus  ftaviis  mcdius  Baker 
Campeche 

Reithrodontomys  gracilis  iusularis  Jones 
Coahiiila 

C ratogeomys  castanops  hullatus  Russell  and  Baker 

Cratugeomys  castanops  jucundiis  Russell  and  Baker 

Craiogeomys  castanops  sordidnhis  Russell  and  Baker 

Crato'^eomys  castanops  nstnlatus  Russell  and  Baker 

Leptonycteris  nivalis  longala  Stains 

Myotis  planiceps  Baker 

Pappogeomys  castanops  elihatus  Russell 

Fappogeomys  castanops  perexiguus  Russell 

Perognathus  flavus  pallescens  Baker 

Peromyscus  clifficilis  petricola  Hoftmeister  and  de  la  Torre 

Peromyscus  melanophrys  coahiiilensis  Baker 

Pipistrellus  subflavus  clams  Baker 

Scalopus  montanns  Baker 

Sylvilagus  floridanns  nelsoni  Baker 

Thomomys  hottae  angustidens  Baker 

Thomomys  hottae  humulis  Baker 

Thomomys  hottae  retractus  Baker 

Thomomys  hottae  villai  Baker 

Durango 

Pappogeomys  castanops  surculus  Russell 
Peromyscus  truei  erasmus  Finley 

Guanajuato 

Pappogeomys  tylorhinus  hrevirostris  Russell 
Jalisco 

Cratogeomys  gymnurus  atratus  Russell 

Cratogeomys  gymnurus  tellus  Russell 

Cratogeomys  zinseri  morulus  Russell 

Cratogeomys  zinseri  zodius  Russell  , 

Cryptotis  euryrhynchis  Genoways  and  Choate 

Molossops  greenhalli  mexicanus  Jones  and  Genoways 

Nelsonia  neotomodon  cliftoni  Genoways  and  Jones 

Pappogeomys  alcorni  Russell 

Pappogeomys  hulleri  infuscus  Russell 

Pa))pogeomys  hulleri  lutulentus  Russell 

Perognathus  flavus  parviceps  Baker 

Peromyscus  melanophrys  micropus  Baker 

Nayarit 

Neotoma  mexicana  eren^ita  Hall 
Quintana  Roo 

Reithrodontomys  spectabilis  Jones  and  Lawlor 
San  Luis  Potosi 

Baiomys  taylori  fuliginatus  Packard 
Sinaloa 

Baiomys  taylori  canutus  Packard 

Sturnira  ludovici  occidentalis  Jones  and  Phillips 

Thomomys  umhrinus  varus  Hall  and  Long 


144  Misc.  Publ.  51,  Una'.  Kansas  Mus.  Nat.  Hist. 

Tamaiilipas 

Dipodomy.s  oidii  paivahuUatiis  Hall 
Lepiis  colifornicus  ctirti  Hall 
Myotis  evotis  auriculus  Baker  and  Stains 
Neotoma  alhigula  siibsolana  AKarez 
Neotoma  angusiapalata  Baker 
Pewtuysciis  ocliravcntcr  Baker 
Sigmoiloit  liispidu.s  solus  Hall 
SpermophiUis  spilosoma  oricohis  Alvarez 
Veracruz 

Heterogeomys  hispid  us  latirostris  Hall  and  Alvarez 
Mephitis  macrouio  eximius  Hall  and  Dalqiiest 
Myotis  argentatus  Dalquest  and  Hall 
Myotis  ck'gans  Hall 

Myotis  nigricans  dalquesti  Hall  and  Alvarez 
Natahis  mexicanus  saturatus  Dalquest  and  Hall 
Peromyscus  angustirostris  Hall  and  Alvarez 
Mozambique 

Nycteris  vinsoni  Dalquest 
ScotO))JiiIus  alvenslehcni  Dalquest 
Nicaragua 

Baiomys  miisculus  puUus  Packard 

Reithrodontomys  fulvescens  meridionalis  Anderson  and  Jones 
United  States 
Alaska 

Microtus  pennsylvanicus  tananaensis  Baker 
Arizona 

Myotis  velifer  hrevis  Vaughan 

Tliomomys  bottae  caneloensis  Lange 
Arkansas 

Feromyscus  maniculatus  ozarkiarum  Black 
Colorado 

Eutamias  umbrinus  montanus  White 

Microtus  pennsylvanicus  tdigocola  Anderson 

Neotoma  mexicana  scopuJoruni  Finley 

Tliomomys  bottae  rubidus  Youngman 
Idaho 

Ochotona  princeps  howelli  Borell 
Kansas 

Antrozous  bunkeri  Hibbard 

Geomys  bursarius  industrius  Villa-R.  and  Hall 

Marmota  monax  hunkeri  Black 

Microtus  ochrogaster  taylori  Hibbard  and  Rinker 

Perognathus  flavescens  cockrunii  Hall 

Perognathus  fiavus  hunkeri  Cockrinn 

Peromyscus  boyUi  cansensis  Long 

Synaptomys  cooperi  pahidis  Hibbard  and  Rinker 

Taxidea  taxus  kanscnsis  Schanz 

Zapus  Jiudsoiiius  ])uUidus  Cockruni  and  Baker 
Montana 

Microtus  montanus  pratincohis  Hall  and  Kelson 
Nebraska 

Microtus  pennsylvanicus  finitus  Anderson 

Syiiaptomys  coo))eri  rehctus  Jones 
New  Mexico 

Pappogeomys  castanops  parviceps  Russell 


Jones  and  Genoways — Holotypes  of  Mammals  145 

Texas 

Dipodomijs  ordii  largus  Hall 
Fappogeomijs  castanops  pratcnsis  Russell 
Fappogeomijs  castanops  tonidus  Russell 

Wyoming 

Etitaiiiias  viininuis  silvatictis  White 
EutuDiias  uiiihiiiitis  frcnionti  White 
Lepus  arnericanus  seclusiis  Baker  and  Hankins 

[=zLcptis  arnericanus  sctzeri  Baker] 
Microtiis  nionta)ius  codiensis  Anderson 
Microtus  niontanus  zt/gornaticus  Anderson 
Microtiis  ))ennsiilvanicus  puUatus  Anderson 
Myotis  tlujsanodes  pahasapensis  Jones  and  Genoways 
Ochotona  princeps  ohscura  Long 
Tlionionu/s  talpoides  attenuatus  Hall  and  Montague 
Tlionioniiis  talpoides  merittis  Hall 
Tlioniontijs  talpoides  rostralis  Hall  and  Montague 
Venezuela 

Eumops  perotis  renatae  Pirlot 


Literature  Cited 

Allen,  J.  A. 

1895.     On   the  species  of   the   genus   Reithrodontomys.    Amer.   Mus.   Nat. 
Hist,  7:107-143. 
Anderson,  S. 

1966.  Taxonomy  of  gophers,  especially  Thomorntis  in  Chihuahua,  Mexico. 
Syst.  Zool,  15:189-198,  6  figs. 

Baker,  R.  H. 

19.59.     Substitute  name  for  Lepus  arnericanus  seclusus  Baker  and  Hankins. 

Jour.  Manini.,  40:145. 
1960.      Mammals   of   the   Guadiana   Lava   Field,   Diuango,    Mexico.     Publ. 
Mus.  Michigan  State  Univ.,  Biol.  Ser.,  l:.30.3-.328,  3  figs. 
Choate,  J.  R.,  C.  J.  Phillips,  and  H.  H.  Genoways 

1967.  Taxonomic  status  of  the  brush  mouse,  Peromifscus  hoylii  cansensis 
Long,  1961.   Trans.  Kansas  Acad.  Sci.,  69:306-313,  2  figs. 

Dalquest,  W.  W. 

1966.  Scotopliilus  alvenslebeni  Dalquest  a  synonym  of  Scotoi^hihis  gigas 
Dobson.    Jour.  Mamm.,  47:134. 

Davis,  W.  B.,  and  D.  C.  Carter 

1962.     Review  of  the  genus  Leptonycteris  ( Mammalia :Chiroptera).    Proc. 
Biol.  Soc.  Washington,  75:193-198. 

DUNMGAN,  P.   B. 

1967.  Pocket  gophers  of  the   genus   Thomomys  of  the   Mexican   state   of 
Sinaloa.    The  Radford  Review,  21:139-168,  4  figs. 

Fixdley,  J.  S. 

1960.     Identity   of  the   long-eared   myotis   of   the   Southwest   and   Mexico. 
Jour.  Mamm.,  41:16-20,  1  pi.,  1  fig. 

Genoways,  H.  H.,  and  J.  K.  Jones,  Jr. 

1969.     Taxonomic  status  of  certain  long-eared  bats   (genus  Myotis)   from 
the   southwestern   United   States   and    Mexico.     Southwestern    Nat., 
14:1-1.3,  5  figs. 
Goldman,  E.  A. 

1914.     A   new   bat   of   the   genus   Mimon   from    Mexico.     Proc.    Biol.    Soc. 
Washington,  27:75-76. 
Goodwin,  G.  G. 

1959.     Bats  of  the  subgenus  Natalus.   Amer.  Mus.  Novit.,  1977:1-22,  2  figs. 


146  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Hall,  E.  R.,  and  E.  L.  Cockrum 

1953.  A  synopsis  of  the  North  American  Microtine  rodents.  Univ.  Kansas 
Pul)l.,  Mus.  Nat.  Hist.,  5:373-498,  149  figs. 

Hall,  E.  R.,  and  K.  R.  Kelsox 

1959.  The  mammals  of  North  America.  Ronald  Press,  New  York,  l-.xxx  -\- 
1-546  +  79  and  2:viii  +  547-1083  +  79,  illustrated. 

Handley,  C.  O.,  Jr. 

1966.  Checklist  of  the  mammals  of  Panama.  Pp.  753-793,  in  Ectoparasites 
of  Panama  (R.  L.  Wenzel  and  V.  J.  Tipton,  eds.).  Field  Mus.  Nat. 
Hist.,  Chicago,  xii  +  861  pp.,  illustrated. 

Jones,  J.  K.,  Jr.,  E.  D.  Fleharty,  and  P.  B.  Dunnigan 

1967.  The  distributional  status  of  bats  in  Kansas.  Misc.  Publ.,  Mus.  Nat. 
Hist.,  Univ.  Kansas,  46:1-33,  11  figs. 

Krutzsch,  p.  H.,  and  T.  A.  Vaughan 

1955.     Additional    data    on    the    bacula    of    North    American    bats.     Jour. 
Mamm.,  36:96-100,  1  fig. 
Long,  C.  A. 

1965a.  The  mammals  of  Wyoming.    Univ.  Kansas  Pub!.,  Mus.  Nat.   Hist., 

14:49.3-758,  82  figs. 
19651).  Taxonomic  status  of  the  snowshoe  rabbit,  Le])us  (imericantis  sechisiis 
Baker  and  Hankins.    Proc.  Biol.  Soc.  Washington,  78:125-126. 

Morse,  R.  C,  and  B.  P.  Glass 

1960.  The  taxonomic  status  of  Antrozous  htmkeri.  Jour,  Mamm.,  41:10- 
15,  2  figs. 

Musser,  G.  G. 

1964.  Notes  on  geographic  distribution,  habitat,  and  taxonomy  of  some 
Mexican  mammals.  Occas.  Papers  Mus.  Zool.,  Univ.  Michigan,  636: 
1-22,  1  fig. 

PaTTON,  J.   L.,   AND  R.   E.   DiNGMAN 

1968.  Chromosome  studies  of  pocket  gophers,  genus  Thomomijs.  I.  The 
specific  status  of  Thomomijs  umhrinus  (Richardson)  in  Arizona. 
Jour.  Mamm.,  49:1-13,  8  figs. 

PiRLOT,  P. 

1968.  Chiropteres  du  Perou,  specialement  de  Haute- Amazonie.  Mam- 
malia, 32:86-96,  2  figs. 

Raun,  G.  G. 

1965.  The  subspecific  status  of  the  cottontail,  Sijlvilagus  ftoridamis,  in 
northern  Coahuila,  Mexico.    Jour.  Mamm.,  46:519-521. 

Russell,  R.  J. 

1968a.  Evolution  and  classification  of  the  pocket  gophers  of  the  subfamily 
Geomyinae.   Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.,  16:473-579,  9  figs. 
1968b.  Revision  of  pocket  gophers  of  the  genus  Pappogeomys.    Uni\-.  Kan- 
sas Publ.,  Mus.  Nat.  Hist.,  16:581-776,  10  figs. 


ECOLOGY  OF  POCKET  GOPHERS  OF 
MESA  VERDE,  COLORADO 

BY 

Charles  L.  Douglas 

The  Mesa  Verde  land  mass  consists  of  about  200  square  miles  of 
plateau  country  in  southwestern  Colorado  near  the  Four  Corners, 
where  Colorado,  Utah,  Arizona,  and  New  Mexico  have  a  common 
boundary.  The  Mesa  Verde,  named  by  early  Spanish  explorers  of 
the  Southwest,  is  the  remnant  of  a  plateau  laid  down  by  late  Creta- 
ceous seas.  Erosion  has  dissected  the  plateau  until  it  is  now  a 
cuesta  consisting  of  long,  finger-like  mesas  joined  at  their  northern 
ends  but  otherwise  separated  from  each  other  by  deep  canyons 
(Figs.  1-3).  In  1906,  part  of  the  land  mass  was  set  aside,  by  Con- 
gress, as  Mesa  Verde  National  Park  in  order  to  preserve  dwellings 
of  prehistoric  Indians,  for  which  the  area  is  famous.  Today,  more 
than  52,000  acres  are  included  within  the  boundaries  of  the  park. 

Elevations  in  Mesa  Verde  National  Park  range  from  8572  feet  at 
Park  Point  down  to  about  6500  feet  at  the  southern  ends  of  the 
mesas,  where  the  canyons  are  from  600  to  900  feet  deep.  Mesa 
Verde  is  in  the  pinyon-juniper  climax  region  that  extends  through 
much  of  the  Southwest;  and  pinyon-juniper  woodland  is  the  domi- 
nant canopy  cover  on  the  tops  of  the  mesas.  Oak  chaparral  and 
mixed  shrubs  occur  on  slopes.  Past  fires  apparently  have  permitted 
establishment  of  a  shrub  zone  at  some  of  the  higher  elevations. 
Sagebrush  grows  in  the  bottoms  of  canyons  and  in  many  of  the 
drainages  on  top  of  the  mesas,  and  also  persists  as  a  late  successional 
stand  on  the  disturbed  soil  around  prehistoric  surface  dwellings. 
Stands  of  Douglas  fir  occurs  on  many  of  the  cooler  north-facing 
slopes  where  micro-climates  are  favorable  for  its  growth.  Indixidual 
trees  and  small  stands  of  ponderosa  pine  are  found  in  sheltered 
places  throughout  the  Mesa  Verde. 

Climatically  the  Mesa  Verde  is  semi-arid.  Precipitation  has 
averaged  about  18  inches  each  year  for  the  past  45  years,  with  July 
and  August  generally  having  the  most  precipitation.  Precipitation  is 
heavier  on  the  higher,  northern  end  of  the  park  than  on  the  southern 
end,  owing  to  the  proximity  of  the  northern  parts  of  the  park  to  the 
nearby  La  Plata  Mountains. 

(147) 


148 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


I08°30' 


I08°25' 


37° 
15' 


37° 
10' 


UTE 


MOUNTAIN       INDIAN      RESERVATION 


10,000     FEET 


MESA  VERDE    NATIONAL    PARK 

AND   VICINITY 

MONTEZUMA     COUNTY 
COLORADO 


I08°30' 


I08°25' 


Fig.  1.    Map  of  Mesa  Verde  National  Park  and  vicinity,  showing  localities  at 
which  specimens  of  pocket  gophers  have  been  collected. 


Interdisciplinary  research  recently  conducted  under  the  auspices 
of  the  Wetherill  Mesa  Archeological  Project  has  resulted  in  reports 
on  mammals  of  the  park  ( Douglas,  1963,  1967a,  1967b ) ,  on  amphib- 
ians and  reptiles  (Douglas,  1966),  on  plants  (Erdman,  1962;  Erd- 
man  et  al,  1962;  Welsh  and  Erdman,  1964;  Fritts  et  ah,  1965;  Doug- 
las and  Erdman,  1967),  and  a  report  on  climatic  conditions  in  the 


Douglas — Pocket  Gophers  of  Mesa  Verde  149 

park  is  in  press.  Earlier  zoological  studies  include  a  report  on  mam- 
mals by  Anderson  ( 1961 ) . 

Pocket  gophers,  Tliomomys  bottae  aureus,  are  found  in  Mesa 
Verde  along  roadsides  where  soil  has  been  disturbed,  in  grassy 
meadows,  in  stands  of  sagebrush,  and  in  the  shrub  zone  along  the 
north  rim  of  the  park.  Gophers  occasionally  are  found  in  openings 
in  the  pinyon-juniper  \\'oodland,  in  oak  brush,  and  in  rocky  areas 
that  do  not  appear  to  offer  adequate  conditions  for  fossorial  animals. 
The  diversity  of  habitats  used  by  gophers  led  me  to  study  their 
ecology  in  Mesa  Verde.  This  report  is  the  first  to  result  from  these 
studies. 

Gophers  were  studied  periodically  from  1962  through  1967. 
Field  notes  were  kept  on  locations  and  activities  of  colonies  and 
indi\iduals  from  1961  through  1963.  Trapping  and  analyses  of 
habitats  was  begun  in  1963  and  continued  through  1967.  Most  of 
the  gophers  trapped  for  use  in  this  study  were  caught  in  1965  and 
1967.  Twenty-eight  months  were  spent  in  residence  in  Mesa  Verde 
National  Park  from  1961  through  1964;  also,  I  visited  the  park  from 
August  10  to  16,  1965,  April  IS  to  May  5,  1967,  and  September  IS 
to  30,  1967,  to  study  pocket  gophers. 

Distribution  of  Gophers  in  Mesa  Verde 

Gophers  have  been  collected  in  Mesa  Verde  at  the  places  shown 
in  Figure  1;  the  localities  are  listed  below  under  "specimens  exam- 
ined" and  "other  locality  records."  Specimens  listed  under  the  latter 
heading  have  not  been  examined  by  me.  Specimens  housed  in 
various  collections  are  designated  as  follows:  Museum  of  Natural 
History,  Uni\'ersity  of  Kansas  (KU);  United  States  National  Mu- 
seum (USNM);  Mesa  Verde  National  Park  (MV);  Museum  of  Verte- 
brate Zoology,  University  of  California  ( M VZ ) .  Specimens  presently 
in  my  personal  collection  are  listed  by  locality,  followed  in  paren- 
theses by  the  number  of  indi\'iduals.  A  total  of  132  specimens  from 
Mesa  Verde  were  examined  and  measured.  To  the  best  of  m\' 
knowledge,  the  specimens  listed  below  represent  all  gophers  that 
ha\'e  been  taken  in  the  park. 

Specimens  examined. — Total  132,  as  follows:  Mancos  River,  6200  ft.  ( KU 
69307-69315);  head  Prater  Canyon,  7700  ft.  (13);  Upper  Well,  Prater  Canyon, 
7575  ft.  (KU  69279);  Prater  Canyon,  7500  ft.  ( KU  7.5977);  ^i  mi.  N  Middle 
Well,  Prater  Canyon,  7,500  ft.  ( KU  69280);  Midd'e  Well,  Prater  Canyon,  7.500 
ft.  (KU  69281-69285);  6.8  mi.  S  North  Rim,  Moccasin  Mesa  ( KU  1020.58); 
8  mi.  S  North  Rim,  Moccasin  Mesa  (20  and  also  KU  1020.59-102061);  9  mi.  S 
North   Rim,   Moccasin   Mesa    (2);    Morfield   Canyon,   7600   ft.    (KU   75978); 


150 


Misc.  PuBL.  51,  Unw.  Kansas  Mus.  Nat.  Hist. 


Fig.  2.    View  of  Navajo  Canyon  looking  northward  toward  the  confluence  with 
Spruce  Canyon.    At  this  point,  the  canyon  bottom  is  600  feet  lower  than  the 

tops  of  the  mesas. 


Park  Point,  8400-8500  ft.  (16  and  also  KU  102047-102048,  102050-102057); 
?i  mi.  S,  1%  mi.  W  Park  Point,  8000  ft.  (KU  69286-69288);  1!^  mi.  S,  1%  mi.  W 
Park  Point,  8000  ft.  (KU  69289);  VA  mi.  S,  2  mi.  W  Park  Point,  8075  ft.  (KU 
69290);  sec.  27,  head  East  Fork,  Navajo  Canyon,  7900  ft.  (KU  69291-69292); 
sec.  27,  head  East  Fork,  Navajo  Canyon,  7875  ft.  ( 1 );  head  West  Fork,  Navajo 
Canyon  (1);  J2  mi.  N  Far  View  Ruins,  7825  ft.  (KU  69293);  Far  View  Ruins, 
7700  ft.  (KU  69294);  100  yds.  W  Far  View  Ruins  (1);  200  yds.  W  Far  \'iew 
Ruins  (6);  300  yds.  W  Far  View  Ruins  (3);  Far  View  Ruins,  )i  mi.  W  Highway 
(1);  Drainage  'A  mi.  S  Far  View  Ruins  (1);  200  yds.  S  Far  View  Ruins  (1); 


Douglas — Pocket  Gophers  of  Mesa  Verde  151 

■4  mi.  S  Far  View  Ruins  Road,  30  ft.,  E  Highway  (2);  1  mi.  SW  Far  View 
Ruins  (2);  1.2  mi.  S  Far  View  Ruins  Road  (1  and  KU  102062);  100  yds.  N 
Wetherill  Lab,  7050  ft.,  Cliapin  Mesa  (2);  Residence  Area,  7000  ft.,  Chapin 
Mesa  (1);  N  Rim,  Wetherill  Mesa,  8239  ft.  (2);  3  mi.  N  Rock  Springs,  8200 
ft.  (KU  69295-69298);  2.2  mi.  N  Rock  Springs  (1);  2M  mi.  N,  M  mi.  W  Rock 
Springs,  8100  ft.  (KU  69299-69301);  2  mi.  N,  Y,  mi.  W  Rock  Springs,  7900  ft. 
(KU  69302-69303);  1  mi.  NNW  Rock  Springs,  7600  ft.  (KU  69304);  Y^  mi. 
NNW  Rock  Springs,  7500  ft.  (KU  69305);  Rock  Springs,  7400  ft.  (1  and 
KU  69306). 

Other  locality  records.— Frater  Canyon,  7600  ft.  ( MVZ  74408-74410);  Far 
View  Ruins,  7700  ft.  (MV  7852/507,  7853/507);  Mesa  Verde,  northern  end, 
8100  ft.  (USNM  149087). 

The  distribution  of  gophers  is  influenced  by  a  complex  inter- 
relationship of  edaphic  and  climatic  factors.  Since  gophers  spend 
most  of  their  lives  in  subterranean  tubes,  it  is  important  that  the 
soil  in  which  they  live  be  of  sufficient  depth  to  permit  development 
of  feeding  tunnels  as  well  as  deeper  living  chambers  where  micro- 
environmental  variables  are  more  constant.  Friability  and  constitu- 
ency of  soils  influence  the  ability  of  gophers  to  dig  new  tunnels  and 
the  amount  of  friability  undoubtedly  affects  gaseous  exchanges  be- 
tween air  in  the  tunnels  and  air  in  the  interstices  of  the  soil  (see 
Kennerly,  1964). 

Soils  on  mesa  tops  are  loessal  in  origin;  the  earliest  preserved 
loess  in  Mesa  Verde  probably  is  Sangamon  in  age  (Arrhenius  and 
Bonatti,  1965 ) .  Soils  in  Mesa  Verde  generally  are  shallow,  although 
deeper  soils  occur  in  some  saddles  between  ridges  along  the  northern 
rim  of  the  land  mass,  and  in  the  bottoms  of  canyons  where  deep, 
sandy  alluvial  terraces  occur. 

It  was  observed  early  in  the  study  that  gophers  occurred  in  a 
variety  of  vegetational  associations  within  Mesa  Verde.  Soils  in 
these  associations  ranged  from  deep,  sandy  alluvium  found  in  drain- 
ages, to  the  shallow  soils  of  the  pinyon-juniper  woodland.  In  some 
places,  gophers  dug  tunnels  through  hard  layers  of  subsurface 
caliche,  and  numerous  pieces  of  it  were  present  in  their  mounds. 
In  a  former  gravel  storage  area,  gophers  dug  to  the  surface  through 
about  one  foot  of  hard-packed  gravel  that  was  almost  impossible  to 
excavate  with  a  shovel.  On  Park  Point,  individuals  dug  through  a 
residual  layer  of  loose  rock  to  deposit  rock-filled  soil  above  ground. 
In  other  areas  on  Park  Point  they  dug  small  contorted  tunnels 
around  and  between  subsurface  boulders  in  areas  that  offered  little 
in  the  way  of  soil  and  vegetation. 

In  order  to  learn  more  about  the  composition  of  soils  in  gopher 
habitats,  samples  of  soil  were  collected  from  the  top  two  inches  of 


152 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


*^*.*  ^'  ..Ml         f'.^'  ■*'       ''•^♦■« 


V 


Fig.  3.   Prater  Canyon  near  upper  well,  looking  southward.    Pocket  gophers  live 
in  the  sedge-  and  grass-covered  soils  of  the  canyon. 

the  surface  in  18  of  the  major  trapping  localities  ( Table  1 ) .  These 
samples  were  analyzed  for  particle  size  by  the  hydrometer  method, 
which  permits  rapid  determinations  of  percentages  of  clay,  silt,  and 
sand  (Bouyoucos,  1967). 

The  results  of  the  hydrometer  analyses  are  given  in  Table  1. 
Each  of  the  values  listed  in  this  table  represents  an  average  of 
duplicate  samples.  The  soils  from  various  gopher  habitats  are 
moderately  fine  to  medium  in  texture,  and  arc  classified  as  silty  clay 
loam  ( 27  to  40  per  cent  clay ) ,  silty  loam  ( 40  per  cent  or  more  silt ) , 
or  as  sandy  clay  loam  ( up  to  39  per  cent  silt ) .  Of  much  more  im- 
portance than  the  classification  of  the  soils  is  the  fact  that  mod- 
erately fine  to  medium  soils  have  excellent  capacities  for  retaining 
moisture,  being  surpassed  only  by  fine  soils  containing  more  than 
40  per  cent  clay.  Available  moisture  is  defined  as  that  percentage 
between  field  capacity  and  wilting  point,  and  is  expressed  in  inches 
of  moisture  per  foot  of  soil.  The  range  and  average  values  of  avail- 
able moisture  for  soils  in  the  state  of  Colorado  are  as  follows: 
moderately  fine,  1.6-2.5,  axerage  2.2;  medium  (40  per  cent  or  more 
silt),  1.6-2.5,  average  2.3;  medium  (0  to  39  per  cent  silt),  1.5-2.4, 
average  1.9  inches  per  foot  (Orville  Parsons,  personal  communica- 
tion ) . 


Douglas — Pocket  Gophers  of  Mesa  Verde  153 

Table  1. — Percentages  of  clay,  silt,  and  sand  occurring  in  soil  samples 

FROM  VARIOUS  LOCALITIES  IN   MeSA   VeRDE   NATIONAL   PaRK,   AS   DETERMINED  BY 

THE  HYDROMETER   METHOD. 

Locality  Clay  Silt  Sand 

Park  Point 23.60  43.10  33.30 

Park  Point,  8500  ft 16.10  29.10  54.80 

Park  Point,  8400  ft.  1 11.60  45.55  42.85 

200  >ds.  W  Far  View  Rnins  16.40  36.85  46.75 

300  yds.  W  Far  View  Ruins  24.15  30.15  45.70 

300  yds.  W  Far  View  Ruins  31.70  32.90  35.40 

1  mi.  SW  Far  View  Ruins 26.10  35.10  38.80 

200  vds.  W  Far  View  Ruins  19.50  39.25  41.25 

North  Rim  Wetherill  Mesa  12.60  48.10  39.30 

2.2  mi.  N  Rock  Springs  15.10  50.60  34.30 

Prater  Canyon,  7600  ft.  17.30  23.90  58.80 

Prater  Canyon,  200  yds.  S  Highway  ._ 30.00  37.00  33.00 

Moccasin  Mesa,  northern  end  11.30  52.70  36.00 

Head  Meadow,  8  mi.  S  North  Rim,  Moccasin  Mesa  20.80  45.75  33.45 

Meadow,  Moccasin  Mesa,  8  mi.  S  North  Rim  25.65  43.05  31.30 

Meadow,  Moccasin  Mesa,  8  mi.  S  North  Rim  27.00  45.20  27.80 

Meadow,  Moccasin  Mesa,  8  mi.  S  North  Rim  24.00  51.20  24.80 

100  yds.,  N  Wetherill  Lab,  Capin  Mesa  18.50  33.80  47.70 

It  was  somewhat  unexpected  to  find  so  much  variation  in  samples 
from  the  same  general  area.  Nevertheless,  soils  in  Mesa  Verde  are 
known  to  occur  in  mozaic  patterns,  and  variation  in  soil  composition 
within  a  valley,  or  between  various  elevations  on  a  ridge,  such  as 
Park  Point,  probably  is  to  be  expected. 

Probably  the  most  important  factor  determining  distribution  of 
gophers  is  vegetational  ground  cover.  Distribution  of  various  plants 
of  importance  to  gophers  is  regulated  by  the  distribution  of  soils 
suitable  for  their  growth  and  by  amounts  of  nutrients  and  moisture 
available  in  such  soils.  Factors  such  as  the  tolerance  of  a  plant  for 
shade  or  sunlight,  the  serai  stage  in  which  the  plant  usually  occurs, 
kinds  of  dispersal  agents,  chemical  factors  in  the  soil,  and  more,  play 
regulatory  roles  in  the  ability  of  plants  to  invade  certain  areas. 
When  species  of  plants  are  heavily  cropped  by  gophers,  the  ability 
to  grow  new  roots  may  be  a  limiting  factor  in  the  survival  of  such 
.species  in  areas  used  by  gophers. 

Some  plants,  such  as  Poa  fendleriona,  are  common  to  various 
^'egetational  associations  within  Mesa  Verde,  whereas  others  are 
restricted  to  a  particular  association.  In  order  to  analyze  the  rela- 
tionship of  plants  to  the  distribution  of  gophers,  the  vegetation  was 
analyzed  at  each  major  trapping  locality.  Nineteen  of  the  trapping 
sites  were  chosen  for  intensi\'e  analysis  of  the  vegetation.  The  sites 
were  selected  to  include  various  associations,  and  all  localities  at 
which  particle  sizes  of  the  soils  had  been  analyzed.   The  dominant 


154 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


and  codominant  species  of  plants  were  determined  at  each  site,  then 
other  species  were  Hsted  in  order  of  their  relative  abundance.  More 
than  70  species  of  plants  were  recorded  from  the  19  sites.  The  most 
abundant  plants  in  gopher  habitats  consisted  of  about  30  species; 
these  are  listed  below  in  order  of  their  relative  number  of  occur- 


rences : 


Foa  fendleriana 
Eriogonum  racemosum 
Pensfeiuon  linarioicles 
Amelanchier  utalicnsis 
Solidago  petradoria 
Pinus  edulis 
Junipenis  osieosperma 
Aster  higelovii 
Chnjsothammis  depressus 
Lupiniis  argenteus 
Astiagahis  scoptdortim 
Artemisia  tridentata 
Artemisia  nova 
Lithosperm iim  niderale 
Purshia  tridentata 


Erigeron  speciosus 
Artemisia  ludoviciana 
Eriogonum  nmhellatum 
SpJiaeraJcea  coccinea 
Tragopogon  pratensis 
Achillea  millefoliurn 
Clienopodittm  sp. 
ChrysotJwmmis  nauseosus 
Comandra  umbellata 
Koeleria  cristata 
Stipa  comafa 
Opuntia  sp. 
Castilleja  chrom,osa 
Erigeron  flagellaris 
Fcndlera  rupicola 


The  ground  cover  at  the  head  of  Prater  Canyon  was  composed 
mostly  of  Carex  praegraciUs,  Equlsetiim  sp.,  and  Agropyron  deser- 
torum.  Although  these  species  were  of  major  importance  to  a  large 
colony  of  gophers  at  this  site,  they  were  found  only  at  this  locality 
and  therefore  did  not  rank  among  the  30  most  abundant  species 
listed  above. 

The  first  10  species  of  plants  listed  above  are  widely  distributed 
in  Mesa  Verde.  Of  these,  six  species  (or  in  two  cases  closely  related 
species )  were  eaten  by  captive  gophers  ( Table  2 ) .  It  seems  reason- 
able to  assume  that  when  a  species  is  found  in  more  than  half  of 
the  19  localities  it  is  utilized  by  gophers.  Plants  such  as  Chnjso- 
thamnus  sp.,  having  numerous,  succulent  roots,  probably  are  utilized 
more  than  plants  such  as  Aster  higelovii,  which  has  only  one  major 
root.  Plants  with  widespreading  root  systems  can  be  cropped  re- 
peatedly whereas  plants  with  fewer  roots  may  be  killed  after  the 
first  severe  cropping. 

Gophers  influenced  changes  in  vegetation  from  year  to  year  by 
their  selection  of  certain  species.  This  was  especially  evident  in  the 
meadow  of  Moccasin  Mesa.  In  1965,  tunnels  of  gophers  were  found 
under  almost  every  clump  of  cactus  {Opuntia  sp.)  occurring  in  the 
meadow.  In  some  cases  the  entire  underground  parts  of  the  plants 
were  eaten  and  the  part  above  ground  was  pulled  partly  into  a 


Douglas — Pocket  Gophers  of  Mesa  Verde                155 

Table   2. — Food   items   consumed   by   captives  of    Tliomomijs    hottae    {-\-, 

EATEX;    O,    NOT    EATEN;    ,    NOT  OFFERED). 

Plant                                       Roots  Stem  Leaves  Other 

Artemisia  frigida  —  +  +  — 

Artemisia  hidoviciana  +  +  0  — 

Aster  higelovii  +  >    +  +  flowers 

Astragalus  seopuJoriim  +  +  o  — 

Castilleja  cliromosa  +  +  +  — 

CJirysopsis  villosa  +  +  +  (basal)        — 

Chnisothamnus  nauseosus  -     +  +  —  — 

Eqtiisetutn  sp.  +  —  —  — 

Fendlera  riipieola  —  +  —  — 

Liipinus  caiidatus  +  +  —  — 

Pcdicidaris   centranthera   _—     +  +  +  — 

Penstcmon  linarioides  +  +  +  — 

Querciis  gamhelii  +  —  —  acorns 

Solidago  petradoria  +  +  —  — 

SpJiaeralcea  coccinea  +  o  o  — 


burrow.  Some  clumps  of  cacti  recovered  by  1967,  but  others  died 
out  completely  and  the  area  was  invaded  by  primary  successional 
species  such  as  Sphaeralcea  coccinea. 

Climatic  factors  affecting  plants  may  also  cause  gophers  to  shift 
their  area  of  activity,  thereby  allowing  recovery  of  the  vegetation  in 
the  former  area  of  concentration.  In  1965,  moisture  from  heavy 
winter  snows  and  spring  rains  promoted  a  luxuriant  growth  of 
grasses  and  herbs  on  Moccasin  Mesa  (see  Fig.  4).  Tn  the  autumn 
of  that  year,  vegetation  in  the  meadow  was  dense  and  gophers  were 
concentrated  around  the  periphery  of  the  meadow  at  the  edges  of 
the  pinyon-juniper  woodland.  Only  a  few  gophers  were  found  in 
the  meadow.  In  this  case,  many  more  roots  were  available  in  the 
meadow  than  in  the  relatively  sparse  ground  cover  of  the  wood- 
land. The  next  year  vegetation  in  the  meadow  was  more  normal 
in  density,  and  gophers  were  again  concentrated  there.  I  suppose 
that  gophers  preferred  the  deeper  soils  and  more  abundant  plants 
in  the  meadow,  but  were  concentrated  in  the  more  sparsely  vege- 
tated areas  when  plants  in  the  meadow  became  too  dense  for 
unhampered  movement  or  vision  above  ground,  because  gophers 
appear  to  avoid  tall,  dense  vegetation. 

Gophers  move  tons  of  soil  each  year  (Kennerly,  1964;  Down- 
hower  and  Hall,  1966).  Mounds  often  cover  the  surface  sufficiently 
to  prevent  further  growth  of  the  underlying  vegetation.  Succession 
of  plants  on  mounds  then  progresses  as  it  would  on  any  denuded 
ground,  but  because  of  the  relatively  small  area  covered  by  each 
mound,  succession  is  fairly  rapid.  Mound  obliteration  in  Mesa 
Verde  customarily  takes  two  or  three  years,  whereas  larger  denuded 


156 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


«««• 


% 


'yt.. 


«*?.-■ 


.•«i..:*R 


it- 


Fig.  4.    Meadow    on  Moccasin  xMesa,  S  mi.  S  i\oith  Rim.    Heavy  precipitation 
in   winter   and   spring   of   1967   promoted   the   unusually   luxuriant   growth    of 

grasses  shown  here. 


areas  require  longer  to  recover.  An  example  of  the  slow  rate  of 
succession  in  Mesa  Verde  is  seen  on  the  northern  part  of  Wetherill 
Mesa,  which  burned  in  1934.  In  1967  this  area  supported  shrubs, 
grasses,  and  herbaceous  species,  but  pinyon  and  juniper  seedlings 
had  not  become  re-established  despite  the  presence  of  unburned 
woodland  adjoining  the  area.  Likewise,  in  1961,  seedlings  of  pinyon 
and  juniper  were  just  becoming  established  in  a  large  area  on 
Chapin  Mesa  that  was  burned  in  1858. 

In  June  of  1964,  stakes  were  placed  beside  each  of  20  new 
gopher  mounds  in  the  meadow  on  Moccasin  Mesa  (see  Fig.  5). 
These  mounds  were  inspected  in  the  autumn  of  1964  and  again  in 
1965  and  1967,  in  order  to  learn  which  species  of  plants  were  early 
invaders  of  such  disturbed  soil.  No  new  vegetation  appeared  on 
the  mounds  in  1964.  In  August,  1965,  Sphaeralcea  coccinea  and 
Tragopogon  pratensis  were  present.  In  May,  1967,  SpJtaeralcea  coc- 
cinea, Tragopogon  pratensis,  Artemisia  dracunculus,  and  Foa  pra- 
tensis grew  on  the  mounds. 


Douglas — Pocket  Gophers  of  Mesa  Verde  157 


». 


P'.' 


A 


...'i\g    •'^'•'       '  '  ■      ',  '.'  ..^*^^" 


'«<,t«»» 


5  "rC  ^ 


•''s';^-1k;. 


Fig.  5.    Photograph  of  gopher  habitat  in  meadow  on  Moccasin  Mesa.    Mounds 
made,  and  staked,  in  1965  were  nearly  obliterated  in  two  years  time   (photo- 
graph taken  in  September,   1967). 


Individuals  of  Thomomys  browse  above  ground  around  the 
openings  of  burrows.  Exidence  of  feeding  usually  is  found  on  plants 
within  a  six-  to  12-inch  radius  around  the  opening  of  the  burrow. 
Such  grazing  was  especially  evident  in  areas  where  rabbit  brush, 
Chrijsoihamnus  nauseosus,  occurred.  Gophers  appear  to  relish  the 
above-ground  parts  of  this  plant  as  well  as  its  roots,  as  was  verified 
by  the  feeding  of  captives  ( Table  2 ) .  Gophers  also  harvested  sweet 
vetch,  Astragalus  scopiilorum,  and  Indian  rice  grass,  Oryzopsis 
hymenoicles,  and  aerial  parts  of  these  plants  were  found  in  tunnels. 
Such  aerial  parts  were  cut  into  lengths  of  six  or  seven  inches  before 
being  taken  below  ground. 

Side  tunnels  of  burrows  often  were  packed  with  small  pieces 
(one  to  two  inches  long)  of  roots  and  grass  stems.  These  accumu- 
lations probably  represent  food  caches  that  are  stored  for  later  use. 
All  such  caches  filled  with  vegetation  and  soil  were  found  within  one 
or  two  feet  of  a  burrow  opening.  Several  cones  and  numerous  nuts 
of  pinyon  pine,  Vinus  eduUs,  were  found  in  the  cache  of  one  gopher. 


158  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

The  cones  had  been  opened  and  most  of  the  nuts  had  been  removed 
and  eaten;  the  chewed  coats  of  seeds  remained  in  the  tunnel.  This 
was  especially  interesting  because  the  nearest  pinyon  pine  was 
about  30  feet  from  the  opening  of  the  burrow;  the  opening  was  also 
at  that  part  of  the  tunnel  nearest  the  woodland. 

Captive  gophers  soon  die  if  they  are  not  given  succulent  plants 
or  roots  to  eat.  On  the  other  hand,  captives  gain  weight  and  thrive 
on  a  diet  of  potatoes  and  rolled  oats,  supplemented  with  fresh  green 
vegetation.  Gophers  probably  do  not  produce  much  if  any,  meta- 
bolic water.  Roots  found  in  cheek  pouches  and  in  tunnels  indicate 
that  they  prefer  non-woody,  succulent  roots  that  can  be  eaten  easily. 
Few  woody  roots  were  recovered  from  cheek  pouches  or  from  tun- 
nels. Captives  chewed  the  softer  parts  of  roots  and  stems,  but  rarely 
bothered  with  woody  roots  when  other  foods  were  available.  All 
stomach  contents  of  Thomomijs  that  I  have  seen  always  were  green 
in  color,  indicating  extensive  foraging  on  aerial  parts  of  plants. 

The  fossorial  mode  of  life  permits  Thomomys  to  avoid  the  prob- 
lems of  thermo-regulation  that  would  be  attendant  with  living  on 
the  surface.  Fossorial  forms  are  known  to  ha\'e  poor  thermo-regula- 
tion and  a  generally  low  tolerance  to  high  temperatures.  The 
maintenance  of  water  balance  also  can  be  a  pressing  problem  facing 
species  Hving  in  arid  environments.  Relative  humidity  is  more  con- 
stant in  burrows  than  it  is  at  the  surface  (Kennerly,  1964);  thus 
living  in  subterranean  tunnels  permits  gophers  to  avoid  desiccation 
and  the  need  for  obtaining  larger  amounts  of  moisture  to  facilitate 
osmoregulation. 

The  storage  of  roots  and  stems  underground  provides  a  source  of 
moist  vegetation  for  gophers.  This  would  be  advantageous  to  them 
in  the  hottest  months  of  summer  and  during  periods  of  drought 
when  aerial  parts  of  plants,  and  perhaps  many  root  systems  are 
reduced  in  moisture  content.  Thomomys  hottae  apparently  is  well 
adapted  to  life  in  arid  environments  because  of  its  behavior,  rather 
than  because  of  its  physiological  adaptations. 

Thomomys  hottae  is  coprophagic,  and  captives  often  were  ob- 
served to  eat  feces.  This  was  accomplished  in  such  a  way  that  it 
has  not  been  possible  to  determine  whether  this  species  produces 
two  kinds  of  droppings,  as  do  rabbits.  At  any  rate,  captive  in- 
dividuals produce  copious  droppings  and,  as  would  be  expected, 
the  caecum  is  large  and  probably  contains  bacteria  that  break  down 
cellulose. 


Douglas — Pocket  Gophers  of  Mesa  Verde 


159 


Reproduction,  Growth,  and  Development 
In  Mesa  Verde,  reproduction  in  Thomomijs  occurs  only  in 
spring.  Pregnant  females  were  taken  in  late  April  and  early  May; 
females  with  enlarged  uteri  were  taken  in  April,  August,  and  Sep- 
tember, indicating  that  reproducti\-e  activity  probably  begins  in 
March,  or  perhaps  late  in  February.  The  gestation  period  for  T. 
hottae  is  about  18  to  19  days  (Schramm,  1961).  Lactating  females 
were  taken  in  early  May,  but  none  of  31  females  taken  in  August  or 
September  was  lactating  (Fig.  6).  Four  females  had  from  two  to 
four  embryos  each  (average  three).    Two  other  females  each  had 


^     10 

< 

Q 
> 

Q 


o 


Pregnant    Q  Lactating      Q  Enlarged 

Uterus 


JJ 


APRIL 


MAY 


]  Non- 
Reproductive 


P — T 

r1-    ,    1 


AUG. 


SEPT. 


W    20- 

u 

Pt 

^ 

kJ 

n 

1- 

iO 

' 

LlI 

h- 

n 

l—, 

u. 

1-, 

o    /o- 

^ 

X 

1- 

^ , 

1 

2 

^ 

LU 

_J         - 

0 

— 

APR.      MAY 


AUG. 


NUMBER 


OF 


SEPT. 
INDIVIDUALS 


Fig.  6.  Condition  of  reproductive  organs  of  dissected  gophers  collected  in 
Mesa  Verde  from  1965  through  1967.  Reproductive  condition  of  females 
(upper)  in  \arious  months;  size  of  testes  of  males  (lower)  collected  in  various 

months. 


160  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

three  placental  scars.  Howard  and  Childs  (1959:295)  reported  an 
average  litter  size  of  4.6  for  T.  h.  meica  from  California.  Young  of 
the  year  appeared  in  traps  in  April,  May,  August,  and  September. 
Unfortunately,  no  gophers  have  been  trapped  in  Mesa  Verde  earlier 
than  April  and  none  has  been  taken  in  June  or  July;  the  data  on 
reproduction  are  therefore  incomplete. 

Eight  males  taken  in  late  April  and  early  May  of  1967  were  in 
breeding  condition.  Howard  and  Childs  (1959:300)  separated  kill- 
trapped  individuals  of  T.  botfae  into  three  age  classes  by  the  size  of 
the  testes.  Testes  of  animals  less  than  five  months  of  age  were 
between  one  and  10  millimeters  in  length  (average  4.5  mm.).  Males 
nine  months  or  older  had  testes  12  millimeters  or  longer.  Figure  6 
shows  the  length  of  testes  in  males  taken  in  Mesa  Verde  in  various 
months  of  the  year.  Testes  of  adults  taken  in  spring  were  engorged, 
highly  vascularized,  and  scrotal,  whereas  those  of  adults  taken  in 
August  or  September  were  flaccid,  wrinkled,  and  less  vasculariza- 
tion was  present.  Testes  of  young  males  were  not  apparent  external- 
ly and  it  is  difficult  to  sex  such  young  animals  without  dissecting 
them. 

Figure  7  shows  the  distribution  of  sizes  of  males  and  females 
taken  from  Mesa  Verde,  and  the  months  in  which  they  were  taken. 
Although  the  samples  are  somewhat  limited,  it  is  apparent  that 
young  of  the  year  grow  rapidly  and  attain  adult  size  by  autumn, 
when  separation  of  young  and  adult  individuals  can  best  be  accom- 
plished by  examining  reproductive  tracts.  Apparently,  few  females 
breed  in  the  first  year,  whereas  all  individuals  a  year  old  or  older 
evidence  reproductive  activity.  This  agrees  with  reports  in  the 
literature  (Howard  and  Childs,  1959:302). 

Table  3  summarizes  measurements  taken  on  122  specimens 
examined  in  this  study.  Measurements  for  young  of  the  year  ( young 
and  subadult  individuals )  are  given  separately  from  those  of  adults. 

Juveniles  are  those  animals  that  are  still  in  the  nest,  or  that  have 
Juvenal  pelage.  These  individuals  can  be  distinguished  easily  from 
adults,  but  seldom  appear  in  traps.  Young  individuals  are  those  that 
are  moving  about  on  their  own,  and  those  in  which  the  juvenal 
pelage  is  being  replaced.  Subadults  are  those  young  of  the  year 
that  have  reached  adult  proportions  but  have  not  yet  bred.  Adults 
are  those  indixiduals  sho\\'ing  signs  of  reproductive  activity.  In  some 
cases,  young  of  the  year  may  breed  and  thus  would  then  be  classified 
as  adults,  but  breeding  in  the  first  season  probably  is  an  unusual 
occurrence  in  Mesa  Verde. 


Douglas — Pocket  Gophers  of  Mesa  Verde 


161 


5 

APRIL 

/ 
/ 

0 

Z 

■  r- 

1 1              ' 

CO      5 

_J 
< 

Q 


> 


0 


MAY 


V 


I 


T 1 r 


/  -■  / 


-a 


T 1  I r- 


Ll_ 

O 


on 

UJ 
QQ 


0 


AUGUST 


10 


5- 


SEPTEMBER 
MALES 

7^   FEMALES 

/ 
/ 

^/ 

*            1          r 1 1      '    ' 

176 


I7S- 
185 


186- 
195 


196- 
205 


206- 
215 


216- 
225 


226- 
235 


LENGTH      IN     MILLIMETERS 


236- 
245 


246- 
255 


Fig.  7.    Size  distril:)ution  of  63  male  and  57   female  pocket  gophers  taken  in 
Mesa  Verde  National  Park  and  vicinity. 


In  young  females  the  pubic  symphysis  is  not  fully  absorbed,  the 
nipples  are  not  enlarged,  and  the  uterus  usually  is  not  dilated  nor 
does  it  show  other  evidence  of  breeding  activity.  Individuals  meet- 
ing these  criteria  were  considered  to  be  young  of  the  year,  irrespec- 
tive of  their  external  size  or  dimensions  of  the  skull.  Three  young 
females  taken  in  late  April  measured  177  to  197  millimeters  in  total 
length  (average  187  mm.),  indicating  the  rapid  growth  achieved  by 


162 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Table  3. — Measurements,  in  millimeters,  of  Thomomijs  bottae.  Adult  in- 
dividuals WERE  THOSE  EXHIBITING  EVIDENCE  OF  REPRODUCTIVE  ACTIVITY; 
YOUNG  INDIVIDUALS  WERE  THOSE  JUVENILES  AND  YOUNG  OF  THE  YEAR  SHOWING 
NO  REPRODUCTIVE  ACTIVITY.  MeAN,  MAXIMUM  AND  MINIMUM  ARE  GIVEN  IN 
THAT    ORDER    FOR    EACH    MEASUREMENT. 


30  adult 

28  young 

37  adult 

27  young 

Measurement 

females 

females 

males 

males 

Total 

length  

..     211 

201 

220 

203 

240 

220 

250 

225 

185 

175 

187 

174 

Length  of 

tail 

.-       57 

57 

62 

60 

66 

71 

79 

83 

42 

48 

48 

50 

Length  of 

hind  foot  

29 

28 

29 

29 

32 

30 

35 

33 

26 

24 

27 

25 

Condylobasal 

length  

..       37.4 

35.8 

39.3 

35.5 

40.6 

38.5 

44.9 

39.9 

34.1 

32.0 

34.4 

30.8 

Palatal 

length  

-       22.0 

21.0 

23.4 

20.9 

23.9 

23.6 

26.7 

23.5 

20.3 

17.9 

19.6 

17.5 

Length  of 

nasal  

._       11.9 

11.2 

12.9 

11.3 

13.4 

12.5 

15.3 

13.8 

10.7 

9.4 

10.2 

8.7 

Zygomatic 

breadth     

..       23.6 

22.6 

24.9 

22.0 

25.6 

24.6 

28.4 

25.5 

21.5 

19.3 

22.6 

19.6 

Mastoidal 

breadth    

.       19.4 

19.1 

20.3 

18.6 

20.9 

20.1 

23.0 

21.0 

18.0 

15.9 

16.8 

16.3 

Least 

interorbital   lireadth   - 

6.3 

6.3 

6.1 

6.3 

6.8 

6.8 

6.9 

6.8 

5.6 

5.7 

5.3 

5.8 

Alveolar  length  of 

maxillary   toothrow  ..- 

7.4 

7.3 

7.7 

7.2 

8.7 

8.3 

8.9 

8.1 

6.6 

6.2 

6.5 

6.5 

Length  of 

mandible    

..       25.3 

24.4 

26.9 

24.6 

28.6 

27.3 

31.8 

27.5 

22.9 

20.1 

22.4 

21.2 

Capacity  of 

braincase   (cc. )  

1.3 

1.2 

1.3 

1.3 

1.5 

1.4 

1.6 

1.4 

1.1 

1.0 

1.1 

1.1 

young  individuals  in  the  first  several  months  of  life.  Twenty-three 
young  females  taken  in  August  and  September  measured  from  178 
to  219  millimeters  in  total  length  (average  202),  whereas  27  adult 


Douglas — Pocket  Gophers  of  Mesa  Verde  163 

females  had  total  lengths  of  185  to  238  millimeters  (average  212). 
It  is  apparent  that  most  young  individuals  attain  adult  size  quite 
rapidly.  Individuals  born  early  in  the  season  may  reach  adult  pro- 
portions by  late  spring  and  certainly  do  so  by  autumn. 

In  young  males  the  testes  are  mucJi  smaller  than  those  of  males 
more  than  one  year  old  (Howard  and  Childs,  1959:300).  Thirty 
young  of  the  year  had  testes  between  three  and  10  millimeters  in 
length  (average  6.0),  and  total  lengths  from  187  to  226  millimeters 
(average  209).  Nineteen  adults  had  testes  that  measured  11  to  20 
millimeters  in  length  (average  15)  and  total  lengths  ranging  from 
187  to  250  (average  223). 

Youngman  (1958:366)  studied  museum  specimens  of  T.  hottae 
from  Colorado,  and  established  three  criteria  for  adulthood: 
".  .  .  (a)  suture  obliterated  between  supraoccipital  and  exoccipital, 
(b)  suture  at  least  partly  obliterated  between  basisphenoid  and 
basioccipital,  (c)  supraorbital  crests  not  widely  separated  and  al- 
most parallel.  ..."  I  examined  26  skulls  from  young  females,  col- 
lected in  the  course  of  the  present  study,  in  order  to  see  if  these 
three  criteria  are  applicable  when  reproductive  condition  is  known. 
The  skulls  were  from  animals  that  were  unquestionably  young  of 
the  year.  These  individuals  exhibited  no  internal  or  external  evi- 
dence of  reproductive  activity,  the  pubic  bar  had  not  been  resorbed, 
and  they  were  among  the  smaller  individuals  taken.  Females  were 
used  because  their  lack  of  reproductive  activity  can  be  determined 
with  much  more  certainty  than  can  that  of  males. 

Of  the  26  young  of  the  year,  16  met  Youngman's  first  criterion 
(see  above),  18  met  the  second  criterion,  and  12  met  the  third. 
Because  of  the  rapid  rate  of  growth  in  young  and  subadult  in- 
dividuals, developmental  characteristics  of  the  skull  may  be  mis- 
leading for  the  establishment  of  age  groupings.  I  do  not  consider 
the  above-listed  criteria  adequate  for  differentiating  young  of  the 
year  from  adult  individuals.  When  studying  museum  specimens, 
the  investigator  must  establish  arbitrary  criteria  for  separating  age 
groupings.  At  present  there  is  little  known  about  rates  of  growth  in 
geomyids,  as  in  other  non-game  species,  and  it  would  be  vii'tually 
impossible  for  an  investigator  to  assign  accurate  age  groupings  on 
the  basis  of  skull  dimensions  alone,  without  having  knowledge  of 
the  condition  of  reproductive  organs,  and  in  the  case  of  females, 
the  condition  of  pubic  bones.  Many  investigators  do  not  dissect 
female  gophers,  or  if  they  do,  the  tendency  seems  to  be  to  make  no 
notation  on  the  specimen  label  unless  embryos  are  found.    There  is 


164 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


more  of  a  tendency  to  measure  and  record  the  size  of  testes.  It  is  to 
be  hoped  that  in  the  future  investigators  will  record  lack  of  embryos 
as  well  as  their  presence;  information  as  to  presence  or  absence  of 
the  pubic  bar  in  females  also  should  be  recorded. 

Climatic  Conditions  Affecting  Gophers  at  Park  Point 
Climatic  conditions  were  studied  at  Park  Point,  (S575  feet  eleva- 
tion (see  Fig.  8),  the  highest  point  in  Mesa  Verde,  as  part  of  an 
environment  measurement  program  of  the  Wetherill  Mesa  Project. 


Fig.  8.  View  southward  from  Park  Point,  the  highest  elevation  in  Mesa  Verde 
(8575  feet).    Gophers  lived  on  this  ridge  from  1961  to  1967  despite  the  rocky 

nature  of  the  soil. 

As  mentioned  earlier,  one  of  the  largest  colonies  of  pocket  gophers 
in  Mesa  Verde  occurs  at  Park  Point.  Thus  the  choice  of  this  site  for 
a  weather  station  was  indeed  fortunate.  The  occurrence  of  gophers 
on  this  rocky  ridge  was  first  noticed  when  the  weather  station  was 
installed.  Gophers  later  were  observed  to  pile  mounds  of  soil  in 
areas  where,  because  of  rocks,  I  had  difficulty  in  obtaining  soil 
samples  at  the  six-  and  12-inch  depths.  I  was  surprised  to  find  a 
sizeable  colony  of  gophers  on  this  ridge  where  the  soil  was  shallow 
and  rocky  and  where  vegetation,   although   not  in   short   supply, 


Douglas — Pocket  Gophers  of  Mesa  Verde  165 

seemed  to  require  such  an  expenditure  of  effort  for  gophers  to 
obtain. 

A  standard  U.S.  Weather  Bureau  shelter  was  installed  at  this  site 
in  September  of  1961  and  instruments  were  run  through  December, 
1963.  Air  temperatures  and  relative  humidities  were  recorded  by 
means  of  a  Bendix  hygrothermograph;  soil  temperatures  were  re- 
corded at  depths  of  two,  six,  and  12  inches  by  a  Kahl  three-pen 
thermograph;  and  precipitation  was  measured  by  means  of  a 
standard  Weather  Bureau  rain  gauge.  Soil  moisture  was  determined 
gravimetric-ally  at  depths  of  two,  six,  and  12  inches  about  twice 
each  month. 

Table  4  summarizes  the  climatic  conditions  at  Park  Point.  Pre- 
cipitation at  this  site  is  the  highest  of  that  at  any  sites  measured  in 
the  Mesa  Verde.  The  higher  parts  of  the  Mesa  Verde  land  mass  are 
first  to  receive  moisture  from  ephemeral  storms  that  originate  in  the 
nearby  La  Plata  Mountains.  Therefore,  the  higher,  northern  ends 
of  the  mesas  are  more  mesic  than  the  southern  ends,  and  winters 
are  the  more  severe  at  the  higher  elevations.  Park  Point  has  the  most 
severe  climate  of  the  Mesa  Verde  as  far  as  cold,  persistent  snows,  and 
frozen  soils  are  concerned.  The  growing  season  at  this  site  is  shorter 
than  that  at  lower  elevations. 

The  range  of  thermal  tolerance  of  Thomomys  hottae  has  not  yet 
been  studied,  but  it  is  apparent  that  this  species  escapes  the  high 
surface  temperatures  and  the  metabolic  stresses  that  would  result 
from  living  at  such  temperatures,  by  living  six  or  more  inches  below 
the  surface. 

Gophers  are  active  in  the  winter  at  Park  Point,  and  in  the  spring 
evidence  of  winter  digging  is  apparent  as  tubes  of  soil  lying  on  the 
surface.  In  winter,  soil  temperatures  at  all  levels  varied  only  a  few 
degrees  on  either  side  of  freezing.  Soil  temperatures  at  the  lower 
levels  were  found  to  be  only  slightly  above  freezing  in  March  when, 
it  is  thought,  most  of  the  young  are  born.  There  is  little  air  move- 
ment in  tunnels  of  TJiomomys;  therefore,  temperatures  in  tunnels 
are  the  same  as  those  in  soils  at  the  same  depths  (Howard  and 
Ghilds,  1959:332;  Wilks,  1963:272;  Kennerly,  1964:408). 

The  tubes  of  soil  evident  on  the  surface  in  spring  represent  tun- 
nels in  the  snow  that  were  filled  with  soil.  There  would  seem  to  be 
little  need  for  gophers  to  move  soil  in  order  to  feed  on  surface 
vegetation  in  the  winter,  because  they  could,  and  do,  tunnel  in  the 
snow  at  the  surface  ( see  Marshall,  1941 ) .   Some  of  the  roots  prob- 


166 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Table  4. — Soil  temperatures  at  depths  of  two,  six,  and  12  inches  and 

PRECIPITATION     RECORDED    AT    PaRK    PoINT,    8500     FT.,     MeSA     VeRDE     NATIONAL 

Park.    The  range  represents   extreme  temperatures   for   each   month; 

MEAN    temperatures     REPRESENT    A    MONTHLY    AVERAGE    OF    DAILY     MAXIMUM 
AND   MINIMUM   READINGS.     An    ASTERISK    INDICATES    INCOMPLETE    DATA. 


Two  inches 

Six  in( 

:hes 

12  inches 

Precipitation 

Date 

Range 

Mean 

Range 

Mean 

Range 

Me.in 

in  inches 

1962° 

April 

29-72 

51.0 

36-59 

48.5 

39-52 

46.5 

0.39 

1962 

May 

33-79 

55.7 

39-67 

53.0 

40-59 

50.9 

0.93 

1962 

June 

42-103 

71.6 

47-81 

66.4 

49-72 

62.7 

1.03 

1962 

July 

55-104 

78.7 

63-83 

72.9 

64-74 

69.7 

2.02 

1962 

August 

52-105 

77.9 

60-85 

73.2 

63-76 

70.2 

1.97 

1962 

September 

38-96 

66.3 

45-77 

63.4 

48-69 

61.8 

2.13 

1962 

October 

34-76 

51.7 

38-64 

50.2 

41-56 

48.8 

2.94 

1962 

November 

30-64 

40.3 

31-54 

39.9 

32-48 

.39.4 

1.50 

1962 

December 

24-41 

29.8 

26-35 

30.4 

26-33 

30.2 

0.90 

1963 

January 

23-31 

27.5 

2,5-32 

29.1 

2.5-34 

29.9 

1..30 

1963 

Feliruary 

26-34 

30.4 

28-35 

31.4 

28-35 

.32.1 

1.20 

1963 

March 

24-64 

33.2 

27-51 

32.6 

28-45 

33.0 

1.00 

1963 

April 

28-73 

47.1 

32-59 

45.2 

35-53 

44.1 

0..55 

1963 

May 

42-95 

67.5 

46-76 

63.1 

48-68 

60.6 

0.16 

1963 

June 

45-111 

76.8 

55-89 

71.2 

59-79 

69.1 

0.07 

1963 

July 

55-110 

81.8 

61-90 

76.7 

66-81 

75.3 

1.97 

1963 

August 

54-110 

74.5 

58-87 

70.7 

63-80 

70.3 

7.62 

1963 

September 

49-99 

72.9 

56-79 

67.8 

61-71 

66.6 

0.45 

1963 

October 

36-100 

64.5 

43-77 

.59.9 

49-70 

.59.8 

2.70 

1963 

November 

24-66 

39.2 

32-54 

38.8 

34-49 

40.5 

1.04 

1963° 

December 

24-44 

33.6 

29-36 

32.7 

32-36 

34.6 

0.75 

1964 

April 

27-70 

43.4 

32-55 

40.6 

35-48 

40.0 

1964 

May 

31-92 

60.0 

35-74 

56.0 

38-67 

54.0 

1964 

June 

40-107 

75.0 

49-82 

68.0 

52-75 

66.0 

1964 

July 

53-112 

82.0 

61-89 

75.0 

65-80 

74.0 

1964° 

August 

47-104 

74.0 

58-83 

70.0 

63-76 

69.0 

ably  are  dug  and  eaten  below  the  surface,  even  though  the  first  few 
inches  at  the  surface  are  frozen.  Mounds  could  not  be  piled  easily 
in  the  winter  due  to  the  pressure  of  overlying  snow,  and  tunnels 
may  be  dug  in  the  snow  to  accept  this  excess  dirt.  The  number  of 
surface  tubes  indicate  that  gophers  do  not  dig  soil  extensively 
throughout  the  winter.  Soil  for  surface  tubes  must  come  from  below 
the  surface,  for  I  have  found  no  evidence  of  surface  gouging  that 
would  be  expected  if  gophers  were  digging  shallow  trenches  to 
obtain  roots  at  the  surface,  and  then  were  filling  the  snow  tunnels 
behind  them.  I  think  that  food  caches  and  surface  feeding  are  the 
most  likely  sources  of  food  in  winter  months  when  the  ground  is 
frozen.  It  seems  likely  that  tubes  of  soil  may  represent  that  removed 
during  tunnel  cleaning,  especially  by  females  preparing  to  bear 
young. 

Gophers  in  Mesa  Verde  have  two  periods  of  increased  mound 
production.  One  period  is  apparent  late  in  the  autumn,  especially  in 


Douglas — Pocket  Gophers  of  Mesa  Verde  167 

October  and  November,  prior  to  the  onset  of  cold  weather.  The 
other  period  is  in  the  spring  after  the  ground  thaws.  In  Mesa  Verde 
the  ground  thaws  and  the  kinds  of  animals  that  hibernate  begin  to 
appear  above  ground  by  about  the  first  of  April.  The  only  significant 
reduction  in  mound  building  activity  occurs  in  June  and  July,  when 
surface  temperatures  are  high.  It  is  not  known  whether  Thomomijs 
aestivates  during  part  of  the  summer,  although  Howard  and  Childs 
( 1959)  have  suggested  this  possibility.  In  Mesa  Verde  there  is  some 
mound  building  throughout  all  months  of  spring,  summer,  and 
autumn. 

Mound  building  has  been  thought  to  be  associated  with  an  in- 
crease in  soil  moisture  (Miller,  1948).  Kennerly  (1964)  has  shown 
that  the  period  of  increased  mound  building  by  Geomys  begins 
before  and  subsides  later  than  the  period  of  sustained  high  soil 
moisture.  He  suggested  that  a  basic  biologic  rhythm  is  involved 
and  that  factors  other  than  soil  moisture  influence  mound  building. 

In  Mesa  Verde,  the  months  generally  having  the  most  precipita- 
tion are  July,  August,  and  February.  Although  the  increase  of 
mound  building  in  the  spring  appears  to  coincide  with  an  increase 
in  soil  moisture,  the  autumn  period  of  activity  comes  after  the  period 
of  increased  soil  moisture. 

Parasitism 

Trap-killed  gophers  were  placed  in  a  plastic  bag  immediately 
upon  removal  from  their  burrows.  Ectoparasites  were  killed  by 
placing  a  piece  of  cotton  soaked  in  carbon  tetrachloride  in  the  closed 
bag  with  the  gopher.  After  several  minutes  the  parasites  could  be 
brushed  from  the  fur  and  collected  on  a  piece  of  paper.  Endopara- 
sites  were  recovered  from  stomachs  and  intestines,  were  stored  in  70 
per  cent  ethanol,  and  were  later  sent  to  experts  for  identification. 

Gophers  in  Mesa  Verde  are  heavily  parasitized  by  fleas.  Several 
individuals  had  more  than  a  dozen  fleas  each;  the  most  heavily 
parasitized  individual,  an  old  male,  had  53  fleas  of  two  species.  This 
male  was  captured  alixe,  and  it  was  apparent  from  his  frequent 
scratching  that  the  fleas  caused  a  great  deal  of  discomfort. 

Only  one  tick,  an  adult  Dermacentor  andersoni,  was  found  on  a 
gopher  collected  in  this  study.  The  tick  was  found  on  the  cheek,  in 
the  middle  of  a  circle  about  one  half  inch  in  diameter  where  the  fur 
had  been  rubbed  off.  Such  aggravation  by  parasites  may  be  linked 
to  the  irregular  molting  patterns  seen  in  Tlwmomijs. 


168  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

It  was  surprising  to  find  gophers  parasitized  by  larvae  of  botflies, 
because  these  flies  lay  eggs  on  the  fur  of  animals,  an  activity  that 
obviously  occurs  above  ground.  The  lar\'ae  found  in  gophers  appear 
to  me  to  be  identical  in  appearance  with  those  of  Cuterehra  cyanella 
Jones  found  in  Peromyscus  from  Mesa  Verde  (Douglas,  1967a). 
Judging  from  larvae  found  under  the  skin  and  from  open  exit  holes 
made  by  larvae,  infected  gophers  have  from  one  to  three  larvae 
each.  The  incidence  of  infestation  was  highest  in  September.  Some 
larvae  were  found  in  August,  but  none  was  observed  in  gophers 
collected  in  spring. 

In  summary,  parasites  presently  known  from  Thomomys  hottae 
from  Mesa  Verde  are  as  follows:  Dermacentor  andersoni  (Acarina, 
Ixodidae);  Cuterehra  cf.  cyanella  (  Diptera,  Cuterebridae ) ;  Trichu- 
ris  fossor  (Nematoda);  DactylopsyUa  (Foxella)  ignota,  Dactylopsylla 
rara,  and  DactylopsyUa  sp.  (Siphonaptera).  In  addition,  specimens 
of  Cestoda,  Mallophaga,  and  mites  currently  are  being  studied  and 
will  be  reported  elsewhere. 

Predation 

Coyotes  are  the  major  carnivores  in  Mesa  Verde  that  are  known 
to  prey  upon  pocket  gophers  (Howard  and  Childs,  1959:337;  Wilks, 
1963:277).  Foxes  also  are  common  in  the  area,  but  appear  to  rely 
more  upon  berries  and  lizards  than  do  coyotes.  I  examined  a  total 
of  114  coyote  scats  that  were  collected  at  Mesa  Verde  each  month 
from  September,  1963,  through  August,  1964.  Numerous  rodents, 
but  no  pocket  gophers,  were  present  in  the  253  food  items  repre- 
sented (Douglas,  1967a).  Sixteen  scats  of  foxes  contained  several 
rodents,  but  no  gophers.  Bobcats  also  are  present  in  Mesa  Verde,  but 
scats  of  bobcats  were  seen  only  rarely,  and  none  has  been  analyzed. 

Hawks,  owls,  and  eagles  live  in  Mesa  Verde.  Red-tailed  hawks 
were  seen  frequently  on  the  burned  area  on  the  northern  end  of 
Wetherill  Mesa.  I  haxe  not  been  successful  in  locating  occupied 
hawk  or  owl  nests  there.  Howard  and  Childs  (1959:337)  reported 
that  gophers  comprise  7.4  per  cent  of  the  diet  of  red-tailed  hawks, 
and  71.4  per  cent  of  the  diet  of  barn  owls  at  the  San  Joaquin  Experi- 
mental Range  in  California.  It  seems  likely  that  owls  may  be  a 
significant  predator  on  gophers  in  Mesa  Verde. 

There  are  two  species  of  snakes,  the  gopher  snake,  Pituophis 
melanoleucus,  and  the  rattlesnake,  Crotalus  viridis,  in  Mesa  Verde 
that  could  be  considered  as  potential  predators  on  gophers  ( Doug- 


Douglas — Pocket  Gophers  of  Mesa  Verde  169 

las,  1966).  Of  the  two,  the  gopher  snake  is  more  abundant.  I  have 
collected  and  examined  numerous  gopher  snakes  and  other  snakes 
in  Mesa  Verde,  but  none  contained  gopher  remains. 

Gopher  remains  were  not  present  in  the  130  scats  of  carnivores 
that  I  examined,  and  none  was  present  in  35  scats  of  coyotes  from 
the  area  examined  by  Anderson  ( 1961 ) .  Therefore,  it  seems  unlikely 
that  carnivores  exert  much  influence  on  population  dynamics  of 
gophers  in  Mesa  Verde. 

Anomalies 

The  following  skeletal  and  dental  anomalies  were  noted  in 
specimens  examined. 

The  lower  jaw  of  an  adult  male  (CLD  SIO)  has  a  deformed 
coronoid  process,  composed  of  two  parts.  The  ventral  part  is  of 
normal  size  in  width  and  depth,  but  the  dorsal  part  is  slender  and 
abnormally  close  to  the  articular  process.  The  coronoid  process 
appears  to  have  been  broken,  then  to  have  ankylosed  where  the 
halves  touched.  The  dorsal  part  of  the  process  barely  clears  the 
squamosal  when  the  jaws  are  articulated  in  occlusal  position  with 
the  skull. 

Another  adult  male  (KU  75977),  taken  on  November  10,  1957, 
in  Prater  Canvon,  7500  ft.,  bv  T.  R.  Alcorn,  has  several  anomalies  of 
the  skull.  A  tumerous  growth  had  deformed  the  zygomatic  plate 
of  the  right  maxilla,  and  also  involved  smaller  bones  in  the  area. 
The  right  zygomatic  plate  is  abnormally  thick,  antero-posteriorly; 
a  bulge  of  bone  protrudes  from  the  posterior  part  of  the  plate  into 
the  orbit.  There  is  a  small  hole  in  the  maxilla  just  anterior  to  the 
lacrimal.  A  wormian  bone  is  present  at  the  junction  of  the  right 
maxilla,  lacrimal,  and  frontal  bones.  The  right  jugal  is  abnormally 
short  and  thick,  and  the  zygomatic  arch  is  somewhat  displaced 
medially  because  of  the  tumerous  growth  on  the  zygomatic  plate 
and  a  bending  of  the  rostrum  to  the  right.  The  posterior,  ventral 
margins  of  the  premaxillae  appear  to  have  broken,  owing  to  the 
torsion  of  the  rostrum.  A  suture  occurs  in  the  left  premaxilla  in  line 
^\'ith  the  posterior  border  of  the  nasals,  and  a  suture  occurs  some- 
what posterior  to  this  in  the  right  premaxilla.  The  torsion  of  the 
rostrum  has  resulted  in  malocclusion  of  the  upper  and  lower  incisors, 
but  the  maxillary  and  mandibular  toothrows  appear  to  be  normal. 

A  young  female  (CLD  959)  has  an  unusual  dental  anomaly. 
The  anterior  prism  of  each  lower  premolar   (p4)   is  unworn  and 


170  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

projects  dorsally  two  and  three  millimeters  above  the  posterior 
prisms  of  the  respective  premolars.  The  reason  for  this  anomaly 
appears  to  be  related  to  the  mobility  of  the  jaws.  The  glenoid  fossa 
in  gophers  is  several  millimeters  in  length  and  is  parallel  to  the 
long  axis  of  the  skull.  This  arrangement  facilitates  propalinal  chew- 
ing with  the  cheekteeth,  and  allows  forward  movement  of  the  jaws 
such  that  the  incisors  occlude  and  can  be  used  when  the  cheekteeth 
are  disengaged.  Normally,  when  the  lower  jaws  are  at  their  pos- 
teriormost  position  and  the  cheekteeth  are  occluding,  the  anterior 
prism  of  each  lower  premolar  lies  in  line  with  the  posterior  prism  of 
each  upper  premolar.  In  this  specimen  the  jaws  cannot  mo\'e  far 
enough  posteriorly  for  such  an  overbite  to  occur.  The  mobility  of 
the  lower  jaws  is  restricted  because  the  distance  between  rami  is 
not  great  enough  at  their  posterior  ends.  Thus,  the  articular 
processes  bind  against  the  medial  sides  of  the  glenoid  fossae  and 
against  the  bullae,  before  reaching  their  posteriormost  position  in 
the  fossa.  As  a  result,  the  anterior  prisms  on  the  lower  premolars 
do  not  occlude,  and  remain  unworn. 

Discussion 

In  Mesa  Verde,  pocket  gophers  live  in  soils  having  less  than  32 
per  cent  clay  and  varying  amounts  of  sand  and  silt.  Ten  of  18 
samples  of  soil  from  areas  inhabited  by  gophers  contained  more  silt 
than  sand,  whereas  the  other  eight  contained  more  sand  than  silt. 
It  seems  likely  that  the  percentage  of  clay  affects  the  ability  of 
gophers  to  use  soils  more  than  do  varying  amounts  of  silt  and  sand. 
Downhower  and  Hall  (1966)  found  Geomijs  in  Kansas  Hving  only 
in  soils  having  less  than  30  per  cent  clay  and  more  than  40  per  cent 
sand.  Attempts  by  these  authors  to  introduce  pocket  gophers  at 
places  where  none  lived  were  least  successful  where  soils  had  a 
relatively  high  clay  content. 

Although  the  percentage  of  clay  in  most  of  the  soils  in  Mesa 
Verde  was  low,  it  was  sufficiently  high  in  most  localities  to  make 
the  setting  of  traps  difficult  when  the  soil  was  dry.  Soils  at  various 
localities  on  the  mesa  were  dry  and  hard  during  much  of  the  sum- 
mer, whereas  those  in  the  bottoms  of  canyons  remained  friable 
throughout  the  year. 

The  largest  sites  of  concentration  of  gophers  in  Mesa  Verde 
occurred  in  areas  having  deep  soils.  Nevertheless,  gophers  were 
taken  from  rocky  areas  and  from  places  in  the  woodland  where  soils 


Douglas — Pocket  Goppiers  of  Mesa  Verde  171 

were  eight  inches  or  less  in  depth.  In  such  cases,  deeper  soils  gen- 
erally were  within  part  of  the  individual's  range.  When  a  gopher 
was  taken  from  a  shallow  burrow  system,  it  invariably  was  a  young 
indixidual.  Burrow  systems  of  adults  were  at  varying  depths  below 
the  surface,  depending  upon  the  depth  of  soil  and  amount  of  rock 
in  the  habitat.  In  deep  sandy  soils,  main  tunnels  of  adults  generally 
were  eight  to  12  inches  below  the  surface;  feeding  tunnels  were 
shallower. 

Gophers  live  in  various  vegetational  associations  in  Mesa  Verde. 
In  some  places  grasses  are  the  most  abundant  plants,  but  in  most 
places  shrubs  and  various  herbaceous  species  also  occur  in  the 
ground  cover.  Gophers  eat  a  variety  of  roots  and  aerial  parts  of 
plants,  as  was  determined  from  feeding  experiments  with  captives 
and  by  inspecting  contents  of  caches  and  cheek  pouches  of  gophers. 

Most  localities  at  which  gophers  have  been  taken  in  Mesa  Verde 
are  on  the  northern  half  of  the  park,  at  the  higher  elevations. 
Gophers  are  most  abundant  along  the  North  Rim  and  in  the  deep 
soils  of  canyons.  In  general,  large  colonies  do  not  occur  on  the 
mesas  at  elevations  of  less  than  7000  feet;  the  lowest  canyon  eleva- 
tions at  which  gophers  were  taken  was  at  6200  feet  along  the 
Mancos  River. 

The  occurrence  of  a  large  colony  of  gophers  on  the  southern  end 
of  Moccasin  Mesa  is  somewhat  unusual.  Gophers  were  not  found  in 
sandy  drainages  on  the  southern  ends  of  other  mesas  in  the  park. 
I  think  that  the  deepness  of  the  soil  in  the  meadow  on  Moccasin 
Mesa  and  the  persistent  moisture  of  lower  depths  helps  to  explain 
the  ability  of  gophers  to  live  there.  This  soil  moisture  is  reflected 
also  in  the  abundance  of  grasses  and  herbaceous  species  in  the 
ground  cover;  a  dense  ground  cover  is  not  common  at  lower  eleva- 
tions on  the  mesas.  Apparently  the  soil  in  the  meadow  is  deep 
enough  over  the  underlying  bedrock  to  retain  runoff  moisture 
effectively. 

Cactus  is  thought  to  be  an  important  source  of  moisture  for 
gophers,  especially  on  the  southern  end  of  Moccasin  Mesa.  Less 
precipitation  occurs  at  the  lower  elevations  on  the  southern  end  of 
Mesa  Verde,  thus  soils  and  plants  tend  to  dry  out  earlier  in  the 
summer  at  these  elevations  than  at  higher,  more  mesic  elevations  on 
the  north.  The  roots  of  cacti  were  cropped  severely  during  summer 
months  by  gophers  on  Moccasin  Mesa.  Such  a  reliance  upon  cacti 
was  not  evident  in  habitats  of  higher  elevations  in  the  park. 


172  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Captive  gophers  were  found  to  prefer  soft,  fleshy  roots  to  hard, 
woody  roots.  Roots  found  in  food  caches  and  in  cheek  pouches  of 
gophers  were  mostly  fleshy  kinds.  Grasses,  bushes,  and  shrubs  have 
numerous  soft  fleshy  roots  that  apparently  can  be  cropped  repeat- 
edly. Plants  having  only  one  major  tap  root,  as  do  many  herbaceous 
species,  can  be  eliminated  rather  quickly  by  having  their  roots 
cropped.  It  is  thought  that  Thomomijs  influences  the  succession  of 
vegetation  by  a  series  of  events.  Initially  the  soil  is  disturbed  by 
mound  building,  thereby  providing  a  seed  bed  for  early  successional 
plants  such  as  globe  mallow  and  other  weedy  species.  Gophers  eat 
roots  of  the  weedy  plants  that  have  become  established  and  the 
evidence  indicates  that  many  plants  are  killed  in  the  process.  In 
about  two  years  time,  the  earliest-built  mounds  have  become 
obliterated.  Because  the  disturbed  area  usually  is  small,  invasion 
by  plants  is  rapid.  The  gopher  thus  changes  the  habitat  temporarily 
insofar  as  the  vegetation  is  concerned.  On  the  other  hand,  gophers 
turn  over  large  quantities  of  soil,  and  their  burrows  permit  percola- 
tion of  moisture  to  lower  depths;  storage  of  roots,  and  deposition  of 
fecal  material  also  raises  the  organic  content  of  the  soil. 

The  two  periods  of  increased  activity  in  mound  building,  occur- 
ring in  spring  and  in  late  autumn,  appear  to  be  related,  respectively, 
to  increasing  and  decreasing  temperatures  at  the  12-inch  depth. 
The  causality  of  this  relationship  can  only  be  inferred.  Gophers  also 
bear  young  in  spring  during  the  period  of  soil  warming,  although 
the  earliest  breeders  must  bear  young  when  the  soil  is  still  rather 
cold. 

Thomomijs  apparently  forages  above  ground  more  extensively 
than  has  been  supposed.  Pine  cones  found  in  a  food  cache  suggest 
that  at  least  one  individual  may  have  foraged  at  a  distance  of  about 
30  feet  from  its  burrow.  Shrubs,  especially  rabbit  brush,  occurring 
within  several  feet  of  fresh  mounds  showed  evidence  of  browsing. 
It  is  assumed  that  the  browsing  was  done  by  gophers,  owing  to  the 
proximity  of  the  shrubs  to  the  mounds  and  because  of  the  fondness 
of  captives  for  stems  of  this  shrub. 

Various  investigators  have  remarked  that  gophers,  when  released 
near  their  burrow,  do  not  alwavs  seem  to  be  aware  of  where  to  enter 
their  tunnels,  and  often  dig  new  holes.  This  may  be  due  to  in- 
dividuals being  released  in  daylight;  gophers  are  known  to  forage 
above  ground  at  night,  and  occasionally  in  the  daytime  when  the 
sky  is  overcast.   Evidence  indicates  that  gophers  forage  around  the 


Douglas — Pocket  Gophers  of  Mesa  Verde  173 

openings  of  their  burrows,  and  that  they  move  about  on  the  surface 
and  perhaps  range  for  some  distance  from  the  openirigs. 

Remains  of  Tlwmomys  have  been  found  in  owl  pellets  (Howard 
and  Child,  1959)  and  these  gophers  also  have  been  taken  in  snap 
traps  set  for  mice  (WilHams,  1955:227).  Larvae  of  bot  flies  were 
found  in  gophers  taken  in  Mesa  Verde,  indicating  that  they  spend 
enough  time  above  ground  for  Hies  to  lay  eggs  on  their  fur. 

The  present  lack  of  knowledge  about  how  to  accurately  deter- 
mine the  age  of  gophers  from  skeletal  characteristics  presents  a  per- 
plexing problem.  Young  of  the  year  were  found  to  attain  adult  size 
at  a  rapid  rate,  and  animals  several  months  old  could  not  be  differ- 
entiated from  adults  on  the  basis  of  skeletal  or  bodily  dimensions. 
Examination  of  reproductix'e  organs  allows  separation  of  young  of 
the  year  from  adults  and  this  is  of  some  value  in  studying  popula- 
tions. It  would  be  invaluable  to  future  studies  of  gopher  populations 
if  an  accurate  method  of  estimating  chronological  age  could  be 
found.  Studies  presently  are  being  conducted  toward  this  objective 
and  these  will  be  reported  elsewhere. 

Relationship  of  Gophers  to  Archeological  Sites 

Mesa  Verde  National  Park  was  formed  to  protect  dwellings  of  prehistoric 
Indians  who  once  lived  in  the  area.  Park  officials  are  responsible  for  protecting 
not  only  these  prehistoric  ruins,  but  also  the  flora  and  fauna.  Burrowing  rodents 
occasionally  cause  damage  to  important  archeological  sites,  and  although 
gophers  are  less  of  a  problem  than  rock  squirrels,  chipmunks,  or  badgers,  they 
can  and  do  damage  sites  by  burrowing  through  such  things  as  trash  mounds, 
which  would  be  far  more  valuable  if  undisturbed.  In  Mesa  Verde  Site  786, 
a  gopher  skeleton  was  recovered  from  a  timnel  that  ended  on  the  floor  of  a 
kiva,  72  inches  below  the  surface  of  the  ground  (Arthur  Rohn,  personal  com- 
munication). The  digging  of  a  tunnel  to  this  deptli  undoubtedly  was  facilitated 
by  the  friable  nature  of  the  fill  in  the  kiva.  Such  burrows  easily  can  allow 
small  artifacts  to  become  juxtaposed  in  the  occupational  strata,  thereby  render- 
ing interpretation  difficult,  if  not  impossible. 

In  order  to  maintain  the  integrity  of  archeological  sites  in  Mesa  Verde  and 
elsewhere,  gophers  and  other  burrowing  rodents  should  be  removed  as  soon  as 
possible  after  they  are  noticed  in  such  places.  Macabee  traps  are  recommended 
for  pocket  gophers  and  are  most  effective  when  used  in  pairs — one  trap  being 
set  in  each  direction  in  a  main  underground  tunnel.  Gophers  are  important 
members  of  the  faunal  assemblage,  however,  and  should  be  removed  only 
when  they  are  distiul)ing  archeological  sites  or  underground  cables. 

Acknowledgments 

These  studies  were  begun  while  I  was  Biologist  for  the  Wetherill  Mesa 
Archeological  Project.  This  and  other  studies  would  not  have  been  possible 
without  the  support  and  encouragement  of  Dr.  Douglas  Osliorne,  Long  Beach 


174  Misc.  Publ.  51,  Uxw.  Kansas  Mus.  Nat.  Hist. 

State  College,  formerly  director  of  the  Wetherill  Project.  I  am  grateful  to 
Robert  R.  Patterson,  The  University  of  Kansas,  who  took  his  vacation  time  to 
assist  me  in  the  field  in  1963  and  1965.  Mr.  Chester  A.  Thomas,  formerly 
Sviperintendent,  and  Mrs.  Jean  Pinkley,  formerly  Chief  of  Interpretation  at 
Mesa  Verde  National  Park,  assisted  me  by  arranging  for  my  NPS  collaborator- 
ship  and  collecting  permits.  Stanley  Welsh,  Brigham  Young  University,  and 
James  A.  Erdman,  United  States  Geological  Survey,  Denver,  proxided  identi- 
fications for  numerous  species  of  plants.  Mrs.  Marilyn  A.  Colyer,  Mancos, 
Colorado,  assisted  me  in  analyzing  vegetation  in  \arious  habitats.  I  am  indebted 
to  John  Ubelaker  for  identifying  endoparasites  and  to  Cluff  Hopla  for  identify- 
ing fleas  from  numerous  gophers.  The  opportunity  to  work  in  the  field  in  1967 
was  made  possible  by  a  grant-in-aid  from  the  Society  of  The  Sigma  Xi. 

The  Wetherill  Mesa  Project  was  an  interdisciplinary  program  of  the  National 
Park  Service  to  which  the  National  Geographic  Society  contributed  generously. 
I  am  indebted  to  the  Society  for  a  major  share  of  the  support  that  resulted  in 
this  report.   This  paper  is  contribution  no.  48  of  the  Wetherill  Mesa  Project. 

Literature  Cited 

Anderson,  S. 

1961.  Mammals  of  Mesa  Verde  National  Park,  Colorado.  Univ.  Kansas 
Publ.,  Mus.  Nat.  Hist.,  14:29-67,  2  pis.,  3  figs. 

Arrhenius,  G.,  and  E.  Bonatti 

1965.  The  Mesa  Verde  loess.   Amer.  Antiquity,  31  (part  2):92-100,  6  figs. 

BouYOucos,  G.  J. 

1962.  Hydrometer  method  improved  for  making  particle  size  analyses  of 
soils.    Agronomy  Jour.,  54:464-465. 

Douglas,  C.  L. 

1963.  Apache  pocket  mouse  found  in  Mesa  Verde  National  Park,  Colorado. 
Southwestern  Nat.,  8:173. 

1966.  Amphibians  and  reptiles  of  Mesa  Verde  National  Park,  Colorado. 
Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.,  15:711-744,  pis.  37-38,  6  figs. 

1967a.  Comparatixe  ecology  f)f  pinyon  mice  and  deer  mice  in  Mesa  Verde 
National  Park,  Colorado.  Ph.D.  dissertation,  Uni\'.  Kansas,  Law- 
rence, 209  pp. 

1967b.  New  records  of  mammals  from  Mesa  Verde  National  Park,  Colorado. 
Jour.  Mamm.,  48:322-323. 

Douglas,  C.  L.,  and  J.  A.  Erdman 

1967.  Development  of  terminal  buds  in  pinyon  pine  and  Douglas-fir  trees. 
Pearce-Sellards  Series,  Texas  Mem.  Mus.,  8:1-19,  5  figs. 

DowNHOWER,  J.  F.,  and  E.  R.  Hall 

1966.  The  pocket  gopher  in  Kansas.  Misc.  Publ.  Mus.  Nat.  Hist.,  Univ. 
Kansas,  44:1-32,  8  figs. 

Erdman,  J.  A. 

1962.  Ecology  of  the  pinyon-juniper  woodland  of  Wetherill  Mesa,  Mesa 
Verde  National  Park,  Colorado.  M.A.  thesis,  Univ.  Colorado, 
Boulder. 

Erdman,  J.  A.,  W.  A.  Weber,  and  J.  M.  Tucker 

1962.     Querciis  ajoensis  in  Colorado.    Southwestern  Nat.,  7:269-270. 

Fritts,  H.  C,  D.  G.  Smith,  and  M.  A.  Stokes 

1965.  The  biological  model  for  paleoclimatic  interpretation  of  Mesa  Verde 
tree-ring  series.    Amer.  Antiquity,  31   (part  2):101-121,  10  figs. 

Howard,  W.  E.,  and  H.  E.  Childs,  Jr. 

1959.  Ecology  of  pocket  gophers  with  emphasis  on  Thomomys  bottae 
mewa.    Hilgardia,  29:277-358,  40  figs. 


Douglas — Pocket  Gophers  of  Mesa  Verde  175 

Kenxerly,  T.  E.,  Jr. 

1964.  Microenvironniental  conditions  of  the  pocket  gopher  l)mr()\v.  Texas 
Jour.  Sci.,  16:395-441,  16  figs. 

Marshall,  \V.  H. 

1941.     Thomomys   as   hurrowers  in   the  snow.    Jour.    Manim.,  22:196-197, 
2  figs. 
Miller,  M.  A. 

1948.  Seasonal  trends  in  burrowing  of  pocket  gophers  (Thomomys).  Jour. 
Manim.,  29:38-44,  1  pi.,  2  figs. 

Schramm,  P. 

1961.     Copulation    and    gestation    in    the    pocket    gopher.     Jour.    Mamm., 
42:167-170. 
Welsh,  S.  L.,  and  J.  A.  Erdman 

1964.  Annotated  checklist  of  the  plants  of  Mesa  Verde,  Colorado.  Brigham 
Young  Univ.  Sci.  Bull.,  Biol.  Ser.,  4:1-32. 

WiLKS,  B.  J. 

196.3.  Some  aspects  of  the  ecology  and  population  dynamics  of  the  pocket 
gopher  (Geomy.s  btirsarius)  in  southern  Texas.  Texas  Jour.  Sci., 
15:241-283,  5  figs. 

Williams,  O. 

1955.  Distribution  of  mice  and  shrews  in  a  Colorado  montane  forest.  Jour. 
Mamm.,  36:221-231,  1  pi.,  1  fig. 

YOUN'GMAX,   P.   M. 

1958.  Ceographic  variation  in  the  pocket  gopher,  Thomomys  bottae,  in 
Colorado.    Univ.  Kansas  Publ.,  Mus.  Nat.   Hist.,  9:363-384,  7  figs. 


COTTON  RATS  OF  THE 
SIGMODON  FULVIVENTER  GROUP 

BY 

RoLLiN  H.  Baicer 

Cotton  rats  of  the  genus  Si^modon  are  dominant  grass-eating 
and  runway-making  rodents  in  most  grassy  habitats  of  south  tem- 
perate and  tropieal  North  and  Middle  America.  In  this  region, 
cotton  rats  play  a  role  somewhat  similar  to  that  of  voles  of  the  genus 
Microtiis  in  north  temperate  and  boreal  parts  of  the  continent.  The 
genus  is  known  from  late  Pliocene  deposits  of  North  America;  one 
present-day  species,  S.  luspidtis,  is  recorded  from  as  far  back  as 
Sangamon  times  in  the  late  Pleistocene.  Today,  rodents  of  this  genus 
are  widely  distributed  (see  Fig.  1)  from  southern  United  States 
(north  to  40°  N  latitude  at  the  Kansas-Nebraska  border)  southward 
to  coastal  Peru  (south  to  about  8°  S  latitude).  Cotton  rats  are  at 
home  at  altitudes  ranging  from  sea  level  in  coastal  marshes  to  more 
than  3200  meters  in  the  highlands  of  the  Trans-Mexican  Volcanic 
Belt.  Ecologically,  they  are  associated  with  grasses — which  may  be 
almost  pure  stands  of  perennial  bunch  grasses,  mixed  grasses  and 
herbs,  or  mixed  grasses  and  shrub  growth — in  a  variety  of  habitats 
and  climates  in  both  xeric  areas  (where  annual  rainfall  may  be  no 
more  than  100  millimeters)  and  mesic  situations  (where  annual 
rainfall  may  be  more  than  500  millimeters ) . 

Cotton  rats  have  been  known  to  science  since  1825,  when  Say 
and  Ord  described  Sigtiwdon  Jiispidus  from  Florida.  Descriptions  of 
cotton  rats  from  other  localities  appeared  in  the  literature  in  the 
middle  and  late  1800's,  and  in  1902  Vernon  Bailey  reviewed  the 
specific  status  of  North  American  cotton  rats  and  established  the  two 
presently  recognized  groups:  the  semi-naked-tailed  S.  Jiispidus  and 
the  hairy-tailed  members  of  the  S.  fidviventer  group.  Bailey's  find- 
ings and  later  taxonomic  accounts,  mostly  by  Nelson  and  Goldman 
and  by  Goodwin,  form  the  basis  for  the  synopsis  found  in  Hall  and 
Kelson  (1959:671-679).  According  to  the  latter,  the  fidviventer 
group  contains  12  species  from  widely-scattered  locations  in  and 
along  the  Cordillera  from  Arizona,  New  Mexico,  and  Texas,  south  to 
the  Mexican  state  of  Oaxaca.  It  is  the  purpose  of  this  report  to 
review  and  revise  this  present  taxonomic  arrangement  in  light  of 

(177) 


178 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Fig.  1.    Geographic  distriliution  of  cotton  rats.    1.  Sigmodon  hispidus.    2.  Sig- 

modon  fulviventer  group. 

additional  field  and  museum  studies  that  have  added  to  our  knowl- 
edge of  the  distribution,  ecology  and  systematics  of  the  S.  fulviventer 
group. 

Specimens  examined  are  listed  in  the  accounts  of  each  taxa  discussed.  The 
number  from  each  locality  is  noted,  using  abbreviations  for  the  museums 
listed  below  to  indicate  where  specimens  are  on  deposit.  I  am  grateful  to  the 
persons  whose  names  appear  below  for  allowing  me  the  privilege  of  examining 
the  specimens:  (AM)  American  Museum  of  Natural  History,  G.  G.  Goodwin, 
R.  G.  Van  Gelder,  and  S.  Anderson;  (CAS)  California  Academy  of  Sciences, 
R.  T.  Orr;  (DMNH)  Dallas  Museum  of  Natural  History,  F.  W.  Miller;  (FM) 
Field  Museum  of  Natural  History,  P.  Hershkovitz;  (KU)  University  of  Kansas, 
Museum  of  Natural  History,  E.  R.  Hall  and  J.  K.  Jones,  Jr.;  (LA)  Los  Angeles 
County  Museum,  K.  Stager  and  C.  A.  McLaughlin;  (LSU)  Louisiana  State 
University,   Museum  of  Zoology,  G.   H.   Lowery,  Jr.;    (MSU)    Michigan   State 


Baker — Sigmodon  fulviventer  Group  179 

UniNersity,  The  Museum,  R.  H.  Baker;  (MVZ)  Uuiveisity  of  California, 
Museum  of  Vertebrate  Zoology,  S.  B.  Benson  and  W.  Z.  Liclicker;  ( OC )  Occi- 
dental College,  J.  W.  Hardy;  (TCWC)  Texas  Cooperative  Wildlife  Collection, 
Texas  A&M  Uni\ersity,  W.  B.  Davis;  (TT)  Texas  Technological  College, 
R.  L.  Packard;  (UI)  University  of  Illinois,  Museum  of  Natural  History,  D.  F. 
Hoflmeister;  (UM)  University  of  Michigan,  Museum  of  Zoology,  W.  H.  Burt 
and  E.  T.  Hooper;  (UNAM)  Instituto  de  Biologia,  B.  Villa-R.;  (US)  U.S. 
National  Museum  including  Biological  Sur\eys  Collection,  R.  H.  Manville  and 
C.  O.  Handley,  Jr. 

Field  data  and  specimens  (preserved  or  brought  liack  alive)  have  been 
obtained  by  summer  expeditions  from  The  Museiuu  at  Michigan  State  Univer- 
sity, beginning  in  1957.  Parties,  headed  by  the  author,  \  isited  known  collecting 
sites  for  \arious  species  of  cotton  rats  and  also  other  likely  places  located 
geographically  bet\veen  these  localities.  Collecting  was  accomplished  each 
summer  from  1957  through  1967,  except  for  1962.  Funds  to  assist  in  defraying 
costs  of  field  work  have  geen  generously  provided  l)y  the  MSU  Development 
Fund,  private  donors  Mrs.  E.  R.  Warren  and  Mr.  Russell  Jameson,  and  the 
National  Science  Foundation  (GB  2227).  In  the  course  of  field  work  25  per- 
sons, including  long-time  associate  Dr.  Robert  G.  Webb,  accompanied  me  and 
must  be  recognized  as  important  contributors  to  the  success  of  the  expeditions. 
These  persons,  most  of  them  at  the  time  advanced  students  in  biology  at 
Michigan  State  University,  are  Bruce  R.  Baker,  Larry  P.  Bowdre,  Daniel  E. 
Boyle,  Bernard  J.  Cripps,  Peter  L.  Dalby,  Julian  P.  Donahue,  Leslie  C.  Drew, 
James  J.  Drake,  Rol:)ert  L.  Fleming,  William  C.  Gasaway,  John  Keever  Greer, 
John  J.  Grost  II,  James  R.  Koschmann,  David  E.  Mohrhardt,  Michael  K.  Peter- 
sen, Carleton  J.  Phillips,  Rudolph  Scheibner,  Henry  L.  Short,  Charles  E.  Smith, 
Thomas  Struhsaker,  Donald  F.  Switzenberg,  Charles  L.  Warner,  Jr.,  Daniel 
Womochel,  Robert  G.  Webb,  and  Frances  E.  Welling. 

I  am  grateful  to  the  many  persons  in  Mexico  who  were  cooperative,  offered 
advice,  or  allowed  our  field  parties  to  \isit  their  properties,  especially  to  Dr. 
Rodolfo  Hernandez  Corzo,  Direccion  General  de  la  Fauna  Silvestre,  for  granting 
scientific  study  permits.  As  always,  I  have  depended  to  a  great  degree  on  the 
council  of  my  good  friend  Dr.  Bernardo  Villa-R.,  whose  vast  knowledge  of 
Mexican  mammals  is  readily  and  generously  shared  with  others.  I  acknowledge 
the  help  of  Dr.  John  Beaman  in  identifying  plants  and  Dr.  Richard  B.  Loomis 
and  Dr.  Robert  Traub  for  identifying  ectoparasites.  Finally,  I  dedicate  this 
modest  report  to  a  truly  great  advisor  and  friend.  Professor  E.  Raymond  Hall, 
with  whom  I  have  spent  many  years  of  pleasant  association  and  to  whom 
American  mammalogy  owes  so  much  for  his  steadfast  determination  to  under- 
stand fully  speciation  in  North  American  mammals. 

All  measurements  in  this  report  are  given  in  the  metric  system.  Capitalized 
color  terms  are  those  of  Ridgway  (1912).  Those  localities  printed  in  Roman 
type  in  the  lists  of  specimens  examined  or  of  other  records  are  recorded  on  the 
distribution  maps  as  either  solid  circles  ( specimens  examined )  or  as  solid 
squares  (other  records).  Localities  printed  in  Italic  type  are  not  mapped  be- 
cause of  their  proximity  to  other  places  that  are  mapped.  All  localities  are 
arranged  in  the  lists  from  north  to  south.  Measurements  of  animals  of  the  two 
sexes  are  included  together  since  sexual  dimorphism  in  the  dimensions  used 
was  found  to  be  negligible,  although  Chipman  (1965)  noted  significant  differ- 


180 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


ences  between  sexes  in  weight  and  length  of  head  and  body  in  S.  hispidus 
from  Louisiana.  Unless  otherwise  credited,  all  photographs  were  taken  by 
the  author. 

Determixatiox  of  Age  in  Cottox  Rats 

Characteristics  of  the  several  species  of  cotton  rats  were  determined  by  both 
gross  and  microscopic  examinations.  The  selecting  of  comparati\e  age-classes 
offered  a  prolilem  because  series  of  museum  specimens  include  mostly  young 
individuals.  This  is  probably  because  most  of  them  have  been  caught  in 
museum  special  traps,  which  are  not  satisfactory  for  taking  animals  larger 
than  60-80  grams.  Experience  shows  that  the  small-  to  medium-sized  animals 
also  make  up  a  major  segment  of  the  trappable  cotton  rat  population,  especially 
in  the  period  of  greatest  breeding  activity  in  the  warm  months  of  the  year, 
during  which  time  most  collecting  parties  are  active.  Cotton  rats  of  each  species, 
born  and  reared  in  captivity  in  the  MSU  Museum  Live  Animal  Colony,  were 
killed  at  intervals  and  measured.  Growth  curves,  which  level  out  at  250-300 
days,  are  shown  in  Figures  2  and  3  (see  also  Hoffmeister,  1963,  and  Chipman, 
1965).    In   laboratory-raised  animals,   females  produced   offspring   at  87   days 


190 


100 


200 


300 


400 


180- 


70 


60 


50, 


500 


600 


S.   fulwiventer 


5.    alleni 


S.   leucotis 


S.     ochrogngthu  s  _ 


100 


200  300  400  500 

AGE         IN         DAYS 


600 


190 


180 


170 


160 


150 


140 


130 


120 


110 


100 


90 


80 


60 


50 


Fig.  2.    Growth  as  judged  by  length  of  head  and  body  in  cotton  rats  of  the 

Sigmodon  fulvivei^ter  group.    Data  from   laboratory  born  and  reared   animals, 

the  exact  ages  of  which  were  known  at  death. 


Baker — Sigmodon  fulviventer  Group 


181 


37.5 


3S.0 


)5.0 


100 

-1— 


200 


300 


400 


SOO 


600 

— T" 


S.   fulviventer 


S.      glleni 


S.    leucotis 
S.     ochrognothus 


37.5 


35.0 


32.5 


30.0 


27.5 


22.5 


20.0 


17.5 


-  18.0 


100 


200 


300 

AGE     IN 


400 

DAYS 


500 


600 


Fig.  3. 
of  the 


Crania]  fiiowth  as  judged  by  condylopremaxillary  length  in  cotton  rat.s 
Sif!.mod()n  fulviventer   group.     Data   from   laboratory   born   and    reared 
animals,  the  exact  ages  of  which  were  known  at  death. 


(S.  alleni),  77  days  (S.  fulviventer),  and  71  days  (S.  hispidus  and  S.  ochrognathus) 
and  would  have  mated  approximately  35  days  earlier.  Cotton  rats  (S.  hispidus) 
in  the  wild  are  known  to  breed  at  40-50  days  of  age  (Odum,  1955).  It  is 
suspected  then  that  the  "normal"  life  span  in  natme  is  no  more  than  six  months 
(Meyer  and  Meyer,  1944,  and  Odum,  1955).  This  would  mean  that  the  trap- 
pable population  would  include  few  truly  "full-grown"  adult  animals  (at  least 
250  days  old)  and  would  help  account  for  the  scarcity  of  such  animals  in 
population  samples  preserved  in  museum  collections. 

Study  specimens  of  colony-raised  animals  purposely  killed  at  100  days  and 
at  200  days  were  compared  with  wild-taken  specimens  with  similar  age  charac- 
teristics. It  was  concluded  that  a  cotton  rat  less  than  75  days  old  is  a  juvenile, 
between  75  and  200  days  old  is  a  young  adult,  between  200  and  300  days  old 
is  an  adult,  and  more  than  300  days  old  is  an  old  adult.  A  detailed  study  of  the 
characteristics  of  aging,  pelage  development,  and  breeding  habits  will  appear 
in  a  later  report. 

Evolution  in  Recent  Species 

Grass-eating  as  an  Isolating  Mechanism 

Probably  an  important  factor  in  speciation  in  cotton  rats  has  been 

the  inabihty  of  different  species  to  Hve  together  in  the  same  habitat. 

On  the  few  occasions  when  two  species  ha^'e  been  collected  within 

the  same  grassy  area,  one  has  been  common  (seemingly  dominant) 


182  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

and  the  other  rare.  Field  observations  lead  me  to  suspect  that  the 
local  distribution  of  individual  species  has  fluctuated  owing  to  this 
competition  in  changing  environments,  especially  due  in  the  past 
century  to  various  land-use  practices  by  man,  including  clearing  and 
lumbering,  grazing  of  livestock,  and  cultivation.  This  activity  has 
caused  environmental  changes  that  have  favored  one  species  o\'er 
another.  Thus,  in  Michoacan,  S.  hispidus  li^'es  in  undisturbed  bunch 
grass  (sacaton),  whereas  S.  ciUeni  lives  nearby  in  heavily-grazed, 
brushy,  fallow  fields. 

The  general  negative  interspecific  reaction  of  species  of  cotton 
rats  is  not  unique  but  seems  typical  as  well  of  other  grass-eating, 
runway-making  rodents,  notably  \'oles  of  the  genus  Micwtus  ( Find- 
ley,  1954;  Anderson,  1959;  Findley  and  Jones,  1962).  In  the  Mexican 
highlands,  cotton  rats,  M.  mexicanus,  and  Neoiomodon  alstoni 
(Davis  and  Follansbee,  1945)  seem  to  avoid  habitats  occupied  by 
the  others.  This  relationship  can  be  found  also  in  such  places  as 
Kansas,  where  the  aggressive  S.  hispidus  probably  is  replacing  M. 
ochrogaster,  appearing  to  have  accomplished  this  already  in  parts  of 
Louisiana,  Texas,  and  Oklahoma.  In  short,  any  given  grassy  habitat 
may  harbor  only  one  (or  one  dominant)  species  of  grass-eating, 
runway-making  rodent,  but  seed-eating  species  (included  in  such 
genera  as  Peromysciis,  Reitlvodonfomys,  Baiomijs,  Liomys,  Perogna- 
tJius)  appear  to  experience  little  evident  incompatibility  in  their 
own  group  or  with  the  grass-eater,  whose  runways  they  use  seem- 
ingly without  restraint. 

That  we  find  several  seed-eating  rodents  but  only  one  grass- 
eating  rodent  in  a  given  grassy  habitat  may  be  because  each  seed- 
eater  is  adapted  to  a  "smaller"  segment  of  the  environment  (and 
thus  allows  space  for  se\'eral  species)  than  is  the  grass-eater,  whose 
habitat  requirements  may  be  too  "broad"  to  allow  for  close  neighbors 
with  similar  food  habits.  Species  segregation  and  replacement,  then, 
can  be  an  important  factor  in  evolution  in  Sigmodon. 

Ancestry 
Cotton  rats  are  presumed  to  be  of  Neotropical  origin,  and  prob- 
ably arose  in  early  or  middle  Pliocene  from  a  yet  unknown,  grass- 
eating  cricetine  ancestor.  This  ancestor  probably  had  a  four-rooted 
first  lower  molar  (C.  W.  Hibbard,  personal  communication)  and 
may  have  developed  as  a  progressive  oftslioot  of  the  main  line  of 
phyllotine  rodent  origin  ( Hershkovitz,  1962:23).  It  is  suspected  that 
cotton  rats  evohed  in  a  Middle  American  pastoral  habitat,  although 
Hershkovitz   (1966)   believes  it  to  be  South  American.    Since  the 


Baker — Sigmodon  fulviventer  Group  183 

^-      hispidus  ^  ^ S.  hispidus 

S.  fulviventer 


Sinaloa  Durango  Zacatecas 


Fig.  4.    Distribution  of  species  of  Sigmodon  along  an  east-west  transect  in 
northwestern  Mexico  between  23°  30'  N  and  24°  30'  N. 

first  terrestrial  rodent  en\'ironments  to  become  established  on  the 
Panamanian  land  bridge  after  its  emergence  in  late  Pliocene  were 
undoubtedly  a  series  of  grassland  successions  leading  ultimately  to 
savannas  and  forests,  cotton  rats  could  have  easily  moved  southeast- 
ward to  populate  coastal  parts  of  northern  South  America  where 
they  occur  today  ( see  Fig.  1 ) .  However,  unlike  other  cricetines 
such  as  the  rice  rats  (genus  Oryzomys),  cotton  rats  appeared  unable 
to  spread  southward  in  South  America  to  reach  such  likely  habitats 
as  the  Argentine  Pampas.  Perhaps  cotton  rats  (for  which  no  South 
American  fossils  are  known)  arri\'ed  later  than  rice  rats  and  faced 
well-entrenched  competitors  or  were  blocked  by  encroaching  in- 
terior and  coastal  forests  in  northern  South  America.  In  North 
America  grasslands,  which  had  their  beginnings  in  the  late  Miocene 
and  their  great  development  in  the  Pliocene,  extended  southward 
into  what  is  now  Mexico  and  offered  a  broad  avenue  to  northward 
movements  of  cotton  rats  (see  Cohn,  1965:138-139).  Throughout 
the  Pleistocene  this  area  was  subjected  to  alternating  mild  and  cool 
periods,  and  to  aridity  (Dillon,  1956:173)  and  subhumid  conditions 
( Hibbard,  1960 ) .  Even  so,  early  cotton  rats,  if  possessed  of  some  of 
the  same  ubiquitious  characteristics  as  their  modern  counterparts, 
should  have  had  no  difficulty  surviving  such  changes  from  the  warm 
climate  of  their  Neotropical  home. 

The  genus  is  first  reported  from  the  Blancan  of  the  late  Pliocene 
(Hibbard,  1960:17)  with  the  modern  species,  S.  hispidus,  known 


184  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

from  what  is  now  Texas  as  early  as  Sangamon  times  (third  inter- 
glacial)  in  the  Pleistocene  (Moore  Pit  local  fauna,  Slaughter,  1966: 
90).  From  the  south-central  United  States,  S.  Jiispidus  was  presumed 
to  have  been  displaced  southward  and  laterally  during  Wisconsin 
glaciation  into  refugia  in  peninsula  Florida  and  the  Southwest 
(alluded  to  by  Blair,  1958:460).  Following  the  melt  of  the  glacia- 
tion, S.  J}ispichis  appeared  again  in  the  southern  Great  Plains  and 
seems  still  to  be  on  the  move  northward  (to  the  southern  border  of 
Nebraska  in  1958 — Jones,  1964:212).  As  mentioned  previously,  S. 
hispidus  is  replacing  Microtus  ochrogaster  as  the  dominant  grass- 
eating,  runway-making  rodent  in  Oklahoma  and  parts  of  Kansas. 

Perhaps  it  was  as  late  as  Wisconsin  times  that  conditions  occurred 
to  account  for  the  evolution  from  S.  hispidus  of  the  distinctive  S. 
fulviventer  group.  Today,  these  species  generally  replace  (ecologic 
segregation)  one  another  in  and  along  the  cordillera  from  Arizona 
and  Texas  southeastward  to  Oaxaca.  Although  there  is  some  overlap 
(see  profile,  Fig.  4),  one  species  dominates  in  each  suitable  montane 
environment:  S.  fulviventer  in  open,  elevated  grasslands;  S.  ochro- 
gnathus  in  bunch  grasses  on  arid,  rocky  slopes;  S.  leucotis  in  mesic, 
boreal  grass-shrub;  and  S.  alleni  in  mesic,  boreal-tropical  grass- 
shrub.  Presumably,  basic  stocks  of  S.  hispidus  occupied  some  of 
these  habitats  in  the  late  Pleistocene  with  climatic,  vegetational,  and 
edaphic  selective  forces  operating  so  as  to  produce  the  generally 
allopatric  Recent  species  ( see  Fig.  5 ) . 

Of  the  species  of  the  S.  fulviventer  group,  S.  alhni  might  have 
been  derived  most  recently  from  a  stock  of  S.  hispidus  (from  the 
westward  in  the  Pacific  coastal  lowlands),  whereas  the  other  three 
species  could  have  been  derived  from  stocks  of  S.  Jiispidus  found 
on  the  Mexican  Plateau.  It  is  thought  that  S.  fulviventer  and  S.  alleni 
are  less  differentiated  from  S.  hispidus  than  are  S.  ochrognathus  and 
S.  leucotis.  The  latter  species  seems  most  remote  and  could  have 
branched  off  even  earlier  than  the  others,  being  set  apart  by  unique 
features  including  rostral  depressions  and  a  reduced  or  absent 
lingual  root  on  the  first  molar  (see  also  Dalby  and  Lillevik,  1969). 

In  summary,  Sigmodon  consists  of  one  wide-spread,  ubiquitous 
species,  S.  hispidus,  and  four  derived  species  that  occupy  compara- 
tively small,  allopatric  ranges,  mostly  in  western  Mexico.  Speciation 
in  and  along  the  cordillera  in  southwestern  North  America  may  have 
come  about  through  ecologic  segregation  and  morphologic  di\'er- 
gence  resulting  from  the  availability  of  arid  uplands,  rocky  slopes, 
and  mesic  boreal  and  mesic  tropical-boreal  habitats   (presumably 


Baker — Sigmodon  fulviventer  Group 


185 


Sigmondon  ochrognathus 

Xeric  hillsides  of 
desert  mountains 


Sigmodon  fulviventer 

Open  grasslands  in 
foothills  and  mountains 


Sigmodon  leucotis 

Boreal  grass-shrub 


Mexican  Plateau  stock 

Sigmodon  hispidus 

Pacific  Lowlands  stock 


Sigmodon  alleni 

Boreal-tropical  grass-shrub 


Fig.  5.    x\dapti\e  relationships  of  Recent  cotton  rats. 

unoccupied  by  grass-eating  rodents)  in  which  the  parent  hispid 
species  lacked  survival  ability.  Selection  for  characters  found  in  the 
species  occurring  in  these  habitats  today  was  the  result.  Segregation 
and  character  displacement  probably  were  intensified  because  of  the 
general  incompatibility  of  different  kinds  of  cotton  rats  in  the  same 
living  places. 

Key  to  Species  of  Cotton  Rats  (  Genus  Sigmodon  ) 

1.  Tail  sparsely  haired  and  scaly  in  appearance,  individual  scales  broad, 
0.75  mm.  wide;  skull  generally  long  and  narrow,  basioccipital  long 

and  ]:)road,  palatal  pits  shallow  Sigrnodon  his))idiis 

1'.  Tail  heavily  haired  and  not  scaly  in  appearance,  individual  scales 
narrow,  0.50  mm.  wide;  skull  generally  short  and  broad,  basioccipital 
either  long  and  narrow  or  short  and  broad,  palatal  pits  deeply 
marked  Sigmodon  fidviventer  group,  2 

2.  Ears  (inside  of  pinnae)  whitish,  in  marked  contrast  to  color  of  dor- 
sum; interparietal  generally  less  than  2.0  mm.  in  length  at  midline; 
upper  part  of  each  premaxillary  with  pronounced  rostral  depression; 


186  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

mesopterygoid  fossa  generally  parallel-sided  at  anterior  end;  lingual 
root  of  first  lower  molar  reduced  in  size  and  sometimes  absent  .._.  S.  leucotis 
2'.  Color  of  ears  not  conspicuously  difl^erent  from  color  of  dorsum; 
interparietal  usually  2.0  mm.  or  more  in  length  at  midline;  upper  part 
of  premaxillary  with  slight  or  no  rostral  depression;  mesopterygoid 
fossa  generally  not  parallel-sided  anteriorly;  lingual  root  of  first  lower 
molar  not  reduced  in  size  3 

3.  Buff  coloring  on  nose  and  around  eye  conspicuous;  adult  size  small, 
length  of  head  and  body  averaging  143  mm.  and  condylopremaxillary 
length  averaging  32.0;  auditory  bullae  small  and  elongate;  median 
keel  on  basioccipital  developed;  lateral  bulge  of  capsular  projections 
of  upper  incisors  pronounced;  interparietal  with  slight  to  marked 
median  posterior  notch;  paraoccipital  process  cur\'ed  (when  \aewed 

from  ]:)elow)  and  notched  on  the  anterior  base  S.  ochrognathiis 

3'.  Buff  color  on  nose  and  around  eye  usually  not  in  marked  contrast 
with  rest  of  dorsum;  adult  size  large,  length  of  head  and  body  averag- 
ing at  least  168  mm.  and  condylopremaxillary  length  averaging  at 
least  34.5;  auditory  bullae  large  and  broad  (relative  to  length  of 
skull);  lateral  bulge  of  capsular  projections  of  upper  incisors  slight  to 
moderate;  interparietal  usually  lacking  any  indication  of  a  median 
posterior  notch;  paraoccipital  process  (when  viewed  from  below) 
generally  straight  or  slightly  hooked  4 

4.  Color  of  dorsum  always  brownish,  undeiparts  washed  with  whitish 
or  pale  buff;  adult  size  medium,  length  of  head  and  body  averaging 
168  mm.  and  condylopremaxillary  length  axeraging  34,5;  skull 
flattened  in  appearance,  long  and  narrow;  incisive  foramina  not 
extending  to  line  drawn  between  anterior  ends  of  first  upper  molars; 
basioccipital  short  and  wide;  mesopterygoid  fossa  broad  anteriorly; 
median  keel  on  palate  slighdy  de\eloped;  palatal  pits  moderately 
deep;  incisors  usually  markedly  recurved  ( opisthodont )  S.  allcni 

4'.  Dorsum  "pepper  and  salt"  in  color,  undeiparts  washed  with  buff; 
adult  size  large,  length  of  head  and  body  averaging  179  mm.  and 
condylopremaxillary  length  averaging  36.5;  skull  arched,  short  and 
broad;  incisive  foramina  extending  to  or  beyond  a  line  drawn  be- 
tween anterior  surfaces  of  the  first  upper  molars;  foramen  ovale  large, 
at  least  three-fourths  diameter  of  third  upper  molar;  basioccipital 
long  and  narrow;  mesopterygoid  fossa  narrow  anteriorK';  median  keel 
on  palate  well  developed;  palatal  pits  markedly  deep;  incisors  not 
highly  recurved S.  fulvivcnter 

Sigmodon  alleni 

Sigmodon  olleni,  the  brown  cotton  rat,  is  the  most  tropical- 
adapted  member  of  the  S.  fulviventer  group.  Its  preferred  Hving 
places  are  in  mixed  grass-herbs-shrub  areas  in  the  Pacific  coastal 
lowlands  and  extending  up  to  the  mesic  boreal-tropical  ecotone  on 
the  Pacific-facing  slopes  of  the  Sierra  Madre  Occidental,  the  Sierra 
Madre  del  Sur,  and  where  these  mountain  chains  join  at  the  western 
end  of  the  Trans-Mexican  Volcanic  Belt  ( see  Fig.  6. )    It  is  recorded 


Baker — Sigmodon  fulviventer  Group 


187 


from  near  sea  level  in  Nayarit,  Colima,  and  Oaxaca  to  as  high  as 
3050  meters  in  Jalisco. 

Uahiiat  and  Habits 

The  brown  cotton  rat  seems  to  pref,er  moist,  vine-  and  shrnb- 
co\ered  slopes  in  tropical  or  mixed  tropical-boreal  situations,  and 
thereby  occupies  parts  of  the  Tropical  Deciduous  Forest  and  humid 
Pine-Oak  Forest  (vegetation  zones  of  Leopold,  1959).   My  acquain- 


FiG.  6.    Geographic  distrilKition  of  the  Isrown  cotton  rat.    1.   Sigmodoii  allcni 
alleni.    2.  Sigmodon  alleni  vulcaui.    3.  Sigmodon  alleni  planifrons. 

tance  with  this  species  is  from  only  four  localities,  although  I  spent 
time  during  each  of  several  summers  trying  to  find  S.  aUeni  at  places 
where  it  had  been  previously  taken  in  Sinaloa,  Nayarit,  Jalisco, 
Michoacan,  Guerrero,  and  Oaxaca.  At  one  locality  (in  Michoacan) 
S.  alleni  was  common,  but  at  three  others  (in  Sinaloa  and  Oaxaca) 
it  was  uncommon.  As  yet  the  habitat  requirements  of  this  species 
are  poorly  known,  although  it  is  surely  more  adapted  to  shrub 
habitat  than  to  open  grasslands.  Specific  trapping  localities  are 
discussed  below. 

Sinaloa. — At  1.7  km.  E  Santa  Lucia,  1720  ni.,  two  1:)rown  cotton  rats  were 
trapped  on  June  23,  1955.  This  locality  is  on  a  Pacific-facing  slope  in  the 
upper  foothills  of  the  Sierra  Madre  Occidental  in  tropical  deciduous  vegetation. 


188 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


just  below  the  pine-oak  belt.  Cotton  rats  were  caught  in  a  narrow  canyon  in 
dense  second-growth  shrub  of  a  fallow  corn  field.  The  writer  visited  this  place 
on  several  other  occasions  but  caught  no  other  S.  alleni,  although  one  S.  hispidiis 
was  lived-trapped  there  in  the  summer  of  1967.  Small  mammal  associates  at 
this  place  included  Liomijs  pictus,  Reitlnodontomys  fulvescens,  Peromyscus 
hoylii  and  Neotoma  mexicana. 

Michoacdn. — Twenty-five  S.  alleni  (12  males  and  13  females)  were  taken 
10  km.  W  Capacuaro,  2360  m.,  in  grazed  and  logged  pasture  with  scattered 
oaks  and  pines  as  the  prominent  woody  vegetation.  The  cotton  rats  li\'ed  in 
short,  closely-grazed,  grassy  areas  with  scattered  forbs  coxering  the  dark 
volcanic  soil.  Some  evidence  of  runways  were  found  under  shrubs  and  immedi- 
ately adjacent  to  rock  walls  (see  Fig.  7).  The  animals  were  active  both  at 
night  and  dming  daylight  hours  in  this  sparse  habitat.  Because  it  rained  during 
most  of  our  successful  trapping  period  at  this  locality,  I  judge  that  cotton  rat 
movements  were  not  curtailed  by  precipitation.  Several  were  seen  moving 
across  open,  grazed  turf  bet\veen  clumps  of  forbs  or  shrubs.  Nests,  presumed 
to  be  constructed  by  cotton  rats,  were  found  in  rotting  pine  logs,  imder  volcanic 
rocks,  beneath  a  pile  of  pine  bark,  and  in  a  rotted  pine  stump,  out  of  which 


Fig.  7.    Habitat  oi  the  brown  cotton  rat,  Sii^iitodoii  allciii,  near  Capacuaro,  2360 
m.,  Michoacan.    Photograph  taken  on  July  22,  1967. 


a  cotton  rat  was  flushed.  Grass  used  as  nest  material  and  also  found  as  cuttings 
along  runways  was  identified  as  Eiagrostis  limhata.  Owing  to  heavy  grazing  by 
cattle,  this  grass  was  able  to  mature  only  in  protected  clumps  of  shrubs  or 
adjacent  to  rock  fences.  Other  plants  collected  were  species  of  the  genera 
Hypiis  and  Ltipiniis,  and  Sohmiim  eJacagnifoliinn.  Of  interest  here  is  that  25 
S.  aUcni  were  ]i\e-trapped  in  two  nights  (July  20-21,  1967)  in  a  sparsely- 
vegetated  area  of  mixed  pasture  and  fallow  fields  of  less  than  25  hectares.    The 


Baker — Sigmodon  fulviventer  Group  1(S9 

only  other  small  mammal  captured  in  the  ]i\e-traps  was  Pcromyscus  hotjlU.  This 
ahility  of  S.  allcni  to  li\e  in  an  open  area  with  \egetati\e  co\er  consisting  of 
forhs  and  shrubs  is  in  marked  contrast  to  the  little-disturbed  grassy  situations 
required  as  living  places  by  S.  fulviventer  and  S.  ochrognathus. 

At  Dos  Aguas,  2135  m.,  Hooper  (  1961,  1962)  found  S.  aUeni  in  growths  of 
grass,  poison  ivy,  other  herbs,  and  shrubs  in  an'  area  of  pine  forest  mixed  with 
some  broad-leafed  oaks,  alder,  and  madrone.  Tree  trunks  and  branches  were 
draped  with  lichens,  bromeliads,  and  orchids. 

Jalisco. — J.  H.  Batty  collected  29  brown  cotton  rats  on  Volcan  de  Fuego 
".  .  .  at  an  altitude  of  about  10,000  feet,  July  10  to  28,  1905"  (Allen,  1906: 
248).  The  exact  hal)itat  from  which  these  rodents  were  taken  is  imknown. 
Previously  in  1892,  Nelson  and  Goldman  (Goldman,  1951:180)  had  ascended 
adjacent  Sierra  Nevada  de  Colima  without  taking  this  cotton  rat;  a  field  party 
from  the  Michigan  State  University  in  196.3  (Baker  and  Phillips,  1965)  also 
climbed  to  the  fir  belt  on  this  mountain  without  finding  S.  alh'ui,  although 
S.  hispidus  was  taken  along  grassy  fence  rows  on  the  lower,  southeastern 
slope.  J.  Knox  Jones,  Jr.  (personal  communication)  reported  that  a  field  party 
from  the  University  of  Kansas  Museimi  of  Natural  History  ascended  the 
Volcan  de  Fuego  in  1966  without  finding  brown  cotton  rats. 

Guerrero. — Nelson  and  Goldman  (Goldman,  1951:152)  obtained  the  brown 
cotton  rat  at  Omilteme  and  subseciuently  named  it  S.  giierrerensis.  In  1964,  a 
field  party  from  Michigan  State  University  failed  to  reach  this  place  owing  to 
rains  in  the  mountains  but  did  collect,  without  finding  brown  cotton  rats,  in 
"similar"  montane  country  near  Xochipila,  somewhat  to  the  north  of  Omilteme. 
After  spending  several  summers  having  little  success  in  finding  cotton  rats  at 
localities  visited  more  than  60  years  earlier  by  Nelson  and  Goldman,  I  have 
concluded  that  the  scarcity  or  absence  of  this  species  is  because  many  of  the 
habitats  in  which  it  lived  have  been  altered  by  man's  land  uses  since  the  turn 
of  the  century.  This  change  has  presumably  "eliminated"  some  of  the  pre- 
ferred living  places  of  these  cotton  rats. 

Oaxaca. — The  brown  cotton  rat  was  live-trapped  on  the  Pacific  side  of  the 
Sierra  Madre  del  Sur  at  approximately  13  km.  SSW  Juchatengo,  1920  m.  This 
locality,  on  the  road  between  Oaxaca  and  Puerto  Escondido,  is  approximately 
200  meters  below  the  seawardmost  crest  of  the  mountains  and  18  kilometers 
east  of  Juquila,  the  type  locality  of  S.  planifron.s.  The  collecting  site  was  in  a 
transition  between  the  pine-oak  montane  forests  and  the  tropical  semi-deciduous 
forests,  the  latter  being  mostly  in  protected,  well-watered  canyons  (see  Baker 
and  Womochel,  1966).  Cotton  rats  were  live-trapped  in  second-growth  vege- 
tation, often  thick  and  tangled,  in  fallow,  hillside  cornfields.  Plants  growing  on 
these  abandoned  fields  belong  to  species  in  such  genera  as  Rttbiis,  Lupinus, 
Adianium,  Castilleia,  Solarium,  Acalypha,  Valeriana,  Geranium,  Cerastium, 
Cyperus,  Hedyotis,  Bomarea,  Arbutus,  and  Tradescantia,  and  the  species 
PJiytolacca  decandra,  Zeugites  mexicana,  Toxidendron  radicans,  and  Ostrya 
virginiana.  Small  mammal  associates  obtained  were  Marmosa  mexicana,  Liomys 
irroratus,  Onjzomys  alfaroi,  Reithrodontomys  sumichrasti,  Peromyscus  megalops, 
Peromyscus  evides,  and  Neotoma  mexicana.  On  returning  to  the  same  locality 
three  years  later  (in  1967),  I  was  unable  to  catch  any  more  cotton  rats. 
Nowhere  in  Oaxaca  have  collectors  obtained  a  large  series  of  this  rodent; 
George  Goodwin    (personal  communication)    reported  six  specimens   taken  at 


190  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Santa  Lucia  as  the  largest  number.  I  took  four  brown  cotton  rats  near 
Juchatengo;  near  Capacuaro,  Michoacan,  we  caught  25  specimens  of  S.  aUeni 
along  with  one  small  mammal  associate,  Peromijscus  botjlii. 

Two  young  S.  alleni  were  captured  ali\'e  in  a  grassy  roadside  ditch  next  to 
a  poorly-drained  palm  jungle  8  km.  ESE  Rio  Grande,  ele\ation  less  than  30  in., 
in  July,  1967.  This  low,  moist  area  adjoined  a  coastal  mangrove  swamp.  Taken 
at  the  same  place  were  Liomtjs  pictiis,  Tijlomijs  niidicaudus,  Peromijscus  mexi- 
canus,  and  Neotoma  mexicana. 

Association  of  brown  cotton  rats  witli  other  species  of  ilie  genus. — There  is 
no  report  of  S.  alleni  li\'ing  in  ecological  association  with  S.  fulviventer,  S. 
leticotis,  or  S.  ochrognathiis.  On  the  western  slopes  of  the  Sierra  Madre  Occi- 
dental in  Durango  (see  Fig.  4)  and  southeastward,  the  ranges  of  S.  alleni  and 
S.  leticotis  might  meet  at  the  junction  of  the  pine-oak  belt  and  the  tropical 
habitat.  It  is  also  possible  that  S.  alleni  and  S.  fulviventer  may  occupy  some- 
what the  same  areas  in  the  vicinity  of  Patzcuaro  in  Michoacan,  where  Nelson 
and  Goldman  caught  a  large  series  of  the  latter  in  the  1890's  and  Hall  and 
Villa-R.  (1949:465)  took  the  former  species,  which  they  erroneously  identified 
as  S.  melanotis.  The  latter  authors  also  took  S.  hispidus  at  the  same  place. 
It  is  suspected  that  in  the  Patzcuaro  area  S.  alleni  occupies  brushy  areas,  and 
S.  fulviventer  and  S.  hispidus  occupy  grassy  situations.  I  failed  to  obtain  any 
of  the  species  directly  west  of  Patzcuaro  in  the  summer  of  1963.  Near  Santa 
Lucia  in  Sinaloa,  S.  hispidus  was  captured  in  a  run\\'ay  under  a  dense  tangle  of 
Acacia  and  Mimosa  on  a  moist,  tropical  hillside.  This  place  is  not  unlike  that 
where  S.  alleni  was  taken  nearby  a  few  years  previously.  Near  Carapan  in 
Michoacan,  S.  hispidus  was  trapped  in  open  sacaton  meadows,  whereas  S.  alleni 
was  caught  in  overgrazed  shrubs  not  many  kilometers  away  near  Capacuaro. 
Since  S.  hispidus  has  proved  to  be  an  aggressive  species  elsewhere,  such  as 
extending  its  range  northward  from  the  Oklahoma-Kansas  line  to  Nebraska 
in  a  space  of  about  56  years  (Cockrum,  1948,  and  Jones,  1964:212),  there  is 
good  reason  to  believe  that  S.  hispidus  may  usurp  habitat  (possibly  altered 
by  man)  formerly  inhabited  by  other  species  of  cotton  rats. 

Specific  Characters  and  Comparisons 
A  rich,  brownish  dorsum  together  with  medium  size  ( for  captive 
animals,  maximum  weights  are  178  grams  for  a  male  and  180  for  a 
non-pregnant  female;  ma.ximum  lengths  of  head  and  body  are  173 
mm.  and  183,  respectively)  and  strongly  recurved  incisors  (opistho- 
dont)  distinguish  S.  alleni  from  other  species  in  the  S.  fulviventer 
group.  Other  unique  characters  include  skull  with  flattened  appear- 
ance when  viewed  laterally,  bulge  of  capsular  projections  for  the 
upper  incisors  slight,  and  paraoccipital  processes,  when  viewed  from 
below,  slightly  hooked  rather  than  straight  or  curved. 

From  S.  fulviventer,  S.  alleni  differs  further  in  having  underparts 
washed  with  \\'hitish  or  pale  buff  rather  than  tawny;  skull  long  and 
narrow  instead  of  short  and  broad;  auditory  bullae  smaller  in  rela- 
tion to  breadth;  basioccipital  short  and  wide  instead  of  long  and 
narrow;  posterior  ends  of  incisive  foramina  usually  not  extending 


Baker — Sigmodon  fulviventer  Group  191 

(rather  than  extending)  to  a  hnc  drawn  between  the  anterior  sur- 
faces of  the  first  upper  molars;  anterior  end  of  mesopterygoid  fossa 
broad  instead  of  narrow;  anterior  Hp  of  foramen  magnum  obviously 
(rather  than  slightly)  notched;  foramen  ovale  small  (no  more  than 
half  width  of  M3)  instead  of  large  (at  least  three-fourths  width  of 
M3);  palatal  pits  shallow  rather  than  deep;  and  median  keel  of 
palate  slightly  developed  instead  of  well  de\  eloped. 

From  S.  leticotis,  S.  alleni  is  further  distinguished  by  hairs  of 
inside  of  ear  not  whitish  in  contrast  to  rest  of  dorsum;  skull  long 
and  narrow  instead  of  short  and  broad;  auditory  bullae  smaller  in 
relation  to  breadth;  basioccipital  short  and  wide  instead  of  long  and 
narrow;  posterior  ends  of  incisive  foramina  usually  not  extending 
(rather  than  extending)  to  a  line  drawn  between  the  anterior  sur- 
faces of  the  first  upper  molars;  anterior  end  of  mesopterygoid  fossa 
broad  rather  than  parallel-sided;  length  at  midline  of  interparietal 
more  (instead  of  less)  than  2  mm.;  anterior  lip  of  foramen  magnum 
obviously  (rather  than  slightly)  notched;  rostral  depressions  on 
sides  of  premaxillae  slight  instead  of  deep  and  pronounced;  angular 
process  of  lower  jaw  rounded  instead  of  slightly  hooked;  lingual 
root  of  first  lower  molar  large  rather  than  reduced  or  absent. 

From  S.  ochrognathus,  S.  aUeni  differs  further  in  nose  lacking 
extensive  and  contrasting  yellow  coloring;  skull  long  and  narrow 
instead  of  short  and  broad;  auditory  bullae  larger  and  broader; 
median  keel  on  basioccipital  slight  instead  of  obvious;  bulges  of 
capsular  projections  for  upper  incisors  slight  rather  than  pronounced; 
median-posterior  area  of  interparietal  lacking  distinct  notch;  and 
paraoccipital  process  (from  ventral  view)  slightly  hooked  rather 
than  curved  with  a  basal  notch. 

From  S.  liispidus,  S.  alleni  is  distinguished  by  small  size  of  tail 
scales  (0.5  mm.  wide  rather  than  0.75  mm.  wide);  tail  heavily  haired 
instead  of  sparsely  haired;  incisors  usually  more  strongly  recurved 
( opisthodont ) ;  basioccipital  short  in  relation  to  breadth  rather  than 
long;  paraoccipital  processes  (from  ventral  view)  slightly  hooked 
instead  of  generally  straight;  and  palatal  pits  moderately  deep  rather 
than  shallow. 

Geographic  Variation 

Brown  cotton  rats  occur  along  the  Pacific-facing  mountains  and 
foothills  from  southern  Sinaloa  southeastward  to  eastern  Oaxaca  at 
the  Isthmus  of  Tehuantepec.  Probably  their  systematic  relationships 
have  remained  unclear  because  of  the  a\'ailability  of  only  a  few 
scattered  specimens  in  museum  collections.  These  have  been  given 
an  assortment  of  no  less  than  six  specific  names  since  1902,  when 


192  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Vernon  Bailey  named  S.  alleni  (see  Hall  and  Kelson,  1959:676-679). 
Like  S.  leucotis,  S.  aUeni  seems  to  occupy  a  disjunct  range  with  field 
collectors  being  especially  lucky  to  pick  up  more  than  one  or  two 
individuals  at  any  locality.  Field  parties  from  the  Michigan  State 
University  worked  the  Pacific  slopes  from  Sinaloa  to  Oaxaca  at- 
tempting to  gather  specimens  from  heretofore  unreported  and  inter- 
mediate localities.  Little  additional  material  was  obtained,  although 
live  animals  from  near  Capacuaro,  Michoacan,  and  from  near  Jucha- 
tengo,  Oaxaca,  were  successfully  bred  as  well  as  crossed  in  the  live 
colony  (the  findings  from  this  study  will  appear  in  a  later  report). 

Herein,  cotton  rats  previously  assigned  to  the  specific  names 
S.  gtierrerensis,  S.  macdoiigaUi,  S.  macrodon,  S.  plonifrons,  and  S. 
vulcani  all  are  arranged  as  a  single  species  under  S.  alleni,  the  oldest 
available  name.  All  share  the  common  specific  characteristics  as 
presented  in  the  prexious  section  and  may  differ  from  one  another 
in  minor  ways  that  demonstrate  intraspecific  geographic  variation. 

In  general,  these  cotton  rats  are  larger  and  darker  in  the  montane 
areas  and  smaller  and  paler  in  coastal  situations.  However,  popula- 
tions in  the  northwestern  part  (assigned  to  S.  a.  alleni)  and  in  the 
southeastern  part  (assigned  to  S.  a.  planifrons)  of  the  range  of  the 
species  are  both  smaller  and  paler  than  populations  from  the  central 
sector  (assigned  to  S.  a.  vulcani)  of  the  species  distribution.  To 
determine  the  degree  of  size  difference,  measurements  of  the  con- 
dylopremaxillary  lengths  of  the  wild-taken  animals  listed  in  Table  1 
were  evaluated  by  analysis  of  variance.  The  overall  differences  be- 
tween the  means  of  this  cranial  dimension  are  just  significant  at  the 
five  per  cent  level  (F^3.72).  The  degree  of  difference  between  the 
indix'idual  means  of  the  four  samples  was  then  determined  by  the 
new  multiple  range  test  (see  Table  2).  Population  means  under- 
scored by  a  common  line  in  the  table  are  not  significantly  different. 
As  suggested  above,  the  smaller  animals  representing  S.  a.  alleni 
from  the  northwestern  part  of  the  range  (Sinaloa,  Nayarit,  and 
northern  JaHsco)  and  representing  S.  a.  planifrons  from  the  south- 
eastern part  of  the  range  (Oaxaca)  are  not  significantly  different 
from  each  other,  but  are  significantly  different  from  the  two  samples 
from  the  central  part  of  the  range  (southern  Jalisco  and  Michoacan) 
that  represent  S.  a.  vulcani. 

Sigmodon  alleni  alleni  Bailey 

Sipmodon  alleni  Bailey,  Pioc.  Biol.  Soc.  Washington,  15:112,  June  2,  1902. 

Type. — Young  adult  male,  skin  and  skull;  no.  88227  U.S.  National  Museum; 
from  San  Sebastian,  Mascota,  Jalisco;  obtained  on  March  15,  1897,  by  E.  W. 
Nelson  and  E.  A.  Goldman,  original  no.  10708. 


Baker — Sigmodon  fulviventer  Group 


193 


g 

173 


< 

a 

z 
o 

H 
H 
O 
U 

Z 

o 

03 
H 

Q 
< 

O 

z 

o 

O 


« 

H 
X 
H 

a 
z 

W 

o 

EC 

w 


z 

H 
< 


Q 
U 
H 

O 


w 

aa 


C  cS  5 

I'd 
< 


tc2 


•S8 


c'V 
N 


IP 

S  s  «i 


o 


ll-S 

■Sft  c 
S  K  c 


II: 


Mug 
1-4  O  C 

re 


O  D-S 

"re  5  * 


5 
o 

Sb 

C/3 


Tj< 

i> 

^ 

o 

lO  ^ 

(M  i3 

1 

05 

coco 

ceo 

irj 

CD 

cc 

^2 


°^co 

00 


CO 

^lO 

ino 

^' 

lO 

co2 

s 

2-t 

a 

'^.  CO     ~ 


CO 


CO 


<N, 


oco 
cot^ 


in, 


CO 


00 


o 


a 

C/3 


fM  > 


'^ 

in 

coed 

T^cd 

COOT 

cdtN 

CO 

CO 

t~ 

^J^ 

-f 

oo2 

2q 

2-1 

2^1 

Ol 


Ol 


GO 

^H 

^2 

2  CO 

23  CO 

ci 

cq 

Xl 

Gi 

^2 

-^2 

So 

2t- 

o 

c? 

t-H 

'^ 

t- 

CD 

coin 

coin 

-t  in 

incq 

in^ 

inVi 

in 

in 

in 

in 

q2 
2^^ 

00 


co' 

<M  rn 

CO  "^ 


cc 


CO 


2co   2^ 

00  CO 


«=>  = 

q2 

2o 

oi 

CO 

CD 

^in 

coin 

ino> 

in'csi 

— t* 

in 

CO 

t- 

S 

oc 

-^2 
2^ 

t^      ~ 


X         ^ 


in  CO 
co^. 


CO 


CO      ?- 


^CO 
CO  H 


Cj       CD 


CO 


CO 


in 


o 

o 

^- 

o 

o 

^— ^ 

CO 

o 

co 

CO 

CC 

o 

CO 

(N 

c^ 

^ 

s 

CO 

oq 

1 

-^CM 

o 

cJ5 

of^i 

6 

1 

<^6 

fM 

c^id) 

C<l 

Ol 

p— ( 

C^  ro 

<M  CO       oi  CO 


CO, 


CO 


CO 


CO 


m 

cr> 

m 

CD 

S-^ 

S-=^ 

in   1 

CO     , 

^o 

— <o 

'^i 

m 

o 

QJ    O 

^§4  M 


C 

a 
o 

o  00 


CO 


O 


O]C0 

CO 


■-H  CO 


«2 

o  c 

CO 
MO 


CO  CO 

o 


CO 


CO 
o 

CO 


ro  CO 

CO 


CO";" 

i-HO 

CO 


„in 

aj  cj 
c   X 


-C-^      =6 


00 

o 
.  n 

Vi      '• 
0)    CO 

"  C5 


C      QJ 


OJ  '^ 

5  E 


Oj 

— 

2 

^ 

s 

c 

h.' 

o 

'3 
> 

-3 

c 

c 

^ 

s 

C 

^" 

.^ 

0) 

c 

.^ 

s 

'3 

0; 

*^ 

o 

p 

1  . 

« 

S 

c 

*c5 

0^ 


~  ^,  a 
=  i!  ~ 
<  2  m 


194  Misc.  Publ.  51,  Unr'.  Kansas  Mus.  Nat.  Hist. 

Range. — Mixed  grass-forb-shrub  areas  on  western  slopes  of  the  Sierra  Madre 
Occidental  and  adjacent  Pacific  coastal  lowlands  from  southern  Sinaloa  south- 
ward to  northwestern  Jalisco   (see  Fig.  6). 

Diagnosis. — Size  small  for  the  species  ( see  Table  1 ) ;  dorsum  intermixed 
with  black  hairs  and  agouti-banded  hairs  of  Ochraceous-Tawny;  underparts 
and  upper  surfaces  of  feet  whitish  or  pale  buff;  base  of  tail  slightly  darker  than 
dorsum;  skull  short  and  broad  with  short  rostnmi  and  small  auditory  bullae. 

Comparisons. — From  S.  a.  vulcani,  S.  a.  alleni  differs  in  size  smaller  (see 
Talile  1 ) ;  color  paler  ( upperparts  Ochraceous-Tawny  rather  than  Clay  Color ) ; 
skull  smaller  with  flatter  cranium,  shorter  rostrum,  and  less  expanded  auditory 
bullae. 

Remarks. — This  northernmost  subspecies  of  S.  alleni  is  characteristically 
pale  in  color  and  smaller  than  adjoining  S.  a.  vulcani.  Specimens  have  been 
taken  at  elevations  from  near  sea  level  at  San  Bias  up  to  1723  meters  near  Santa 
Lucia.  In  its  coastal  habitat,  S.  a.  alleni  probably  comes  into  some  contact  with 
S.  hispidiis.  At  Santa  Lucia,  alleni  was  caught  in  hillside  fallow-field  habitat 
in  1955  and  hispidus  was  taken  in  1967  in  approximately  the  same  place. 

S))ecimens  examined  (20). — Sixaloa:  2  km.  E  Santa  Lucia,  1723  m.,  2 
(KU);  Copala,  4  (LACM).  Nayarit:  San  Bias,  1  (US);  Tepic,  2  (US); 
Valle  de  Banderas,  1  (US).   Jalisco:    San  Sebastian,  10  (US). 

Sigmodon  alleni  vulcani  J.  A.  Allen 
Sigmodon  vulcani  J.  A.  Allen,  Bull.  Amer.  Mus.  Nat.   Hist.,  22:247,  July  25, 

1906. 
Sigmodon  guerrerensis  Nelson  and  Goldman,  Proc.  Biol.  Soc.  Washington,  46: 

196,  October  26,  1933,  type  from  Omilteme,  2440  m.,  Guerrero. 

Type. — Young  adult  female,  skin  and  skull;  no.  26310  American  Museimi  of 
Natinal  History;  from  Volcan  de  Fuego,  3050  m.,  Jalisco;  obtained  on  July  25, 
1905,  by  J.  H.  Batty,  original  no.  2330. 

Range. — Western  slopes  of  Sierra  Madre  Occidental  and  Sierra  Madre  del 
Sur  and  coastal  plain  from  southwestern  Jalisco  southeastward  to  southern 
Guerrero  (see  Fig.  6). 

Diagnosis. — Size  large  for  the  species  ( see  Table  1 ) ;  dorsum  intermixed 
with  black  hairs  and  agouti-banded  hairs  of  Clay  Color;  underparts  and  upper 
surfaces  of  feet  whitish  or  pale  buff;  base  of  tail  slightly  darker  than  dorsum; 
skull  large  and  broad  with  cranium  abruptly  expanding  posteriorly;  auditory 
bullae  large. 


Table  2. — Result  of  new  multiple  range  test  on  condylopremaxillary 

LENGTHS    OF    CRANIA    OF    SAMPLES     OF     Sigmodon    olleili.       POPULATION     MEANS 
UNDERSCORED  BY  A  COMMON  LINE  ARE   NOT   SIGNIFICANTLY   DIFFERENT. 


s. 

a.  planifrons 

S.  a.  alleni 

S.  a. 

vulcani 

Locality 

Tehuan- 
tepec, 
Oaxaca 

Sinaloa,  Naya- 
rit, northern 
Jalisco 

Volcan  de 
Fuego, 
Jalisco 

Michoacan 

Means  of  condylo- 
jremaxillary 
engths 

30.80 

31.10 

32.18 

32.53 

Baker — Sigmodon  fulviventer  Group  195 

Comparisons. — For  comparison  with  S.  a.  alleni,  see  account  of  that  sub- 
species. From  S.  a.  planifrons,  S.  a.  vtilcani  dilFers  in  size  larger;  color  of 
upperparts  darker  (near  Clay  Color  rather  than  Cinnamon-Buff);  skull  larger 
and  broader;  skull  flatter;  nasals  longer;  incisors  less  recur\ed. 

Remarks. — This  large,  dark  subspecies  occurs  from  near  sea  level  in  Colima 
to  a])o\e  3000  meters  in  Guerrero.  It  occupies  a  variety  of  hal^itats  from  mesic, 
l)orcal-tropical,  montane  forest-shrub  to  coastal  shrub-vine  areas.  In  many 
places  its  distribution  seems  to  be  interdigitated  with  populations  of  S.  hispidus. 

Speciryiens  examined  (67). — Jalisco:  10  km.  SSW  Autlan,  1372  m.,  2 
(UM);  9  km.  NNW  Barro  de  Navidad,  2  (KU);  Volcan  de  Fuego,  3050  m., 
36  (AMNH).  Colima:  3  km.  E  Santiago,  1  (KU).  Michoacax:  3  km. 
W  Patzcuaro,  2380  m.,  2  (MVZ);  6  km.  S  Pdtzcuaro,  2350  m.,  2  (MVZ); 
14  km.  E  on  road  from  Angahuan,  2300  m.,  2  (UM);  i.5  km.  E  on  road  from 
Anfiahuan,  2300  m..  2  (UM);  10  km.  W  Capacuaro,  2059  m.,  3  (MSU);  9  km. 
N  Uruapan,  1  (UM);  7  km.  N  and  2  km.  W  Uruapan,  1937  m.,  1  (KU); 
Urtiapan,  Cupatitzio  National  Park,  1  (UM);  23  km.  W  Dos  Aguas,  2135  m., 
5  (UM);  Dos  Aguas,  2135  m.,  3  (UM).  Guerrero:  Omilteme,  2440  m., 
1  (UM),2  (US). 

Other  record. — Michoacax:  Tancitaro,  1830  m.  (Hall  and  Villa-R.,  1949: 
465). 

Sigmodon  alleni  planifrons  Nelson  and  Goldman 
Sigmodon  planifrons  Nelson  and  Goldman,  Proc.  Biol.  Soc.  Washington,  46:197, 

October  26,  1933. 
Sigmodon  planifrons  minor  Goodwin,  Amer.  Mus.  Novit.,  1705:1,  Fel)ruary  4, 

1955,  type  from  Santa  Lucia,  1220  m.,  12  km.  NE  Tenango,  Tehuantepec, 

Oaxaca.  Not  Sigmodon  minor  Gidley,  1922. 
Sigmodon  macdougalli  Goodwin,  Amer.  Mus.  Novit.,  1705:3,  February  4,  1955, 

type  from  Santo  Tomas  Teipan  (rain  forest  abo\e  village),  2135  m.,  12  km. 

S  San  Bartolo  Yautepec,  Yautepec,  Oaxaca. 
Sigmodon  macrodon  Goodwin,  Amer.  Mus.  No\it..   1705:4,  February  4,   1955, 

type  from  Cerro  San  Pedro  (rocky  summit),  1098  m.,  20  km.  W  Mixtequil- 

la,  Tehuantepec,  Oaxaca. 
Sigmodon  planifrons  setzeri  Goodwin,  Jour.  Mamm.,  40:447,  August  20,  1959, 

a  renaming  of  S.  p.  minor  Goodwin. 

Type. — Young  adult  female,  skin  and  skull;  no.  71918  U.S.  National  Mu- 
seum; from  Juquila,  1525  m.,  Oaxaca;  olitained  on  February  28,  1895,  by 
E.  W.  Nelson  and  E.  A.  Goldman,  original  no.  7569. 

Range. — Western  slopes  of  the  Sierra  Madre  del  Sur  and  adjacent  Pacific 
coastal  lowlands  of  southern  Oaxaca  eastward  to  the  Isthmus  of  Tehuanatepec 
(see  Fig.  6). 

Diagnosis. — Size  small  for  the  species  ( see  Table  1 ) ;  dorsum  intermixed 
with  Ijlack  hairs  and  agouti-banded  hairs  of  near  (a)  Cinnamon-Buft';  imder- 
parts  and  upper  surfaces  of  feet  whitish  or  pale  buff;  base  of  tail  slightly  darker 
than  dorsum;  skull  small  and  flat;  rostrum  depressed;  incisors  markedly  recurved. 

Comparisons. — For  comparison  with  S.  a.  vidcani,  see  account  of  that  sub- 
species. 

Remarks. — This  subspecies  is  smaller  (especially  those  from  Tehuantepec) 
and  .shghtly  less  richly  colored  than  S.  a.   vulcani,  its  relative  directly  to  the 


196  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

northwest.  Although  much  of  the  available  study  material  consists  of  young 
animals,  the  few  adults  present  show  that  there  is  considerable  variation 
between  populations  in  coastal  Oaxaca.  Animals  from  lower  elevations  are 
smaller  and  those  from  mesic  tropical  situations  on  the  higher  slopes  seem 
larger.  Goodwin  ( 1955 )  detected  some  of  this  \'ariation  and  quite  reasonably 
named  three  taxa  from  the  area  of  Tehuanatepec.  However,  a  comparison  of 
his  material  with  the  recently-obtained  and  laboratory-raised  animals  from 
near  Juchatengo  shows  that  the  differences  are  not  of  sufficient  magnitude  to 
warrant  retention  of  these  taxa  and  that  all  of  the  populations  should  be 
arranged  imder  the  oldest  name,  S.  a.  planifroiifi. 

The  arid  tropical  shrub  habitat  in  coastal  areas  of  Oaxaca  is  chiefly  the 
home  of  S.  hispidus,  with  grayish  S.  h.  mascotensis  to  the  westward  and 
brownish  S.  h.  ischyrus  to  the  eastward.  The  distributional  relationship  between 
S.  alleni  and  S.  hispidus  in  this  area  is  obscure,  although  S.  allcni  in  the  low- 
lands can  be  expected  in  highly  mesic  situations,  such  as  palm  forests  near 
mangrove.  The  brownish  S.  h.  ischyrus  resembles  S.  alleni  in  color  and  in  tail 
hair,  but  is  larger  with  a  longer  tail  and  shorter,  thinner  pelage. 

Specimens  examined  (29). — Oaxaca:  Santo  Tomas  Teipan,  21.35  m.,  12 
km.  S  San  Bartolo  Yautepec,  Yautepec,  1  (AMNH);  Arroyo  Palmar  (Tequi- 
sistldn),  Tehuantepec,  3  (AMNH);  Arroyo  Arenal,  Tehuantepec,  3  (AMNH); 
Cerro  PoUe,  Tehuantepec,  3  (AMNH);  Cerro  del  Chorro,  Tehuantepec,  1 
(AMNH);  Cerro  de  Pastle,  Tenango,  Tehuantepec,  1  (AMNH);  Cerro  Ocate, 
Tenango,  Tehuantepec,  1  (AMNH);  Llano  de  Ocate,  Tehuantepec,  2  (AMNH); 
Tres  Cruces,  Tehuantepec,  5  (AMNH);  Santa  Lucia,  1220  m.,  1  (AMNH); 
Cerro  San  Pedro,  1921  m.,  20  km.  W  Mixtequilla,  1  (AMNH);  40  km.  SSE 
Miahuatlan,  1647  m.,  1  (CAS);  2  km.  NNW  Soledad,  14.33  m.,  2  (KU);  13 
km.  SSW  Juchatengo,  1921  m.,  1  (MSU);  Juquila,  1525  m.,  2  (US);  8  km. 
ESE  Rio  Grande,  30  m.,  1  (MSU). 

Sigmodon  fulviventer 

Habitat  and  Habits 
The  tawny-bellied  cotton  rat  (Fig.  8)  is  an  inhabitant  of  the 
mesquite-grassland  that  occurs  in  a  north-northwest  to  south-south- 
east direction  along  the  eastern  base  of  the  Sierra  Madre  Occidental 
(see  Leopold,  1959:27-28).  This  range  extends  northward  to  central 
New  Mexico  and  Arizona  and  southward  to  the  northern  base  of  the 
Transverse  Volcanic  Belt  in  Jalisco  and  northern  Michoacan  (see 
Fig.  15).  To  the  westward,  the  mesquite-grassland  merges  with  the 
pine-oak  of  the  higher  elevations  of  the  mountains  at  approximately 
2000-3000  meters.  To  the  eastward  the  growing  aridity  of  the  lower 
elevations  causes  a  gradual  change  to  the  desert.  Leopold  (loc.  cit.) 
noted  that  the  mesquite-grassland  has  been  the  major  Mexican 
habitat  for  the  pronghorn  (Antilocapra  americana),  the  mule  or  burro 
deer  (Oclocoileus  hemionus),  the  white-sided  jackrabbit  (Lepus  callo- 


Baker — Sigmodon  fulviventer  Group  197 


•*4«*i^ 


Fig.  8.  A  tawny-bellied  cotton  rat,  Sigmodon  fulviventer,  captured  near  Bo- 
quilla,   Durango,  on  July   10,    1965.    Photograph   taken  in   October,    1965,   by 

Robert  Brown. 

tis),  and  for  such  rodents  as  the  banner-tailed  kangaroo  rat  (Dipo- 
domys  spectabilis),  the  grasshopper  mouse  (Onijchomys  torridusj, 
the  northern  pygmy  mouse  (Baiomys  taylori),  the  spiny  pocket 
mouse  (Liomys  irroratus),  the  hispid  pocket  mouse  (Perognathus 
hispidus),  and  the  tawny-belHed  cotton  rat  (S.  fulviventer).  How- 
ever, as  Leopold  stated,  much  of  this  fine  grassland,  interspersed 
with  mesquite,  acacias,  cacti,  agaves,  and  herbaceous  plants,  has 
been  depleted  by  overgrazing.  Today,  habitats  suitable  for  cotton 
rats  consist  of  scattered  plots,  which  have  been  protected  from 
heavy  grazing  by  cattle,  goats  and  sheep,  along  fence  lines,  at  edges 
of  cultivated  fields,  along  rights-of-way  of  highways  and  railroads, 
and  in  situations  where  low  thorny  shrubs  protect  grasses  under- 
neath. The  reduction  in  the  quality  of  the  habitat  for  cotton  rats 
and  other  grassland  mammals  is  not  a  recent  development;  E.  A. 
Goldman  (1951:291),  who  went  to  Zacatecas  in  December,  1902, 
hoping  to  collect  topotypes  of  S.  fulviventer,  wrote:  "The  season 
had  been  a  dry  one  and  at  the  time  of  my  visit  the  hills,  overgrazed 


198 


Misc.  Publ.  51,  Uxiv.  Kansas  Mus.  Nat.  Hist. 


Fig.   9.     Runway   of   the    tawny-bellied    cotton    rat,    Sigmodon   fulvicenter,    in 
grass  belonging  to  species  of  the  genera  Bouteloua  and  Muhlenhcrgia.    Photo- 
graph taken  near  Atotonilco,  2040  m.,  Durango,  on  July  11,  1967. 


by  goats,  were  nearly  bare,  leaving  no  local  habitat  suitable  for  the 
cotton  rat." 

Tawny-bellied  cotton  rats  have  been  found  in  association  with 
bunch  grasses.  Their  well-worn  runways  may  be  completely  hidden 
in  thick,  grassy  cover  or  may  be  periodically  exposed  (see  Fig.  9) 
with  segments  of  trails  covering  distances  over  bare  ground  of  as 
much  as  a  third  of  a  meter  between  grassy  clumps.  Plant  and 
animal  associates  are  described  in  the  following  paragraphs. 

CliihuaJuia. — Ta\\ny-bellied  cotton  rats  were  found  in  open,  moderately- 
grazed  grasslands.  One  such  place  (2  km.  N  Gallego,  1610  m. )  had  hea\'y 
grass  ( Miihlenbergia  sp. )  in  poorly-drained  areas  \\  ith  niunerous  nmways 
paralleling  roadside  or  railside  ditches  (see  Fig.  10).  Other  plants  collected 
were  KiiJjnia  chlorolepis,  Engclmannia  pinnatifida,  Aphanostcplius  ramosissimiis, 
Allionia  sp.,  Baileya  mtiltiradiata,  Jatropha  niacruJUza,  Hedijotis  rubra,  and 
Evolvidtis  sericeus.  Small  mammals  obtained  in  association  with  S.  fulvivei^ter 
at  the  Gallego  locality  were  Dipodomijs  spectabdis,  Pcrognaditts  liispidiis, 
Pcrognathtis  flaviis.  Pcrognaditis  nclsnni,  Baiomys  tot/Iori.  Onychomys  torridus, 
Rcithrodontoimjs  megaloti.s,  RcitIuodo)iloi)iyf>  niontaniis,  Peroinyscus  inanicu- 
latus,  and  Neotoma  albigula.  Although  this  locality  was  \isited  on  two  occasions 
(in  1965  and  1967),  only  one  S.  liispidiis  was  captured.  This  indicates  that 
at  least  two  cotton  rats  were  present,  I)ut  S.  liispidus  \\  as  scarce  and  S.  fidviven- 
ter  was  common,  based  on  trap  success. 


Baker — Sigmodon  fulviventer  Group 


199 


il 


Fig.  10.   Ele\ated  grasslands  near  Gallego,  1610  m.,  Chihuahua.   The  dry  bunch 
grass    is    inhabited    by    die    tawny-bellied    cotton    rat,    Sigmodon    fulviventer. 

Photograph  taken  on  July  4,   1965. 


Durango. — At  3.3  km.  NE  Boquilla,  1952  m.,  tawny-bellied  cotton  rats  were 
caught  in  the  level  part  of  a  mountain  valley  floor  in  mixed  grassland,  mesquite, 
acacia,  juniper,  and  oaks  (see  Fig.  11).  Grasses  were  mostly  Miihlenbergia  sp. 
and  Aristida  sp.  Composites  present  were  Erigcron  sp.,  LeucoJene  cricoidcs, 
and  Sanvitalia  sp.  Nightshade  (Solanum  sp.)  also  was  present  in  the  vicinity 
of  runways.  Small  mammal  associates  included  Perognathus  nelsoni  and,  on 
adjacent  hillsides,  S.  ochrognathus.  At  12  km.  NNE  Boquilla,  1890  m.,  a 
locality  just  abo\e  the  Durangan  grassland  plains  to  the  eastward  but  \\ithin 
a  scattered  stand  of  juniper,  acacia,  Mahonia  thfoliaia,  and  mesquite,  S.  ftd- 
viventer  was  captured  in  runways  in  clumps  of  grass  ( Mtdilenbergia  sp. )  and 
scattered  composites,  Baileya  sp.  and  Haplopapptis  spinulosus.  Small  mammal 
associates  included  Notiosoiex  craicfordi  (see  Baker,  1966:345),  Perognathus 
jiavus,  Perognathus  nelsoni,  Perognathus  hispidus,  Reithrodontonnjs  megalotis, 
Onycliomys  torridus,  Baiomys  taylori,  and  (on  adjacent  hillsides)  S.  ochrogna- 
thus. At  9  km.  NNW  Canatlan,  1952  m.,  S.  fidviventer  was  caught  in  runways 
in  a  dense  stand  of  grasses,  Bouteloua  gracilis  and  Muhlenbergia  sp.,  which 
were  protected  from  intense  grazing  by  an  overhead  growth  of  low  acacia. 
Also  present  were  Verbena  sp.,  Perymenium  sp.,  and  Haplopappus  spinulosus. 
Runways  led  from  one  clump  of  grass  to  another,  tlirough  bunches  of  prickl>' 
pear  (Opuntia  sp.),  among  rocks,  and  between  roots  of  huisache  and  acacia. 
One  S.  fidviventer  was  dug  out  of  a  burrow,  which  was  appro.ximately  three- 
fourths  of  a  meter  long  with  a  tunnel  diameter  of  45  mm.  Other  mammals 
caught  in  these  runways  were  Liomys  irroratus,  Perognathus  nelsoni,  Reithro- 


200 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Fig.    11.     Field   camp   in   oak   grove   near   Boquilla,    1952   m.,    Durango.     The 

sparsely-vegetated  hillsides  are  occupied  by  the  yellow-nosed  cotton  rat,   Sig- 

yuodou    ochrognathus,   and   the   \alley   floor   by    the    tawny-bellied   cotton   rat, 

Sigmodon  fidviventer.    Photograph  taken  on  July  1,  1965. 


dontomtj.s  megalotis,  Peromyscus  pectoralis,  Baiomys  taylori,  and,  on  adjacent 
hillside.s,  S.  ochrognaihus. 

In  the  vicinity  of  Hacienda  Coyotes,  2475  ni.,  tawny-bellied  cotton  rats 
were  trapped  in  open  grassy  meadows  surrounded  by  pine-oak  vegetation 
characteristic  of  the  higher  parts  of  the  Sierra  Madre  Occidental.  These 
rodents  and  their  runways  were  most  common  in  bunch  grasses  adjacent  to 
rock  fences,  and  seemed  to  a%oid  stands  of  tall  sacaton  grass,  in  which 
Microtus  mcxicanus  was  the  only  grass-eating  rodent  captured.  The  meadow 
was  dotted  with  plants  belonging  to  species  of  the  genera  Verbena,  Cnapha- 
lium.  Achactogcron.  Commcliua,  Hctcrodieca,  and  Ranuucultis,  and  to  Achillae 
lamdosa  and  Taraxacum  ufficinale.  Other  small  mammals  caught  in  cotton  rat 
runways  were  Reithrodontomijs  megalotis,  Peromyscus  melanotis,  and  Pero- 
myscus boylii.  Sigmodon  leucotis  and  Microtus  mexicanus  were  present,  but 
not  trapped  in  the  same  areas,  and  were  presumably  ecologically  segregated. 

At  5  km.  SE  Tepehuanes,  1789  m.,  one  S.  fidviventer  was  trapped  in  a 
small  patch  of  Johnson  grass,  Sorghum  halepense,  in  an  apple  orchard  that  was 
protected  from  grazing  by  a  rock  fence.    This  patch  also  produced  a  catch  of 


Baker — Sigmodon  fulviventer  Group 


201 


Fig.  12.    Mixed  desert  shrub  and  mesquite  grassland  near  Atotonilco,  2040  m., 

Durango.     Here,   the    tawny-ljellied   cotton    rat,   Sif^moclon    fulviventer,    is   the 

dominant  grass-eating  rodent  and  the  hispid  cotton  rat,  Sigmodon  Jiispidiis,  is 

less  common.    Photograph  taken  on  August  11,  1967. 


Liomys  inoratus,  Ferognathus  flavus,  Rcithrodoutomijs  megahtis,  Peiomij.scus 
hoylii,  and  Pewwysciis  truei.  The  latter  two  species  were  taken  adjacent  to  a 
high  canyon  wall  that  bordered  the  orchard  on  one  side. 

At  Hacienda  Atotonilco,  2040  m.,  tawny-bellied  cotton  rats  were  found  in 
distinct,  well-used  runways  (see  Figs.  9  and  12)  in  mixed  grass,  weeds,  and 
shrubs  along  an  arroyo.  Clumps  of  grass  were  identified  as  blue  gramma, 
Botitelotia  gracilis,  and  Muhlenbergia  sp.  Runways  also  led  through  tall  weeds 
of  the  genus  Viguiera.  Verbena  hipinnatifida  also  was  collected  in  the  cotton 
rat  area.  Thorn  shrubs  of  the  genus  Mimosa  made  up  the  woody  cover  along 
the  arroyo;  pods  emptied  of  the  beans  w^ere  found  in  small  piles  along  runways. 
It  was  supposed  that  S.  fulviventer  or  other  rodents  used  this  food.  Other  small 
mammals  captured  in  runways  and  adjacent  habitats  included  Notiosorex 
crawfordi,  Liomys  irroratus,  Perognatlius  flavus,  Perognatlius  hispidus,  Reith- 
rodontomys  fulvescens,  Baiomys  taylori,  and  Sigmodon  hispidus.  S.  hispidus 
was  uncommon,  and  S.  fulviventer  seemed  to  dominate  as  the  grass-eating 
njdent  in  the  community. 

Zacetccas. — One  S.  fidviventer  was  hand-caught  in  mid-afternoon  in  a  rim- 
way  in  tall  grass,  Muhlenbergia  sp.,  at  13  km.  S  Villanue\'a,  2090  m. 

Guanajuato. — Runways  of  S.  fulviventer  were  noted  in  elevated  grassland 
sinroimded  by  oak-covered  valleys  in  the  mountains  of  extreme  northwestern 
Guanajuato  at  8  km.  SW  Ibarra,  2500  m.  (see  Fig.  13).  Tliis  cotton  rat  was 
taken  in  runways  in  open  country  and  near  the  edge  of  the  oak  timber  in  mixed 
grass  and  low-growing  scrub  oak,  Quercus  sp.  Most  runways  were  through 
grasses  belonging  to  species  of  the  genera  Muhlenbergia,  Stipa  and  Bouteloua. 


202 


Misc.  Publ.  51,  Unw.  Kansas  Mus.  Nat.  Hist. 


Fig.  13.  Open  grassland  and  oak-filled  canyons  near  Ibarra,  2500  m.,  Guana- 
juato. The  mixed  grass,  forbs,  and  scrub  oak  clumps  in  the  foreground  are 
occupied  by  the  tawny-bellied  cotton  rat,  Sigmodon  fidviventer;  the  rocky 
canyon  slide  in  the  background  is  the  habitat  of  the  white-eared  cotton  rat, 
Sigmodon  leucotis.    Photograph  taken  on  July  19,  1966. 


Moderate  grazing  by  cattle  seemed  not  to  be  a  limiting  factor  on  cotton  rat 
distribution  over  this  broad  grassland.  Other  plants  identified  were  Euphorbia 
sp.,  Oenothera  sp.,  Milla  hiflora.  Dahlia  coccinea,  OxaUs  sp.,  Sisyrinchium  sp., 
Verhesina  sp.,  Calochartus  sp.,  Zemenia  sp.,  Hijpoxis  .sp.,  Castilleia  sp.,  Achae- 
togeron  sp.,  Baccharis  ghitinosa,  Linum  sp.,  Bouvardia  sp.,  and  Giitierrezia 
ghttinosa.  Other  small  mammals  taken  in  runways  were  Peromyscus  manicidatus 
and,  on  adjacent  hillsides,  S.  leucoiis.  The  skull  of  a  cotton  rat,  disgorged  by 
a  gopher  snake,  Pituophis  sp.,  was  obtained  at  this  locality.  A  nest  containing 
four  young  S.  fulviventcr  (with  eyes  not  yet  open)  was  found  on  July  21  imder 
a  large  rock  (30  by  60  centimeters),  adjacent  to  a  rock  wall  in  an  abandoned 
corral  (see  Fig.  14).  The  nest  was  of  shredded  grass  about  75  mm.  in  circum- 
ference and  50  mm.  high,  and  situated  in  a  shallow  depression  about  25  mm. 
deep  and  100  mm.  across. 

Jalisco. — Taw n\'-bellied  cotton  rats  were  captured  (mostly  by  hand)  in 
mi.xed  \egetation  in  a  fence  row  bordering  a  field  of  carrots  at  2  km.  NW 
La  Barca,  1525  m.  Grass  and  weeds  were  growing  in  piles  of  mesquite  branches, 
which  had  been  placed  over  the  vegetation  on  the  field  side  of  the  barbed-wire 
fence.  The  runways,  mostly  directly  under  the  fence  wire,  led  through  thick 
Bermuda  grass  (Cynodon  dactylon),  with  some  morning  glory  {Ipomoca  sp.) 
growing  on  the  fence.  Cuttings  of  both  morning  glory  and  Bermuda  grass 
were  found  in  rodent  trails;  a  nest  constructed  entirely  of  Bermuda  grass  \\'as 
found  along  one  runway.  The  field  part\-  visited  this  place  between  2:00  and 
4:00  p.m.  on  Jul>'  14,  1966,  set  Sherman  li\e-traps  in  runways  at  one  end  of 
the  fence  line  and,  starting  at  the  other  end,  dro\  e  any  animals  using  the  run- 


Baker — Sigmodon  fulviventer  Group 


203 


ways  toward  the  traps.  This  action  was  repeated  se\eral  times.  Either  captured 
or  obser\ed  were  the  anipliihians,  Rana  pi))icns  and  Hyhi  cxiivia,  the  reptiles, 
Sceloponts  toiqiiatiis,  Cnemidophortis  scalaris,  and  Thamnophis  sp,  and  Baiomys 
taylori.  Sigmodon  hispidus  was  captured  nearby,  within  the  \egetable  field  but 
not  in  the  fence  row. 

At  2/2  km.  W  Mazamitla,  Hooper  (1955:21)  captured  one  tawny-bellied 
cotton  rat  in  a  plant  community  consisting  of  grass,  weeds,  cactus,  scrub  oak, 
and  a  few  shrubs.  This  place  was  in  the  oak  belt,  altitudinally  just  below  the 
pine  forest.  Less  than  two  kilometers  away  he  captured  S.  Jiisjtidus  in  under- 
storx'  plants  including  grass,  blackberry,  and  Salvia. 


Fig.  14.    Nest  of  young  of  the  tawny-bellied  cotton  rat,  Sifj.inodoii  fidviventer, 
found  under  a  rock  near  Ibarra,  2500  m.,  Guanajuato.    Photograph   taken  on 

July  19,  1966. 


New  Mexico. — Mohlhenrich  ( 1961 )  found  taw  ny-bellied  cotton  rats  at  ele- 
vations of  1200  to  1920  meters,  mostly  associated  with  piiion,  oak,  and  juniper, 
and  often  found  in  swales  containing  thick  grasses  such  as  Hilaria  jamesii. 

Association  of  iaicity-hellicd  cotton  rats  witli  other  species  of  tlie  gemis. — 
Sigmodon  fulviventer  has  been  taken  in  association  with  or  immediately  adjacent 
to  S.  hispidus,  S.  leucotis,  and  S.  ochrognathus  (see  Fig.  4).    Probably  S.  /(//- 


204  Misc.  Publ.  51,  Unr'.  Kansas  Mus.  Nat.  Hist. 

viventer  and  S.  alleni  do  not  co-exist  liecause  of  their  differing  ecological 
preferences.  In  the  grasslands  on  deep  alluvial  soils  along  the  eastern  base  of 
the  Sierra  Madre  Occidental,  S.  fulviventer  dominates  or  presumably  excludes 
other  cotton  rats.  When  found  in  the  xicinity  of  either  S.  leticotis  or  S. 
oclirognathiis,  S.  fulviventer  occupies  the  level,  open  areas  lea\ing  the  periphe  al 
rocky,  sparsely-vegetated  slopes  to  the  former.  However,  where  .S.  fulviventer 
is  absent,  S.  leucotis  (in  Durango  west  of  San  Luis  and  southwest  of  ^'icente 
Guerrero)  and  S.  ochrognathus  (in  Coahuila  in  the  Sierra  del  Carmen  and 
southwest  of  Ocampo  and  in  Texas  in  the  Chisos  and  the  Davis  mountains) 
may  take  over  all  available  grassy  habitats.  Whereas  S.  fulviventer  comes  in 
contact  with  S.  leucotis  and  S.  ochrognathus  in  the  higher  foothills  and  in 
mountain  meadows  of  the  Sierra  Madre  Occidental,  S.  fulviventer  is  more  apt 
to  li\e  in  some  degree  of  association  with  S.  hi.spidus  in  the  lower,  eastern  parts 
of  its  range,  especially  where  the  mesquite  grasslands  and  the  desert  habitat 
interdigitate.  In  New  Mexico,  Mohlhenrich  ( 1961 )  found  hispid  cotton  ^ats 
more  common  in  plant  commimities  containing  Cottonwood  trees,  joint-fir,  cat- 
tail, saltbush,  mesquite,  creosote-bush,  choDa,  and  prickly  pear  and  tawny- 
bellied  cotton  rats  more  common  in  communities  containing  pinon,  juniper, 
evergreen  oak,  and  grasses  such  as  Hilaria  jamesii.  In  areas  where  both  species 
occur  in  New  Mexico,  S.  fulviventer  is  found  at  higher  elevations  with  lower 
temperatures  and  more  vegetatixe  cox'er,  whereas  S.  hispidus  occurs  at  lower 
elevations  with  higher  temperatures  and  more  sparse  vegetative  cover.  Tliis 
same  situation  seems  to  prexail  where  the  distribution  of  these  species  adjoin 
along  the  western  border  of  the  Mexican  Plateau.  In  the  northern  part  (in 
Chihuahua  near  Gallego  and  in  Durango  at  Atotonilco)  S.  fulviventer  appeared, 
at  least  from  our  trapping  records,  to  be  more  abundant  than  S.  hispidus.  In 
more  humid  Jalisco  (near  La  Barca),  however,  we  captured  almost  equal 
nimibers  of  each  (fixe  S.  fulviventer  and  four  S.  hispidus),  although  each 
species  was  from  a  different  part  of  a  cultivated  field.  It  would  seem  that 
the  ftdviventer  group  of  cotton  rats  saturates  most  suitable  habitats  in  the 
western  part  of  the  Mexican  Plateau,  in  the  adjacent  Sierra  Madre  Occidental, 
and  in  montane  grasslands  from  Trans-Pecos  Texas  and  Coahuila  westward  to 
Arizona,  Chihuahua,  and  northern  Durango.  In  southern  Zacatecas,  Jalisco, 
and  Michoacan,  where  rainfall  increases  markedly,  S.  hispidus  takes  over  more 
of  the  available  grassland  habitat,  apparently  restricting  greatly  the  areas  used 
by  S.  fulviventer  and  S.  leucotis.  In  these  states,  grassy  places  that  in  Durango 
might  ordinarily  harbor  S.  fulviventer  contain  only  S.  hispidus. 

Elevated  grasslands  in  eastern  Chihuahua,  western  Texas,  and  western 
Coahuila  seem  entirely  suitable  for  occupancy  by  S.  fulviventer,  although  now 
are  included  in  the  ranges  of  S.  ochrognathus  (in  the  higher  areas)  and  S. 
hispidus  (in  the  lower  areas).  It  would  appear  diat  S.  fulviventer  has  been 
unable  to  "cross"  desert  country  in  the  \icinity  of  the  Rio  Grande  in  New 
Mexico  and  Texas  and  in  the  Bolson  de  Mapimi  in  Chihuahua  and  Coahuila 
to  reach  these  areas  to  the  eastward.  The  success  of  S.  ochrognathus  and  the 
ubiquitous  S.  hispidus  in  occupying  these  desert  mountains  suggests  that  they, 
as  species,  can  survive  more  xeric  conditions  than  S.  fulviventer. 

Parasites. — Dr.  Robert  Traub  identified  from  S.  fulviventer  the  fleas,  Poly- 
genis  martinez-haezi   Vargas,    1951,   and  Archopcas   Icucopus  ssp.   from   near 


Baker — Sigmodon  fulviventer  Group  205 

Boquilla,  and  Picochaetis  sp.  from  near  Hda.  Coyotes.  Doran  (1955:164) 
recorded  die  nematode,  Litomosoidcs  carinii  (Travassos,  1919),  from  S.  /. 
melanotis. 

Specific  Characters  and  Comparisons 

The  "pepper  and  salt"  colored  upper  parts  and  the  tawny  under- 
parts,  coupled  with  large  size  (for  captive  animals,  maximum 
weights  are  222  grams  for  a  male  and  206  for  a  non-pregnant  female; 
maximum  lengths  of  head  and  body  are  197  mm.  and  200,  respec- 
tixely)  distinguishes  S.  fulviventer  from  other  members  of  the  S. 
fulviventer  group.  Other  unique  characters  include:  anterior  end 
of  mesopterygoid  fossa  narrow;  foramen  ovale  large  ( at  least  three- 
fourths  width  of  M3);  palatal  pits  markedly  deep;  and  median  keel 
on  palate  well  developed. 

From  S.  alleni,  S.  fulviventer  is  further  distinguished  by  skull 
more  arched,  short  and  broad  rather  than  long  and  narrow;  incisors 
less  recurved;  basioccipital  more  elongate  in  comparison  to  width; 
auditory  bullae,  in  relation  to  size  of  skull,  larger;  posterior  ends  of 
incisive  foramina  extending  (instead  of  not)  to  or  beyond  a  line 
drawn  between  the  anterior  surfaces  of  the  first  upper  molars;  and 
paraoccipital  process,  when  viewed  from  below,  straight  instead  of 
hooked. 

From  S.  leucotis,  S.  fulviventer  differs  further  in  interparietal 
more  than  (instead  of  less  than)  2  mm.  long;  rostral  depressions  on 
upper  premaxillary  slight  instead  of  pronounced;  angular  process  of 
lower  jaw  more  rounded  and  less  hooked;  and  lingual  root  of  first 
lower  molar  normal,  not  reduced  or  absent. 

From  S.  ochrognathus,  S.  fulviventer  is  further  distinguished  by 
whitish  tips  of  hairs  on  inner  side  of  pinnae  that  do  not  contrast 
with  color  of  pelage  of  rest  of  head;  auditory  bullae  large  and  broad 
rather  than  small  and  elongate;  median  keel  on  basioccipital  slight 
instead  of  obvious;  anterior  lip  of  foramen  magnum  not  obviously 
notched;  bulge  of  capsular  projections  of  upper  incisors  moderate 
rathc-r  than  pronounced;  median-posterior  notch  in  interparietal 
absent  instead  of  present;  and  paraoccipital  process  (in  \entral 
view)  straight  instead  of  curved  with  basal  notch. 

From  S.  hispidus,  S.  fulviventer  differs  in  small  size  of  tail  scales 
(0.5  mm.  wide  rather  than  0.75  wide);  tail  heavily  haired  instead  of 
sparsely  haired;  posterior  ends  of  incisive  foramina  extending  (in- 
stead of  usually  not )  to  or  beyond  a  line  drawn  between  the  anterior 
surfaces  of  the  first  upper  molars;  foramen  ovale  larger,  at  least 


206  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

three-fourths  instead  of  one-half  the  width  of  M3;  palatal  pits  more 
deeply  marked;  and  median  keel  on  palate  better  developed. 

GeograpJiic  Variation 

Tawny-bellied  cotton  rats  were  recognized  by  Bailey  (1902)  as 
belonging  to  three  distinct  species:  S.  minimus  Mearns  1894,  with 
type  locality  on  the  Mexican-New  Mexico  boundary  and  southern- 
most collecting  locality  at  Casas  Grandes,  Chihuahua;  S.  ftilviventer 
Allen,  1889,  with  type  locality  at  Zacatecas  and  northernmost  col- 
lecting locality  at  Durango,  about  800  kilometers  southward  of  Casas 
Grandes;  and  S.  melanotis  Bailey  1902,  with  type  locality  at  Patz- 
cuaro,  400  kilometers  to  the  south  of  Zacatecas.  Hall  and  Kelson 
(1959:676-677)  retained  Bailey's  arrangement  and  noted  that  S. 
minimus  was  polytypic  by  listing  subspecies  named  for  New  Mexi- 
can populations  (also  see  Findley  and  Jones,  1963)  and  citing 
records  for  S.  m.  minimus  from  northern  Durango,  thus  bridging  the 
gap  considerably  between  known  collecting  stations  of  S.  jninimus 
and  S.  fulviventer  (to  about  300  kilometers — Rosario  to  Durango). 
They  also  reported  additional  collecting  localities  to  shorten  the  gap 
between  S.  fulviventer  and  S.  melanotis  to  about  325  kilometers 
(Zacatecas  to  Mazamitla,  Jalisco). 

Baker  and  Greer  (1962:121-123)  concluded  that  S.  minimus  and 
S.  fulviventer  were  conspecific  and  arranged  taxa  previously  desig- 
nated as  S.  minitntis  under  the  older  name,  S.  fulviventer.  My  own 
field  work  in  southern  Zacatecas,  northwestern  Guanajuato,  and  in 
the  vicinity  of  Lago  de  Chapala  in  Jalisco  has  provided  specimens 
from  additional  collecting  stations  to  reduce  the  gap  between  S. 
fulviventer  and  S.  melanotis  to  less  than  175  kilometers  and  to 
demonstrate  that  tawny-bellied  cotton  rats  live  in  grassy  upland 
habitats  in  more  or  less  continuous  fashion  all  the  way  south  to  the 
northern  foothills  of  the  Trans-Mexican  Volcanic  Belt.  Since  char- 
acters distinguishing  S.  melanotis  from  S.  fulviventer  (see  accounts 
of  subspecies)  are,  in  magnitude,  no  greater  than  those  distinguish- 
ing one  subspecies  of  S.  fulviventer  from  another,  S.  melanotis  is 
arranged  as  a  subspecies  with  the  older  name,  S.  fulviventer,  being 
used  to  designate  all  of  these  tawny-bellied  cotton  rats.  Subspecies 
now  recognized  are:  Sig,moclon  fulviventer  fulviventer  J.  A.  Allen, 
Sigmodon  fulviventer  goldmani  Bailey,  Sigmodon  fulviventer  mini- 
mus Mearns,  and  Sigmodon  fulviventer  melanotis  Bailey.  S.  /. 
goldmani  is  not  treated  in  the  accounts  beyond. 


Baker — Sigmodon  fulviventer  Group 


207 


Fig.    15.    Geographic  distribution   of  tawny-bellied   cotton   rats.     1.    Sigmodon 
fulviventer  goldmani.    2.  Sigmodon  fulviventer  minimus.    3.  Sigmodon  fulviven- 
ter fulviventer.    4.  Sigmodon  fulviventer  melanotis. 

In  their  study  of  this  species  (listed  as  the  least  cotton  rat, 
Si<!,modon  niinimiis)  in  New  Mexico,  Findley  and  Jones  (1963) 
found  a  slight  cline  in  color  in  at  least  the  population  inhabiting  the 
upper  Rio  Grande  valley  but  no  definite  patterns  in  the  southern  part 
of  the  state.  Along  the  eastern  face  of  the  Sierra  Madre  Occidental 
from  Chihuahua  and  northeastern  Sonora  southeastward  to  Michoa- 
can,  there  is  a  conspicuous  color  cline  with  pale,  buff-colored  popula- 
tions in  the  arid  grasslands  of  the  north  grading  into  rich,  tawny- 
colored  populations  in  the  more  humid  country  in  the  northern  foot- 


208  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

hills  of  the  Trans-Mexican  Volcanic  Belt.  Logical  breaks  in  the  color 
pattern  in  northern  Durango  and  northern  Guanajuato-Jalisco  are 
aligned  with  changes  in  other  external  and  cranial  characteristics. 

There  is  no  attempt  here  to  compare  cranial  characters  of  the 
various  wild-taken  Mexican  .specimens  of  S.  fulviventer.  The  few 
museum  specimens  of  sufficient  maturity  to  show  wear,  as  illustrated 
in  figure  8  of  Findley  and  Jones  (1963:313),  present  a  similar  dis- 
cordance in  the  geographically  varying  features,  as  found  for  New 
Mexican  cotton  rats  of  this  species  by  Findley  and  Jones.  It  is  likely 
that  their  specimens  included  some  of  the  same  variations  due  to 
differences  in  age  as  does  the  Mexican  material.  In  Table  3  are 
presented  average  and  extreme  measurements  of  selected  external 
and  cranial  dimensions  of  100-day-old,  laboratory-born  S.  fulviven- 
ter, whose  parents  were  wild-caught  at  localities  in  each  of  four 
Mexican  states  ( from  north  to  south — Chihuahua,  Durango,  Guana- 
juato, and  Jalisco).  These  animals  were  raised  in  the  same  room 
and  in  similar  cages  and  fed  similar  amounts  and  the  same  kind  of 
food  ( Purina  Mouse  Chow ) .  Even  though  it  might  be  argued  that 
cotton  rats  whose  ancestors  came  from  an  arid  grassland  plain  north 
of  the  Tropic  of  Cancer  (at  29°  17'  N  in  Chihuahua)  might  develop, 
as  individuals,  more  slowly  than  cotton  rats  whose  ancestors  came 
from  a  much  more  humid  situation,  south  of  the  Tropic  of  Cancer 
(at  20°  17'  N  in  JaHsco),  the  dimensions  of  these  100-day-old  ani- 
mals are  used  in  this  taxonomic  comparison  along  with  those  of 
offspring  of  similar  age  resulting  from  a  cross  between  parents  from 
Chihuahua  and  from  Jalisco.  Laboratory-raised  cotton  rats  200-210 
days  old  whose  parents  came  from  Gallego  and  who  were  the  off- 
spring of  the  above-mentioned  cross  also  were  available  in  sufficient 
numbers  to  be  inserted  in  the  table  for  growth  comparisons.  It  will 
be  noted  that  200-day-old  cotton  rats  are  considerably  larger  than 
100-day-old  specimens.  There  is  no  attempt  to  compare  laboratory- 
raised  animals  with  animals  snap-trapped  in  nature,  but  it  is  possible 
from  this  demonstrated  growth  that  the  comparison  of  dimensions 
(except  for  alveolar  length  of  maxillary  toothrow)  of  "adult"  wild- 
taken  animals  may  not  be  highly  meaningful  from  a  taxonomic 
point  of  view.  Because  of  this,  comparisons  of  the  dimensions  of 
laboratory-raised  cotton  rats  of  known  ages  are  made  to  assist  in 
appraising  geographic  variation  in  S.  fulviventer. 

A  study  of  Table  3  shows  that  100-day-old  samples  of  S.  fulviven- 
ter range  in  size  both  externally  and  cranially  from  small  in  Chihua- 
hua to  large  in  Jalisco.    To  evaluate  the  extent  of  the  variation 


Baker — Sigmodon  fulviventer  Group 


209 


1=1 

C   E   r* 

^  o 

<; 


bC  en 

c  2 
►J 


So 
.S8 


N 


§11 


4-  c-c 
<u  -•  R 


fit  .  I    I  '*H 


St  v£ 


ra      y5      ^H 

M  O  f^ 


CO 

00 

^CD 

<©«? 

(OCO 

CO'* 

0 

CD 

■^-' 

^_^ 

^ 

1> 

05 

x2 

CO 

:::o'i 

1— 1 

co»o 

00 

Tt< 

^2 

c<i 

2^. 

:i^. 

05 

0 

t>io 

0^ 

Tf  CO 

10  CD 

-t 

-r 

3  00 


u 


O.  CO 


o 


C5 


O' 


CO 


10  Ol 


OcO 

CM 

CO 


.— 1 

C-1 

^co 

OCO 

^6 

cot^ 

CO 

01 

"■^ — ' 

""^ 

_,_^ 

10 

0 

10 

CO^ 

t-^ 

"^  J, 

10    , 

,-hoo 

CO 

— 1  01 

in 


o 

o 


03 

O 
O 
fM 


in 

■^  CD 

COCO 

CO 


-«  CO 

CO   r-H 


<M, 


00 

.n2 
2-. 

o 


05 


Pg 


0 

i-H 

c 

rt 

^"^ 

;-< 

3 

,^— ... 

Q 

ic 

_  c^ 

vT 

if^co 

0 

4-< 

S3  CO 

0 

CO    '• 

c 

CO 

U 

""^ — ' 

CO 
oico 

CO    ' 


o 

^^ 

IC     I 

in 


o 
o 


o 

CS 


c 
2 

6 


CO 

,CD 


CO' 


CO 


CO 


110 

'd 


05 

in»« 

-* 


00 

COOQ 

d  CT> 

05 


^H  CO 

CO 


o 
o 


CT> 
lOCO 

coco 

CO 


00 


10 

CO 


o 


ffl 


(N 


d_^ 

I — I  •. 

o 


CO 
^  CD 


05 


CD 
COro 

CO 
CO 


qa 


o 


—^     o 


CM 

a> 

00  tf 

00  CO 

dco 

cdt^ 

CO 

CO 

<M 


0}  . 

CO  . 


fM 


8    ocii 

.:2    -"d 


in  X! 


CO 


Co  05 


M      -ICO 


S     CO 


CO 
CO 


I  CM 


J2     <>ick 


c 


iniS 


^ 


in 


o 
o 


o 
o 


00 


CO 

pin 
in  CD 


CO 
»  CSl 

CM 


inco 

^^. 
m 

CO 


(Mas 


CD 

CD 

OJ 

m 

in 

-tCO 

in 

CO 

CO 

CO 

roCO 

c^oi 

CO 

1 

CO 

^H 

CO  Jh 

CO 

CO 

0 

CO 

CO 

^ 

^ 

0 

^ 

CD 

00 

(J 

t- 

CO 

t^ 

CO 

w 

0 

^u 

00 

1 — 1 

Ol 

r^ci 

CO     1 

t- 

CD 

I 

t^  -A 

^cn 

. — 1 

oi 

. — 1 

0 

-H  CD 

00 

o 

CO 
CO 

CO 


s 
o 


-a 


in 


Oi 


c« 
O 

o 

(M 


210  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Table  4. — Result  of  new  multiple  range  test  on  con dyloprem axillary 
LENGTHS  OF  CRANIA  OF  SAMPLES  OF  Sigmodoji  fiilvivetiter.    Population  means 

ITNDERSCORED  by  a  common   line   ARE   not   significantly   DIFFERENT. 


Subspecies 

S.f. 

Gallego, 
Chihuahua 

fu 

s.f. 

Ivwenter 

s.f. 

melanotic 

S.  f.  minimus- 
S.  f.  melanotis 

Locality 

Coyotes, 
Durango 

Ibarra, 
Guanajuato 

La  Barca, 
Jalisco 

Cross 
Chihuahua- Jalisco 

Means  of 
condylopr 
maxillary 
lengths 

e- 
30.7 

32.5 

34.0 

34.8 

33.9 

between  the  samples,  analysis  of  variance  was  applied  to  the  con- 
dylopremaxillary  lengths.  The  overall  differences  between  the 
means  of  this  dimension  are  highly  significant  (F=r  18.36).  A  new 
multiple  range  test  was  then  made  to  determine  the  degree  of  dif- 
ference between  the  means  from  the  selected  localities  (see  Table 
4).  Population  means  underscored  by  a  common  line  are  not  sig- 
nificantly different.  As  expected,  the  sample  from  Chihuahua  (rep- 
resenting S.  /.  minimus)  is  significantly  different  from  all  other 
samples;  the  samples  from  Durango  and  Guanajuato  (representing 
S.  /.  fulviventer)  are  not  significantly  different  from  each  other;  the 
samples  from  Guanajuato,  from  Jalisco  (representing  S.  /.  melanotis) 
and  from  a  cross  between  animals  from  Chihuahua  and  from  Jalisco 
also  are  not  significantly  different  from  one  another.  This  test 
demonstrates  that  there  are  subspecific  differences  between  these 
Mexican  populations  of  tawny-bellied  cotton  rats  and  there  is  logic 
for  arranging  them  under  three  subspecific  designations. 

Sigmodon  fulviventer  minimus  Mearns 
Sigmodon  minima  Mearns,  Proc.  U.S.  Nat.  Mus.,  17:130,  July  19,  1894. 
Sigmodon  fulviventer  minimus  Baker  and  Greer,  Michigan  State  Univ.,  Publ. 
Mus.,  Biol.  Ser.,  2:123,  August  27,  1962. 

Tijpc. — Young  adult  male,  skin  and  skull;  no.  21187/37291  U.S.  National 
Museum;  from  near  Monument  no.  40,  1500  m.,  Hidalgo  Co.,  New  Mexico,  on 
the  Mexican  boundary  line,  166  km.  W  initial  monument  on  west  bank  of  Rio 
Grande;  olitained  on  April  26,  1892,  by  Edgar  A.  Mearns  and  Frank  X. 
Holzner,  original  no.  1704. 

Range. — Elevated  grasslands  of  southeastern  Arizona,  central  and  south- 
western New  Mexico,  northeastern  Sonora,  westcentral  Ghihuahua  and  north- 
central  Durango  (.see  Fig.  15). 

Diagnosis. — Size  small  for  the  species;  dorsum  intermixed  with  l^lack  hairs 
and  agouti-banded  hairs  near  (c)  Pinkish  Buff;  underparts  and  upper  surfaces 
of  feet  and  tail  with  this  same  pale  coloring;  base  of  tail  slightly  darker,  hairs 
dark  Pinkish  Buff.    No  measurements  of  wild-taken  specimens  are  given;  con- 


Baker — Sigmodon  fulviventer  Group  211 

suit  Talile  3  for  listing  of  selected  measurements  of  laboratory-raised  animals 
from  north-central  Chihuahua    (Gallego). 

Comparisons. — To  compare  S.  /.  goldmani  of  New  Mexico  with  S.  /.  mini- 
nttts,  the  reader  may  refer  to  the  account  by  Findley  and  Jones  (1963).  From 
S.  /.  fitlviventci\  found  directly  to  the  southward,  S.  /.  minimtis  differs  in  size 
smaller  (see  Table  3);  color  paler  (near  Pinkish  Bull  rather  than  Cinnamon- 
Buff)  both  a!io\e  and  below;  zygomatic  arches  noticeably  wider  in  relation  to 
length  of  skull. 

Remarks. — Sigmodon  fuhivcnicr  minimus  is  the  smallest  and  palest  of  the 
Mexican  tawny-bellied  cotton  rats.  This  subspecies  is  identifiable  as  far  south- 
ward as  north-central  Durango  where  the  color  of  specimens  from  near  Boquil- 
las  is  almost  identical  to  that  of  animals  from  Gallego.  Specimens  from  near 
Canatlan  in  west-central  Durango  show  an  intermediate  coloring  between  the 
paler  S.  /.  tninimus  and  the  darker  S.  /.  fulviventer,  but  are  assigned  to  the 
latter  subspecies. 

Specimens  examined  (42). — Soxora:  Los  Nogales,  1  (US),  Santa  Cruz 
River,  near  Monument  no.  Ill,  6  (US).  Chihuahua:  Casas  Grandes,  2135 
m.,  1  (US);  2  km.  N  Gallego,  1366  m.,  2  (MSU).  Duraxgo:  Ro.sario,  9 
(AMNH);  Rio  Sestin,  10  (AMNH);  Guanacevi,  1  (AMNH);  Rancho  Bailon, 
5  (AMNH);  11  km.  NNE  Boquilla,  1952  m.,  7  (MSU). 

Sigmodon  fulviventer  fulviventer  J.  A.  Allen 
Sigmodon  fulviventer  J.  A.  Allen,  Bull.  Amer.  Mus.  Nat.  Hist.,  2:180,  October 

21,  1889. 

Type. — Young  adult  male,  skin  and  skull:  no.  1975/1244  American  Museum 
of  Natural  History;  from  Zacatecas,  Zacatecas;  obtained  on  August  17,  1889,  Ijy 
Audley  Buller,  original  no.  59. 

Range. — Grasslands  on  eastern  foothills  of  the  Sierra  Madre  Occidental 
from  central  Durango  southeast\vard  to  western  Guanajuato  and  southern 
Zacatecas  (see  Fig.  15). 

Diagnosis. — Size  medium  for  the  species;  dorsum  intermixed  with  black 
hairs  and  agouti-banded  hairs  of  Cinnamon  Bufl";  underparts  and  upper  sur- 
faces of  feet  and  tail  washed  with  similar  coloring;  base  oi  tail  darker,  near 
( a )  Clay  Color.  Average  and  extreme  external  and  crania!  measurements  of 
nine  wild-trapped  adults  (measurements  compare  closely  with  those  of  100- 
day-old,  laboratory-raised  animals  from  the  same  locality,  see  Table  3)  from 
Hda.  Coyotes,  Durango,  are  as  follows:  length  of  head  and  body,  150  ( 138-164); 
length  of  hind  foot,  28  (26-30);  height  of  ear  from  notch,  21.6  (20-22);  con- 
dylopremaxiUary  length,  32.5  (31.2-33.7);  zygomatic  breadth,  19.7  (18.8-20.6); 
least  interorbital  constriction,  4.9  (4.6-5.3);  interparietal  breadth,  10.8  (10.0- 
11.4);  depth  of  cranium,  10.9  (10.7-11.1);  length  of  nasals,  12.1  (10.6-12.7); 
and  alveolar  length  of  maxillary  toothrow,  6.5  (6.2-6.9). 

Comparisons. — For  comparison  with  S.  /.  minimus,  see  account  of  that  sub- 
species. From  S.  /.  melanotis,  S.  f.  fulviventer  differs  in  size  smaller  (see 
Table  3);  color  less  rich  (Cinnamon-Buff  rather  than  near  (a)  Clay  Color) 
both  above  and  below;  skull  less  massive  in  appearance  and  less  arched  in  the 
interorbital  area. 

Remarks. — Sigmodon  fulviventer  fulviventer  is  intermediate  in  both  size  and 
coloration  between  S.  /.  minimus  to  the  northward  and  S.  /.  melanotis  to  the 


212  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

southward.  There  is  evidence  of  intergradation  witli  the  former  subspecies  in 
specimens  taken  near  Canatlan  in  Durango.  These  are  more  or  less  intermediate 
in  color  but  allocated  to  S.  /.  fulviventer.  Specimens  from  an  elevated  grass- 
land near  Ibarra  in  Guanajuato  are  paler  than  typical  S.  /.  ftdvivet^ter,  but  in 
terms  of  size  show  a  tendency  toward  the  larger  S.  /.  melanotis  of  nearby 
Jalisco.  As  Table  3  shows,  the  laboratory-raised  animals  from  Ibarra  also  are 
more  or  less  intermediate  in  size  between  animals  from  Hda.  Coyotes  in 
Durango  and  from  near  La  Barca  in  Jalisco. 

Specimens  examined  (42). — Durango:  5  km.  SE  Tepehuanes,  1780  m., 
1  (MSU);  9  km.  NNW  Canatlan,  1950  m.,  2  (MSU);  4  km.  SE  Atotonilco, 
2037  m.,  1  (MSU);  5.5  km.  SE  Atotonilco,  2037  m.,  1  (MSU);  10  km.  N 
Durango,  2  (UI);  Durango,  2  (US);  Hda.  Coyotes,  2475  m.,  1  (CAS),  12 
(MSU).  Zacatecas:  Laguna  Valderama,  67  km.  W  Fresnillo,  2380  m.,  5 
(CAS);  Zacatecas,  3  (AMNH);  13  km.  S  Villanueva,  2090  m.,  1  (MSU). 
Guanajuato:  8  km.  SW  Ibarra,  2500  m.,  9  (MSU);  12  km.  SW  Ibarra,  2623 
m.,  3  (MSU). 

Sigmodon  fulviventer  melanotis  Bailey 
Sigmodon  tnelanotis  Bailey,  Proc.  Biol.  Soc.  Washington,  15:114,  June  2,  1902. 
Type. — Adult   female,   skin   and   skull;   no.   50190   U.S.   National   Museum; 
from  Patzcuaro,  2135  m.,  Michoacan;  obtained  on  July   15,    1892,  by  E.   W. 
Nelson,  original  no.  2834. 

Range. — Mesic  grasslands  of  the  southwestern  part  of  the  Mesa  Central 
( southern  part  of  the  Mexican  Plateau )  bordering  and  in  the  foothills  of  the 
northwestern  slopes  of  the  Trans-Mexican  Volcanic  Belt  (see  Fig.  15).  Dis- 
tribution in  grasslands  and  mixed  agricultural  country  spotty,  presumably  be- 
cause of  competition  with  S.  hispidus. 

Diagnosis. — Size  large  for  the  species;  dorsum  intermixed  with  black  hairs 
and  agouH-banded  hairs  near  (a)  Clay  Color;  underparts,  upper  surfaces  of 
feet  and  tail  washed  with  similar  rich  coloring;  base  of  tail  slightly  darker.  No 
measurements  of  wild-taken  specimens  are  given;  consult  Table  3  for  selected 
measurements  of  laboratory-raised  animals  from  Jalisco. 

Comparisons. — From  S.  /.  fulviventer,  which  occurs  directly  to  the  north- 
ward, S.  /.  melanotis  is  larger  and  more  richly  colored  as  indicated  in  the 
accoimt  of  the  former  subspecies. 

Remarks. — Until  now,  the  systematic  status  of  S.  melanotis  has  remained 
unchanged  since  first  described  by  Bailey  in  1902.  Its  rich  coloring  (especially 
conspicuous  on  the  underparts  of  the  specimens  from  Patzcuaro,  Michoacan) 
and  its  characteristic  "pepper  and  salt"  dorsal  appearance  set  it  apart  from 
other  cotton  rats.  Moreoxer,  its  seemingly  spotty  distribution  has  been  a 
deterrent  to  collectors.  Aside  from  the  extensive  type  series  taken  in  the 
I890's,  from  the  vicinity  of  Patzcuaro,  there  were  no  sizeable  series  assigned 
to  this  taxa  extant,  imtil  members  of  a  Michigan  State  University  Museum  field 
party  saw  a  tawny-bellied  cotton  rat  cross  the  road  near  La  Barca,  Jalisco,  in 
July  of  1966,  and  subsequently  caught  it  and  four  others.  In  other  cases  (see 
specimens  examined),  only  a  few  examples  of  these  rats  from  localities  in 
Michoacan  and  Jalisco  occur  in  Mexican  and  American  museums.  The  Patz- 
cuaro series  is  the  richest  in  color;  specimens  from  near  Zamora  and  in  the 
vicinity  of  Lago  de  Chapala  are  slightly  paler   ( tending  in  color  to  be  closer 


Baker — Sigmodon  fulviventer  Group  213 

to  Cinnainon-Bufl  than  to  Clay  Color),  which  shows  evident  relationship  with 
the  more  northward  and  paler  S.  /.  fulviventer.  Specimens  assigned  to  the 
latter  from  Ibarra  in  Guanajuato  also  are  somewhat  intermediate  in  size 
between  the  larger  S.  /.  melanotis  and  its  smaller  relatives  to  the  northward. 
Specimens  examined  (18). — Jalisco:  2  km.  NW  La  Barca,  1525  m.,  1 
(MSU);  2  km.  N  Mazamitla,  1  (UM).  Michoacax:  5  km.  S  Cumuato,  1 
(UNAM);  3  km.  E  La  Palma,  SE  side  Lago  de  Chapala,  1  (MSU);  18  km. 
E  Zamora,  1   (TCWC);  Patzcuaro,  2135  m.,  12   (US). 

Sigmodon  leucotis 

Siiimodon  leucotis,  the  white-eared  cotton  rat,  occupies  montane 
habitats  in  a  Y-shaped  distributional  pattern  (see  Fig.  16)  from 
appro.ximately  25°  N  latitude  in  the  Sierra  Madre  Occidental  (in 
the  west)  and  the  Sierra  Madre  Oriental  (in  the  east)  southeast- 
ward into  the  central  part  of  the  Trans-Mexican  Volcanic  Belt  and 
culminating  in  the  Sien-a  Madre  del  Sur  in  Oaxaca.  This  cotton  rat 
lives  in  a  comparable,  but  more  mesic,  habitat  than  does  S.  ochro- 
gnatJnis,  which  occupies  montane  areas  generally  northward  of 
25°  N  latitude. 

Hahitot  and  Habits 

The  white-eared  cotton  rat  is  strictly  a  montane  species  and 
associated  chiefly  with  mesic  pine-oak  habitat.  In  such  areas  the 
species  seems  most  adapted  to  mixed  grass  and  shrub  cover  on  shal- 
low, rocky  soils,  although  animals  also  have  been  taken  in  grassy 
meadows,  adjacent  to  streams,  in  "sacaton"  meadows,  and  in  scat- 
tered clumps  of  bunch  grass  on  dry,  rocky  slopes.  The  latter  situa- 
tion is  much  like  that  preferred  by  S.  ochrognathns,  although  these 
species  ha\e  never  been  found  together  where  their  ranges  overlap 
(in  latitude)  in  the  Sierra  Madre  Occidental  of  central  Durango. 
In  grassy  patches,  S.  leucotis  can  be  taken  in  well-used  runways, 
typical  of  other  cotton  rats.  In  low,  shrub  cover,  runways  are 
obscure  or  absent;  little  or  no  sign  may  be  present  and  the  detection 
of  the  presence  of  the  species  in  such  areas  can  be  difficult.  Such 
places,  with  exposed  rocky  ledges  and  shallow  soils,  often  appear 
to  be  more  suitable  for  species  of  Pewmyscus  and  Neotoma  than 
for  Sigmodon.  Like  S.  ochrognatlius,  S.  leucotis  does  not  seem  to 
exist  in  such  dense  populations  as  does  S.  hispidus  or  S.  fulviventer. 
Descriptions  of  collection  stations  are  given  below. 

Durango. — In  boreal,  mesic,  pine-oak  forest  30  km.  SSW  of  Tepehuanes, 
2500  m.,  two  white-eared  cotton  rats  were  trapped  in  dense  shrubs  on  a  hill- 
side coxered  with  mixed  pine,  oak,  man/anita,  and  juniper.  The  catches  were 
made  under  clumps  of  Ceanothiis  fendleri  and  scrub  oak,  Quercus  sp.  Other 
plants    in    this    thicket    included    bunch    grass    {Muhlenbergia    sp.),    Senccio 


214 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Fig.   16.    Geographic  distriljution  of  the  white-eaied  cotton  rat.    1.  Sigmodon 
leucotis  leucotis.    2.  Sigmodon  leucotis  alticola. 

actinella,  and  species  of  the  genera,  Geranium,  Cosmos,  Valeriana,  CastiUcia, 
Achaetogeron,  Lupinus,  Tradescantia  and  Cologania.  At  ground  level  there 
was  no  e\idence  of  runways  among  the  woody  stems  of  the  shruljs  or  at  the 
bases  of  grass  clumps.  No  burrows  were  found  although  one  surface  nest  of 
grass,  thought  to  be  that  of  a  cotton  rat,  was  located  at  the  base  of  a  manzanita 
bush,  ArctostapJu/los.  The  nest  measured  130  mm.  in  circumference  and  80 
mm.  in  depth.  Although  no  other  small  manunals  were  captured  in  this  bushy 
habitat,  Peromyscus  hoijlii  and  Pcromyscus  melanotis  were  taken  in  adjacent 
woody  cover  and  Thomomys  innbrinus,  Reitlirodontomy.s  megalotis,  and  Micro- 
ttis  mexicanus  in  wet  montane  meadows  dotted  with  clumps  of  sacaton  grass. 

Within  3  km.  of  San  Luis,  2300  m.,  S.  leucotis  was  found  in  an  open 
meadow  along  a  cold,  mountain  tributary  of  the  Rio  Piaxtla  (see  Fig.  17).  Tliis 
area  was  surrounded  by  boreal  pine-oak,  fir-aspen  forest.  Traps  set  in  runways 
also  captured  Reithrodontomys  megalotis  and  Peromyscus  melanotis.  Microtus 
mexicanus  was  present,  but  only  in  the  open  grass,  whereas  S.  leucotis  occupied 
the  shrub  zone. 

At  Hacienda  Coyotes,  2475  m.,  white-eared  cotton  rats  li\ed  on  partly- 
bare,  rocky  slopes  imder  patches  of  low  vaccinium  and  manzanita,  just  at  the 
edge  of  dense  pine-oak  forest.  No  runways  or  cuttings  were  e\'ident,  but  holes 
under  rock  ledges  were  conspicuous.  Traps  placed  in  front  of  these  holes 
captured  most  of  the  animals.  No  other  small  mammals  were  taken  in  associa- 
tion with  S.  leucotis,  although  Reithrodontomys  megalotis,  Sigmodon  fulviven- 
ter,  and  Microtus  mexicanus  occupied  adjacent  wet  meadows  in  bunch  grasses 
and  sacaton. 


Baker — Sigmodon  fulviventer  Group 


215 


Fig.  17.    Riparian  grass-shrub  habitat  surrounded  by  pine-fir  boreal  forest  near 

San  Luis,  2300  m.,  Durango.    The  Mexican  vole,  Microtus  mexicanus,  uses  the 

open  grass,  whereas  the  white-eared  cotton  rat.   Sigmodon   Icucotis,  seems   to 

prefer  the  areas  of  mixed  shrubs.    Photograph  taken  on  July  17,  1957. 


On  a  canyon  side,  approximately  3.3  km.  N  of  Pueblo  Nue\o,  1S30  m.,  a 
white-eared  cotton  rat  was  shot  (by  headlight  at  night)  in  an  area  of  large 
rocks  in  an  abandoned  weedy  and  brushy  peach  orchard  ( see  description  in 
Webb  and  Baker,  1962:328).  This  locality  is  situated  on  the  west  side  of  the 
Sierra  Madre  Occidental  in  mixed  boreal-tropical  habitat.  Intensi\e  trapping 
in  the  place  where  the  one  cotton  rat  was  shot  produced  no  more  indix'iduals. 
Other  mammals  taken  there  were  Thomomijs  umbrinus,  Peromyscus  hoijlii, 
and  Neotoma  mexicana.  This  is  the  most  "tropical"  habitat  in  which  S.  Icucotis 
has  been  found. 

In  southern  Durango  on  the  Rancho  Las  Margaritas  (47  km.  S  and  28  km. 
W  Vicente  Guerrero,  2545  m.),  S.  leucotis  was  captured  in  pine-oak  habitat 
in  a  narrow  valley  in  obscure  runways  in  scattered  bunch  grasses  (see  Drake, 
1958).  Other  small  mammals  in  this  association  were  Thomomys  umbrinus, 
Eutamias  bulleii,  Reithiodontomys  megalotis,  and  Peromyscus  boylii.  In  nearby 
canyon-side  habitat  our  field  party  caught  Peromyscus  difficilis,  Peromyscus 
truei,  Neotoma  mexicana,  and  Nelsonia  neotomodon. 

Zacatecas. — At  approximately  13  km.  W  of  Milpillas,  2530  m.,  J.  Dan 
Webster  (personal  communication)  reported  catching  S.  leucotis  in  a  wet 
meadow  surrounded  by  boreal  pine-oak  forest.  At  15  km.  W  Zacatecas,  2135 
m.,  white-eared  cotton  rats  were  found  at  a  creek  border  in  willow  trees  and 
baccharus  bushes. 

Guanajuato. — At  the  Rancho  La  Puerta  Guadalupe  ( 8  km.  SW  Ibarra,  2500 
m. )  S.  leucotis  occupied  obscure  runways  in  grass  leading  between  rocks  and 
clumps  of  low-growing  (40  centimeters  high)   scrub  oak   (see  Fig.   13).    The 


216 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Fig.  18.  Open  \alley  meadow  surrounded  l)y  oak  torest  near  Ibarra,  2590  m  , 
Guanajuato.  The  sacaton,  Briza  rotundata.  is  occupied  by  the  white-eared 
cotton   rat,   Sigmodon  leucotis,  and   three  species  of  Peromyscus.     Photograph 

taken  on  July  22,  1966. 


bunch  grasses  included  representatives  of  the  genera  Muhlenbergia,  Stipa,  and 
Bouteloua.  This  location  was  on  a  sloping  hillside  with  large  oaks  and  manza- 
nita  in  the  shallo\\'  \alley  below.  Abo\e  the  slope  the  le\'el  open  grasslands 
were  occupied  by  Sigmodon  ftdviventer.  In  runways  used  by  S.  leucotis,  we 
found  grass  cuttings,  the  husks  of  acorns  and  manzanita  berries,  and  small 
piles  of  lily  bulbs.  These  foods  might  ha\e  been  used  by  the  white-eared 
cotton  rat  or  by  other  small  mammals  found  in  the  runways  (Pcromijscus  hoijlii, 
Pcromijscus  maniculatus,  and  Pcromijscus  truei). 

At  13  km.  S  Ibarra,  2590  m.,  S.  leucotis  was  captured  in  a  small  meadow 
of  about  two  hectares  along  a  cold  stream  and  surrounded  entirely  by  an  oak 
forest  (see  Fig.  18).  This  opening  contained  short  grass,  Piptochaetium  sp., 
with  scattered  clumps  of  bimch  grass,  Muhlenbergia  sp.,  and  sacaton,  Briza 
rotundata.  Eaten-out  areas  at  the  bases  of  these  grasses  plus  a  few  leaf  cuttings 
were  the  only  evidences  of  sign  attributable  to  cotton  rats.  One  cotton  rat  was 
taken  in  front  of  a  hole  beneath  a  prickly  pear  plant,  in  mixed  oak  and 
manzanita.  There  was  no  exidence  that  either  S.  fulviventer  or  Microtus  mexi- 
canus  were  present,  although  the  moist  meadow  seemed  ideal  for  the  Mexican 
vole.  Small  mammal  associates  in  the  sacaton  area  were  Thomomys  umbrinus, 
Peromyscus  dijficilis,  Peromyscus  boylii,  and  Peromyscus  truei. 

Morelos. — In  a  moist,  open  meadow  6.5  km.  NW  Huitzilac,  2800  m.,  one 
white-eared  cotton  rat  was  captured  in  an  eaten-out  area  at  the  base  of  a 
large  clump  of  sacaton,  Muhlenbergia  macroura  (see  Fig.  19).  This  clump 
was  one  of  a  large  number  that  surrounded  an  open  meadow  containing  such 
plants  as  Senecio  pinnatisectus.  Ranunculus  sp.,  and  Taraxacum  officinale.    Sur- 


Baker — Sigmodon  fulviventer  Group 


217 


rounding  the  meadow  was  a  lioreal  forest  of  pine,  oak,  fir,  and  other  montane 
vegetation.  Small  mammal  associates  were  Reithrodontomijs  mef^alotis,  Pcro- 
mysciis  inelanotis,  Ncotoiiiodou  alstoni,  and  Microtus  mexicantis.  In  the  case 
of  the  latter  two,  both  grass-eating  "competitors"  of  the  cotton  rat,  M.  mexi- 
canus  was  most  abundant  in  the  short  grass-herb  meadow  and  N.  ahtoni  (see 
Davis  and  Follansbee,  1945)  dominated  the  sacaton  clumps.  On  the  basis  of 
these  observations,  S.  leucotis  exddently  was  rare  and  perhaps  at  an  ecological 


Fig.  19.  Open  valley  meadow  surrounded  by  pine-fir  boreal  forest  near  Huit- 
zilac,  2800  m.,  Morelos.  The  open  meadow  is  occupied  by  the  Mexican  vole, 
Microtus  mexicanus;  the  \olcano  mouse,  Neotomodon  alstoui,  is  the  dominant 
grass-eating  rodent  in  the  large  clumps  of  sacaton  w  here  the  white-eared  cotton 
rat,  Siginodon  leucotis,  is  unconunon.   Photograph  taken  on  July  24,  1964. 


disadxantage.  This  same  locality  was  \isited  again  three  years  later  (in  1967) 
and  the  entire  area  was  intensely  live-trapped  for  three  days,  without  obtaining 
other  S.  leucotis.  Davis  (1944:399),  caught  five  animals  in  a  meadow  with 
Microtus  at  Monte  Rio  Frio  in  the  state  of  Mexico,  but  subsequently  failed  to 
catch  others.  Field  parties  from  the  Michigan  State  Unixersity  Museum  also 
failed  to  find  cotton  rats  in  montane  habitats  in  the  vicinity  of  Oaxaca  in  the 
state  of  Oaxaca  and  in  the  vicinity  of  Pinal  de  Amoles  in  Queretaro;  specimens 
were  obtained  at  both  places  by  E.  W.  Nelson  and  E.  A.  Goldman  in  the 
1890's  (Bailey,  1902:116).  The  fact  that  S.  leucotis  may  live  in  shrub  vege- 
tation, especially  where  bunch  grass  habitat  is  grazed  ofi^,  makes  the  presence 
of  the  species  difficult  to  determine. 

Association  of  white-eared  cotton  rats  with  other  species  of  the  genus. — 
Sigmodon  leucotis  has  not  been  taken  in  company  with  S.  alleni,  S.  hispidtis,  or 
S.  ochrogiuithus.  Perhaps  S.  leucotis  and  S.  (illeni  might  associate  in  such  situa- 
tions as  foimd  in  the  \icinity  of  Pueblo  Xuexo  in  Durango  where  the  former 
species  has  been  taken  in  a  mixed  boreal-tropical  habitat,  not  unlike  some  of 


218  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

the  places  in  Michoacan  and  Oaxaca  where  S.  alleni  occurs.  In  the  mountains 
of  west-central  Durango,  S.  leucotis  and  S.  ochrognathus  occur  in  the  same 
latitude  (see  Fig.  4)  but  at  different  elevations,  with  the  former  in  higher 
more  mesic  areas  and  the  latter  on  the  lower  slopes  and  foothills  in  arid  oak- 
juniper  areas. 

The  two  species,  S.  leucotis  and  S.  ftilviventer,  were  found  in  the  same 
areas  in  Durango  (vicinity  of  Hacienda  Coyotes)  and  in  Guanajuato  (8  km. 
SW  Ibarra).  At  both  of  these  places,  S.  ftilviventer  dominates  the  open  grass- 
lands whereas  S.  leucotis  lives  on  the  rocky,  brushy,  well-drained  slopes.  The 
meeting  ground  of  these  two  species  appears  to  be  at  the  junction  of  these  two 
en\ironments.  In  somewhat  similar  situations,  where  S.  fulviventer  is  absent 
(as  at  San  Luis  and  the  Rancho  Las  Margaritas  in  Durango),  .S.  leucotis  takes 
over  both  habitats.  Also,  S.  leucotis  seems  to  avoid  extensive  association  with 
other  grasseaters,  Microtus  mexicanus  and  Neotonwdon  ahtoni,  when  S.  ful- 
viventer is  absent  (as  at  Huitzilac  in  Morelos).  When  S.  leucotis,  S.  fulviventer 
and  M.  mexicanus  are  in  the  same  area  (as  at  Hacienda  Coyotes  in  Durango), 
S.  leucotis  occupies  the  forest-meadow  ecotone  (mixed  brush  on  rocky  slopes). 
S.  fulviventer  occupies  the  bunch  grass  flats,  and  Microtus  mexicanus,  although 
occasionally  taken  in  runways  with  S.  fulviventer,  occupies  exclusi\ely  the 
poorly-drained  parts  of  the  montane  meadow  that  is  co^'ered  with  clumps  of 
sacaton. 

Parasites. — Dr.  Robert  Traub  identified  the  flea,  Pohjgenis  martinez-haezi 
Vargas,  1951,  from  S.  leucotis  from  near  Hda.  Coyotes.  Dr.  Richard  B.  Loomis 
identified  the  chiggers,  Hijponeocula  argenicola  and  Fonsecia  sp.,  from  S.  leu- 
cotis from  the  same  area. 

Specific  CJuiraciers  and  Comparisons 

Conspicuous  whitish  ears  in  contrast  to  a  brownish-gray  head 
and  body  together  with  small  to  medium  size  (for  captive  animals, 
maximum  weights  are  131  grams  for  a  male  and  140  for  a  non- 
pregnant female;  maximum  lengths  of  head  and  body  are  160  mm. 
and  169,  respectively),  pronounced  premaxillary  depressions  on 
each  side  of  the  rostrum,  and  highly  reduced  or  absent  lingual  root 
on  the  first  lower  molar  distinguish  S.  leucotis  from  other  species  in 
the  S.  fulviventer  group,  and,  with  the  exception  of  size,  from  S. 
hispidus  as  well.  Other  unique  characters  are:  anterior  portion  of 
mesopterygoid  fossa  parallel-sided;  interparietal  short,  length  at 
midline  less  than  2  mm.;  angular  process  of  lower  jaw  slightly 
hooked  rather  than  rounded.  The  reduction  or  absence  of  the  lin- 
gual root  on  the  first  lower  molar  is  possibly  the  most  distinctive 
character  and  sets  this  species  apart  as  perhaps  the  most  liighly 
evolved  in  the  genus  ( see  also  Dalby  and  Lille\ik,  1969 ) . 

From  S.  alleni,  S.  leucotis  is  further  distinguished  by  skull  short 
and  broad  rather  than  long  and  narrow;  dorsal  profile  of  skull  more 
arched  than  flattened;  upper  incisors  less  instead  of  more  recurved 


Baker — Sigmodon  fulviventer  Group  219 

( opisthodont ) ;  auditor)  bullae  large  instead  of  small  in  relation  to 
size  of  skull;  basioccipital  long  and  narrow  rather  than  short  and 
wide;  posterior  ends  of  incisixe  foramina  extending  (rather  than 
not  extending)  to  a  line  drawn  between  the  anterior  surfaces  of  the 
first  upper  molars;  anterior  lip  of  foramen  magnum  slightly  instead 
of  obx'iously  notched;  and  paraoccipital  process  (from  ventral  view) 
straight  rather  than  slightly  hooked. 

From  S.  fulviventer,  S.  leucotis  differs  in  color  of  under  parts 
(usually  whitish  not  buff),  and  in  that  the  foramen  ovale  is  small 
(no  more  than  half  width  of  M3)  instead  of  large  (at  least  three- 
fourths  width  of  M3 ) . 

From  S.  ochro^^nathus,  S.  leucotis  is  further  distinguished  by 
nose  usually  lacking  extensive,  contrasting  yellow  coloring;  auditory 
bullae  large  and  broad  instead  of  small  and  elongate;  basioccipital 
long  and  narrow  instead  of  short  and  wide;  median  keel  on  basioc- 
cipital slight  rather  than  obvious;  posterior  ends  of  incisive  foramina 
extending  (rather  than  not  extending)  to  a  line  drawn  between  the 
anterior  surfaces  of  the  first  upper  molars;  notch  on  anterior  lip  of 
foramen  magnum  slight  instead  of  ob\'ious;  bulge  of  capsular  projec- 
tions for  upper  incisors  moderate  rather  than  pronounced;  inter- 
parietal without  (instead  of  having)  medium-posterior  notch;  and 
paraoccipital  process  (from  ventral  view)  straight  rather  than 
curved  with  a  basal  notch. 

From  S.  hispidus,  S.  leucotis  differs  in  small  size  of  tail  scales 
(0.5  mm.  wide  rather  than  0.75  mm.  wide);  tail  hea\ily  haired 
instead  of  sparsely  haired;  skull  short  rather  than  long;  palatal  pits 
deep  as  opposed  to  shallow;  and  median  keel  on  palate  conspicuous 
rather  than  slight. 

Geographic  Variation 

Although  certain  cranial  and  dental  characters  seem  to  set  S. 
leucotis  apart  as  the  most  distinctixe  Recent  species  in  the  genus, 
the  apparent  lack  of  geographic  \'ariation  in  this  montane  species  is 
surprising  and  can  be  compared  with  the  condition  found  in  mono- 
typic  Peromyscus  melanotis,  which  occupies  some  of  the  same  habi- 
tat in  the  same  Y-shaped  distributional  pattern  in  the  Sierra  Madre 
Occidental,  the  Trans-Mexican  Volcanic  Belt  (part),  and  the  Sierra 
Madre  Oriental.  On  the  southward  side  of  the  watershed  of  the 
Rio  Balsas,  the  more  richly-colored  S.  leucotis  in  Puebla  and  Oaxaca 
can  be  easily  distinguished  at  the  subspecific  level. 

Herein,  cotton  rats  pre\'iously  assigned  to  S.  leucotis  and  to  S. 
alticola  are  arranged  as  belonging  to  the  same  species,  with  the 


220  Misc.  Publ.  51,  UiNiv.  Kansas  Mus.  Nat.  Hist. 

former  name  being  adopted  as  the  specific  name  because  of  page 
priority.  Both  species  were  named  by  Vernon  Bailey  in  his  1902 
paper  with  the  description  of  S.  leucotis  appearing  on  page  115  and 
that  of  S.  alticola  on  page  116.  After  examining  and  comparing 
recently-obtained  material  from  localities  in  Aguascalientes  and 
Guanajuato,  which  are  intermediate  between  places  from  where 
S.  leucotis  and  S.  alticola  were  previously  reported  (see  Hall  and 
Kelson,  1959:678),  it  was  readily  evident  that  these  montane  cotton 
rats  in  central  and  southern  Mexico  belong  to  one  species. 

Cotton  rats  of  this  species  were  obtained  in  abundance  (as 
based  on  series  in  museum  collections)  only  at  a  few  places:  in 
Durango  near  San  Luis,  at  Hda.  Coyotes,  and  southwest  of  Vicente 
Guerrero;  in  Zacatecas  in  the  Valparaiso  Mountains;  in  Aguascali- 
entes near  Cerro  del  Jagiiey;  in  Guanajuato  near  Ibarra;  in  the  state 
of  Mexico  at  Monte  Rio  Frio;  and  in  Morelos  near  Huitzilac.  Field 
parties  from  the  Michigan  State  University  Museum  either  failed  to 
obtain  any  or  could  not  get  adequate  series  of  animals  in  numerous 
"likely"  montane  localities  from  west-central  Durango  and  central 
Nuevo  Leon  south  to  Oaxaca,  including  such  out-of-the-way  places 
as  Pinal  de  Amoles  in  Queretaro.  Furthermore,  live  animals  (from 
Hda.  Coyotes  and  Ibarra)  brought  back  to  the  MSU  Museum  Live 
Animal  Colony,  unlike  the  other  species  of  cotton  rats,  produced 
few  offspring  under  captive  conditions.  It  is  my  opinion  that  this 
animal  may  be  highly  adapted  to  certain  montane  mixed  grass  and 
brush  areas,  but  presumably  in  many  such  situations  gives  way  to 
other  grass-eating  "competitors"  including  Microtiis  mexicanus  and 
possibly  Neotomodon  alstoni  and  S.  fiilviventer.  At  least  this  is  one 
way  to  explain  the  disjunct  distribution,  because  on  many  occasions 
S.  leucotis  was  not  taken  in  places  that  looked  almost  "identical"  to 
other  sites  where  the  animals  were  easily  obtained. 

Sigmodon  leucotis  leucotis  Bailey 
Sigmoclou  leucotis  Bailey,  Proc.  Biol.  Soc.  Wasliington,   15:115,  June  2,  1902. 
Sigmodon  alticola  amoles  Bailey,  Proc.  Biol.  Soc.  Washington,  15:116,  June  2, 

1902,  type  from  Pinal  de  Amoles,  Queretaro. 

Type. — Young  adult  female,  skin  and  skull;  no.  92001  U.S.  National  Mu- 
seum; from  Valparaiso  Mountains,  2653  m.,  Zacatecas;  obtained  on  December 
2,  1897,  by  E.  W.  Nelson  and  E.  A.  Goldman,  original  no.  11812. 

Range. — Montane  grass-brush  habitats  from  approximately  25°  N  latitude 
in  the  Sierra  Madre  Occidental  and  Sierra  Madre  Oriental  southward  to  the 
Trans-Mexican  Volcanic  Belt  in  the  states  of  Morelos  and  Mexico  (.see  Fig.  16). 

Diagnosis. — Size  large  for  the  species;  dorsinu  intermixed  with  black  hairs 
and  agouti-banded  hairs,  Pinkish  Buff  or  slighth'  darker,  near  (c)  Cinnamon- 


Baker — Sigmodon  fulviventer  Group 


221 


.s 

Co 

^  fc, 

^S^ 

o 

^Sg 

y 

b  xj3 

a  (3*3 

S 

"pSg 

'— ^ 

1^- 

,^ 

5 

< 

-c 

o 

r^ 

Mh 

a 

1^ 

•«-rt 

C/3 

!i 

S 

h4 

< 

a 

z 

o 

H 

■£  3 

0 

ac 

U 

Q 

<u  a 

QS 

U 

P5 

< 

•— ■  :3 

1 

Ǥ 

» 

■sS-B 

H 

K  Pi: 

S 

•SS 

^ 

h 

«; 

g-s 

r  PJ 

o 

asi 

z 

>.Q 

P 

N 

o 

>< 

>, 

b 

u, 

O 

o^^ 

75 

S 

U2i 

H 

a 

rt 

H 

X 

^*^  _ 

H 

21^ 

Q 

J3  S  o 

< 

0)  a  C 

K« 

en 

W 

O 

< 

ri* 

J513 

u 

«.s^ 

> 

C^  0 

•< 

v^^ 

iJ   0 

»2 

H 

Z 

>. 

U 

^7313 

M 

*-  n:  o 

i< 

^o| 

u 

^^1 

c« 

< 

u 

M 

< 

a 

» 

H 

U 

W 

l; 

1-1 

3.S 

a 

^ 

lo 

u 

pq 

< 

c-— 

H 

hJ 

cJ5 


00 

in 

00 

-^o 

-f<  o 

lOi^ 

CD 

(N 

CD  -H 

co4 

Co'^t" 

CO 

CD 

d 

CD 

CO 

CO 


05 


CD 


lO 


-H  ^  ;z;  in  ;z,  CD 

1— I     1— I     o 


d 


CO 

CD   o' 
CD 

CO 

2'  ^: 


M 


o^2 

ic2 

co^ 

t- 

05 

::'^ 

2-+. 

2-^ 

d 
1—1 

CJ< 

o 

o 

o 

CO 

-t 

1— 1 

^i6 

—I  in 

o  in 

-t 

^H 

G> 

lO  t^ 

ino 

in  CD 

in 

in 

-r 

•r 

in 

-V 

CD 

in 

-V 

2 
o 

CJ 

_d 

co2 

00 

00 

ci 
f— ( 

d 

1— H 

d 

T-H 

-H       O 


5 
o 


m  m 


CO 


CO 

<^ieo 

co^. 

CO 


CO 


CO 


CO 


CM 


00 


05 


CD 


CO 


CO 


o 


CO 


o 


CO 


o 

o 
N 


CO 


^    ^ 


o 


a; 


3 


c 

OS 


lO 

in 

co' 

in 

o 

d 

Ol 

^ 

r 

o 

t- 

o 

•«?». 

d 

^H 

^ 

^ 

CO 

'>z 

CO 

o 

CO 

sj 

•V 

CO 

S 

tc 


00 


05 

^^ 

o 

t- 

Ol 

CO 

oco 

CD(M 

05 

a> 

t^ 

CD 

^d 

«>iin 

C] 

o\ 

Cl 

oo' 

CI 

in 

o 

CD 

m 

^:!; 

2- 

t- 

CO 

00 

1— 1 

-r    1 

-^  , 

-f 

in 

-r 

^ 

cq 

■-HCO 

^  -T 

1 — 1 

1 — ( 

^H 

-t 

CO 

-t 

-r 

-r 

Cl 


O  O  U 


o 


in 

co" 


CO 


00 

d 


in 


d 
d 


in 

oo' 

00 


05 


CO 

d 

CO 
CO 


Ol 

in 


oq 


00 

o 

CO 


0,05 

S  -I 


c  2 


0:5 

'■^ 

o 

CM 

>^ 

d" 

.1 

CO 

u 

a 

cr> 

E 

-3 

CO 

f7! 

o  ■' 
c  c 

3  S 


-<  s 


222  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

BufF;  base  of  tail  darker;  under  parts  whitish  but  sometimes  washed  with  bufF; 
skull  wide  in  interorbital  area,  heavily  ridged,  and  with  noticeably  large  audi- 
tory bullae.    Selected  external  and  cranial  measurements  are  gi\en  in  Table  5. 

Comparisons. — From  S.  /.  alticola,  S.  I.  leiicotis  differs  in:  size  larger  (see 
Table  5);  color  paler  (near  Pinkish  Buff  rather  than  Clay  Color);  imder  parts 
usually  not  washed  with  pale  buff;  skull  more  massive,  with  greater  width  in 
interorbital   area;    auditory  bullae   larger;   nasals  broader;   palatal   pits   deeper. 

Remarks. — Sigmodon  leucotis  leucotis  is  slightly  larger  and  much  paler  in 
color  than  S.  /.  alticola  from  the  Sierra  Madre  del  Sur,  southward  of  the  Rio 
Balsas,  whose  watershed  seems  to  have  effectively  barred  the  north-south 
passage  of  many  mammalian  species  (see  Baker,  1963:245).  Specimens  of 
S.  /.  leucotis  from  the  northern  limits  of  its  range  ( Durango,  Zacatecas,  Quere- 
taro,  and  Nuevo  Leon)  are  palest  (Pinkish  Buff),  whereas  specimens  from 
further  to  the  south  are  slightly  more  richly  colored    (near  Cinnamon-Buff). 

Sjjecimens  examined  (112). — Nuevo  Leox:  20  km.  SSW  Galeana,  1891 
m.,  1  (CAS).  DuRAXGo:  30  km.  SSW  Tepehuanes,  2500  m.,  2  (MSU);  2.5 
km.  W  San  Luis,  2303  m.,  2  (MSU);  San  Luis,  1  (AMNH);  1  km.  E  San  Luis, 
2.348  »i.,  6  (UM),  1  (UNAM);  25  km.  ENE  Coyotes,  2544  m.,  1  (MSU); 
7  km.  N  from  higlncaij  on  road  to  San  Luis  via  Coyotes,  2  (CAS);  92  km.  W 
Durango,  on  road  to  El  Salto,  2407  m.,  7  (CAS);  Hda.  Coyotes,  2477-2501  m., 
4  (CAS),  8  (MSU);  El  Salto,  2318-2440  7n.,  15  (FM),  2  (US);  3  km.  N 
Pueblo  Nuevo,  1830  m.,  1  (MSU);  43  km.  S  and  30  km.  W  Vicente  Guerrero, 
2547  m.,  9  (MSU).  Zacatecas:  13  km.  S  Chalchuites,  2623  m.,  4  (CAS); 
13  km.  W  Milpillas  (or  100  km.  W  Fresnillo),  2531  m.,  1  (CAS);  27  km. 
W  Mdpillas  (or  112  km.  W  Fre.millo),  2531  m.,  1  (CAS);  Valparaiso  Mountains, 
2653  m.,  10  (US);  15  km.  W  Zacatecas,  2135  m.,  1  (CAS);  17  km.  S  Pinos, 
2165  m.,  1  (UNAM).  Aguascaliextes:  5  km.  N  Cerro  del  Jagiiey,  2501  m., 
Sierra  Fria,  15  (MVZ);  Rio  de  San  Pedro,  NW  side  San  Antonio,  1   (MVZ); 

1  km.  S  La  Labor,  1830  m.,  1  (MVZ);  7.5  km.  NW  Calvillo,  1830  m.,  1 
(MVZ).  Guanajuato:  Puerta  de  Guadalupe,  2196  m.,  8  km.  W  Ibarra,  1 
(OC);  J3  km.  SW  Ibarra,  2592  m.,  1  (MSU).    Queretaro:    Pinal  de  Amoles, 

2  (US).  Me.xico:  Monte  Rio  Frio,  45  km.  ESE  Mexico,  5  (TCWC);  Hda. 
Cordoba,  2600  m.,  1  (UM).  Morelos:  4  km.  N  Tres  Cumbres,  .3202  m.,  2 
(TCWC);  3  km.  W  Huitzilac,  3050  m.,  4  (TCWC);  7  km.  W  Huitzilac,  2806 
m.,  1   (MSU). 

Sigmodon  leucotis  alticola  Bailey 

Sigmodon  alticola  Bailey,  Proc.  Biol.  Soc.  Washington,  15:116,  June  2,   1902. 

Type. — Young  adult  male,  skin  and  skull;  no.  68231  U.S.  National  Museum; 
from  Cerro  San  Felipe,  3050  m.,  Oaxaca;  obtained  on  March  15,  1894,  by 
E.  W.  Nelson  and  E.  A.  Goldman,  original  no.  6624. 

Range. — Montane  grass-shrub  areas  in  parts  of  the  Sierra  Madre  del  Sur  of 
Puebla  and  Oaxaca  (see  Fig.  16). 

Diagnosis. — Si^e  medium;  dorsum  intermixed  with  black  hairs  and  agouti- 
banded  hairs.  Clay  Color  becoming  more  Sayal  Brown  on  rump  and  at  base  of 
tail;  under  parts,  tops  of  hind  feet,  and  tail  faintly  washed  with  buff  (Pale 
Pinkish  Bufi);  skull  slender,  lightly  constructed,  and  with  narrow  interorbital 
space;  small  auditory  bullae;  and  shallow  palatal  pits.  Measurements  are 
given  in  Table  5. 


Baker — Sigmodon  fulvivexter  Group  223 

Comparisons. — For  comparison  wilh  S.  /.  kitcotis,  see  account  of  that  sub- 
species. 

Remarks. — Again  the  paucity  of  study  material  made  an  analysis  of  geo- 
graphic  \ariati()n  difficult,  but  this  sul)species  is  much  more  richly  colored 
tlian  S.  /.  Icucotis.  A  sulnidult  from  near  Acatzingo,  Puebia,  is  less  richly 
colored  above  but  has  the  distinctive  ])uffy  wash  on  the  underparts.  The  wide- 
spread occurrence  and  abundance  of  voles  (genus  Microtus)  in  lioreal  grassy 
areas  in  Oaxaca  lead  me  to  wonder  if  the  white-eared  cotton  rat  in  Oaxaca 
plays  a  secondary  role  and  is  highly  restricted  ecologically. 

Specimens  examined  (6). — Puebla:  15  km.  NE  Acatzingo,  1  (KU). 
Oaxaca:  25  km.  W  Oa.xaca,  2897  m.,  2  (US);  Cerro  San  Felipe,  2200  m., 
2  (UM),  1  (US). 

Sigmodon  ochrognathus 

Sipnodon  ochrognathus,  the  yellow-nosed  cotton  rat,  is  the  most 
xerophilous  species  in  the  genus.  It  is  adapted  to  the  dry,  rocky 
slopes  ( oak-pinon- juniper  habitat)  of  the  eastern  side  of  the  Sierra 
Madre  Occidental  from  Arizona  and  New  Mexico  south  to  central 
Durango  and  in  the  widely-scattered  desert  ranges  of  the  Mesa  del 
Norte  (northern  part  of  the  Mexican  Plateau)  in  Trans-Pecos  Texas, 
western  Coahuila,  northeastern  Durango,  and  probably  eastern 
Chihuahua  (see  Fig.  20). 

Habitat  and  Habits 
The  yellow-nosed  cotton  rat  is  an  inhabitant  of  the  desert  moun- 
tains of  the  American  Southwest.  It  li\es  chiefly  on  rocky  slopes 
with  scattered  clumps  of  grasses,  mostly  in  oak-juniper  habitat  ( see 
Figs.  11  and  21),  although  it  occupies  grassy  montane  flats  in 
localities  where  other  species  of  Sigmodon  are  not  present.  Un- 
doubtedly, the  preferred  habitat  of  S.  ochrognathus  has  been  altered 
where  grazing  by  livestock,  especially  goats,  has  been  severe.  How- 
e\'er,  this  cotton  rat  will  persist  on  rocky  hillsides  where  only  sparse 
grass  occurs.  Its  runways  are  often  well-marked  in  thick  grass  but 
are  rarelv  visible  on  bare  hillsides,  where  the  animals  dart  from  one 
rock  shelter  to  another.  On  several  occasions  animals  were  observed 
(especially  in  the  vicinity  of  Boquilla,  Durango)  moving  across 
such  open  spaces  of  as  much  as  one  meter  in  distance.  Cuttings  of 
grass  blades  and  piles  of  fecal  droppings  are  generally  conspicuous; 
openings  into  burrows,  many  being  excavated  by  pocket  gophers 
(Tliomomys  timbrinus),  may  be  obscured  behind  or  at  the  sides  of 
rocks.  These  cotton  rats  will  often  feed  at  the  base  of  a  clump  of 
grass.  Since  the  grass  droops  down,  it  is  necessary  for  one  to  raise 
up  the  dead  grass  to  find  the  rat's  secluded  chamber,  circling  around 
the  base  of  the  clump. 


224 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Fig.   20.    Geographic   distril^ution    of   the   yellow-nosed   cotton   rat,    Sigmodon 

ochrop.nathus. 

As  a  species,  this  cotton  rat  lives  under  more  xeric  conditions 
than  any  of  the  other  species  in  the  genus.  Succulent  plants  may 
provide  necessary  moisture  on  dry  hillsides,  because  it  is  doubtful 
that  S.  ochrognathus  has  much  opportunity  to  find  surface  water  in 
such  well-drained  situations,  even  if  rain  were  more  prevalent  than 
is  actually  the  rule  in  these  foothills.  Although  no  quantitative  data 
are  available,  one  obtains  the  impression  that  S.  ochrognathiis  occurs 
in  less  concentrated  numbers  than  do  other  species  of  cotton  rats. 
Perhaps  the  food  supply  is  generally  less  abundant  on  the  rocky 
hillsides  than  on  deep  alluvial  valley  soils  where  S.  ftdvwenfer  and 
S.  luspidiis  live.  The  seemingly  lower  carrying  capacity  of  the  hill- 
side habitat  may  cause  yellow-nosed  cotton  rats  to  be  more  widely 
spaced  in  nature  than  other  cotton  rats.  This  suggestion  is  based 
on  the  fact  that  it  is  unusual  to  catch  more  than  one  yellow-nosed 
cotton  rat  at  any  one  trap  station,  although  there  was  no  reason  to 
indicate  that  the  rodents  did  not  share  common  runwavs.   Hoffmeis- 


Baker — Sigmodon  fulviventer  Group 


223 


Fig.  21.  Hill.side  occupied  by  the  yellow-nosed  cotton  rat,  Sigmodon  ochrogna- 
thiis,  near  Fort  Davis,  1610  m.,  JefF  Davis  Co.,  Texas.  The  \egetative  co\er 
consists  of  scattered  shrubs  of  species  belonj^ing  to  the  genera  Jiiiui)cnis,  Rhus, 
and  Mimosa  and  grasses  of  the  genera  Aiuhopogoit,  Botitelotia,  Miililenhcrgia, 
Ehjoimnis,  Aragrostis,  Setaria,  and  Panicinu.    Photograph  taken  on  August  17, 

1967. 

ter  ( 1963),  who  described  the  habits  of  S.  ochrognathus  in  Arizona, 
found  nests,  usually  of  grass,  in  thickets  of  grasses  or  drooping 
Nolina  and  Agave.  Captive  rats  sometimes  made  nests  of  cotton 
either  inside  or  outside  of  refuge  cans  placed  in  their  cages.  Hoff- 
meister  as  well  as  Baker  and  Greer  (1962:125)  have  given  dimen- 
sions of  runway  systems  and  burrows.  Descriptions  of  several  col- 
lecting localities  are  given  below. 

Diirango. — x\t  12  km.  NNE  Boquilla,  1890-1965  ni.,  yellow-nosed  cotton 
rats  were  captured  on  two  occasions  (10  July  1965  and  8  July  1967).  Most 
animals  were  taken  on  a  north-facing  slope  on  the  eastern  foothills  of  the  Sierra 
Madre  Occidental  (see  Fig.  11).  The  slope  was  rocky  and  covered  with 
scattered  shrubs — algerita  (Mahouia  trifoliata),  cliff  rose  (Cowania  mexicana), 
and  catclaw  {Mimosa  sp.).  Moderately-grazed  chunps  of  tall  grass  of  the  genus 
Mnhlenbergia,  and  composites,  Haplopapptis  spinulosus  and  Baileija  sp.,  were 
scattered  on  this  hillside.  Small  mammal  associates  taken  with  S.  ochrognathus 
include  Perognathiis  nelsoni,  Reithrodontomijs  fidvescens,  Reithrodontomijs 
megalotis,  and  Baiomys  iaylori.  Sigmodon  ftdviventer  lived  on  the  grassy  valley 
floor  which  adjoins  the  hillsides. 

At  9  km.  NNW  Canatlan,  1950  m.,  yellow-nosed  cotton  rats  were  taken  on 
a  rocky  hillside  in  runways  through  clumps  of  grasses,  Bouteloua  gracilis  and 
Muldenhergia  sp.,  mostly  protected  from  grazing  by  thick  over-head  cover  of 
Acacia,  Mimosa  and  prickly  pear  (Optintia).  Associated  small  mammals  ob- 
tained in  the  area  were  Perognathtis  nelsoni,  Liomys  irroratus,  Thomomys  urn- 


226  Misc.  Publ.  51,  Unw.  Kansas  Mus.  Nat.  Hist. 

hrinus,   Reithrodoniomijs  fulvescens,   Peronnjsctis   pectoralis,    Baionujs    laylori, 
and,  on  adjacent  flats,  Si^nwdon  fulviventer. 

Coahuila. — Baker  (1956:278)  found  yellow-nosed  cotton  rats  in  runways 
in  bunch  grass,  prickly  pear,  and  scrub  oak  at  1616  meters  in  elevation  (in  the 
Sierra  de  la  Madera),  and  in  thick  grass  in  a  narrow  valley  floor  with  runs 
leading  to  burrows  under  small  oaks  at  2135  meters  in  elevation  (in  the  Caiion 
del  Hillcoat  in  the  Sierra  de  la  Encantada).  Taylor  et  al.  (1945:26)  found 
cuttings,  burrows,  and  piles  of  earth  in  grass  in  the  Sierra  del  Carmen. 

Texas. — A  north-facing,  rocky  hillside  3  km.  NW  Fort  Davis,  1610  m.,  in 
Jeff  Davis  County  (see  Fig.  21)  was  covered  with  scattered  cedars,  Junipertis 
sp.,  and  shrubs,  Rhus  trilohata,  Rhus  microphtjUa,  and  Mimosa  sp.  Yellow- 
nosed  cotton  rats  were  caught  in  obscure  runways  in  clumps  of  grasses  includ- 
ing Bonteloua  gracUis,  Boiiteloua  curtipendula,  Andropogon  saccharoides,  Mtih- 
Icnhergia  spp.,  Ehjomirus  harbictdmis,  Aragrostis  pilosa,  and  Setaria  macro- 
stachija.  Leaf  cuttings,  presumed  to  have  been  left  in  the  rimways  by  S. 
ochwgnathiis,  of  Andropogon  saccharoides,  Setaria  macrostachija,  and  Panicum 
sp.  were  identified.  Small  mammal  associates  were  Perognathus  nelsoni  and 
Peromijscus  pectoralis.  In  Brewster  County,  Denyes  (1956)  found  S.  ochrogna- 
thiis  in  such  plant  associations  as  sotol-sachuiste,  oak  chaparral,  gramma-blue- 
stem,  and  feathergrass-grama. 

New  Mexico. — Findley  and  Jones  (1960)  found  S.  ochrognathus  in  an  alti- 
tudinal  range  from  1160  to  2560  meters,  from  the  upper  limits  of  the  grassland 
into  the  pine-oak  forests  on  rocky  slopes.  In  higher  areas  this  cotton  rat  is 
associated  with  pines,  juniper  and  oaks;  at  lower  elevations  it  lives  in  bunch 
grass,  Yucca,  Agave,  Opuntia,  and  beargrass  (Nohna).  These  authors  also 
observed  the  rodents  in  association  with  plants  belonging  to  such  genera  as 
Dasyhrion,  Fouquieria,  Prosopis,  Mimo.m,  Acacia,  Arctostaphtjios,  and  Cero- 
carpus. 

Arizona. — Hoffmeister  (1963)  recorded  the  yellow-nosed  cotton  rat  on 
grassy,  rocky  slopes  near  or  within  the  oak  belt.  He  listed  plant  associates  as 
Quercus,  Agave,  Nolina,  Cowanio  mexicana.  Mimosa  hiuncifera,  Opuntia, 
Yucca,  Rhus  ovata,  and  Dasyhrion  wheeleri.  The  sparse  cover  used  by  S. 
ochrognathus  included  Bonteloua  gracilis,  Bouteloua  curtipendula,  Aristida  sp., 
Heteropogon  contortus,  Muhlenbergia  sp.,  Senecio  longilobus,  Grindelia 
aphanactis,  and  Eleocharis  sp.  Small  mammal  associates,  according  to  Hoff- 
meister, were  Thomomys  umbrinus,  Reithrodontomys  fulvescens,  Peromyscus 
hoylii,  Peromyscus  eremicus,  and  Ncotoma  albigula. 

Association  of  yellow-nosed  cotton  rats  with  other  species  of  the  genus. — 
The  yellow-nosed  cotton  rat  occupies  rocky,  bunch-grass  slopes  from  just  below 
to  within  the  pink-oak-juniper  belt.  Where  it  is  the  only  cotton  rat  present 
(as  in  southwestern  Texas  and  western  Coahuila),  S.  ochrogtmthus  also  occurs 
on  grassy  montane  "flats"  or  alluvial  fans  where  deep  soils  and  few  rocks  occur. 
In  most  of  its  range  in  New  Mexico,  Arizona,  Chihuahua,  and  Durango,  how- 
ever, this  cotton  rat  is  absent  or  occurs  only  peripherally  in  the  latter  habitats 
because  these  areas  are  occupied  by  S.  fulviventer,  although  Hoft'meister  (loc. 
cit.)  found  S.  ochrognathus  and  S.  hispidus  at  one  place  in  Arizona.  Near 
Canatlan,  Durango,  S.  ochrognathus  was  captured  in  bunch  grass  and  shrulis 
on  a  hillside  slope  of  approximately  20  degrees,  whereas  S.  fulviventer  was 
trapped  in   more  or  less  identical,   but   less   rocky,   cover   at   the   base  of  the 


Baker — Sigmodon  fulviventer  Group  227 

slope  where  tlie  alhuial  fan  l)egan  to  le\el  out.  Here,  the  two  species  were 
not  taken  at  the  same  trap  stations  and  according  to  Held  notes,  were  not 
trapped  closer  together  than  18  meters.  At  this  locality,  in  July  of  1965, 
17  S.  ocIuo^natJuts  and  three  S.  fiilviocntcr  were  caught.  The  larger  catch  of 
S.  ocJuo^iuitlius  probably  is  the  result  of  trap  placement,  more  on  the  rocky 
slopes  than  on  the  level  base.  In  the  \icinity  of  Boqnilla,  Durango,  we  also 
caught  both  species  of  cotton  rats;  S.  ocJirofinatJiiis  was  taken  on  rocky,  bunch- 
grass  slopes  and  S.  fulviventer  in  grassy  areas  on  "flat"  hilltops  and  in  the  deep 
soils  of  valley  floors.  In  one  narrow  intermontane  valley  ( 3.3  km.  NE  Bociuilla ) 
containing  bunch-grass,  a  few  scattered  rocks  and  reddish  "clay"  soil,  I  caught 
in  one  live-trap  set  in  a  nmway,  one  S.  fulviventer  on  the  night  of  8  July  1965 
and  one  S.  ochrog.nathus  the  following  day.  Here  was  one  instance  where 
both  species  "occupied"  the  same  nmway  and  were  caught  at  the  same  trap 
station.  At  12  km.  NNE  Boqnilla  tawny-bellied  cotton  rats  seemed  entirely 
restricted  to  a  grassy,  hilltop  flat,  and  yellow-nosed  cotton  rats  occupied  the 
rocky  slopes  almost  entirely  around  the  hill. 

Although  the  ranges  of  S.  ochro^nathus  and  S.  leucotis  "overlapped"  in 
latitude  in  the  Canatlan-Tepehuanes  area  of  west-central  Durango  (see  Fig.  4), 
the  yellow-nosed  cotton  rat  seemed  confined  to  the  lower^  dry  slopes  of  the 
foothills  (no  higher  than  1950  meters  in  elevation),  whereas  S.  leucotis  occupied 
higher  mesic  areas  of  mixed  grass,  brush  and  rocks  (at  an  elevation  of  2500 
meters),  well  within  the  montane  boreal  forest  of  the  Sierra  Madre  Occidental. 
It  would  appear  that  as  a  species,  S.  ochrognathus  is  the  cotton  rat  mostly 
highly  adapted  to  the  extreme  aridity  of  the  lower  slopes  of  the  "desert" 
mountains  of  the  northern  part  (Mesa  del  Norte)  of  the  Mexican  Plateau. 
This  kind  of  habitat,  in  the  more  mesic  southern  part  (Mesa  Central)  of  the 
Mexican  Plateau  in  such  states  as  Aguascalientes  and  Guanajuato,  is  occupied 
by  S.  leucotis. 

Parasite. — Dr.  Robert  Traub  identified  the  flea,  Pohjfienis  martinez-haezi 
Vargas,  1951,  from  S.  ochrognathus  from  near  Canatlan,  Durango. 

Specific  Characters  and  Comparisons 

The  drab  gray  dorsum  contrasting  with  an  ochraceous-colored 
nose  and  eye-ring  together  with  small  size  (for  captive  animals, 
maximum  weights  are  130  grams  for  a  male  and  133  for  a  non- 
pregnant female;  maximum  lengths  of  head  and  body  are  154  mm. 
and  149,  respectively)  distinguish  S.  ochroiinaihus  from  other 
species  in  the  S.  fulviventer  group.  Other  uni(|ue  characters  in- 
clude: small  and  elongate  auditory  bullae;  an  ob\ious  median  keel 
on  the  basioccipital;  pronounced  lateral  bulges  of  the  capsular 
projections  of  the  upper  incisors;  a  median-posterior  notch  on  the 
interparietal,  and  curxed  paraoccipital  processes  with  distinctive 
basal  notches. 

From  S.  alleni,  S.  oclirognatlius  further  differs  in  having  a  short 
and  broad  skull  rather  than  a  long  narrow  one,  and  slightly  recurved 
incisors  rather  than  pronounced  recurved    (opisthodont)    incisors. 


228  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

From  S.  fidvive titer,  S.  ochrognatlnis  is  further  distinguished  by 
whitish  instead  of  tawny  underparts;  less-arched  skull;  short  rather 
than  long  (in  relation  to  width)  basioccipital;  posterior  ends  of 
incisive  foramina  not  extending  (rather  than  extending)  to  a  line 
drawn  between  the  anterior  surfaces  of  the  first  upper  molars; 
anterior  end  of  mesopterygoid  fossa  broad  instead  of  narrow;  an- 
terior lip  of  foramen  magnum  obviously  notched  rather  than  not; 
and  foramen  ovale  small  (no  more  than  half  width  of  M3)  instead 
of  large  (at  least  three-fourths  width  of  M3). 

From  S.  leucotis,  S.  ochrognatlnis  is  further  distinguished  by  less- 
arched  skull;  short  and  broad  rather  than  long  and  narrow  basioc- 
cipital; posterior  ends  of  incisive  foramina  not  extending  (rather 
than  extending)  to  a  line  drawn  between  interior  surfaces  of  the 
first  upper  molars;  anterior  end  of  mesopterygoid  fossa  expanded 
instead  of  parallel-sided;  anterior  lip  of  foramen  magnum  obviously 
notched  rather  than  not;  length  at  midline  of  interparietal  more 
instead  of  less  than  2  mm.;  rostral  depressions  on  sides  of  premaxil- 
laries  slight  rather  than  pronounced;  angular  process  of  lower  jaw 
rounded  instead  of  slightly  hooked;  and  lingual  root  of  first  lower 
molar  large  instead  of  reduced  or  absent. 

From  S.  hispidus,  S.  ochrognathus  difi^ers  in  narrower  tail  scales, 
0.50  mm.  rather  than  0.75;  heavily  haired  instead  of  sparsely  haired 
tail;  short  and  broad  rather  than  long  and  narrow  skull;  short  instead 
of  long  (in  relation  with  width)  basioccipital;  less  arched  skull; 
deep  rather  than  shallow  palatal  pits;  and  conspicuous  instead  of 
slight  median  keel  on  the  palate. 

Geograph ic  Va riatio n 

Yellow-nosed  cotton  rats  have  been  considered  as  being  sep- 
arable into  three  subspecies  (Hall  and  Kelson,  1959:677-678) :  S.  o. 
madrensis  Goldman  and  Gardner,  S.  o.  montanus  Benson,  and  S.  o. 
ochrognathus  Bailey.  Later  S.  o.  madrensis  was  placed  in  synonymy 
under  S.  o.  hadeiji  J.  A.  Allen,  which  had  been  previously  regarded 
as  a  subspecies  of  S.  hispidus  (Baker  and  Greer,  1962:125).  The 
presence  of  distinctive  geographic  variation  in  this  species  with  its 
isolated  desert-mountain  populations  seems  at  first  logical,  although 
Findley  and  Jones  (1960),  after  a  thorough  examination  of  most 
museum  specimens  then  extant,  came  to  the  conclusion  that  no 
significant  geographic  variation  was  discernible.  They  thought  that 
the  presently  disjunct  environment  in  which  S.  ochrognatlnis  lives 
has  not  long  been  separated.    This  is  in  line  with  the  findings  of 


Baker — Sigmodon  fulviventer  Group  229 

Wells  (1966)  that  xerophilus  woodland  ( oak-pinon-juniper )  vege- 
tation, in  which  associations  S.  ochrognathus  lives  today,  occurred 
as  much  as  800  meters  lower  in  elevation  perhaps  11,560  to  more 
than  40,000  years  B.  P.  (during  the  Wisconsin  pluvial)  than  today. 
This  would  then  mean  that  these  disjunct  populations  have  not 
been  separated  for  much  more  than  10,000  years.  Furthermore,  it  is 
likely  that  this  xerophilus  woodland  maintained  its  continuity  in 
the  Mesa  del  Norte  up  to  a  postglacial  period  of  heavy  rains,  per- 
haps between  6210  and  7756  B.P.  (Findley  and  Jones,  1960:468), 
and  that  the  present  disjunct  montane  distribution  may  have  re- 
sulted from  ensuing  desiccation  after  the  above  dates.  Findley  and 
Jones  felt  that  the  disjunct  habitat  of  S.  ochrognathus  may  be  so 
newly  developed  that  isolation  has  not  been  long  enough  to  allow 
for  a  discernible  degree  of  geographic  variation.  Relying  in  part  on 
the  findings  mentioned  above  and  in  part  on  study  of  material  from 
Durangan  localities  from  which  S.  ochrognathus  lias  not  been 
reported  previously,  I  am  inclined  to  consider  yellow-nosed  cotton 
rats,  for  the  present,  as  belonging  to  a  monotypic  species. 

Sigmodon  ochrognathus  Bailey 

Sifimodon  ochroffiiafliiis  Bailey,  Proc.  Biol.   Soc.  Washington,   15:115,  June  2, 

1902. 
Sigmodon  baileyi  J.  A.  Allen,  Bull.  Amer.  Mus.  Nat.  Hist.,  19:601,  November 

12,  1903,  type  from  La  Cienega  de  las  Vacas,  2990  m.,  Durango. 
Sigmodon  ochrognaihus  montanus  Ben.son,  Proc.  Biol.  Soc.  Washington,  .5.3:1.57, 

December    19,    1940,   type    from    Peterson's   Ranch   "Sylvania"),    1860    m., 

3  km.  N  Sunnyside,  Huachuca  Mts.,  Cochise  Co.,  Arizona. 
Sigmodon  ochrognathus  madrensis  Goldman  and  Gardner,  Jour.  Mamm.,  28:58, 

February  17,  1947,  type  from  foothills  of  Sierra  Madre  Occidental,  50  km. 

NW  Parral,  1890  m..  Chihuahua. 

Table  6. — Selected  measurements  (averages  and  extremes)  of  wild- 
caught  AND  OF  laboratory-raised  (100  DAYS  OLD )  Signiodon  ochrognathus. 

9  km.  NXW  Canatlan,  .3  km.  NE  Boquilla, 
,,                     .                                                                                     Durango  Durango 

Measurement  (8  specimens,  (6  specimens, 

wild-taken)  laboratory-raised) 

Length  of  head  and  body  1.39  (1.32-144)  1.38(1.30-144) 

Length  of  hind  foot 28  (25-29)  29  (29-30) 

Height  of  ear  from  notch 21  ( 20-22  )  19  (  18-19 ) 

Condvlopremaxillary   length  31.2(30.2-32.0)  29.4(28.6-30.1) 

Zygomatic  breadth  18.7  (18.3-19.4)  18.1  (17.6-18.4) 

Least  interorbital  constriction  4.8  (4.6-4.9)  4.8  (4.6-5.1) 

Depth  of  cranium*  10.0(9.6-10.3)  9.5(9.4-9.7) 

Length  of  nasals  11.7  (11.0-12.5)  11.5  (11.2-12.1) 

Alveolar  length  of  max.  toothrow 6.3  (6.1-6.4)  6.1  (6.0-6.2) 


Measurement  taken  as  described  by  Findley  and  Jones   (1963:308). 


230  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Type. — Young  adult  female,  skin  and  skull;  no.  110333  U.S.  National 
Museiun;  from  Chisos  Moimtains,  2840  m.,  Brewster  Co.,  Texas;  obtained  on 
June  13,  1901,  by  Vernon  Bailey,  original  no.  7681. 

Range. — Montane  habitats,  mostly  arid,  rocky,  bunch-grass-covered  slopes 
in  pinon-oak-juniper  in  southern  Arizona,  southwestern  New  Mexico,  Trans- 
Pecos  Texas,  Chihuahua,  Coahuila,  and  Durango  (see  Fig.  20). 

Diagnosis. — See  account  of  specific  characters  and  comparisons. 

Remarks. — In  Talkie  6  are  presented  selected  external  and  cranial  measure- 
ments of  young  adult,  wild-taken  specimens  and  of  laboratory-raised,  100- 
day-old  specimens,  all  from  Durango.  The  two  samples  originated  from 
localities  approximately  175  km.  apart.  The  100-day-old  individuals  are  smaller 
and  less  mature  than  the  wild-taken  sample,  the  latter  judged  to  be  almost  adult. 

Specimens  examined  (41). — Chihuahua:  near  Parral,  1890  m.,  1  (US). 
Coahuila:  Tinaja  de  Telles,  1464  m..  El  Jardin  Ranch,  Sierra  del  Carmen, 
1  (TCWC);  Juarez  Caiwn,  1464  m.,  Sierra  del  Carmen,  16  (DMNH),  3  (US); 
33  km.  S  and  7  km.  W  Ocampo,  1616  m.,  3  (KU).  Durango:  Rancho  San- 
tuario,  4  (AMNH);  La  Cienega  de  las  Vacas,  1  (AMNH);  Arroyo  de  Bucij, 
1  (AMNH);  11  km.  NNE  Boquilla,  1952  m.,  2  (MSU);  2  km.  NE  Boquilla, 
1890  ;».,  1  (MSU);  3  km.  NE  Boquilla,  1952  m.,  2  (MSU);  9  km.  NNW 
Canatlan,  1950  m.,  11  (MSU);  2  km.  ESE  Atotonilco,  2043  m.,  1  (MSU). 

Other  records. — Chihuahua:  5  km.  SW  Pacheco;  13  km.  NE  Laguna,  2211 
m.;  3  km.  W  Minaca,  2104  m.;  Cherry  Ranch,  18  km.  NW  Cocomorachi 
(Findley  and  Jones,  1960:468).  Duraxgo:  Guanacevi  (Baker  and  Greer, 
1962:126). 

Literature  Cited 

Allen,  J.  A. 

1906.  Mammals  from  the  states  of  Sinaloa  and  Jalisco,  Mexico,  collected 
bv  J.  H.  Battv  during  1904  and  1905.  Bull.  Amer.  Mus.  Nat.  Hist., 
22:191-262,  pis.  20-23. 

Anderson,  S. 

1959.  Distribution,  variation,  and  relationships  of  the  montane  vole, 
Microtus  montanus.  Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.,  9:415- 
511,  12  figs. 

Bailey,  V. 

1902.  Synopsis  of  the  North  American  species  of  Sigmodon.  Proc.  Biol. 
Soc.  Washington,  15:101-116. 

Baker,  R.  H. 

1956.     Mammals    of   Coahuila,    Mexico.     Univ.    Kansas    Publ.,    Mus.    Nat. 

Hist.,  9:125-335,  75  figs. 
1963.     Geographical  distribution  of  terrestrial  mammals  in  Middle  America. 

Amer.  Midland  Nat.,  70:208-249,  1  fig. 
1966.     Further  notes  on  the  manmials  of  Durango,  Mexico,    four.  Nhunm., 

47:344-345. 
Baker,  R.  H.,  and  J.  K.  Greer 

1962.     Mammals  of  the  Mexican  state  of  Durango.    Michigan  State  Univ., 

Publ.  Mus.,  Biol.  Ser.,  2:25-154,  4  pis.,  6  figs. 

Baker,  R.  H.,  and  C.  J.  Phillips 

1965.  Mammals  from  El  Ne\'ado  de  Colima,  Mexico,  four.  Mamm.,  46: 
691-693. 

Baker,  R.  H.,  and  D.  Woxiochel 

1966.  Mammals  from  southern  Oaxaca.    Southwest.  Nat.,  11:306. 


Baker — Sigmodon  fulviventer  Group  231 

Blaih,  W.  F. 

1958.  Distributional  patterns  of  vertebrates  in  the  soutliern  United  States 
in  relation  to  past  and  present  en\'ironments.  Pp.  433-468  in  Zoo- 
geography (C.  L.  Hubbs,  ed.),  AAAS  Publ.,  51:.x  +  509  pp., 
ilhistrated. 

Chipman,  R.  K. 

1965.  Age  determination  in  the  cotton  rat  (Sigmodon  hispidiis).  Tulane 
Studies  Zool.,  12:19-38,  15  figs. 

COCKRUM,  E.    L. 

1948.  The  distribution  of  the  hispid  cotton  rat  in  Kansas.  Trans.  Kansas 
Acad.  Sci.,  51:306-312,  3  figs. 

CoHx,  T.  J. 

1965.  The  arid-land  katydids  of  the  North  American  genus  Neohancttia 
(Orthoptera:  Tettigoniidae ) :  their  systematics  and  a  reconstruction 
of  their  history.  Misc.  Publ.  Mus.  Zool.,  Univ.  Michigan,  126:1-179, 
frontispiece,  24  figs. 

Dalijy,  p.  L.,  and  H.  a.  Lillevik 

1969.  Taxonomic  analysis  of  electrophoretic  blood  serimi  patterns  in  the 
cotton  rat,  Sigmodon.  Michigan  State  Univ.,  Publ.  Mus.,  Biol.  Ser., 
4:65-104,  4  pis.,  11  figs. 

Davis,  W.  B. 

1944.  Notes  on  Mexican  mammals,    jour.  Mamm.,  25:370-403. 

Davis,  W.  B.,  axd  L.  A.  Follaxsbee 

1945.  The  Mexican  volcano  mouse,  Neotomodon.  Jour.  Mamm.,  26:401- 
411. 

Denyes,  H.  a. 

1956.  Natural  terrestrial  communities  of  Brewster  County,  Texas,  with 
special  reference  to  the  distribution  of  the  mammals.  Amer.  Mid- 
land Nat.,  55:289-320,  3  figs. 

DiLLOX,  L.  S. 

1956.  Wisconsin  climate  and  life  zones  in  North  America.  Science,  123: 
167-176,  13  figs. 

DOHAN,   D.   J. 

1955.  A  catalogue  of  the  Protozoa  and  helminths  of  North  American 
rodents.    III.  Nematoda.  Amer.  Midland  Nat.,  53:162-175. 

Drake,  J.  J- 

1958.  The  brush  mouse  Peromijscus  hoijlii  in  southern  Durango.  Michigan 
State  Univ.,  Publ.  Mus.,  Biol.  Ser.,  1:97-132,  3  pis.,  8  figs. 

FiXDLEY,  J.   S. 

1954.  Competition  as  a  possible  limiting  factor  in  the  distribution  of 
Microtus.   Ecology,  35:418-420,  1  fig. 

FiXDLEY,  J.   S.,  AXD  C.  J.  JOXES 

1960.  Geographic  \ariation  in  the  yellow-nosed  cotton  rat.  Jour.  Mamm., 
41:462-469,  1  fig. 

1962.  Distribution  and  \ariation  of  \oles  of  the  genus  Microtus  in  New 
Me.xico  and  adjacent  areas.    Jour.  Mamm.,  43:154-166,  5  figs. 

1963.  Geographic  variation  in  the  least  cotton  rat  in  New  Mexico.  Jour. 
Mamm.,  44:307-315,  9  figs. 

Goldman,  E.  A. 

1951.  Biological  investigations  in  Mexico.  Smithson.  Misc.  Coll.,  115:.\iii 
+  476  pp.,  frontispiece,  70  pis.,  1  map. 

Goodwin,  G.  G. 

1955.  Three  new  cotton  rats  from  Tehuantepec,  Mexico.  Amer.  Mus. 
Novit.,  1705:1-5. 

Hall,  E.  R.,  and  K.  R.  Kelsox 

1959.  The  mannnals  of  North  America.  Ronald  Press  Co.,  New  York. 
2:\iii  +  547-1083  +  79  pp.,  figs.  313-553,  maps  321-500. 


232  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Hall,  E.  R.,  and  B.  Villa-R. 

1949.  An  annotated  check  list  of  tlie  mammals  of  Michoacan,  Mexico. 
Univ.  Kansas  Publ,  Mus.  Nat.  Hist.,  1:431-472,  2  pis.,  1  fig. 

Hershkovitz,  p. 

1962.  Evolution  of  Neotropical  cricetine  rodents  (Muridae)  with  special 
reference  to  tlie  phyllotine  group.  Fieldiana:  Zool.,  46:1-524, 
frontispiece,  123  figs. 

1966.  Mice,  land  l)ridges  and  Latin  American  faunal  interchange.  Pp.  725- 
751  in  Ectoparasites  of  Panama  (R.  L.  Wenzel  and  V.  J.  Tipton, 
eds.).  Field  Mus.  Nat.  Hist.,  xii  +  861  pp.,  93  pis.,  154  figs.,  1  map. 

HiBBARD,   C.    W. 

1960.  An  interpretation  of  Pliocene  and  Pleistocene  climates  in  North 
America.  62nd  Ann.  Rep.,  Michigan  Acad.  Sci.,  Arts,  Letters, 
1959-60,  pp.  5-30,  2  figs. 

HOFFMEISTER,    D.    F. 

1963.  The  yellow-nosed  cotton  rat,  Sigmodon  ochrognathus,  in  Arizona. 
Amer.  Midland  Nat.,  70:429-441,  7  figs. 

Hooper,  E.  T. 

1955.  Notes  on  mammals  of  western  Mexico.  Occas.  Papers  Mus.  Zool., 
LTniv.  Michigan,  565:1-26. 

1961.  Notes  on  mammals  from  western  and  southern  Mexico.  Jour. 
Mamm.,  42:120-122. 

1962.  The  glans  penis  in  Sigmodon,  Sigmomys,  and  Rcitlirodon  ( Roden- 
tia,  Cricetinae).  Occas.  Papers  Mus.  Zool.,  Univ.  Michigan,  625: 
1-11,  2  figs. 

Jones,  J.  K.,  Jr. 

1964.  Distribution  and  taxonomy  of  mammals  of  Nebraska.  Univ.  Kansas 
Publ.,  Mus.  Nat.  Hist.,  16:1-356,  pis.  1-4,  82  figs. 

Leopold,  A.  S. 

1959.  Wildlife  of  Mexico  the  game  birds  and  mammals.  Univ.  California 
Press,  Berkeley  and  Los  Angeles,  xiii  +  568  pp.,  194  figs.,  illustrated. 

Meyer,  B.  J.,  and  R.  K.  Meyer 

1944.  Growth  and  reproduction  of  the  cotton  rat,  Sigmodon  liispidiis 
hispidus.  under  laboratory  conditions.  Jour.  Mamm.,  25:107-129, 
2  pis. 

Mohlhenrich,  J.  S. 

1961.  Distribution  and  ecology'  of  the  hispid  and  least  cotton  rats  in  xNew 
Mexico.    Jour.  Mamm.,  42:12-24,  3  figs. 

Odum,  E.  p. 

1955.  An  ele\'en  year  history  of  a  Sigmodon  population.  Jour.  Mamm., 
36:368-378,  1  pi,  4  figs. 

Ridgway,  R. 

1912.  Color  standards  and  color  nomenclature.  Privately  printed,  Wash- 
ington, D.C.,  iii  -f  43  pp.,  53  colored  pis. 

Slaughter,  B.  H. 

1966.     The   Moore   Pit   local   famia;   Pleistocene   of  Texas.     Jour.   Paleont., 
40:78-91,  6  figs. 
Taylor,  W.  P.,  W.  B.  McDougall,  C.  C.  Presnall,  and  K.  P.  Schmidt 

1945.  Preliminary  ecological  sur\ey  of  the  northern  Sierra  del  Carmen, 
Coahuila,  Mexico.    Mimeo.,  48  pp.,  16  figs.,  1  map. 

Webb,  R.  C,  and  R.  H.  Baker 

1962.  Terrestrial  vertebrates  of  the  Pueblo  Nuevo  area  of  southwestern 
Durango,  Me.xico.    Amer.  Midland  Nat.,  68:325-333,  1  fig. 

Wells,  P.  V. 

1966.  Late  Pleistocene  vegetation  and  degree  of  pknial  climatic  change 
in  the  Chihuahuan  Desert.    Science,  153:970-975,  2  figs. 


CONE  CACHES  AND  MIDDENS  OF  TAMIASCIURUS 
IN  THE  ROCKY  MOUNTAIN  REGION 


BY 


Robert  B.  Finley,  Jr. 

The  red  squirrel  (Tamiasciurus  hudsonicus)  of  the  Rocky  Moun- 
tain region  is  an  animal  of  the  coniferous  forests.  Its  conspicuous 
middens  of  cone  debris  are  easily  recognized  throughout  the  Rocky 
Mountains  from  Alaska  to  Arizona.  In  this  \ast  region  deciduous 
hardwood  forests  are  of  minor  extent  and  Httle  importance  as  pro- 
ducers of  food  for  squirrels.  Tamiasciurus  is  well  adapted  to  these 
conditions  and  is  able  to  subsist  on  almost  any  of  the  conifer  seed 
crops  that  may  be  a\'ailable.  In  the  West,  red  squirrels  are  able  to 
survi\'e  a  winter  of  complete  cone  crop  failure  (M.  C.  Smith,  1968) 
but  rarely,  if  ever,  occupy  hardwood  stands  without  conifers. 

In  the  eastern  United  States  and  adjacent  Canada  red  squirrels 
are  likewise  primarily  dependent  on  conifer  seed  crops,  but  they 
also  make  considerable  use  of  hardwood  mast  where  available  and 
are  able  to  exist  in  some  deciduous  forests  without  conifers  (Hatt, 
1929:43). 

As  to  be  expected  of  a  species  so  wide  ranging,  the  red  sc^uirrel 
has  a  wide  adaptability  to  different  habitats  and  food  supplies.  It  is 
not  known  to  what  extent  this  adaptability  represents  genetic  dif- 
ferences between  populations  and  to  what  extent  it  is  behavioral 
response  to  one  environmental  condition  or  another.  Much  informa- 
tion has  been  published  on  the  life  history  and  behavior  of  red 
squirrels  of  the  northeastern  (Klugh,  1927;  Hatt,  1929;  Layne, 
1954)  and  northwestern  (Shaw,  1936;  C.  C.  Smith,  1968)  United 
States.  Because  much  less  has  been  reported  for  the  Rocky  Moun- 
tain region,  a  contribution  to  knowledge  of  red  squirrel  habitat 
relations  is  presented  here. 

The  caching  habit  of  red  squirrels  has  long  been  known  and 
exploited  in  the  West  by  foresters  and  nurserymen  as  a  source  of 
conifer  seeds  for  planting  (Cox,  1911:17).  It  is  usually  the  cheapest 
source  for  large  quantities  of  seed  of  the  commercially  and  horti- 
culturally  desirable  spruce,  fir,  and  pine  trees,  and  the  seeds  thus 
harvested  have  high  \iability  ( Lavender  and  Engstrom,  1956 ) . 
However,  some  timber  managers  accuse  the  squirrels  of  harvesting 
such  a  high  percentage  of  cones  of  certain  species  as  to  prevent 

(233) 


234  Misc.  Publ.  51,  Uxiv.  Kansas  Mus.  Nat.  Hist. 

adequate  natural  forest  regeneration.  Inasmuch  as  there  seems  to 
be  Httle  information  in  the  hterature  on  the  economic  status  of 
Tamiasciurus  in  the  Rocky  Mountains,  some  observations  on  this 
aspect  are  also  given. 

This  paper  is  based  on  examination  of  sexeral  hundred  middens 
and  caches  of  Tamiasciurus  hudsonicus  fremontii,  T.  li.  mogoUonen- 
sis,  and  T.  h.  richardsoni  in  Colorado,  northern  Arizona,  and  western 
Montana  in  the  period  1963  to  1968.  Approximately  60  of  these 
were  excavated  for  cone  caches,  and  34  of  the  more  noteworthy 
middens  were  described  in  some  detail.  Many  of  the  middens  in 
the  Front  Range  of  Colorado  were  examined  repeatedly  in  different 
years. 

Much  general  information  was  gained  from  Harrx'  M.  Swift, 
nurseryman  and  owner  of  Colorado  Evergreens,  Inc.,  of  Golden, 
Colorado,  who  has  had  many  years  of  experience  collecting  and 
marketing  conifer  seeds  from  squirrel  caches  in  Colorado.  I  am 
indebted  to  Curtis  H.  Halvorson  for  many  stimulating  discussions 
of  his  red  squirrel  studies  in  Montana  and  for  reviewing  a  draft  of 
this  paper.  William  B.  Finley,  U.S.  Forest  Service,  first  acquainted 
me  with  the  remarkably  productixe  cone  caches  on  the  Kaibab 
Plateau. 

Characteristics  and  Variation  of  Middens  and  Caches 
The  importance  of  conifer  seed  in  the  diet  of  the  red  squirrel  is 
strikingly  evident  to  anyone  who  examines  one  of  the  great  midden 
heaps  under  a  dense,  shady  stand  of  blue  spruce  (Picea  pungens) 
alongside  a  mountain  stream  in  Colorado.    The  larger  middens  are 
freciuently  20  to  30  feet  across,  one  to  one  and  a  half  feet  deep  in  the 
center,  and  carpet  the  ground  to  the  exclusion  of  all  living  plants. 
The  surface  is  usually  littered  with  fresh  cone  scales   and  cores 
dropped  by  the  squirrels.    The  material  below  is  loose  and  damp, 
easily  dug  into  with  the  bare  hands.    As  one  digs  into  the  deeper 
deposits  the  material  at  lower  lexels  is  found  to  be  older  and  more 
decomposed,  forming  a  rich  mulch  in  contact  with  the  mineral  soil. 
Large  middens  must  be  decades  old  and  represent  the  accumula- 
tions of  many  successive  generations  of  red  squirrels.    Such  a  mid- 
den is  an  example  of  ecological  homeostasis,  for  its  owner  can  easily 
store  in  it  a  much  larger  supply  of  food  in  good  condition  than  a 
squirrel  could  store  under  a  similar  tree  stand  not  previously  oc- 
cupied by  red  scjuirrels.   Cache  holes  are  much  harder  to  dig  in  soil 
and  the  cones  dry  out  (juicker  if  not  covered  with  cone  litter. 


FiNLEY — Cone  Caches  of  Tamaisciurus  235 

The  continuing  utility  of  such  a  long-used  midden  is  illustrated 
by  a  medium-sized  one  I  saw  in  1963  under  a  recently  dead  spruce 
tree  57  feet  tall  near  Corral  Park,  8600  feet,  in  the  Rio  Grande 
National  Forest,  Colorado.  The  midden  was  exposed  in  the  sunlight 
but  still  in  use  by  the  squirrel  chattering  from  a  dead  limb  above. 
Almost  five  years  later  the  midden  was  still  acti\e,  as  evidenced  by 
an  accumulation  of  fresh  cone  scales  and  cores  of  blue  spruce 
around  the  base  of  the  dead  tree.  The  midden,  on  a  north-facing 
slope,  was  still  damp  below  a  depth  of  three  inches  in  mid-June. 
Sparse  grasses  and  blooming  red  columbines  (Aquilegia  elegantula) 
were  growing  on  its  outer  portion.  There  was  a  grass  nest  about  30 
feet  high  in  a  live  blue  spruce  standing  20  feet  upslope  from  the 
dead  tree. 

Active  middens  remain  fairly  loose,  with  fresh  or  discolored  cone 
scales  and  cores  on  the  surface,  but  old  inactive  middens  become 
mixed  with  needle  litter,  compacted,  weathered,  and  decomposed. 
Abandoned  middens  eventuallv  become  covered  with  a  laver  of 
forest  duff;  if  exposed  to  some  sunlight,  they  gradually  become 
coxcred  o\'er  with  encroaching  grasses,  forbs,  mosses,  kinnikinnick 
(Arctostapljylos  tivo-tirsi),  or  low  shrubs. 

Kinds  of  Cones  Cached 

Although  cones  of  nearly  all  species  of  needle-leaved  trees  in 
Colorado  are  cached  for  food,  there  are  differences  in  preference. 
In  general,  red  squirrels  seem  to  prefer  cones  of  blue  and  Engel- 
mann  spruce  (Picea  en<i,ebnonni)  and  Douglas-fir  (Pseudotsuga  men- 
ziesii).  Cones  of  ponderosa,  limber,  and  bristlecone  pines  (Pinus 
ponderosa,  P.  fexilis,  and  P.  aristata)  are  also  abundanth'  cached  at 
times,  but  seem  to  be  somewhat  less  preferred.  Cones  of  lodgepole 
pine  (Pinus  contorfa)  provide  a  staple  food  supply  when  other  cone 
crops  fail,  but  are  often  passed  over  when  one  or  more  of  the  above- 
named  cone  crops  is  good. 

Use  of  Engelmann  spruce  and  lodgepole  pine  by  squirrels  en- 
ables them  to  occupy  thousands  of  square  miles  of  montane  forest 
in  which  other  cone  supplies  are  absent  or  erratic.  Two  reasons  for 
this  are  the  great  extent  of  these  two  forest  types  in  the  Canadian 
Life-Zone  and  the  relatively  high  reliability  of  their  cone  crops. 
Lodgepole  pines  produce  cones  almost  every  year,  and  in  this  region 
a  high  percentage  of  the  cones  are  serotinous  (persist  on  the 
branches  unopened  throughout  the  year),  thus  proxiding  an  excel- 
lent reserve  food  supply.  Engelmann  spruce  in  Colorado  produces 
a  good  or  moderate  cone  crop  two  out  of  three  years,  on  the  average. 


236  Misc.  Publ.  51,  Uxiv.  Kansas  Mus.  Nat.  Hist. 

Douglas-fir,  limber  pine,  and  bristlecone  pine  are  much  more 
limited  in  distribution  and  produce  a  good  cone  crop  in  Colorado 
only  one  year  out  of  four  or  five.  But  when  cones  of  these  species 
are  produced  at  all,  they  are  invariably  cached,  perhaps  because  of 
their  unusually  large  seeds.  The  ponderosa  pine  is  a  somewhat 
erratic  cone  producer.  Its  cones  are  usually  cached  to  some  extent 
when  available,  but  may  be  ignored  if  a  good  crop  of  spruce  or 
Douglas-fir  is  at  hand. 

There  is  one  other  species  of  pine  native  to  Colorado,  the  pinyon 
(Pinus  echilis),  but  I  have  no  knowledge  of  the  use  of  its  cones  by 
Tamiasciurus.  Although  the  pinyon  usually  occurs  zonally  below 
the  red  squirrel  and  on  dry  sites,  in  some  places  it  grows  rather 
high  on  south  slopes  and  in  contact  with  other  forest  types  on  north 
slopes  occupied  by  the  squirrels. 

Although  the  pinyon  produces  a  good  seed  crop  only  once  in 
several  years,  its  seeds  in  these  years  are  large  and  heavily  con- 
sumed by  mice,  chipmunks,  and  other  animals.  Failure  of  red 
squirrels  to  utilize  pinyon  seed  crops  is  probably  explained  primarily 
by  the  dryness  of  the  pinyon-juniper  woodland  type.  It  would  be 
almost  impossible  to  find  a  damp  caching  site  under  pinyon  trees 
except  under  water  in  the  few  permanent  streams. 

According  to  H.  M.  Swift,  cones  of  all  three  kinds  of  true  fir  in 
Colorado  are  cached — the  white  fir,  subalpine  fir,  and  corkbark  fir 
(Abies  concolor,  A.  lasiocarpa,  and  A.  lasiocarpa  var.  arizonica).  I 
have  seen  a  few  caches  of  subalpine  fir  cones,  which  were  stored  in 
somewhat  smaller  amounts  than  cones  of  the  Engelmann  spruce  in 
the  same  stand.  White  fir  cones  are  heavily  cached  by  red  squirrels 
on  the  Kaibab  Plateau,  Arizona,  according  to  W.  B.  Finley. 

Junipers  {Juniperus  sp.)  are  present  in  most  of  the  forest  types 
of  Colorado,  up  to  elevations  of  10,000  to  12,000  feet,  but  I  have  not 
found  any  juniper  seeds  or  berries  in  midden  caches,  nor  has  Mr. 
Swift  seen  any  use  of  juniper  seed. 

In  spite  of  the  importance  of  conifer  seeds  as  winter  food,  the 
red  squirrel  is  able  to  survive  a  year  of  complete  cone  crop  failure. 
In  Colorado  there  are  few  forests  in  which  this  is  likely  to  occur, 
because  of  the  wide  availability  of  Engelmann  spruce  and  lodgepole 
pine.  However,  in  southern  Colorado  red  squirrels  occur  in  forests 
40  miles  or  more  from  the  nearest  lodgepoles.  When  cones  failed 
there,  Mr.  Swift  found  the  squirrels  had  cached  small  piles  of  kin- 
nikinnick  berries,  bird  cherries  (Primus  pennsylvanicus),  and  squaw 
currants  (Ribes  cereum). 


FiNLEY — Cone  Caches  of  Tamaisciurus  237 

Kinds  of  'Midden  Sites  Used  for  Caches 

Earlier  observers  have  described  the  cache  sites  of  red  squirrels 
on  or  in  the  ground  in  various  kinds  of  situations.  Nearly  all  agree 
tliat  caches  are  characteristically  in  damp,  shady  locations,  such  as 
beneath  a  dense  tree  canopy  or  in  a  spring  or  boggy  area.  The 
importance  of  moisture  for  preserving  the  cones  in  closed  condition 
was  briefly  mentioned  by  Grinnell  and  Storer  (1924:206).  Shaw 
(1936)  demonstrated  the  ability  of  cones  of  white  fir  (Abies  con- 
color)  and  white  pine  (Finns  monticola),  after  storage  for  two  years 
in  \\  ater,  to  open  and  begin  to  shed  their  seeds  after  eight  days  of 
exposure  to  dry  air.  The  high  adaptive  value  of  this  caching  habit 
is  illustrated  by  the  varying  degrees  to  which  red  squirrels  in 
different  climatic  regimes  and  site  situations  restrict  their  choice  of 
cone  caching  sites. 

Caches  on  the  high  plateaus  of  the  Colorado  f-lockies  often  cover 
extensive  level  ground  under  stands  of  spruce,  fir,  and  lodgepole 
pine.  Cones  are  buried  in  shallow  cone  litter,  duff,  or  soft  soil,  or 
stuffed  under  the  edges  of  logs  and  roots  of  stumps.  Such  sites 
remain  damp  most  of  the  summer  and  autumn  because  of  high 
rainfall  and  the  early  arrival  of  snow  cover. 

At  lower  elevations  (below  9000  feet)  on  the  east  side  of  the 
Continental  Divide  the  squirrels  rarely  cache  cones  in  soil  or  thin 
duff,  but  almost  solely  in  the  deeper,  looser  cone  debris  of  the 
middens  left  by  their  own  feeding  activities. 

Mountain  slope  in  relation  to  insolation  and  density  of  tree 
canopy  are  also  vital  factors  at  lower  elevation,  severely  restricting 
the  sites  that  are  utilized  by  red  squirrels  for  cone  caching.  Middens 
are  almost  totally  absent  from  the  ponderosa  pine  stands  on  south- 
facing  slopes,  and  are  sparse  and  small  under  Douglas-fir  and  spruce 
stands  on  north-facing  slopes.  They  are  largest  and  closest  together 
under  blue  spruces  along  drainage  bottoms  at  the  foot  of  north- 
facing  slopes  ( see  Fig.  1 ) .  Protection  against  the  sun,  either  by  tree 
crowns  or  topographic  exposure,  seems  to  be  of  great  importance 
below  9000  feet  in  Colorado,  except  where  hillside  drainage  brings 
water  directly  into  the  midden. 

The  life  form  of  a  tree  has  an  important  bearing  on  its  suitability 
for  providing  a  good  midden  site.  Spruces  retain  their  lower 
l^ranches  almost  to  the  ground,  thus  shading  the  ground  beneath 
the  tree  through  almost  all  hours  of  the  day.  In  dense  stands,  the 
lower  branches  are  lost,  but  on  such  sites  the  ground  is  shaded  any- 


238 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Fig.  1.    Blue  spruce  stand  by  Redskin  Creek,  Pike  National  Forest,  Colorado. 

This  low,  dense  tree  crown  provides  maximum  shade  for  red  squirrel  caches 

in  the  Transition  Life-Zone  below  9000  feet. 


way  by  the  closed  crown  canopy.  Spruces  provide  ideal  midden 
sites. 

Douglas-firs  retain  a  fairly  dense  crown  well  down  the  trunk  in 
open  stands,  but  the  ground  level  is  usually  fairly  open  and  exposed 
to  sunlight  unless  the  trees  are  in  a  closed  stand.  In  closed  stands, 
these  trees  provide  adequate  midden  sites,  but  a  single  tree  rarely 
shelters  a  midden  in  Colorado. 

Ponderosa  pines  have  few  limbs  on  the  lower  portion  of  the 
trunk.  In  Colorado  they  nearly  always  occur  in  fairly  open  stands, 
or  are  mixed  with  Douglas-fir.  The  ground  under  these  pines  is 
nearly  always  open  to  sunlight  most  of  the  day.  This  growth  form, 
combined  with  the  dry  sites  occupied,  makes  the  ponderosa  pine 
quite  unsuitable  as  cover  for  squirrel  caches  (see  Fig.  2).  Conse- 
quently they  can  provide  a  winter  food  supply  only  at  the  edges  of 
the  pine  forest  where  they  grow  close  to  other  trees  that  produce 
better  shade. 

Lodgepole  pines  replace  ponderosa  pines  in  the  Canadian  Life- 
Zone  at  higher  elevations.  Lodgepoles  likewise  have  few  branches 
on  the  lower  trunk,  but  they  usually  occur  in  dense  even-age  stands 


FiNLEY — Cone  Caches  of  Tamaisciurus 


239 


Fk;.  2.  Ponderosa  pine  stand  on  dry  open  slope  with  southerly  exposure  near 
Redskin  Creek,  Pike  National  Forest.  This  type  of  tree  form  does  not  provide 
enouj:;h  shade  and  soil  moisture  for  cone  caching  in   the  Colorado  Transition 

Zone. 

with  a  closed  canopy  that  provides  a  moderate  amount  of  shade. 
The  sites,  however,  are  usually  dry,  and  middens  in  lodgepole  pine 
stands  are  usually  small  and  dry.  Because  the  cones  are  commonly 
serotinous,  their  seeds  remain  available  throughout  the  year. 

Oddly  enough,  red  squirrels  seem  to  know  that  lodgepole  cones 
do  not  ha\'e  to  be  cached  in  wet  places  to  remain  closed,  for  the 
scjuirrels  commonly  \eave  them  on,  or  close  to,  the  surface  of  a 
midden,  or  pile  them  in  Httle  heaps  against  logs,  stumps,  or  tree 
trunks  (see  Fig.  5).  Because  of  the  greater  and  more  continuous 
snow  cover  at  higher  elevations,  middens  in  lodgepole  stands  may 
remain  damp  near  the  bottom,  but  such  levels  are  not  much  used 
to  cache  lodgepole  cones. 

Bristlecone  pines  and  limber  pines  also  occur  at  high  elevations 
and  on  usually  poorer  sites  than  lodgepoles,  growing  even  up  to 
timberline.  In  Colorado,  the  high  precipitation  and  nearly  con- 
tinuous winter  snow  cover  at  these  elevations  allow  middens  to 
remain  fairly  damp  under  moderately  dense  limber  and  bristlecone 
stands. 


240 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Fig.  3.    Part  of  a  large  complex  Taniiasciurus  midden  in  an  Engelmann  spruce 

forest  near  Arapaho  Basin,  Summit  County,  Colorado.    A  lied  of  spruce  cone 

litter  covers  the  entire  forest  floor  in  the  foreground. 

The  true  firs  (Abies),  like  the  spruces,  retain  branches  close  to 
the  ground  and  provide  ample  shade  for  middens.  In  addition,  they 
occur  on  moist  sites  and  at  high  elevation  where  middens  and  caches 
under  them  have  little  opportunity  to  dry  out.  Abies  cones  may  be 
found  piled  under  the  edges  of  logs,  inside  rotten  logs,  or  even  piled 
on  the  surface.  Though  they  are  usually  buried  in  the  wet  midden, 
those  left  exposed  to  the  air  soon  become  covered  under  snow  and 
are  thus  protected  against  drying. 

In  view  of  the  wide  variety  of  middens  and  cone-caching  situa- 
tions used,  I  have  selected  a  few  examples  illustrating  this  diversity 
from  among  the  34  studied  in  detail.  Descriptions  of  these  are  given 
in  the  following  paragraphs.  Tree  measurements  were  made  with  a 
clinometer  and  a  tree  tape.  Timber  stand  densities  were  estimated 
from  three  stem  counts  by  angle  gauge,  expressed  in  square  feet  of 
basal  area  per  acre. 

Engelmann  spruce — snhalpine  fir  foiesi. — One  of  the  largest  middens  I 
have  seen  was  characteristic  of  those  found  at  higher  elevations  in  the  spruce- 
fir  forests  (see  Fig.  3).  It  was  imder  a  stand  of  Engelmann  spruce  in  a  small 
swale  at  10,000  feet  elevation  on  a  lower  south-facing  slope  3  mi.  W  Arapaho 
Basin,    Summit    County,    Colorado.     The    midden    was    roughly    triangular    in 


FiNLEY — Cone  Caches  of  Tamaisciurus  241 

shape,  measuring  60,  58,  and  54  feet  along  the  three  sides,  with  1440  square 
feet  of  surface  area.  Outside  this  area  there  were  several  other  small  middens 
and  feeding  stations  covered  with  cone  scales  around  individual  trees.  When 
I  first  examined  the  midden  in  October  of  1963,  it  was  crammed  full  of 
Engelmann  cones  stuffed  in  little  pits  two  to  three  inches  in  diameter  that 
riddled  a  layer  of  cone  debris  covering  the  ground.  The  midden  was  about 
10  to  15  inches  deep  over  the  western  half  of  the  area,  decreasing  in  depth 
toward  the  east  side.  I  dug  six  bushels  of  cones  out  of  it  without  exhausting 
the  supply.  On  August  .3,  1968,  the  same  midden  had  changed  considerably 
in  appearance,  the  west  side  being  nearK'  inacti\e  and  having  become  more 
compacted  and  reduced  to  depths  of  only  three  to  10  inches.  The  area  of 
greatest  acti\ity  had  shifted  to\\ard  the  east  and  south  sides  where  most  of 
the  recent  cone  debris  was  accumulating.  The  cone  crop  of  1968  promised  to 
be  almost  a  complete  failure,  and  the  only  cached  Engelmann  spruce  cones 
I  foimd  in  the  midden  were  a  few  wet,  black,  but  still  sound  cones  of  the 
1967  crop.  The  midden  was  still  damp  at  depths  below  six  or  seven  inches, 
but  dry  in  the  upper  6-inch  layer,  except  in  open  areas  where  the  top  two  to 
three  inches  were  wet  from  recent  rains.  Most  of  the  fresh  cone  debris  on 
the  surface  was  dry,  light  brown  scales  of  1967  cones,  many  of  which  still 
could  be  seen  in  the  tree  tops.  They  probably  retained  only  a  small  percentage 
of  the  original  seed  content.  A  further  indication  of  food  shortage  in  1968 
was  the  large  number  of  lodgepole  pine  cones  and  cone  debris  in  ( and  on )  the 
topmost  layer  of  the  active  midden.  Almost  no  lodgepole  cones  were  cached 
in  this  midden  in  1963,  a  good  seed  year  for  spruce. 

Sometime  in  the  winter  or  spring  of  1968  an  Engelmann  spruce  18  inches 
d.b.h.  (diameter  at  breast  height)  and  89  feet  tall  blew  over  on  the  south 
side  of  the  midden  downslope  from  the  area  of  principal  activity  of  the  squirrel 
in  1963.  This  event  may  have  influenced  the  squirrel  to  shift  its  center  of 
feeding  activity,  for  the  deepest  piles  of  fresh  cone  scales  in  August  of  1968 
were  around  the  stiunp  and  under  the  butt  end  of  the  tree.  The  lower  10  feet 
of  the  log  was  much  used  as  a  feeding  area.  The  foliage  was  still  fairly  green 
and  the  top  branches  still  held  many  open  and  partly  open  cones. 

The  spruce  stand  was  fairly  open  oxer  the  midden,  averaging  140  square 
feet  basal  area  per  acre.  It  enclosed  14  dominant  and  codominant  trees  from 
seven  to  19  inches  d.b.h.,  the  largest  being  81  feet  tall,  and  eight  trees  of  pole 
and  sap'ing  size.  Two  of  the  codominant  trees  were  subalpine  firs  with  eight 
and  nine  inches  d.b.h.  Many  small  lodgepole  pines  stood  on  a  low  ridge 
beyond  the  east  end  of  the  midden.  The  light  coming  through  the  tree  canopy 
oxer  the  midden  xxas  sufficient  for  a  thin  groimd  coxer  of  Arnica,  Rosa,  Salix, 
and  Louiccia  invohicrata.  Green  foliage  cuttings  of  spruce  lay  scattered  over 
the  midden.  The  southxx'est  half  of  the  midden  xvas  littered  xvith  hundreds  of 
slender  dead  xvilloxv  branches  from  dead  xxilloxv  climips. 

In  August  of  1968,  I  counted  13  grass  nests  in  trees  xxithin  the  periphery 
of  the  midden,  the  largest  number  I  haxe  seen  at  any  single  midden.  There 
xvere  five  in  a  large  spruce  standing  directly  over  the  middle  of  the  fallen  tree. 
The  highest  nest  was  34  feet  up  in  a  subalpine  fir.  The  loxvest  xvas  58  inches 
aboxe  the  ground  in  drooping  iManches  of  a  small  spruce  located  six  feet  from 
the  fallen  spruce  stump.  This  nest  measured  13  inches  xvide  and  16  inches  deep. 
It  xxas  a  nearly  solid  ball  of  dry  grass  xxdth  a  fex\'  forb  leaves,  strands  of  xxilloxv 
bark,  and  some  fine  dusty  debris  in  the  center. 


242  Misc.  Publ.  51,  Unw.  Kansas  Mus.  Nat.  Hist. 

Many  similar,  but  smaller,  middens  were  scattered  through  the  spruce-fir 
forest  on  adjacent  mountainsides.  Crown  canopy  closure  and  shade  on  the 
ground  seemed  to  be  adequate  for  some  caching  almost  anywhere  in  the  forest. 
Presence  of  good  seed-producing  trees  and  logs  or  stumps  for  feeding  posts 
may  have  been  factors  initiating  the  accumulation  of  middens  at  certain  sites. 
Engelmann  spruce  cones  had  been  cached  abundantly  in  most  of  these  middens 
in  previous  years,  but  were  scarce  in  August,  1968,  whereas  some  lodgepole 
cones  were  commonly  seen  on  the  middens  or  piled  against  logs.  One  large 
midden  (50  by  47  feet)  among  many  rotten  logs  contained  some  1967  cones 
of  subalpine  fir  in  addition  to  those  of  lodgepole  pine.  It  was  the  only  midden 
at  which  I  ha\'e  seen  a  grass  ball  nest  on  the  groimd,  nearly  hidden  under  a 
large  rotten  log. 

Blue  spruce  in  valley  bottoms. — A  large  midden  in  a  typical  situation  for 
the  lower  elevations  east  of  the  Continental  Di\ide  was  first  examined  in 
October  of  1963,  and  examined  again  in  ]u\\e  of  1968.  It  was  imder  a  dense 
stand  of  large  blue  spruce  trees  on  the  south  side  of  Beaver  Creek,  8300  feet 
elevation,  below  Beaver  Creek  Campground,  Rio  Grande  County,  Colorado. 
The  spruces  stood  on  an  alluvial  flat  in  a  ravine  about  40  feet  deep,  cut  by 
the  creek  through  volcanic  rock.  The  creek  was  about  20  feet  wide  where  it 
passed  the  spruce  flat.  The  steep  slope  north  of  the  creek  as  well  as  the  upper 
flats  both  north  and  south  of  the  ravine  were  grassy,  with  scattered  mature 
ponderosa  pines.  A  few  Douglas-firs  and  ponderosa  pines  as  well  as  spruces 
grew  on  the  50  degree  north-facing  slope  of  the  ra\ine. 

The  squirrel  midden  was  a  low,  oval  mound  of  cone  litter  under  and  around 
three  blue  spruce  trees,  two  of  which  were  among  the  tallest  in  the  stand, 
which  extended  about  50  yards  along  the  creek.  There  were  52  trees  of  about 
10  or  more  inches  d.b.li.  in  this  stand.  The  basal  area  at  three  sample 
points  in  the  stand  averaged  167  square  feet  per  acre.  This  density  of  timber 
was  enough  to  exclude  almost  all  understory  vegetation,  there  being  little 
ground  cover  except  needle  litter  and  cone  debris.  The  dominant  and  co- 
dominant  trees  were  75  to  90  feet  tall. 

The  midden  was  34  feet  long  and  27  feet  wide,  co\ering  an  area  of  approxi- 
mately 722  square  feet,  with  a  maximum  depth  of  16  inches.  The  contents 
were  a  mixture  of  cone  debris  of  blue  spruce  and  ponderosa  pine. 

The  largest  blue  spruce,  under  which  recent  cone  debris  was  piled  highest, 
measured  22  inches  d.b.h.  and  105  feet  liigh.  The  second  tree  on  the  midden 
was  nearly  as  large,  and  the  third  was  12  inches  d.b.h.  and  85  feet  tall.  Two 
spruces  of  even  larger  diameter  (25  and  29  inches)  stood  on  the  creek  bank 
about  15  feet  north  of  the  midden.  There  were  also  t\vo  cut  10-inch  stumps 
in  the  midden  that  were  used  as  feeding  posts,  and  a  tangle  of  dead  branches 
between  the  trees  and  the  steep  ravine  bank.  Two  grass  nests  could  be  seen 
in  the  trees  over  tlie  midden.  One  was  at  a  height  of  23  feet  and  the  other  at 
43  feet  in  the  foliage  of  the  main  blue  spruce  in  the  center  of  the  midden. 

There  was  much  cone  litter  scattered  elsewhere  on  the  flat  and  fairly  recent 
accumulations  under  a  few  other  spruce  trees  in  die  stand,  but  there  were  no 
middens  deep  enough  to  provide  good  cone  caching  sites. 

In  October,  1963,  I  dug  about  \\\'o  bushels  of  ponderosa  pine  and  blue 
spruce  cones  ( mostly  the  former )  out  of  the  midden  on  the  east  and  south 
sides.  Much  of  the  rest  of  the  midden  had  already  been  excavated,  probably 
having  yielded  at  least  six  more  bushels.    In  June  of  1968,  some  sound  blue 


FiNLEY — Cone  Caches  of  Tamaisciurus 


243 


1^ 


Vk..  4.   A  large  midden  of  lodgepole  cone  debris  in  a  dense  stand  of  lodgepole 

pine,  Gilpin  County,  Colorado.    A  stump  at  the  peak  of  the  midden  served  as 

the  principal  squirrel  feeding  station. 

spruce  cones  of  the  1967  crop  were  still  cached  in  the  midden,  and  a  few 
sound  1967  pine  cones.  Most  pine  cones  found  were  rotting,  from  years 
previous  to  1967,  indicating  a  poor  pine  cone  crop  in  1967.  Nearly  all  of  the 
recent  midden  material  consisted  of  cone  scales  and  cores  of  blue  spruce.  In 
June,  the  top  two  to  four  inches  of  midden  were  dry,  but  material  below  that 
level  was  still  damp  enough  to  keep  cones  closed. 

A  few  feet  east  of  the  midden  I  found  a  water  cache  of  ponderosa  pine  cones 
described  below  under  a  separate  heading. 

Lodgepole  pine  forest. — Cone  caches  in  a  pure  stand  of  lodgepole  pine  bear 
little  resemblance  to  those  described  above.  Shade  and  moisture  are  not 
required,  and  piles  of  closed  lodgepole  cones  lie  exposed  on  the  surface  of  the 
midden.  I  have  seen  many  such  caches  and  middens  on  a  fairly  open  south- 
facing  slope  northeast  of  Blackhawk,  at  9300  feet,  Gilpin  County,  Colorado. 
In  October,  1965,  t\vo  rather  large  mounds  of  cone  debris  were  heaped  around 
an  old  stump  and  an  uprooted  snag.  These  feeding  posts  had  more  than  a 
bushel  of  closed  lodgepole  cones  piled  under  and  around  the  snag  and  buried 
in  the  top  layer  of  fresh  cone  scales.  The  cones  covered  by  cone  scales  were 
found  not  concentrated  in  cache  holes,  but  scattered  or  loosely  aggregated 
through  the  material  as  though  incidentally  covered  by  accumulation  of  newly 
dropped  cone  debris. 

A  much  larger  midden  nearby  was  centered  around  another  stump  and 
formed  a  low  mound  of  lodgepole  cone  scales  and  cores  18  feet  long  and   15 


244  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

feet  across.  Extending  out  from  this  on  the  lower  side  was  a  tail  of  older, 
more  weathered  cone  debris  and  needle  litter  reaching  as  a  carpet  for  another 
18  feet  from  the  active  mound. 

The  lodgepole  stand  was  of  closely-spaced,  predominantly  pole- sized  trees 
from  two  to  11  inches  d.b.h.  and  nearly  all  of  about  the  same  height,  estimated 
to  be  approximately  30  feet.  Few  other  kinds  of  trees  were  seen  within  100 
yards  of  the  midden — only  a  single  ponderosa  pine,  widely  scattered  Douglas- 
firs,  and  a  small  group  of  blue  spruces  in  a  wet  spring  area  perhaps  farther 
away.  On  and  around  the  big  midden,  the  tree  spacing  varied  from  four  to 
54  inches  measured  from  any  tree  to  its  nearest  neighbor.  There  were  1 1  lodge- 
poles  of  three  to  five  and  a  half  inches  d.b.h.  standing  in  the  midden,  as  well 
as  t\vo  dead  poles  still  standing.  The  largest  opening  between  tree  stems  on 
the  midden  was  seven  and  a  half  feet  across.  The  crown  canopy  was  mostly 
closed,  and  lower  branches  and  understory  vegetation  were  almost  totally 
absent. 

The  midden  material,  consisting  almost  entirely  of  the  scales  and  cores  of 
lodgepole  cones,  was  heaped  up  around  and  almost  covering  the  stump  (see 
Fig.  4),  the  tip  of  which  stood  22  inches  above  the  ground  and  served  as  the 
primary  feeding  perch  on  the  midden.  Six  poles  lay  across  the  mound  and 
others  were  partly  buried  in  it.  When  the  midden  was  examined  and  measured 
in  June  of  1968,  stored  cones  were  few,  and  there  were  tunnels  into  the  cone 
scales  under  the  stimip  and  between  its  root  snags.  I  counted  72  open  cache 
holes  in  the  midden.  The  material  was  mostly  damp  or  wet  below  the  top 
layer  of  fr^vo  or  three  inches,  but  1  found  no  cones  cached  in  the  wet  material. 
There  were  a  few  closed  lodgepole  cones  lying  on  the  midden,  mainly  on  the 
higher  parts  and  near  the  main  feeding  post;  few  were  buried  in  the  debris. 
A  few  green  lodgepole  needle  cuttings  lay  on  the  edges  of  the  midden  and  on 
the  ground  nearb\'.  Groimd  co\'er  outside  the  midden  consisted  almost  entirely 
of  dry  needle  litter  with  a  few  low  clumps  of  junipents  communis  and  many 
patches  of  the  mat-forming  kinnildnnick. 

Numerous  smaller  middens  in  the  vicinity  were  centered  around  single  trees 
that  served  as  the  main  feeding  stations.  Carpets  of  kinnikinnick  grew  around 
several  of  these  trees,  encroaching  a  foot  or  two  over  the  outer  layer  of  old 
decomposing  midden.  One  small  midden  had  accumulated  so  rapidly  that  a 
patch  of  kinnikinnick  with  bright  green  leaves  was  cox'ered  completely  b>' 
fresh  cone  scales. 

Aliout  a  quarter  mile  away  on  a  north-facing  slope  I  found  the  largest  lodge- 
pole midden  measured.  It  was  39  feet  long  and  18  feet  across  and  was  shaded 
by  more  widely  spaced  and  larger  trees,  up  to  12  inches  d.b.h.  A  noteworthy 
feature  was  \\\o  piles  of  closed  lodgepole  cones  heaped  up  under  trees  on  the 
lower  end  of  the  midden  (see  Fig.  5).  The  larger  pile  contained  1072  cones 
(approximately  two-thirds  of  a  bushel).  Within  a  radius  of  25  yards  of  the 
center  of  this  midden  were  seven  smaller  middens,  five  of  which  had  cones 
piled  on  the  surface.  Because  of  their  close  spacing,  it  is  likely  that  all  these 
cone  piles  were  made  by  the  same  squirrel. 

On  the  alluvial  flat  of  Gold  Creek,  4300  feet  elevation,  25  mi.  NE  Missoula, 
Montana,  there  were  numerous  small  middens  under  lodgepole  pines  in  August 
1963.  Such  creek  \alley  bottoms  in  Colorado  are  usually  occupied  by  spruce 
stands.  The  pine  cone  litter  was  dry  on  the  surface  but  wet  below,  from  the 
frequent   summer   thundershowers.     There   were    a   few    closed   cones   on    the 


FiNLEY — Cone  Caches  of  Tamaisciurus 


245 


)!,«»■•"    »^^ 


Fic.  5.    A  Miiiace  caclif  ol  lodgt-pole  cones  (hall  bushel)  on  a  midden  ol  cone 
scales  and  cores  against  the  base  of  a  lodgepole  pine. 


middens,  but  no  piles  were  in  evidence.  I  trapped  red  squirrels  on  some  of 
these  middens  and  saw  others  in  the  larch — Douglas-fir  forests  on  adjacent 
slopes. 

Ponderosa  pine  forest,  Kaibab  Plateau. — Although  I  have  never  seen  a  sub- 
stantial midden  or  cache  solely  under  ponderosa  pines  in  Colorado,  sizable  ones 
occur  on  the  Kaibab  Plateau  in  northern  Arizona.  I  examined  such  a  midden  in 
August  of  1968  under  two  pines  on  a  south-facing  slope  at  8700  feet,  about 
12  mi.  S  Jacob  Lake.  It  was  in  an  open  ponderosa  forest  bordered  on  the 
lower  side  with  aspen.  The  stand  density  was  110  square  feet  per  acre,  basal 
area.  White  fir  and  Douglas-fir  grew  at  the  foot  of  the  slope  about  100  yards 
a\\'ay  from  the  midden.  Both  pines  on  the  midden  were  infested  with  mistletoe 
in  the  lower  branches,  which  induced  the  heavier-than-normal  foliage  on 
branches  near  the  ground.  The  two  pines  were  21  and  16  inches  d.b.h.,  the 
larger  one  57  feet  tall,  with  two  secondary  stems  arising  from  a  mutilated 
mid-section. 

The  midden  measured  31  by  19  feet,  encompassing  a  stump  and  rotten  wood 
as  well  as  the  two  trees.  Moisture  distribution  in  the  midden  was  erratic.  At 
different  points  the  dry  sinface  layer  of  cone  debris  \aried  from  one  to  five 
inches  in  depth.  Below  that  it  was  more  or  less  wet,  depending  on  the  degree 
of  shade  or  shelter.  Pine  needles  were  mixed  in  the  midden,  and  in  some  parts 
layered,  probably  indicating  a  period  of  alxindonment  when  no  squirrel  dug 
into  it  and  mixed  the  needle  fall  with  the  cone  litter.  All  cone  debris  seen 
was  of  ponderosa  pine,  mostly  old  but  with  a  few  fresh  green  scales  and  cores 


246  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

on  the  surface.  I  dug  about  30  green  closed  pine  cones  out  of  some  of  the 
many  cache  holes  scattered  o\er  the  midden,  plus  one  old  closed  pine  cone 
and  several  of  Douglas-fir  from  the  prexious  year's  crop.  One  grass  nest  was 
about  halfway  up  in  the  smaller  pine. 

About  2  mi.  SE  Jacob  Lake,  at  8100  feet  elevation  in  a  ponderosa  pine 
forest,  there  were  a  few  widely  scattered  middens  imder  single  large  pines  and 
around  logs  and  stimips  on  a  north-facing  slope.  The  largest  and  most  active 
midden,  measuring  23  by  11  feet,  was  around  a  pine  22  inches  d.b.h.  and 
91  feet  tall,  and  extended  under  nearby  logs  and  large  limbs.  The  greatest 
depth  of  the  midden  was  about  one  foot.  It  contained  no  cone  debris  other 
than  of  ponderosa  pine,  and  no  cached  cones.  The  top  two-inch  layer  was 
dry;  below  that  it  was  wet.  The  density  of  trees  aroimd  the  midden  was 
137  square  feet  per  acre.  Large  aspens  were  scattered  through  the  pines, 
but  I  saw  no  other  species  of  conifer  within  a  mile  or  more  of  the  site. 
Because  the  new  pine  seed  crop  was  light  at  best,  the  squirrels  were  facing 
a  tight  winter. 

Near  Jacob  Lake,  I  dug  into  two  mounds  of  duff  around  the  bases  of  large 
pines.  They  were  wet  below  t\vo  inches  and  had  six  to  eight  inches  of  loose, 
wet  bark  litter,  needles,  and  decaying  cones  and  twigs.  They  were  ideal  sites 
for  cone  caching  but  contained  no  squirrel  food  litter  of  any  kind.  Elsewhere 
in  the  pine  forest,  I  saw  a  few  inactive,  deteriorated  middens  in  seemingly 
suitable  habitat. 

Water  Caches 

Since  middens  must  remain  damp  to  be  of  value  for  cone  storage, 
it  is  not  uncommon  to  find  them  extending  into  springs,  boggy  areas, 
and  creek  bottoms.  Shaw  (1936:340)  described  and  illustrated  such 
wet  cache  sites  in  the  Pacific  Northwest.  He  made  little  mention 
of  caching  in  midden  material,  possibly  because  soil  and  leaf  mulch 
were  soft  and  easy  to  cache  in.  In  Colorado  red  squirrels  cache  in 
wet,  saturated  middens  extending  out  onto  creek  bottoms,  and  also 
make  "water  caches,"  aggregations  of  cones  on  the  bottoms  of  small 
pools  of  water  and  under  roots  and  undercut  stream  banks.  Some- 
times cones  may  be  more  widely  scattered  in  small  groups  between 
rocks  on  the  bottom  of  a  flowing  stream. 

Cones  that  rest  on  the  bottom  are  all  closed  and  must  have  been 
cut  from  the  branches.  The  more  scattered  ones  may  have  been  cut 
and  dropped  directly  from  overhanging  tree  branches.  At  first  I 
suspected  that  the  cones  in  pools  came  to  rest  in  aggregations  by 
the  actions  of  currents  and  eddies.  However,  I  found  many  in  small 
isolated  pools  with  no  water  flow,  and  many  stuffed  in  holes  and 
crannies  under  water.  The  caches  were  often  of  such  size  (one  to 
two  bushels)  as  to  make  accidental  accumulation  under  water 
almost  inconceivable.  Red  squirrels  haxe  been  reported  by  many 
authors  to  be  good  swimmers  (Seton,  1929,  4:120),  and  Mr.  Swift 
also  reported  to  me  that  they  swim  under  v/ater  both  to  cache  cones 


FiNLEY — Cone  Caches  of  Tamaisciurus  247 

and  to  retrieve  them.  He  told  me  of  finding  a  large  midden  filled 
with  ponderosa  cones  under  spruce  trees  on  one  side  of  a  swift 
stream  several  yards  across.  The  only  ponderosa  pines  were  on  a 
dry  slope  on  the  opposite  side  of  the  stream,  which  could  not  be 
crossed  except  by  swimming. 

In  November,  1963,  I  found  a  "water  midden"  on  a  small  creek 
bottom  near  Wellington  Lake,  Park  County,  Colorado.  A  bed  of 
cone  debris  had  built  up  under  the  branches  of  a  blue  spruce  much 
used  as  feeding  posts.  It  formed  a  mound  14  inches  deep  like  a 
small  dam  co\'ering  the  mud  creek  bottom,  through  which  only  a 
trickle  of  water  was  flowing.  The  top  three  inches  of  fresh  midden 
were  dry,  and  the  bottom  portion  was  blackened  and  saturated  by 
creek  water,  which  percolated  out  of  the  midden  at  the  base  of  an 
undercut  bank.  Many  ponderosa  cones  were  dug  out  of  this  midden 
in  1963,  but  in  June  of  1968  it  contained  only  some  blue  spruce 
cones. 

Another  w^ater  midden,  on  a  tributary  of  Redskin  Creek,  Park 
County,  covered  the  bottom  of  a  shallow  creek  three  to  seven  feet 
wide  that  flowed  without  entrenched  channel  through  a  dense  blue 
spruce  grove  between  hillsides  clothed  in  ponderosa  pine.  The 
midden  was  38  feet  long  on  one  side  of  the  creek  and  extended  as 
an  apron  across  the  mud  creek  bottom  and  beyond.  A  prime  reason 
for  the  large  amount  of  cone  litter  on  the  creek  bottom  was  a  big 
spruce  log  that  had  fallen  across  the  creek  and  was  heavily  used  by 
squirrels  as  a  feeding  station  and  bridge.  The  water  line  on  the 
midden  was  marked  by  a  sharp  change  in  color  from  light  brown 
fresh  cone  litter  above  to  black  discolored  litter  under  water.  In 
September,  1965,  this  midden  was  a  rich  source  of  ponderosa  cones 
buried  in  litter  both  above  and  below  the  water  hne.  It  contained 
mostly  blue  spruce  cones  in  1967. 

A  creek  bed  was  used  as  a  water  cache  without  a  midden  on  Cub 
Creek  southwest  of  Evergreen,  Colorado,  in  October,  1965.  The 
creek  passed  through  a  group  of  eight  large  blue  spruce  trees  50  to 
80  feet  tall  and  13  to  18  inches  d.b.h.  There  was  no  midden  on  either 
bank  deep  enough  to  cache  in.  I  gathered  about  two  bushels  of 
blue  spruce  cones  and  a  bushel  of  Douglas-fir  from  along  the  stony 
creek  bottom,  many  submerged  on  the  bottom  in  water  two  to  six 
inches  deep,  some  floating  in  backwater  pools,  and  many  stuffed  in 
holes  or  hollows  under  the  steep  banks.  The  creek,  three  to  eight 
feet  wide,  was  flowing  rather  swiftly,  and  many  of  the  cones  were 
on  the  bottom  well  out  from  the  edge  of  the  water. 


248  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

On  Beaver  Creek,  in  1963,  I  found  a  much  more  concentrated 
water  cache  associated  with  the  large  midden  under  a  blue  spruce 
stand  described  above  ( in  the  section  entitled  "Blue  spruce  in  valley 
bottoms").  It  was  in  a  pool  of  water  about  four  feet  long  and  10 
inches  deep  on  the  shore  of  the  main  stream.  The  pool  was  in  gently 
flowing  water  where  the  soil  had  washed  away  under  and  between 
roots  of  one  of  the  big  spruces.  The  bottom  of  the  pool  was  covered 
with  green  closed  ponderosa  pine  cones  and  many  were  lodged 
among  the  exposed  roots  under  the  bank.  Some  closed  cones  were 
floating  but  were  kept  from  drifting  downstream  by  a  large  root  and 
a  few  fallen  branches.  This  pool  yielded  more  than  a  bushel  of 
pine  cones.  About  a  mile  farther  upstream  a  similar  pool  against 
the  bank  of  the  main  creek  contained  over  a  bushel  of  blue  spruce 
cones. 

Caching  Behavior  and  Seed  Consumption 

In  Colorado,  red  squirrels  begin  cutting  new  spruce  and  Douglas- 
fir  cones  in  August  while  the  seeds  are  still  in  the  milk  stage  and  not 
ripe  enough  to  store.  Such  cones  are  eagerly  stripped  of  scales  and 
the  seeds  consumed  as  a  welcome  addition  to  the  current  food 
supply.  Large  numbers  of  green,  unopened  cones  can  be  seen 
scattered  on  the  ground  where  squirrels  have  dropped  them  from 
the  branches  above.  The  manner  and  vigor  \\ath  which  this  cone- 
cutting  and  tossing  is  done  was  described  by  Bell  in  1898  ( Klugh, 
1927:19),  and  numerous  later  observers.  Shaw  (1936:348)  timed  a 
squirrel  cutting  white  spruce  cones.  In  two  one-minute  intervals  it 
cut  and  dropped  28  and  29  cones. 

Although  a  few  of  the  early-cut  green  cones  may  be  stuffed  into 
shallow  pockets  on  the  midden  or  ground  surface,  they  are  not 
cached  in  quantity  until  the  seeds  have  reached  the  dough  stage, 
beginning  in  late  August  or  early  September  in  Colorado.  At  this 
time  feeding  on  the  new  cone  crop  continues,  and  immense  num- 
bers of  ripe  but  still  closed  cones  are  cut  from  the  trees,  gathered 
from  the  ground,  and  cached  in  the  midden  for  future  use.  Harvest- 
ing of  cones  is  the  major  activity  of  red  squirrels  during  daylight 
hours  through  the  months  of  September  and  October.  The  numbers 
of  cones  that  can  be  stored  away  in  this  period  are  truly  prodigious. 
In  October,  1963,  I  removed  about  a  bushel  of  Engelmann  spruce 
cones  from  a  midden  5  mi.  W  Arapaho  Basin,  Summit  County, 
Colorado.  Two  days  later  I  visited  the  same  midden  again  and 
found  that  the  owner  had  akeady  refilled  most  of  the  cache  holes  I 


FiNLEY — Cone  Caches  of  Tamaisciurus  249 

had  emptied.  The  sHghtly  drier  eondition  of  the  newly  cached  cones 
suggested  that  they  had  been  gathered  from  the  ground,  not  freshly 
cut. 

Cones,  particularly  those  of  Douglas-fir,  are  sometimes  cached 
after  they  have  lain  on  the  ground  long  enough  for  the  scales  to  open 
partially.  Such  cones  seem  to  have  shed  few,  if  any,  seeds.  Cones 
that  have  dried  enough  for  the  scales  to  open  fully  are  not  cached. 
But  in  June,  when  cones  were  scarce,  I  saw  a  squirrel  pick  up  an 
open  blue  spruce  cone  lying  on  a  midden  and  carry  it  up  into  a  tree. 

Most  of  the  caching  in  Colorado  is  done  in  small  holes  dug  into 
the  midden  material  of  loose  cones,  scales,  and  cores.  After  the 
caches  have  been  exhausted,  the  surface  of  the  midden  remains 
riddled  with  these  little  pits.  They  are  usually  shallow,  four  to  eight 
inches  deep,  but  may  extend  as  far  as  20  inches  under  buried  logs, 
roots  or  other  solid  objects.  Such  cache  holes  are  sometimes  dug 
into  soft  soil,  but  they  usually  stop  when  mineral  soil  is  reached. 
The  number  of  cones  in  a  single  pit  may  vary  from  only  two  or 
three  pines  cones  to  as  many  as  50  or  more  of  the  small  cones  of 
Engelmann  spruce.  In  October,  after  a  heavy  cone  crop  has  ripened, 
an  active  midden  is  usually  crammed  with  cached  cones.  Many 
cache  holes  are  stuffed  full  to  the  surface,  with  clusters  of  fresh 
cones  projecting  out  of  the  midden.  But  most  of  the  cache  pits  are 
in\'isible  on  the  surface,  having  become  covered  with  a  seemingly 
undisturbed  layer  of  cone  debris. 

During  the  winter  and  following  spring  the  squirrel  digs  up  the 
cached  cones  and  carries  them  to  logs,  branches  or  other  feeding 
stations,  where  it  gnaws  the  scales  off  the  cores  to  get  at  the  seeds. 
The  accumulation  of  cone  scales  and  cores  around  preferred  feeding 
stations  and  trees  provides  the  source  of  new  midden  material, 
which  subsequently  becomes  spread  more  widely  over  the  area  as 
the  squirrel  digs  into  and  reworks  the  material  year  after  year. 

In  areas  of  prolonged  snow  cover,  squirrels  dig  tunnels  into  the 
snow,  and  under  it  on  the  surface  of  the  midden  to  reach  cached 
cones.  A  few  such  tunnels  kept  open  provide  essential  access  to 
the  midden  when  it  is  covered  with  heavy,  crusted  snow. 

EdihiJitij  and  ViahiUtij  of  Stored  Seed 

Loss  of  the  potential  winter  seed  supply  can  result  from  spoilage 

as  well  as  from  opening  of  cones  and  shedding  of  seeds  in  the  tree 

tops.    Insofar  as  edibilit)'  is  concerned,  seeds  of  conifers  store  well 

when  the  cones  are  kept  wet.   I  have  dug  cones  of  blue  spruce  and 


250  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Douglas-fir  out  of  wet  caches  in  June  and  found  them  to  be  black 
but  firm  and  still  retaining  some  pitch.  These  cones  were  probably 
nearly  a  year  old,  and  still  contained  many  sound  seeds.  However, 
cones  removed  from  the  same  caches  and  kept  wet  at  room  tempera- 
ture in  June  had  most  of  the  seeds  spoiled  after  less  than  a  month. 
Apparently  seeds  must  be  kept  cool  to  be  preserved  in  summer. 
Wet  middens  excavated  in  June  were  distinctly  cool  in  Colorado, 
some  at  10,000  feet  elevation  even  retaining  frost  near  the  bottom. 
However,  by  the  end  of  the  summer,  middens  in  most  of  this  region 
probably  reach  temperatures  too  high  for  preservation  of  the  seed 
as  food. 

Cones  of  Douglas-fir  and  blue  spruce  excavated  from  caches  in 
October  and  November  are  easily  separable  into  sound  cones  of  the 
year  and  older  rotting  cones.  The  sound  cones,  although  generally 
brown,  retain  a  little  straw  or  greenish  color  at  the  tips  of  the  scales. 
Older  cones  are  completely  brown  or  black,  have  looser  cone  scales, 
and  can  easily  be  twisted  in  two  with  the  hands.  Much  of  the  woody 
material  has  rotted  and  the  seeds  are  black  or  completely  decom- 
posed. 

Many  of  the  current  year's  cones  cached  within  the  preceding 
four  to  six  weeks  may  be  enveloped  in  a  mesh  of  fungal  mycelia, 
which  also  penetrates  the  midden  matrix.  The  fungus  seems  to  have 
no  effect  on  the  edibility  or  viability  of  the  seeds,  at  least  during 
the  first  several  months. 

I  doubt  that  the  squirrels  in  Colorado  can  get  many  edible  seeds 
out  of  cones  cached  longer  than  a  year,  but  they  are  reported  to  dig 
them  up  and  use  them  in  Montana  (Halvorson,  unpublished)  and 
Alaska  (M.  C.  Smith,  1968).  Perhaps  in  those  areas  summer  temper- 
atures do  not  cause  as  much  spoilage  of  seed. 

Red  squirrels  rarely  cache  cones  that  are  too  green  to  yield  ripe 
seeds  or  that  contain  predominantly  unfilled  or  insect-infested  seeds. 
Nevertheless,  a  significant  but  highly  variable  percentage  of  seeds 
in  all  cones  is  empty  or  incompletely  filled.  In  general,  in  years  of 
poor  cone  crops  the  cones  have  lower  percentages  of  sound  seeds 
(Toumey  and  Korstian,  1942:105),  which  further  reduces  the  avail- 
able squirrel  food.  Viability  for  germination  is  highly  variable,  even 
for  seeds  solidly  filled  and  apparently  sound,  but  many  filled  seeds 
remain  edible  even  when  no  longer  capable  of  germination. 

In  spite  of  the  high  viability  of  seeds  in  caches  for  nearly  a  year, 
there  is  little  opportunity  for  such  seeds  to  germinate  naturally  and 
produce   successfully   established    seedlings.     Stored   seeds,    if   not 


FiNLEY — Cone  Caches  of  Tamaisciurus  251 

eaten  usually  decompose  in  the  cone.  In  July  I  have  found  some 
germinated  seedlings  emerging  from  cones  in  wet  middens,  but  all 
died  in  the  midden  without  establishing  roots  in  the  soil  below. 
If  the  midden  dries  enough  for  the  cone  scales  to  open,  the  seeds 
are  apt  to  be  released  into  an  unsuitable  sprouting  medium,  too  far 
above  the  mineral  soil  and  where  the  chance  of  reaching  sunlight 
is  slight.  Seedlings  of  conifers  are  almost  never  found  on  squirrel 
middens,  except  occasionally  in  marginal  situations  of  more  open 
tree  canopy  and  where  cones  are  cached  in  damp  soil.  I  think  Bailey 
(1931:79)  exaggerated  the  value  of  red  squirrels  as  agents  in 
natural  forest  regeneration. 

Comparative  hiterpretafions  of  Cone  Caclie  Ecology 

Cone-caching  behavior  of  the  species  as  a  whole  is  highly  diverse 
and  seems  to  indicate  merely  a  wide  range  of  tolerance.  However, 
when  geographic  and  ecological  xariations  of  caches  and  middens 
are  compared  locally  as  well  as  regionally,  some  explanations  of  the 
pattern  seem  to  make  sense. 

The  range  of  local  variation  in  cache  sites  is  not  so  great  as  the 
regional  variation  because  each  local  area  offers  only  a  limited  set 
of  forest  conditions  suitable  for  red  squirrels.  Kinds  of  trees  used 
for  midden  sites  and  kinds  of  topographic  situation  occupied  in 
different  areas  are  related  to  differences  in  temperature  and  moisture 
at  the  ground  level,  hence  to  the  microclimate  as  well  as  the  regional 
climate. 

Tamiasciurus  is  primarily  a  boreal  genus.  Its  special  adaptations 
for  feeding  on  stored  conifer  seeds  require  cool  moist  sites  where 
cones  can  be  stored  without  opening  and  without  spoilage  for  a  year 
or  more.  Such  conditions  are  easy  to  find  almost  anywhere  in  the 
boreal  conifer  forests.  Hence  red  squirrels  are  continent-wide  in 
distribution  in  the  Canadian  Taiga  and  occur  southward  throughout 
the  Canadian  Life-Zone  in  the  higher  mountains.  In  western  Mon- 
tana they  occur  as  low  as  3000  feet  elevation,  but  in  Colorado  come 
down  only  to  about  7000  feet  in  the  Transition  Life-Zone.  From 
about  9500  feet  up  to  timberline  in  Colorado  the  spruce-fir  type, 
both  on  flats  and  steep  slopes,  provides  plenty  of  cool  moist  cache 
sites. 

The  driest  conditions  are  in  lodgepole,  limber,  and  bristlecone 
pine  forests,  mostly  on  south-facing  slopes,  where  conditions  for 
caching  are  marginal.  At  Geneva  Basin,  Colorado,  for  instance, 
large  middens  were  occupied  at  10,000  feet  in  the  shadier  parts  of 


252  Misc.  Publ.  51,  UxNiv.  Kansas  Mus.  Nat.  Hist. 

a  fairly  open  limber  pine — bristlecone  forest,  and  the  middens  were 
still  partly  damp  in  June  on  southwestern  slopes.  Smaller  middens 
in  more  open  situations  had  already  dried  out.  Some  lodgepole 
stands  as  low  as  9000  feet  are  occupied  by  red  squirrels  because  of 
their  perpetual  supply  of  closed  cones,  but  this  habitat  seems  to  be 
less  than  optimum. 

At  lower  elevations  the  drier  forests,  mainly  ponderosa  pine,  are 
more  extensive.  In  this  zone  only  the  blue  spruce  stands  occupying 
the  wettest  and  deepest  soils  along  the  xalley  bottoms  provide 
enough  shelter  to  retain  cool  moist  conditions  through  most  of  the 
year.  Hence  the  red  squirrels  are  restricted  to  within  "commuting 
distance"  of  the  blue  spruce  stands.  Middens  under  such  stands 
usually  contain  ponderosa  cones  if  there  is  a  good  ponderosa  seed 
crop  on  the  adjacent  sunny  slope. 

Tamiasciunis  is  absent  from  the  Black  Forest,  a  flat  upland 
divide  about  7000  feet  in  elevation  between  the  South  Platte  and 
Arkansas  drainages  northeast  of  Colorado  Springs.  It  is  a  solid 
stand  of  ponderosa  pines  on  dry  sandy  soil  and  lacks  blue  spruce  in 
the  small  tributary  canyons  leading  from  it. 

Farther  to  the  south,  the  altitudinal  restriction  of  red  scjuirrels  in 
Arizona  and  New  Mexico  is  more  limiting.  The  Canadian  Life-Zone 
is  less  extensive  than  in  Colorado  and  the  squirrels  are  scarcely 
found  below  about  8000  feet,  except  in  shady  ravines.  In  New 
Mexico,  according  to  Bailey  (1931:76):  "They  rarely  come  down 
into  the  edge  of  the  yellow  pine  belt  farther  than  the  spruces  extend 
on  cold  slopes." 

On  the  Kaibab  Plateau  in  Arizona,  a  marked  difference  in  use  of 
forest  types  is  evident.  The  higher  parts  of  the  plateau,  above 
about  8400  feet,  are  covered  with  a  mixed  conifer  type  composed  of 
blue  spruce,  white  fir,  Douglas-fir,  and  some  ponderosa  pine.  Below 
8400  feet  the  pure  ponderosa  type  extends  down  to  about  7000  feet 
where  it  gives  way  to  pinyon-juniper.  The  ponderosa  pine  forests 
of  the  Kaibab  bear  little  resemblance  to  those  of  Colorado  east  of 
the  Continental  Divide.  On  the  Kaibab  the  trees  are  much  taller, 
larger,  and  faster-growing  and  cover  the  bottoms  of  the  draws  as 
well  as  the  slopes  and  upland  flats.  Blue  spruce  does  not  extend 
down  the  valleys  through  the  pine  type,  perhaps  because  of  the 
limestone  parent  material  and  subsurface  drainage. 

Red  squirrel  middens  on  the  Kaibab  occur  in  great  abundance 
throughout  the  mixed  conifer  type  and  in  much  less  abundance  in 
the  upper  part  of  the  ponderosa  pine.    The  middens  under  large 


FiNLEY — Cone  Caches  of  Tamaisciurus  253 

mature  pines  and  around  logs  and  stumps  in  pine  forests  were 
unlike  any  seen  in  the  eastern  Transition  Zone  in  Colorado.  But  on 
the  Kaibab,  in  spite  of  the  fairly  open  tree  stand  and  lack  of  low 
branches,  the  ground  has  a  good  layer  of  forest  litter  that  retains 
moisture  a  few  inches  below  the  surface.  Middens  under  pines 
were  wet  in  August,  and  the  abundant  down-timber  provided  plenty 
of  good  feeding  stations,  some  of  which  served  as  the  nuclei  of 
middens  even  without  any  tree  overhead.  In  general,  ponderosa 
middens  on  the  Kaibab  resembled  bristlecone  middens  more  than 
any  other  kind  in  Colorado. 

Middens  higher  up  in  the  mixed  conifer  type  were  mostly  under 
large  white  fir  ( Fig.  6 ) ,  Douglas-fir,  and  blue  spruce  trees,  as  well 
as  against  logs  and  stumps.  Some  of  the  fir  middens  were  on  fairly 
open  sites  much  like  those  in  the  ponderosa  type,  but  middens  were 
scarce  under  pines  in  the  mixed  forest,  perhaps  because  the  pines 
were  often  on  the  drier  sites  at  the  higher  elevation.  In  general,  in 
both  forest  types  the  squirrels  preferred  locations  with  more  than 
average  shade  or  shelter  of  some  kind. 

To  provide  moisture  in  the  period  of  cone  caching  and  until 
the  arrival  of  winter  snows,  good  rainfall  in  late  summer  and  autumn 
is  required.  Total  rainfall  on  the  Kaibab  Plateau  is  not  particularly 
high  but  has  a  peak  in  late  summer,  as  shown  by  the  graph  for 
central  Arizona  in  figure  S-2  of  Barrett  (1962:348).  The  axerage 
precipitation  for  the  four  months  of  July  through  October  for 
nine  stations  in  Coconino  County,  Arizona,  was  8.88  inches,  whereas 
the  average  for  nine  stations  in  the  eastern  Transition  Zone  of  Colo- 
rado was  7.45  inches  (Greening,  1941:761;  Gittings,  1941:798). 

Although  moisture  conditions  were  adequate  in  August,  1968, 
for  cone  caching  in  the  ponderosa  forest  on  the  Kaibab,  I  noted  a 
diff^erent  deficiency.  The  second  of  two  successive  pine  seed  failures 
had  left  the  squirrels  with  their  caches  exhausted  and  no  alternative 
kind  of  cones  to  harvest.  In  the  mixed  conifer  forest,  however,  some 
white  fir  cones  were  available  to  eke  out  a  generally  poor  cone  year. 
The  presence  of  a  low  density  of  middens  in  the  pine  forest  and 
a  high  density  in  the  mixed  conifers  is  compatible  with  my  inter- 
pretation that,  on  the  Kaibab,  midden  sites  are  adequate  though  not 
abundant  in  the  pine  type,  but  the  unreliability  of  the  pine  cone 
crop  does  not  permit  a  good  population  to  build  up  there. 

Rasmussen  (1941:262),  in  his  study  of  the  communities  of  the 
Kaibab  Plateau,  wrote  of  the  red  squirrel:  "Its  distribution  is  con- 
fined to  areas  where  Abies,  Pseudotsuga,  or  Picea  occur.  .  .  .  But 


254 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Fig.  6.    A  simple  Tamiasciunis  midden  around  the  base  of  a  white  fi 
Kail:)ab  Plateau,  Arizona.    The  midden  is  composed  of  cone  scales  and 
white  fir,  Douglas-fir,  and  ponderosa  pine. 


r  on  the 

cores  of 


at  .  .  .  lower  elevations  the  squirrel  middens  show  great  preponder- 
ance of  yellow  pine  [ponderosa]  cones."  He  thought  that  some  fac- 
tor other  than  food  prevented  red  squirrels  from  occupying  the  pure 
pine  forest,  but  he  seems  not  to  have  observed  any  years  of  pine 
seed  failure  on  the  Kaibab. 

C.  H.  Halvorson  has  seen  cone  caches  in  middens  under  pon- 
derosa pines  about  40  mi.  E  Lewistown,  Montana.  It  was  in  an  area 
of  low  wooded  hills  and  ridges  bordering  sagebrush  plains,  with 
wet  meadows  along  the  narrow  creek  bottoms.  Middens  and  caches 
were  seen  only  on  the  slopes  under  pines  or  against  logs  and  stumps. 

Red  Squirrels  and  Silviculture 

Caches  as  a  Source  of  Seed  for  Silviculture 

Since  the  early  years  of  reforestation  on  the  national  forests,  red 
squirrel  cone  caches  have  been  utilized  as  a  source  of  seed  for 
planting.  Cox  (1911)  gave  a  good  account  of  early  reforestation 
work  and  described  some  of  the  cone  collecting  and  processing 
operations  in  the  western  states.  He  wrote  (p.  17):  "Squirrels' 
caches  are  often  excellent  places  from  which  to  get  cones  ...  it  is 
not  uncommon  to  find  in  a  single  one  of  their  caches  from  S  to  12 


FiNLEY — Cone  Caches  of  Tamaisciurus  255 

bushels  of  good  cones,  though  the  a\'erage  quantity  is  about  2 
bushels.  These  caches  are  located  by  old  rotten  logs,  in  springy 
places  and  muck,  and  in  duff,  sometimes  at  a  considerable  depth, 
as  well  as  under  bushes  and  felled  tree  tops,  along  streams,  and 
beneath  overhanging  stream  banks.  .  .  .  Among  the  species  of  cones 
which  are  often  obtained  from  caches  are  Douglas-fir,  Engelmann 
spruce,  western  yellow  pine,  lodgepole  pine,  and  western  white 
pine.  Usually,  however,  the  cones  of  but  one  species  are  foimd  in  a 
single  cache.  In  collecting  from  sciuirrels'  hoards  it  is  well  to  have 
a  pack  horse  along  for  immediate  transportation,  since  if  cones  are 
dug  out  and  left  on  the  ground  for  any  length  of  time  they  will  be 
carried  away  and  cached  again  by  the  industrious  animals." 

Korstian  and  Baker  (1925:2)  reported  that  in  the  intermountain 
region,  lying  between  the  Sierra  Nevada  and  the  Rocky  Mountains, 
"seed  collecting  has  been  confined  almost  entirely  to  squirrel  caches, 
because  logging  operations  have  never  been  large  enough  to  make 
the  usual  method  of  collecting  seed  from  felled  trees  economical." 
Cones  were  commonly  bought  by  the  Forest  Service  directly  from 
individual  cone  collectors  scattered  over  wide  areas  and  were  proc- 
essed in  central  seed  extraction  plants.  Seeds  of  nearly  all  the 
major  timber  species  were  collected  by  this  method  in  quantities  as 
high  as  6000  pounds  in  a  single  operation. 

Some  of  the  long  recognized  advantages  of  collecting  seed  from 
squirrel  caches  are  ease  of  collection,  high  quality  of  seed  ( ripeness, 
viability,  and  freedom  from  insect  or  disease  damage),  and  the 
extended  period  during  which  closed  cones  can  be  gathered  after 
cones  on  the  trees  have  opened  and  shed  their  seeds.  Seed  collecting 
from  caches  also  has  some  disadvantages,  such  as  occasional  scarcity 
of  squirrels  and  caches  in  forests  from  which  seed  is  desired,  and 
inability  to  restrict  harvesting  to  trees  of  selected  seed-producing 
characteristics. 

Some  Forest  Service  silviculturists  have  expressed  dissatisfaction 
with  the  poor  quality  of  seed  obtained  from  squirrel  cached  cones, 
but  their  problem  is  mainly  one  of  finding  reliable  private  cone 
collectors  who  know  squirrel  caching  habits  and  will  sack  up  only 
the  sound,  freshly  cached  cones.  If  large  quantities  of  cones  are 
bought  from  individuals  who  irresponsibly  gather  up  old  rotting 
cones  as  well  as  dry  open  cones  from  the  surface  of  the  ground,  it  is 
not  realistic  to  blame  the  poor  quality  of  the  cones  on  the  squirrels. 

Private  nurserymen  and  seed  dealers  also  have  made  large-scale 
use  of  squirrel  caches  in  the  western  states  as  a  source  of  supply. 


256  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

They  collect  cones  themselves  or  purchase  cones  from  many  in- 
dividual collectors.  Two  nurserymen  known  to  me  buy  many 
thousand  bushels  of  squirrel-cached  cones  each  year  at  prices 
ranging  from  $1.50  to  $3.00  per  bushel.  They  buy  only  sound, 
closed  cones  of  the  current  year's  crop  and  obtain  high  per  cent 
germination.  Since  the  kinds  and  amounts  of  cones  vary  greatly 
from  year  to  year,  depending  on  the  crops,  cone  buyers  try  to 
process  and  store  an  excess  supply  during  good  seed  years,  in  order 
to  carry  them  over  the  years  of  seed  failure.  Most  of  the  commercial 
market  seems  to  be  for  horticultural  use,  though  seed  is  also  sold 
for  reforestation.  Mr.  Swift  buys  cones  of  most  kinds  of  Colorado 
conifers,  but  relatively  few  of  lodgepole  pine  and  subalpine  lir, 
because  of  their  low  horticultural  value. 

In  recent  years  there  has  been  increasing  interest  by  the  Forest 
Service  and  the  timber  industry  in  obtaining  higher  quality  seed 
for  reforestation.  There  also  has  been  an  increase  in  research  in 
tree  genetics  and  the  development  of  tree  seed  farms,  particularly 
in  the  South  and  Pacific  Northwest.  But  tree  seed  farms  and  direct 
picking  of  cones  in  the  West  still  provide  only  a  small  part  of  the 
seed  needed  for  forest  planting  and  seeding.  More  effort  is  being 
made  in  some  areas  to  obtain  seed  from  selected  forest  trees  by 
direct  picking  from  standing  or  felled  trees.  Efforts  are  also  made 
to  obtain  seed  from  the  same  latitudinal  and  elevational  tree  seed 
zone  where  it  is  to  be  used,  or  even  from  within  the  same  national 
forest. 

Present  methods  of  obtaining  tree  seed  practiced  by  the  Forest 
Service  vary  in  different  regions,  depending  on  the  local  conditions 
and  seed  years.  Some  seed  is  collected  by  Forest  Service  personnel, 
but  most  is  purchased  from  individual  private  collectors,  or  from 
commercial  seed  dealers.  However  it  may  be  obtained  by  the  user, 
the  great  bulk  of  conifer  seed  used  in  the  Pacific  Northwest  and  the 
Northern  and  Southern  Rocky  Mountain  regions  is  initially  har- 
vested by  Tamiasciums.  There  seems  to  be  no  other  source  at 
present  from  which  the  required  amounts  can  be  obtained  at  reason- 
able cost.  The  continued  existence  of  large  and  healthy  populations 
of  red  squirrels  is  a  major  asset  not  widely  appreciated. 

Seed  Biomass  in  Relation  to  Squirrel  Energy  Requirements 
The  amount  of  conifer  seed  cached  by  red  squirrels  and  the 
amount  actually  consumed  are  of  interest  from  several  aspects.    Is 
the  available  cone  crop  a  limiting  factor  on  the  population  or  the 


FiNLEY — Cone  Caches  of  Tamaisciurus  257 

productivity  of  red  squirrels?  Do  the  squirrels  ever  harvest  enough 
cones  to  reduce  natural  forest  regeneration?  Does  the  amount  of 
cones  gathered  from  caches  by  people  affect  a  squirrel's  chance  of 
sur\  ival?  Clear  cut  answers  to  these  questions  cannot  be  given,  but 
enough  information  is  available  to  justify  some  discussion  and 
tentati\'e  conclusions.  Perhaps  presenting  them  may  stimulate  the 
collection  of  needed  facts. 

Although  a  great  deal  of  work  has  been  done  on  the  silviculture 
of  the  major  timber  trees  used  by  squirrels,  little  has  been  published 
on  the  biomass  of  seed  produced  by  most  species.  Seed  productivity 
fluctuates  widely  from  year  to  year  and  also  from  tree  to  tree  even 
in  a  "bumper"  seed  year.  Cone  and  seed  production  of  a  given  tree 
is  influenced  by  age,  dominance  in  the  stand,  site  quality,  pre\'ailing 
weather  conditions,  inheritance,  and  factors  still  unknown  ( Toume)^ 
and  Korstian,  1947:272).  Available  data  cannot  be  regarded  as 
definiti\'e  or  "average"  for  any  part  of  the  country,  but  for  the  sake 
of  discussion  I  have  used  whatever  information  I  could  get,  to  see 
if  calculations  and  reasoning  based  on  these  data  would  lead  to 
plausible  conclusions.  The  same  may  be  said  for  data  on  the  energy 
budget  and  behavior  of  red  squirrels,  but  here  we  seem  to  be 
dealing  with  less  wide-ranging  variables. 

Data  I  ha\'e  been  able  to  find  on  seed  producti\ity  of  several 

species  of  Rocky  Mountain  conifers  are  compiled  in   Tables    1-5. 

Sources  are  indicated,  many  of  which  are  for  studies  outside  the 

geographic  region  covered  in  this  report.    It  should  be  remembered 

that  information  obtained  for  Douglas-fir  and  ponderosa  pine,  and 

perhaps  other  species,  in  the  Pacific  Northwest  will  not  be  strictly 

applicable  to  the  Rocky  Mountains,  because  of  considerable  intra- 

specific  difi^erences  between  trees  of  these  regions.   For  some  species, 

such  as  the  blue  spruce,  I  have  been  able  to  find  few  data.    Seed 

production  data  based  on  cone  counts  on  trees  may  give  estimates 

of  production  prior  to,  or  after,  cone  cutting  by  squirrels;  whereas 

data  on  seed  fall  per  acre,  based  on  use  of  seed  traps,  give  only 

estimates  of  production  after  squirrels  have  taken  their  toll. 

Red  squirrel  energy  requirements. — C.  C.  Smith  (1968)  studied  the  food 
consumption  of  Tamiasciurus  hudsonicus  strcatori  in  relation  to  liehavior  and 
territoriality  in  southern  British  Cohunbia.  He  arri\ed  at  the  follo\\'ing  Nalues 
for  ingested  energy  of  a  few  individual  squirrels:  an  adult  male,  117  kg.  calories 
per  day;  a  female  at  the  height  of  lactation,  322  kg.  calories  per  day;  and  six 
ju\eniles  (134  to  170  grams)  80.5  to  95.4  kg.  calories  per  day.  From  extremely 
limited  samples  of  seed  he  obtained  the  energy  content  (dry-weight  basis)  of 
several  kinds  of  conifer  seed:    Douglas-fir,  7131  calories  per  gram;  Engelmann 


258  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Table  1. — Seed  production  data  reported  for  Douglas-fir. 


Amount 


Area 


Reference 


Thousand 
seeds/pound  42 

44 

24-49 
Seeds/cone       44 

45 

Pounds 

seed/bushel  .76  (.36-1.33) 

1.0 
Cones/bushel    1584 
Cones/tree         1J26 

1300  (1000-4000; 

Bushels 

cones/tree      1 .5 

2.5 
Pounds 

seed/ tree        .1-1.0 
1. 

Thousand 

seeds/acre     34-123 

168-1.500 
291 
Pounds 

seed/acre       .85-3.1 
Seed 

trees/acre      10 

Cal/g  dw 

whole  seed    5998 

Cal/g  dw 

kernel  7131 

Kernel 

corr  factor     .686 


Central  Rockies 
Pacific  Northwest 
Utah  and  Idaho 
British  Columbia 
Utah 


Colorado 
Pacific  Northwest 
British  Columbia 

Pacific  Northwest 


Pacific  Northwest 

Pacific  Northwest 
British  Columbia 
New  Mexico 


Anonymous,  1948 

Isaac,  1943 

Korstian  and  Baker,  1925 

Carman,  1951 

Hay  ward,  1940 

Korstian  and  Baker,  1925 

Tillotson,  1917 

R.  B.  Finley 

Winjum  and  Johnson,  1964 

Carman,  19.55 

Isaac,  1943 
Anonymous,  1948 

Fowells,  1965 
Isaac,  1943 

Isaac,  1943 
Carman,  1951 
Krauch,  1945 


Pacific  Northwest  Isaac,  1943 
Co.x,  1911 
Long,  1934 
C.  C.  Smith,  1968 
R.  B.  Finley 


spruce,  7107  calories  per  gram;  lodgepole  pine,  6827  calories  per  gram;  and 
ponderosa  pine,  7558  calories  per  gram.  These  \alues  are  for  only  those  parts 
of  the  seed  consumed  by  squirrels — the  endosperm  and  embryo.  Hence,  an 
adult  male  would  require  42,700  kg.  calories  per  year  which,  if  provided 
entirely  by  Douglas-fir  seed,  would  amount  to  5.99  kilograms  (dry  weight)  of 
seed  per  year. 

Douglas-fir. — Korstian  and  Baker  (1925:3)  give  the  pounds  of  clean  seed 
per  bushel  for  Douglas-fir  as  from  0.36  to  1.33  pounds  (averaging  0.76).  But 
these  values  are  for  seed  with  the  seed  coat,  live,  as  used  for  planting.  I 
dissected  the  edible  parts  out  of  the  hulls  of  100  seeds  and  weighed  them  in 
10  lots  of  10  seeds  each.  The  endosperms  and  embryos  combined  weighed 
71.2  per  cent  of  the  total  seed  weights.  Since  the  moisture  content  of  com- 
mercially stored  seed  is  usually  si.x  to  10  per  cent,  I  assumed  eight  per  cent, 
converted  the  fresh  weights  to  equivalent  dry  weights,  and  computed  a  cor- 
rected value  for  the  edible  parts  of  68.6  per  cent  of  total  seed  weight.  Assuming 
the  same  moisture  content  in  the  seed  weights  per  bu.shel,  and  applying  the 


FiNLEY — Cone  Caches  of  Tamaisciurus  259 

68.6  per  cent  correction  factor  to  con\'ert  to  edible  seed  weight,  one  obtains 
0.52  pound,  or  236  grams  of  food  (dry  weight)  per  a\'erage  bushel  of  Douglas- 
fir  cones.  Hence,  Smith's  adult  male  would  require  25.4  bushels  of  Douglas-fir 
cones  to  meet  his  energy  requirements  for  an  entire  year.  1  ha\e  not  applied 
any  correction  for  per  cent  of  somid  or  filled  seed  in  cones  because  this  is  highly 
\ariable  and  no  data  are  a\'ailable  for  cones  cut  by  squirrels.  This  factor  \\'ould 
tend  tf)  increase  the  required  number  of  cones  in  poor  seed  years  more  than 
it  would  in  good  years,  when  high  percentages  of  seed  are  sound. 

According  to  Isaac  (1943:16),  "the  average  forest-grown  tree,  which  has  a 
narrow  crown,  yields  about  I/2  bushels  dming  a  good  seed  year.  This  amount 
of  cones  produces  about  one  poimd  of  cleaned  seed."  At  this  producti\ity  the 
25  bushels  of  cones  required  by  one  male  sciuirrel  would  be  provided  liy  17 
"a\erage"  trees.  But  the  number  of  cones  produced  per  tree  is  notoriously 
\ariable.  Winjum  and  Johnson  (1964)  studied  variation  in  Douglas-fir  cones 
and  seeds  in  yoimg,  open-grown  Douglas-fir  trees  in  Oregon  and  Washington 
and  found  from  151  to  6000  (average  1126)  cones  per  tree.  (I  counted  1584 
cones  in  a  bushel. )  Because  trees  in  closed  stands  produce  far  fewer  cones,  a 
squirrel  would  probably  be  able  to  har\est  less  than  a  bushel  per  average 
tree.  However,  in  most  situations  cone  cutting  is  concentrated  on  those  few 
trees  in  a  stand  that  are  the  best  seed  producers.  Cox  (1911:13)  estimated 
10  trees  per  acre  bearing  seeds  in  appreciable  quantities. 

A  perhaps  better  way  to  judge  the  seed  crop  requirement  of  a  squirrel  is 
on  the  basis  of  seeds  produced  per  acre  of  forest.  Isaac  (1943:17)  reported 
the  seed  fall  in  a  100-year-old  stand  to  \ary  from  34,000  to  123,000  seeds 
per  acre.  If  we  assume  126,000  seeds  per  acre  and  accept  the  value  of  42,000 
seeds  per  pound  (Anon.,  1948:291),  this  amounts  to  three  pounds  of  whole 
seed  per  acre.  Applying  the  68.6  per  cent  correction  factor  for  edible  dry 
weight,  we  get  2.06  pounds  (934  grams)  of  dry  squirrel  food  per  acre.  Hence, 
to  obtain  5.99  kilograms  of  seed  in  a  year,  our  male  squirrel  would  require 
6.4  acres  of  Douglas-fir  forest  producing  126,000  seeds  per  acre.  Unfortunately, 
in  the  Rocky  Mountain  region  the  Douglas-fir  would  probably  produce  this 
gf)od  a  crop  only  two  or  three  years  out  of  10. 

I  ha\e  not  found  more  than  two  bushels  of  Douglas-fir  cones  in  a  cache, 
and  they  have  always  been  mixed  with  blue  spruce,  or  sometimes  ponderosa 
pine  cones.  H.  M.  Swift  has  found  six  to  eight  bushels  of  Douglas-fir  cones  in 
a  single  mixed  cache — still  considerably  below  the  theoretical  maximum  require- 
ment of  25.4  bushels  a  year.  The  seeming  discrepancy  bet\veen  the  require- 
ment and  the  magnitude  of  caches  usually  found  is,  no  doubt,  explained  by 
the  fact  that  squirrels  do  not  subsist  entirely  on  Douglas-fir  cones,  or  even  on 
cones  of  all  species  available.  Conifer  seeds  are  the  mainstay  in  winter  months, 
but  in  summer  large  amounts  of  other  kinds  of  plant  food  are  eaten,  particularly 
fungi  (Buller,  1920;  Hatt,  1929;  and  C.  C.  Smith,  1968).  A  more  realistic 
estimate  of  conifer  seed  required  by  a  red  squirrel  would  be  40  or  50  per  cent 
of  the  theoretical  yearly  requiremnt;  and  in  most  localities  and  years  only 
part  of  the  conifer  seed  would  be  pro\ided  by  Douglas-fir.  Fifty  per  cent  of 
the  yearly  requirement  would  be  proxided  by  12.7  bushels  of  cones,  which 
might  be  produced  by  about  nine  mature  trees,  or,  by  the  alternative  estimate, 
Ijy  3.2  acres  of  forest. 

Eufielinann  spruce. — Seed  productivity  for  the  Engelmann  spruce  is  less 
well  known  than  for  the  more  commercially  \aluable   Douglas-fir.     However, 


260 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Table  2. — Seed  production  data  reported  for  Exgelmaxx  spruce. 


Amount 

Area 

Reference 

Thousand 

seed.s/poiind 

135  (69-200) 

Anonymous,  1948 

175 

Nordiern  Rockies 

McKeever,  1942 

69-135 

Utah  and  Idaho 

Korstian  and  Baker, 

1925 

Pounds 

seed/bushel 

.50  (.40-1.) 

Korstian  and  Baker, 

1925 

.8-1.2 

Montana 

Lowdermilk,   1925 

Cones/ljushel 

2800 

Colorado 

R.  B.  Finley 

Bushels 

cones/tree 

1.25 

Cox,  1911 

Thousand 

seeds/acre 

570-760 
200-2000 

Montana 

Lowdermilk,  1925 
Roe,  1967 

Pounds 

seed/acre 

12 

Cox,  1911 

Seed 

trees/acre 

12 

Cox,  1911 

Cal/g  dw 

kernel 

7107 

C.  C.  Smith,  1968 

Kernel 

corr  factor 

.770 

R.  B.  Finley 

the  Engelmann  is  the  most  important  tree  in  this  region  for  the  red  squirrel 
and  commonly  provides  practically  its  entire  winter  cone  suppl>'.  To  supplement 
the  data  on  Engelmann  spruce  in  Table  2,  I  dissected  and  weighed  three  lots 
of  10  seeds  as  was  done  for  Douglas-fir.  The  weight  of  live  endosperms  and 
embryos  averaged  78.9  per  cent  of  the  whole  seed  weight  and  the  corrected 
value  for  dry  weight  was  77.0  per  cent.  Applying  this  conversion  factor  to  the 
0.50  pound  of  seed  per  bushel  in  Table  2  gives  0.385  pound  of  dry  food  per 
bushel,  or  175  grams.  An  adult  male  requiring  42,700  kg.  calories  per  year 
would  thus  need  6.01  kilograms  of  Engelmann  seed,  or  34.3  bushels  of  cones. 
If  50  per  cent  of  the  squirrel's  diet  were  provided  by  other  foods,  this  would 
still  lea\'e  17  bushels  of  cones  needed.  The  only  figure  available  for  cones  per 
tree  is  1.25  bushels,  by  Cox,  for  a  "good  crop."  Assuming  such  a  cone  produc- 
tion, 14  good  seed  trees  would  provide  50  per  cent  of  the  squirrel's  annual 
energy  requirement. 

For  an  estimate  per  acre  of  forest  we  can  use  135,000  seeds  per  poimd  and 
600,000  seeds  per  acre  (Table  2)  and  compute,  as  before,  3.42  pounds  (1.55 
kilograms)  of  dry  food  per  acre,  or  3.9  acres  to  meet  the  full  energy  needs  for 
one  year.  Only  2.0  acres  would  be  needed  to  provide  the  17  bushels  of  cones 
for  half  the  diet.  Mr.  Swift  has  collected  as  much  as  15  bushels  from  a  single 
midden,  and  many  bushels  may  have  been  left  because  of  the  small  size  of 
Engelmann  cones  and  the  tediousness  of  digging  them.  However,  most  middens 
yield  much  less. 

Blue  spruce. — The  blue  spruce  is  the  most  dependable  seed  producer  for 
red  scjuirrels  in  the  Transition  Zone  of  the  Colorado  East  Slope.  Unfortunately, 
little  information  has  been  published  on  its  cone  and  seed  production.  If  we 
assume  the  same  values  as  used  for  Engelmann  spruce,  for  the  energy  content 


FiNLEY — Cone  Caches  of  Tamaisciurus 


261 


Table  3. — Seed  production  data  reported  for  blue  spruce. 


Amount 


Area 


Reference 


Thousand 
seeds/pound 

Pounds 

seed/bushel 

Cones/bushel 

Cones/tree 


106  (80-163) 
80-85 

1.0  (.75-1.2.5) 
989 


Utah  and  Idal 


Colorado 


500  ( 200-2000 )     Colorado 


Anonymous,  1948 
Korstian  and  Baker,  1925 

Korstian  and  Baker,  1925 
R.  B.  Finley 
R.  B.  Finley 


of  seed  and  the  correction  factor  for  edible  dry  weight,  the  error  should  not 
exceed  10  per  cent.  I  counted  989  cones  in  one  bushel.  In  the  \icinity  of 
Evergreen,  Colorado,  I  made  some  rough  visual  estimates  of  cones  on  trees, 
and  concluded  that  they  averaged  about  500  per  tree  for  a  good  seed  year. 
With  these  figures  and  the  few  data  in  Talile  3,  we  can  come  up  with  the 
following  for  blue  spruce:  0.77  pound,  or  350  grams,  of  dry  food  per  bushel; 
6.01  kilograms  of  dry  seed,  or  17.2  bushels  of  cones  to  fill  the  total  annual 
need;  0.506  bushel  of  cones  per  tree;  and  34  trees  to  fill  the  annual  need. 
The  more  likely  consumption,  for  50  per  cent  of  the  diet,  would  be  8.6  bushels 
from  17  trees. 

I  have  no  figures  for  seed  fall  per  acre,  but  we  may  reasonably  assume 
10  good  seed  trees  per  acre,  in  which  case  1.7  acres  would  provide  enough 
blue  spruce  seed  for  50  per  cent  of  the  annual  diet.  Mr.  Swift  reports  that  he 
has  obtained  as  many  as  13  liushels  of  blue  spruce  cones  from  a  single  midden, 
considerably  more  than  the  8.6  needed  for  half  the  diet. 

Ponderosa  pine. — Although  the  ponderosa  pine  is  a  less  important  food 
source  for  red  squirrels  than  the  preceding  tree  species,  its  seed  productivity  is 
one  of  the  better  known  (.see  Table  4).  I  dissected  and  weighed  si.x  lots  of 
five  seeds  each  and  found  that  the  live  kernels  weighed  57.6  per  cent  of  the 
whole  seed,  and  the  corrected  dry  weight  was  53.9  per  cent  of  the  whole 
seed  weight.  Accepting  1.32  pounds  of  seed  per  bushel  and  1.5  bushels  of 
cones  per  tree  from  the  table,  we  obtain  0.71  pound,  or  323  grams,  of  dry  food 
per  bushel;  5.65  kilograms  of  dry  seed,  or  17.5  bushels  of  cones  for  a  year's 
supply;  and  12  "good  seed  trees"  to  yield  this  many  cones.  The  seed  study  by 
Powells  and  Schubert  (1956)  is  one  of  the  most  detailed  yet  published. 
It  analyzes  seed  data  obtained  over  a  28-year  period  from  several  species  in 
mixed  stands  in  the  pine  region  of  California.  From  their  data  on  ponderosa 
pine  we  obtain  71  seeds  per  cone  (p.  10)  and  150  cones  per  tree  (p.  21), 
which,  using  9700  seeds  per  pound,  we  can  con\ert  to  0.593  pound  or  269 
grams  (dry  weight)  per  tree,  and  21  "dominant  trees"  to  produce  a  one  year 
supply. 

Of  the  several  reported  measurements  of  ponderosa  pine  seed  fall,  we  can 
take  205,000  seeds  per  acre  from  Roeser  (the  average  of  two  good  seed  years 
out  of  nine  in  Colorado)  and  his  14,700  seeds  per  pound  to  compute  7.49 
pounds  or  3.40  kilograms  (dry)  per  acre,  and  1.66  acres  to  meet  the  needs  of 
a  full  year,  or  0.83  acre  for  half  that  need.  Because  ponderosa  pine  produces 
notoriously  erratic  seed  crops  in  Colorado,  in  most  years  an  acre  of  timber  would 
produce  considerably  less  seed  than  required  to  proxide  half  the  energy  needs 
of  a  squirrel. 


262  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Table  4. — Seed  productiOiX  data  reported  for  ponderosa  pine. 


Amount 

Area 

Reference 

Thousand 

seeds/pound 

14.7 

Colorado 

Roeser,  1941 

16  (13-19) 

Utah 

Korstian  and  Baker,  1925 

9.7  (8.3-12) 

Idaho 

Korstian  and  Baker,  1925 

Seeds/cone 

64-92 

Anonymous,  1948 

69-73 

California 

Powells  and  Schubert,  1956 

Pounds 

seed/l)ushel 

1.32  (.75-2.) 

Korstian  and  Baker,  1925 

1.5 

Cox,  1911 

Cones/bushel 

200-300 

Anonymous,  1948 

218 

Idaho 

Miller  and  Lemmon,  1943 

Cones/tree 

200 

Fowells,  1965 

50-2250 

California 

Fowells  and  Schubert,  1956 

Bushels 

cones/tree 

1-1.5 

Rudolf,  1961 

4 

Co.x,  1911 

Pounds 

seed/tree 

2 

Southwest 

Pearson,  1950 

Thousand 

seeds/acre 

205 

Colorado 

Roeser,  1941 

164 

California 

Fowells  and  Schubert,  1956 

123-345 

Idaho 

Curtis  and  Foiles,  1961 

Pounds 

seed/acre 

30 

Cox,  1911 

Seed 

trees/acre 

5 

Co.x,  1911 

Cal/g  dw 

whole  seed 

5626 

Long,  1934 

Cal/g  dw 

kernel 

7558 

C.  C.  Smith,  1968 

Kernel 

corr.  factor 

.539 

R.  B.  Finley 

Lodgepole  pine. — Lodgepole  pine  seed  is  not  a  preferred  food  of  the  red 
squirrel  but  its  widespread  and  almost  continuous  a\'ai]ability  make  it  an 
important  reser\e  or  emergency  food.  Using  the  data  in  Table  5,  and  assuming 
the  same  correction  factor  for  edible  dry  weight  as  with  ponderosa  pine  (which 
may  be  considerably  in  error  because  of  the  great  differences  in  seed  size), 
we  can  compute  the  following  estimates  for  lodgepole  pine:  0.291  pound  or 
132  grams  of  dry  food  per  l^ushel,  and  6.26  kilograms  of  dry  seed,  or  47.4 
bushels  of  cones,  to  fill  the  annual  requirement,  which  might  be  produced  by 
95  trees.  From  this  we  see  that  almost  three  times  as  many  bushels  of  cones 
are  required  as  for  ponderosa.  This  is  a  consequence  of  the  small  number  of 
lodgepole  seeds  per  cone  and  their  very  small  size. 

If  we  use  an  estimated  annual  yield  in  Colorado  of  320,000  lodgepole  seeds 
per  acre,  this  amounts  to  1.69  pounds  or  0.77  kilograms  (dry)  per  acre.  The 
6.26  kilogram  annual  requirement  would  be  produced  by  8.2  acres  of  lodgepole 
forest.  Lotan's  estimate  of  3,228,000  seeds  in  Montana  was  based  on  counts  of 
total  serotinous  cones  on  branches  and  is  not  annual  production.    It  is  a  measure 


FiNLEY — Cone  Caches  of  Tamaisciurus 


263 


Table  5. — Seed  production  data  reported  for  lodgepole  pine. 


Amount 

Area 

Reference 

Thousand 
seeds/pound 

102 

Anonymous,  1948 

120 

Co.x,  1911 

85-160 

Bates,  1930 

Seeds/cone 

21.1 

Montana 

Lotan,  1967 

1-50 

Fowells,  1965 

Pounds 

seed/bushel 

..54  (..34-1.) 

Korstian  and  Baker,  1925 

.40 

Tillotson,  1917 

Cones/liushel 

1778 

Colorado 

R.  B.  Finley 

1500-2000 

Bates,  1930 

Bushels 
cones/tree 

.5 

Cox,  1911 

Thousand 
seeds/acre 

.320  ( 30-700 ) 

Colorado 

Bates,  1930 

3228 

Montana 

Lotan,  1967 

Pounds 
seed/acre 

3  (.3-7.5) 

Bates,  19.30 

Seed 

trees/acre 

40 

Co.x,  1911 

Cal/g  d\v 
\\  hole  seed 

5989 

Long,  1934 

Cal/g  dw 
kernel 

6827 

C.  C.  Smith,  1968 

of  total  stored  seed  per  acre  a\'ailal:)le  in  cones  of  the  crops  of  many  years. 
A  squirrel  obliged  to  subsist  for  a  year  entirely  on  such  a  diet  could  get  by  on 
only  0.8  acre,  but  could  do  it  for  only  one  year.  Lodgepole  stands  differ  widely 
in  proportion  of  serotinous  cones.  Bates  (1930:11)  reported  about  three  times 
as  much  seed  retained  in  older  persistent  cones  as  was  produced  in  the  current 
cone  crop. 

Cox  (1911:18)  reported  as  much  as  seven  bushels  of  lodgepole  cones  col- 
lected from  a  single  cache  in  Wyoming.  Although  this  is  far  below  the  24 
bushels  providing  half  the  annual  requirement,  the  squirrels  are  not  dependent 
on  cached  lodgepole  cones.  Indeed,  such  industrious  activity  seems  to  serve 
little  more  than  convenience  and  an  urge  for  "busy  work." 

Estimated  seed  requirements. — To  summarize  and  compare  the  estimated 
seed  requirements  of  a  squirrel  for  half  a  year,  the  amounts  calculated  from 
Tables  1-5  as  described  above  are  presented  in  Table  6.  Requirements  ha\e 
also  been  expressed  in  numbers  of  fresh  whole  seeds  by  using  the  correction 
factors  for  edible  dry  weight,  and  the  numbers  of  seeds  per  poimd. 

It  must  be  emphasized  that  a  wide  degree  of  variation  enters  into  most  of 
the  parameters  of  Table  6,  particularly  those  expressed  in  trees  or  acres.  Never- 
theless, rough  though  these  estimates  are,  they  provide  a  quantitative  basis  for 
consideration  of  several  questions  of  red  squirrel  ecology. 

Rate  of  midden  accumulation. — How  many  squirrel-years  of  cone-stripping 
were  required  to  produce  a  midden  of  the  magnitude  of  those  described?  The 
blue  spruce  midden  by  Beaver  Creek  was  sufficiently  well  demarcated  to  permit 


264  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

easy  nieasuiement.  From  its  surface  area  and  30  depth  measurements,  I 
estimated  its  total  volume  to  be  602  cubic  feet,  or  484  bushels.  Three  samples 
of  mixed  midden  material  totaling  one  bushel  were  taken  from  three  parts  of 
the  midden  and  examined  for  content.  The  bushel  contained  the  cores  of  239 
blue  spruce,  52  ponderosa,  and  57  Douglas-fir  cones,  which  were  equivalent 
to  uneaten  closed  cones  amoimting  to  0.24  bushel  blue  spruce,  0.17  bushel 
ponderosa,  and  0.036  bushel  Douglas-fir.  These  total  0.45  bushel  of  mixed 
cones  from  which  the  bushel  of  midden  was  deri\ed.  This  ratio  indicates  that 
the  484  bushels  of  midden  were  deri\ed  from  218  bushels  of  fresh  cones.  If 
the  fractions  of  each  kind  of  cone  in  the  sample  bushel  are  the  same  as  for 
the  entire  midden,  they  represent  117  bushels  of  blue  spruce,  83  bushels  of 
ponderosa,  and  17/2  bushels  of  Douglas-fir  cones. 

This  amount  of  cones  could  have  been  consumed  by  one  adult  male  squirrel 
in  a  25-year  period  (13.6  plus  9.6  plus  1.3  years,  respectively),  based  on  the 
estimates  in  Table  6.  Five  squirrels  could  have  stripped  that  amount  of  cones 
in  only  fi\e  years.  If  the  squirrel  lixed  on  cones  for  substantially  more  than 
50  per  cent  of  his  diet,  that  amount  might  ha\  e  been  consumed  by  one  squirrel 
in  only  15  or  20  years.  I  have  no  information  on  the  rate  of  weathering  and 
decomposition  of  squirrel  middens,  l:)ut  have  seen  noticeable  deterioration  of 
a  few  middens  in  a  fi\e-year  period  when  these  were  little  used.  Still,  since 
that  stand  of  blue  spruces  was  undoubtedly  occupied  by  squirrels  for  many 
decades,  five  years  seems  to  me  an  unbelievably  short  period  in  which  all 
preceding  debris  should  have  disappeared.  These  considerations  are  compatible 
with  the  \'iew  that  such  a  midden  is  utilized  by  no  more  than  a  single  family 
group  of  squirrels  at  one  time,  and  probably  for  most  of  the  time  l)y  only  a 
single  territorial  indi\'idual,  as  believed  by  numerous  previous  workers. 

Effects  of  Cache  Raiding  by  Man  on  Red  Squirrels 
Does  the  practice  of  raiding  cone  caches  for  nursery  seed  reduce 
the  squirrels'  chances  of  survival  or  affect  their  reproduction?  The 
amount  cached  in  a  single  midden  varies  greatly.  As  much  as  15 
bushels  has  been  reported  from  a  single  cache  (Toumey  and  Kor- 
stian,  1942:116),  but  the  average  amount  is  probably  between  two 
and  five  bushels.  This  almost  never  makes  up  the  total  seed  supply 
in  the  midden,  because  most  cones  are  well  concealed  and  the 
average  person  digging  them  out  mo\es  on  to  another  midden  when 
the  work  becomes  less  productive.  I  have  no  basis  for  judging  what 
percentage  of  cached  cones  is  found  by  the  average  "cone-digger." 
When  amounts  usually  gathered  from  single  middens  are  com- 
pared with  estimated  squirrel  requirements  ( Table  6 ) ,  the  amounts 
lost  by  a  squirrel  seem  to  be  a  substantial  fraction  of  its  total  winter 
needs.  If  its  needs  are  as  great  as  indicated  in  Table  6,  it  seems 
evident  that  they  would  not  always  be  satisfied  by  a  single  cache. 
The  common  occurrence  of  small  to  medium-size  middens  and 
caches,  often  close  together,  leads  me  to  believe  that  indi\'idual 
squirrels  often  control  or  utilize  more  than  one  midden,  especially 


FiNLEY — Cone  Caches  of  Tamaisciurus  265 

Table  6. — Estimated  amounts*  of  seed  needed  to  provide  21,350  kg.  cal, 

HALF  OF   THE   ANNUAL   ENERGY   REQUIREMENT   OF    AN    ADULT    TaiuiaSCiuniS. 

Douglas-      Engelniann  Blue  Ponderosa       Lodgepole 

fir  spruce  spruce  pine  pine 

Thousands  fresh 

whole  seeds  400  1200  910  170  1300 

Ky  of  fresh 

whole  seeds  4.4  3.9  3.9  5.2  5.8 

Kg  of  dry 

seed  kernels  3.0  3.0  3.0  2.8  3.1 

Number  of 

cones    9200  ....._  . 2400  62,000 

Bushels  of 

cones    13  17  8.6  8.7  24 

Number  of 

good  seed  trees  -        8.5  14  17  6  47 

Acres  of 

forest    3.2  2.0  1.7  .8  4.1 

"  The  values  in  this  table  are  internally  inconsistent  because  they  are  derived  from  several 
independent  sources. 

when  populations  are  low.  If  so,  the  squirrel  is  less  likely  to  lose 
most  of  his  winter  supply  to  eone  eollectors. 

The  seriousness  of  the  cone  loss  varies  with  the  site  and  cone 
crop.  Loss  of  a  cache  of  ponderosa  or  Douglas-fir  cones  from  a 
blue  spruce  midden  in  the  Colorado  Transition  Zone  would  be  of 
little  consequence  because  the  squirrels  are  limited  by  cache  sites, 
not  cones,  plenty  more  of  which  can  be  gathered  farther  up  the 
hillside.  On  the  other  hand,  in  a  year  of  seed  failure  in  these  species 
and  when  blue  spruce  cones  are  being  cached,  loss  of  the  latter 
could  be  critical,  because  blue  spruce  stands  and  cache  sites  are 
restricted  to  narrow  dendritic  distribution  patterns  easily  accessible 
and  sought  out  by  cone  collectors.  At  higher  elevations,  in  Engel- 
mann  spruce  forests,  distribution  of  squirrels  and  caches  is  more 
widespread  and  unrestricted  by  suitable  midden  sites.  In  such 
forests  a  squirrel  that  loses  most  of  its  winter  supply  may  be  unable 
to  obtain  replacement  cones  from  trees  closer  to  the  middens  of  its 
neighbors  and  may  be  obliged  to  fall  back  on  lodgepole  cones  or 
other  second-choice  foods.  Hence,  the  seriousness  of  cache  loss  in 
the  spruce-fir  forest  may  depend  on  the  population  densit)'  and 
territorial  pattern  of  the  sfjuirrels  in  the  local  area. 

Foresters  and  silviculturists  have  expressed  the  view  that  red 
squirrels  are  not  adversely  affected  by  cone  gathering  (Cox,  1911: 
17;  Baldwin,  1942:48).  This  belief  is  supported  by  the  general 
obser\  ation  that  in  good  seed  years  squirrels  cache  far  more  cones 
than  they  need,  and  that  excess  cones  a  year  or  more  old  are  com- 
monly found  in  later  years  by  cone  collectors.    They  also  note  that 


266  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

squirrels  survive  bad  eone  years  in  substantial  numbers  and  are  on 
hand  to  exploit  the  next  good  crop  that  appears  (Carman,  1955:9). 
I,  too,  have  sometimes  found  excess  cones  cached  in  large  numbers 
from  previous  years,  but  I  have  also  found  in  other  years  and  locali- 
ties almost  total  exhaustion  of  cones  from  caches,  even  those  a  year 
or  more  old.  The  ability  of  red  squirrels  to  survive  cone  crop  failures 
is  aided  by  their  remarkable  ability  to  locate  and  utilize  such  old 
"excess"  cones  (M.  C.  Smith,  1968:308)  as  well  as  their  ability  to 
subsist  on  dried  fungi,  cambium,  buds,  and  other  plant  parts. 

Halvorson  (unpublished  data)  found  evidence,  however,  that 
squirrels  surviving  a  winter  of  cone  crop  failure  have  lower  repro- 
ductive success  than  in  years  when  a  good  crop  of  conifer  seed  is 
available.  C.  C.  Smith  (1968:53)  stated  that  during  the  height  of 
lactation  a  female  would  have  been  unable  to  extract  and  consume 
enough  lodgepole  pine  seed  to  equal  the  amount  of  energy  she 
obtained  from  Douglas-fir  seed.  Any  inadequacy  of  lodgepole  seed 
is  caused  by  the  time  and  energy  required  to  strip  the  cones,  not 
the  total  amount  of  seed  available.  It  is  perhaps  possible  that 
dependence  on  lodgepole  seed  would  reduce  the  survival  of  young 
if  other  good  food  sources  were  not  available,  but  squirrel  popula- 
tions seem  able  to  maintain  themselves  in  pure  lodgepole  forests  in 
Colorado.  Cathering  of  lodgepole  cones  from  such  caches  would 
have,  in  any  case,  no  noticeable  effect  on  either  survival  or  repro- 
duction of  squirrels  in  the  Central  Rockies,  because  in  this  region 
a  high  percentage  of  the  axailable  seed  always  remains  in  the  sero- 
tinous cones  on  the  branches.  The  circumstance  most  likely  to 
affect  squirrel  reproduction  would  be  loss  of  a  large  spruce  cache 
after  the  remaining  cones  on  the  trees  have  opened,  and  in  a  year 
preceded  by  a  year  of  cone  crop  failure. 

Effects  of  Red  Squirrels  on  Natural  Forest  Regeneration 
The  \'alue  of  red  squirrels  as  har\esters  of  conifer  seed  for  tree 
planting  must  be  weighed  against  the  harm  they  do  in  cutting  down 
cones  that  would  otherwise  shed  their  seeds  and  contribute  to  the 
replacement  of  the  forest  stand.  The  actual  harm  done  is  not  easy 
to  assess  because  only  a  tiny  fraction  of  any  year's  seed  crop,  or 
often  none  at  all,  survives  to  produce  mature  trees.  A  large  propor- 
tion of  the  seed  crop  is  normally  consumed  by  birds  and  mammals, 
without  preventing,  under  natural  conditions,  the  successful  replace- 
ment of  old  forests  with  new.  But  under  the  impact  of  logging  and 
the  demand  for  rapid  regeneration  of  the  forest,  the  problem  of 
seed  loss  becomes  more  acute. 


FiNLEY — Cone  Caches  of  Tamaisciurus  267 

A  few  intensive  studies  in  the  Pacific  Northwest  have  clearly 
defined  and  quantified  the  large  number  of  environmental  factors 
that  cut  down  the  survival  of  Douglas-fir  seeds  and  seedlings  ( Isaac, 
1943;  Carman,  1955;  Cashwiler,  1967).  Other  studies  on  regenera- 
tion of  ponderosa  pine  ( Roe  and  Squillace,  1950;  Foiles  and  Curtis, 
1965)  have  described  the  infrequent  and  improbable  circumstances 
required  for  the  replacement  of  ponderosa.  From  Munger's  (1930) 
\  i\  id  review  essay  on  forest  regeneration,  it  is  clear  that  densities 
of  seed  fall  in  the  tens  or  e\'en  hundreds  of  thousands  of  seeds  per 
acre  are  required  even  in  favorable  years  if  nature  is  to  have  a  good 
chance  to  achieve  a  satisfactory  stocking  of  young  trees  within  a 
few  years.  In  view  of  the  high  seed  fall  required,  it  is  important  to 
know  whether  Tamiascitirus  is  capable  of  cutting  such  a  high 
proportion  of  a  good  cone  crop  that  enough  seeds  are  not  left  to 
oxercome  the  extremely  high  odds  against  surxdval. 

In  Table  6  the  figures  gixen  for  acres  of  forest  of  the  various 
timber  types  are  estimates  of  area  required  to  provide  half  of  the 
squirrel's  needs,  assuming  that  the  seed  fall  is  of  the  densit}'  specified 
under  that  tree  species  in  the  section  on  seed  biomass.  One  may 
estimate  the  area  of  forest  required  under  different  conditions  of 
seed  fall  by  assuming  any  seed  fall  figures  from  Table  6,  or  any 
desired  density  of  seed  fall,  and  dividing  it  into  the  number  of  fresh 
whole  seeds  required,  as  given  in  Table  6.  For  example,  the 
1,500,000  Douglas-fir  seeds  reported  for  a  virgin  stand  in  British 
Columbia,  dixided  into  400,000  seeds  required,  gi\e  0.27  of  an  acre 
per  squirrel.  It  is  extremely  unlikely  that  a  population  density  as 
high  as  this  could  occur  and  consume  all  of  such  a  bumper  seed  crop. 

Squirrel  densities  as  high  as  the  figures  in  Tables  1-5  (from  0.8  to 
four  acres  per  squirrel)  are  remarkably  similar  to  squirrel  popula- 
tion estimates  reported  by  several  workers  (Crinnell  and  Storer, 
1924:207;  Seton,  1929,  4:119;  Hatt,  1929:50;  C.  C.  Smith,  1968:55). 
Indeed,  they  are  probably  no  more  than  coincidence,  for  it  is 
obvious  that  red  squiiTels  do  not  consume  all  of  the  conifer  seed 
produced.  Forests  with  squirrel  populations  do  regenerate  success- 
fully, and  an  abundance  of  cones  is  often  left  in  the  trees  to  open 
and  shed  seed,  even  in  areas  where  many  squirrels  are  actively 
harvesting  cones.  Nevertheless,  these  calculations  and  comparisons 
indicate  that  a  high  population  of  red  squirrels  may  be  capable  of 
harxesting  practically  an  entire  crop  of  cones  even  in  a  good  seed 
X  ear.  The  fact  that  a  fairly  good  seed  fall  does  ocur  now  and  then 
can  be  explained  by  txvo  xariables;  on  the  one  hand,  bumper  seed 


268  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

crops  sometimes  produce  far  more  seed  than  shown  in  Tables  1-5, 
and  on  the  other  hand,  squirrel  population  densities  are  often  much 
lower  than  the  estimates  given  above. 

A  great  many  shaky  assumptions  and  narrowly  based  estimates 
entered  into  the  calculations  of  seed  production  and  food  require- 
ments presented  in  Tables  1-6.  The  variables  range  so  widely  that 
some  of  the  averages  could  be  in  error  by  two  or  three  times.  Much 
better  biomass  and  consumption  data  are  needed,  in  particular,  for 
years  and  forests  of  interest. 

The  theoretical  deduction  reached  above,  that  red  squirrels  may 
be  capable  of  harvesting  an  entire  cone  crop,  is  supported  by  a 
number  of  reports  on  the  magnitude  of  depletion  of  cone  crops  in 
particular  situations.  In  California,  during  a  year  when  sugar  pines 
were  the  only  trees  bearing  cones,  Tamiasciunis  douglasii  cut  down 
896  (54  per  cent)  of  an  original  crop  of  1656  sugar  pine  cones  on 
20  marked  trees  (Tevis,  1953:130).  On  the  Stanislaus  Experimental 
Forest  in  1952  many  ponderosa  pines  had  more  than  50  per  cent  of 
their  cones  cut;  one  tree  that  started  with  926  cones  lost  93  per  cent 
( Schubert,  1953 ) .  In  a  study  of  cone  losses  on  ponderosa  pines  on 
the  Kootenai  National  Forest,  Montana,  Squillace  (1953:2)  found 
that  60  to  89  per  cent  of  the  cones  produced  in  poor  and  fair  seed 
years  were  cut  by  squirrels. 

Carman  (1955:9)  reported  that  in  1948  at  the  Cowichan  Lake 
Experiment  Station,  British  Columbia,  Tamiasciunis  was  very  active 
in  24-inch  Douglas-firs  after  two  years  of  poor  cone  crops.  Four 
trees  that  had  an  average  crop  of  3000  cones  in  July  were  remeas- 
ured  in  September;  by  that  time  the  squirrels  had  taken  70  per  cent 
of  the  cones.  According  to  Carman:  "A  good  crop  may  coincide 
with  a  low  level  of  squirrel  activity  and  its  abundance  be  far  more 
effective  than  another  crop  with  similar  potential  but  subject  to 
intensive  depredation." 

In  summary,  I  do  not  agree  with  the  view  of  Crinnell  and  Storer 
(1924:207),  also  accepted  by  Hatt  (1929:132),  when  they  wrote: 
"It  would  appear  that  the  squirrels  merely  harvest  a  surplus."  The 
evidence  that  I  can  bring  together  on  the  subject  seems  to  fit  better 
the  view  of  Isaac  ( 1943:23) :  'The  rodents  and  birds  together  prac- 
tically clean  up  the  seed  in  years  of  light  or  medium  crops;  appar- 
ently a  surplus  is  left  for  germination  only  when  there  is  a  heavy 
crop."  It  may  well  be  that  the  habit  of  producing  a  "bumper"  seed 
crop  only  once  in  many  years  is  an  adaptive  response  to  populations 
of  squirrels  and  other  seed  eaters.    This  view  was  expressed  more 


FiNLEY — Cone  Caches  of  Tamaisciurus  269 

than  50  years  ago  by  a  forester  (Cox,  1911:16),  who  wrote:  "It  is 
only  by  the  production  of  occasional  or  periodic  crops  of  seed  that 
conditions  are  made  favorable  for  the  natural  reproduction  of  the 
tree  species." 

This  situation  poses  no  threat  to  the  regeneration  of  forests  uncut 
by  man,  because  delays  of  10,  20,  or  even  40  years  in  the  establish- 
ment of  a  stand  of  young  trees  in  a  forest  opening  or  burn  are 
normal  features  of  plant  succession  and  provide  a  beneficial  degree 
of  ecological  diversity.  When  it  is  desirable  to  speed  up  the  replace- 
ment of  mature  trees  on  logged  or  burned  areas,  the  potential 
effects  of  red  squirrels  on  natural  seed  dissemination  should  be 
taken  into  consideration. 

Summary 

Red  squirrels  in  the  Rocky  Mountain  region  are  well  adapted 
to  the  montane  coniferous  forests  by  their  habit  of  storing  cones  in 
damp  places  for  winter  use.  They  are  able  to  subsist  on  seed  of  any 
species  of  spruce,  fir,  or  pine  available.  Cones  of  the  Engelmann 
spruce  and  lodgepole  pine  are  most  important  because  of  the  wide 
distribution  of  these  trees  in  the  Rockies  and  the  relative  depend- 
ability of  their  seed  crops. 

In  this  semiarid  region,  cones  are  cached  mainly  in  the  large 
middens  of  cone  litter  that  accumulate  under  trees  where  squirrels 
feed.  Such  shady  sites  provide  the  moisture  required  to  preserve 
cones  in  the  closed  condition.  A  large  midden  commonly  covers  an 
area  of  forest  floor  20  to  30  feet  across  and  may  contain  as  many  as 
five  to  10  bushels  of  cached  cones.  Amounts  up  to  a  bushel  are  often 
cached  under  water  in  shallow  pools,  bogs,  or  mountain  streams. 

Not  all  conifers  provide  adequate  shade  for  cone  storage.  The 
Engelmann  and  blue  spruces  are  ideal  because  of  their  dense  lower 
branches,  which  often  extend  to  the  ground.  The  ponderosa  pine, 
having  fewer  lower  branches  and  a  more  open  crown,  rarely  pro- 
vides suitable  cone-caching  sites.  Lodgepole  pines,  although  pro- 
viding poor  cache  sites,  retain  a  seed  supply  in  persistent  unopened 
cones.  Lodgepole  cones  are  most  heavily  harvested  when  other 
conifers  have  suffered  poor  cone  crops. 

The  kinds  of  middens  and  caching  sites  used  vary  with  latitude 
and  elevation.  Middens  are  widely  distributed  on  almost  any  kind 
of  terrain  in  the  spruce-fir  forests  of  the  Canadian  Life-Zone.  At 
lower  elevations  and  latitudes  middens  and  caches  are  restricted  to 
the  wettest,  coolest  sites,  mainly  under  spruce  stands  along  stream 
bottoms. 


270  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Cone  caches  of  Tamiasciurus  have  been  used  by  foresters  and 
nurserymen  as  the  main  source  of  conifer  seed  in  the  western  states 
since  the  beginning  of  large-scale  reforestation.  Many  thousands 
of  bushels  of  cones  are  harvested  each  year  by  raiding  squirrel 
caches.  Successful  exploitation  of  this  source  of  supply  requires  a 
good  knowledge  of  the  caching  behavior  of  squirrels  as  well  as  the 
site  requirements  and  cone-bearing  characteristics  of  the  various 
tree  species  in  the  area. 

Published  information  on  the  seed  productivity  of  live  conifer 
species  and  the  energy  requirements  of  Tamiasciunis  served  as  the 
basis  for  computing  the  seed  requirements  of  the  squirrel  and  its 
impact  on  the  seed  crop  of  the  forest.  These  theoretical  calculations 
indicate  that  a  single  male  may  require  from  eight  to  24  bushels  of 
cones  to  meet  energy  requirements  for  half  a  year.  Depending  on 
the  tree  species,  such  quantities  of  cones  might  be  produced  by 
eight-tenths  of  an  acre  to  four  acres  of  forest  in  a  good  seed  year. 
Measurement  and  sampling  of  the  cone  litter  in  one  midden  of 
602  cubic  feet  indicated  that  this  midden  could  have  been  accumu- 
lated from  the  feeding  activity  of  one  squirrel  in  a  period  of  25  years. 

Comparisons  of  reported  seed  productivity  of  various  conifers 
with  the  seed  requirements  of  Tamiasciunis  suggest  that  these 
rodents  may  be  capable  of  harvesting  an  entire  cone  crop,  at  least 
in  poor  or  moderate  seed  years.  Although  red  squirrels  pose  no 
threat  in  the  long  term  to  natural  forest  regeneration,  they  may 
harvest  enough  cones  to  delay  for  several  years  an  adequate  natural 
reseeding  of  burned  or  cutover  forest  land.  Nevertheless,  a  good 
population  of  Tamiasciurus  is  an  invaluable  asset  wherever  man 
must  collect  seed  for  artificial  seeding  or  planting  of  conifers. 

Literature  Cited 

Anonymous 

1948.     Woody-plant  seed  manual.    USDA  Forest  Serv.,  Misc.  Publ.,  654: 
vi  +  1-416,  281  figs. 

Bailey,  V. 

1931.     Mammals  of  New  Mexico.    N.  Amer.  Fauna,  53:1-412,  22  pis.,  58 
figs. 

Baldwin,  H.  I. 

1942.     Forest  tree  seed  of  the  north  temperate  regions.    Chronica  Botanica 
Co.,  Waltham,  Mass.,  xvi  +  240  pp.,  illustrated. 

Barrett,  J.  W. 

1962.     Regional  silviculture  of  the  United  States.    Ronald  Press,  New  York, 
\'ii  +  610  pp.,  illustrated. 

Bates,  C.  G. 

19.30.     The  production,  extraction,  and  germination  of  lodgepole  pine  seed. 
USDA  Tech.  Bull.,  191:1-92,  3  pis.,  21  figs. 


FiNLEY — Cone  Caches  of  Tamaisciurus  271 

BULLER,  A.  H.  R. 

1920.  The  red  squirrel  of  North  America  as  a  niyeophagist.  Trans.  Britisli 
Mycological  Soe.,  6:355-362. 

Cox,  W.  T. 

1911.  Reforestation  on  the  national  forests.  USDA  Forest  Serv.  Bull., 
98:1-57,  7  pis. 

Curtis,  J.  D.,  and  M.  W.  Foiles 

1961.  Ponderosa  pine  seed  dissemination  into  j^ronp  clearcuttings.  Jour. 
Forestry,  59:766-767. 

Foiles,  M.  W.,  and  J.  D.  Curtis 

1965.  Natural  regeneration  of  ponderosa  pine  on  scarified  group  cuttings 
in  central  Idaho.    Jour.  Forestry,  63:530-535. 

FowELLS,  H.  A.  (compiler) 

1965.     Sih'ics  of  forest   trees  of  the   United   States.     USDA   Forest   Serv., 
Agric.  Handbook,  271:  vi  +   1-762,  illustrated. 
FowELLS,  H.  A  ,  AND  G.  H.  Schubert 

1956.  Seed  crops  of  forest  trees  in  the  pine  region  of  California.  USDA 
Tech.  Bull.,  1150:1-48,  13  figs. 

Carman,  E.  H. 

1951.  Seed  production  by  conifers  in  the  coastal  region  of  British  Colum- 
bia related  to  dissemination  and  regeneration.  British  Columbia 
Forest  Serv.  Tech.  Publ.,  T  35:1-47,  illustrated. 

1955.  Regeneration  problems,  and  their  silvicultural  significance,  in  the 
coastal  forests  of  British  Columbia.  British  Columbia  Forest  Serv. 
Tech.  Publ,  T  41:1-67,  illustrated. 

Gashwiler,  J.  W. 

1967.  Conifer  seed  survival  in  a  western  Oregon  clearcut.  Ecology,  48: 
431-438. 

GiTTINGS,   E.   B. 

1941.  Climate  of  Colorado.  Pp.  798-808,  in  Climate  and  man,  USDA 
Yearbook  of  Agricultiue,  xii  -|-  1248  pp.,  illustrated. 

Greening,  G.  K. 

1941.  Climate  of  Arizona.  Pp.  761-772,  in  Climate  and  man,  USDA 
Yearbook  of  Agriculture,  xii  +   1248  pp.,  illustrated.  . 

Grinnell,  J.,  and  T.  I.  Storer 

1924.  Animal  life  in  the  Yosemite.  Univ.  California  Press,  Berkeley,  xvii 
+  752  pp.,  60  pis. 

Hatt,  R.  T. 

1929.  The  red  squirrel;  its  life  history  and  habits,  with  special  reference  to 
the  Adirondacks  of  New  York  and  the  Harvard  Forest.  Roosevelt 
Wild  Life  Ann.,  2:10-146,  52  figs. 

Hayward,  C.  L. 

1940.     Feeding  habits  of  the  red  squirrel.    Jour.  Mamm.,  21:220. 

Isaac,  L.  A. 

1943.  Reproductive  habits  of  Douglas-fir.  Charles  Lathrop  Pack  Forestry 
Found.,  Washington,  107  pp.,  illustrated. 

Klugh,  a.  B. 

1927.     Ecology  of  the  red  squirrel.   Jour.  Mamm.,  8:1-32,  12  figs. 

Korstian,  C.  F.,  and  F.  S.  Baker 

1925.  Forest  planting  in  the  Intermountain  Region.  USDA  Bull.,  1264: 
1-56,  7  pis.,  7  figs. 

Krauch, H. 

1945.  Influence  of  rodents  on  natural  regeneration  of  Douglasfir  in  the 
Southwest.    Jour.  Forestry,  43:585-589. 

Lavender,  D.  P.,  and  W.  H.  Engstrom 

1956.  VialMlity  of  seeds  from  squirrel-cut  Douglas  fir  cones.  Oregon  State 
Board  of  Forestry  Research  Note,  27:1-19,  illustrated. 


272  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Layxe,  J.  N. 

1954.     The   biology   of   the   red   squirrel,   Tamiasciurus  hudsonicus   loquax 
(Bangs),  in  central  New  York.    Ecol.  Monogr.  24:227-267. 
Long,  F.  L. 

1934.  Application  of  colorimetric  methods  to  ecological  research.  Plant 
Physiol.,  9:323-337. 

LOTAX,   J.   E. 

1967.  Cone  serotiny  of  lodgepole  pine  near  West  Yellowstone,  Montana. 
Forest  Sci.,  13:55-59,  1  fig. 

LOWDERMILK,  W.   C. 

1925.     Factors   affecting  reproduction  in   Engelmann   soruce.     Jour.   Agric. 
Res.,  30:995-1009,  6  pis. 
McKeever,  D.  G. 

1942.  Results  of  direct  seeding  of  western  redcedar  and  Engelmann  spruce 
in  the  northern  Rocky  Mountain  Region.  Nortliern  Rocky  Mountain 
Forest  and  Range  Expt.  Sta.,  Research  Note,  21:1-9,  niimeo. 

Miller,  H.  W.,  and  P.  E.  Lemmon 

1943.  Processing  cones  of  ponderosa  pine  to  extract,  dewing,  and  clean  the 
seed.   Jour.  Forestry,  41:889-894. 

Munger,  T.  T. 

1930.     Ecological  aspects  of  the  transition  from  old  forests  to  new.    Science, 
72:327-332. 
Pearson,  G.  A. 

1950.  Management  of  ponderosa  pine  in  the  Southwest.  USFS  Agric. 
Monogr.,  6:  viii  +  1-218,  64  figs. 

Rasmussen,  D.  I. 

1941.  Biotic  communities  of  Kaibab  Plateau,  Arizona.  Ecol.  Monogr. 
11:229-275,  20  figs. 

Roe,  a.  L. 

1967.  Seed  dispersal  in  a  bumper  spruce  seed  year.  U.S.  Forest  Serv. 
Research  Paper  INT-39:1-10,  1  fig. 

Roe,  a.  L.,  and  A.  E.  Squillace 

1950.     Can   we   induce   prompt   regeneration   in    selectively-cut   ponderosa 
pine   stands?     Northern    Rocky    Mountain    Forest   and    Range   Exp. 
Sta.,  Research  Note,  81:1-7,  mimeo. 
Roeser,  J. 

1941.  Some  aspects  of  flower  and  cone  production  in  ponderosa  pine.  Jour. 
Forestry,  39:534-536. 

Rudolf,  P.  O. 

1961.     Collecting  and  handling  seeds  of  forest  trees.    Pp.  221-226,  in  Seeds, 
USDA  Yearbook  of  Agriculture,  xv  -{-  591  pp.,  illustrated. 
Schubert,  G.  H. 

1953.     Ponderosa  pine  cone  cutting  by  squirrels.    Jour.  Forestry,  51:202. 
Seton,  E.  T. 

1929.     Lives  of  game  animals.    Doubleday,   Doran   &   Co.,  Garden  City, 
New  York,  4:  xxii  -f  1-440,  illustrated. 
Shaw,  W.  T. 

1936.     Moisture  and  its  relation   to  the  cone-storing  habit  of  the  western 
pine  squirrel.   Jour.  Mamm.,  17:337-349,  4  figs. 
Smith,  C.  C. 

1968.  The  adaptive  nature  of  social  organization  in  the  genus  of  tree 
squirrels  Tamiasciurus.    Ecol.  Monogr.,  38:31-63. 

Smith,  M.  C. 

1968.  Red  squirrel  responses  to  spruce  cone  failure  in  interior  Alaska. 
Jour.  Wildlife  Mgt.,  32:305-317,  1  fig. 

Squillace,  a.  E. 

1953.  Effect  of  squirrels  on  the  supply  of  ponderosa  pine  seed.  Northern 
Rocky  Mountain  Forest  and  Range  Exp.  Sta.,  Research  Note,  131: 
1-4,  niimeo. 


FiNLEY — Cone  Caches  of  Tamaisciurus  273 

Tevis,  L.  p. 

1953.     Effect  of  N'eitebrate  animals  on  seed  crop  of  sngar  pine.   Jour.  Wild- 
life Mgt.,  17:128-131,  1  fig. 

TiLLOTSOX,    C.    R. 

1917.     Reforestation   on    the   national   forests.     USDA    Forest   Serv.    Bull., 
475:1-63,  illustrated. 

TOUMEY,  J.   W.,   AND  C.    F.    KORSTIAX 

1942.     Seeding  and  planting  in  the  practice  of  forestry.    John  Wiley  &  Sons, 

New   York,  3rd  ed.,  xxii  +  520  pp.,  1.52  figs. 
1947.     Foundations  of  silviculture  upon  an  ecological  basis.    John  Wiley  & 

Sons,  New  York,  2nd  ed.,  xxi  +  468  pp.,  20  figs. 

WiNJUM,   J.    K.,    AND   N.   E.   JOHN-SOX 

1964.     Difierences  in  cone  numbers,  lengths,  and  cut-counts  in  the  crowns 
of  young  open-grown  Douglas  fir.    Jour.  Forestry,  62:389-391. 


LATE  CENOZOIC  BATS 

(SUBFAiVIILY  NYCTOPHYLINAE )  FROM  THE 

ANZA-BORREGO  DESERT  OF  CALIFORNIA 


BY 


John  A.  White 

In  the  summer  of  1967  almost  six  tons  of  matrix  were  washed 
using  modifications  of  the  techniques  originated  by  Hibbard  (1949) 
and  McKenna  (1962).  The  matrix  was  obtained  from  15  sites  in 
tliat  part  of  the  Pahn  Spring  Formation  ( Dibblee,  1954;  Woodring, 
1931;  and  Woodard,  1963)  in  the  Fish  Creek- Vallecito  Creek  areas  in 
the  Anza-Borrego  Desert  State  Park,  San  Diego  County,  Cahfornia. 
Washing  was  done  under  the  auspices  of  the  Los  Angeles  County 
Museum  of  Natural  History  and  supervised  by  Theodore  Downs 
and  George  J.  Miller.  Locality  LACM  65S3  proved  to  be  especially 
rich  in  small  vertebrate  bones,  and  except  for  one  specimen  from 
locality  LACM  6552,  it  is  from  this  site  that  all  specimens  discussed 
herein  were  recovered. 

I  thank  Theodore  Downs  for  his  continued  support,  advice,  and 
encouragement.  David  E.  Fortsch  and  George  J.  Miller  critically 
read  the  manuscript.  I  am  indebted  to  J.  Knox  Jones,  Jr.,  and 
Richard  G.  Van  Gelder  for  valuable  suggestions,  including  the  pro- 
\iding  of  clues  to  the  literature  on  Recent  Antrozoiis\  and  to  Lisa  A. 
Hansen,  who  made  the  illustrations.  The  cooperation  of  the  per- 
sonnel and  management  of  the  Anza-Borrego  Desert  State  Park  is 
gratefully  acknowledged.  This  research  was  supported  by  the 
National  Science  Foundation  under  grant  GB-5116.  For  permission 
to  use  comparative  specimens  I  thank  J.  Knox  Jones,  Jr.,  Museum  of 
Natural  History,  University  of  Kansas  (  KU ) ,  Richard  G.  Van  Gelder, 
American  Museum  of  Natural  History  ( AMNH),  and  Edson  Fichter, 
Idaho  State  University  Museum  (ISUM).  Theodore  Downs  and 
J.  R.  Macdonald,  Los  Angeles  County  Museum  of  Natural  History 
( LACM )  made  the  fossil  bat  material  available  for  study.  Measure- 
ments of  specimens  were  made  with  a  Gaertner  measuring  micro- 
scope to  the  nearest  micron.  Symbols  used  here  to  identify  indi- 
vidual teeth  are  as  follows:  in  the  upper  half  of  the  dentition 
indi\  idual  teeth  are  identified  with  capital  letters  and  numbers,  thus 
12  refers  to  the  upper  second  incisor;  teeth  in  the  lower  half  of  the 
dentition  are  identified  with  lower  case  letters  and  numbers,  thus 

(275) 


276  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

i3  refers  to  the  third  lower  incisor;  L  or  R  indicates  left  or  right  side 
of  the  jaw. 

The  genus  Antrozous  is  currently  subdivided  into  two  subgenera 
— Aiitrozous  and  Bauerus  (Van  Gelder,  1959).  There  are  two 
species  in  the  former  (A.  pallidiis  and  A.  koopmani),  which  are 
closely  related  (Orr  and  Silva  Taboada,  1960),  and  only  one  species 
in  Bane t us  (A.  dubiaquerciis). 

The  nyctophiline  bats  from  the  early  Pleistocene  of  the  Anza- 

Borrego  Desert  are  morphologically  distinct  from  the  two  currently 

recognized  subgenera  in  North  America  to  the  same  extent  that 

these  two  differ  from  one  another.   It  is  for  this  reason  and  on  data 

presented  subsequently  in   this  paper,  that   a  new   genus  is  here 

established,    and    Bauerus   is   regarded   as    a    genus    distinct    from 

Antrozous. 

Anzanycteris,  new  genus 
figures  1-5 

Type  species. — Anzanycteris  anzensis  (new  species). 

Diagnosis. — Mandible  with  il  and  i2  crowded  together,  i2  markedly  re- 
duced, with  simple  crown,  and  appressed  into  indentation  near  base  of  lower 
canine;  angular  process  slender  and  projecting  almost  horizontally  and  posteriad; 
weakly  developed  cingulum  on  lingual  edge  of  C-M3. 

Anzanycteris  anzensis,  new  species 

Holotype. — LACM  19300,  skull  with  jaws  in  articulation,  with  posterior 
part  of  cranium  missing  from  the  postglenoid  process  posteriad;  skull  crushed 
laterally;  a  left-lateral  and  two  medial  incisors  missing  from  mandible;  alveoli 
for  missing  incisors  visible. 

Diagnosis. — Same  as  for  the  genus. 

Type  locality  and  stratigraphy. — Locality  LACM  6583,  Upper  Tapiado 
Wash,  Badlands  in  Anza-Borrego  Desert  State  Park,  San  Diego  County,  Cali- 
fornia; approximately  4850  feet  (1478  meters),  stratigraphically  below  top  of 
Palm  Spring  Formation  in  the  Diablo  member  of  the  formation;  late  Blancan 
(early  Pleistocene)  in  age,  Arroyo  Seco  Fauna. 

Referred  specimens. — Topotypes:  LACM  19301,  fragmentary  cranium  with 
incisors  and  right  canine  missing,  RP4  and  LMl  broken,  posterior  end  of 
craniimi  missing,  fragmentary  right  petrous  portion  of  inner  ear  imbedded  in 
the  matrix  near  forward  end  of  braincase;  LACM  19303,  maxillary  fragment 
with  RM2  and  broken  RMl;  LACM  19304,  fragmentary  left  dentary  with  m2 
and  broken  ml  and  m3;  LACM  19305,  fragmentary  right  dentary  with  m2, 
aheoli  for  c,  p3,  p4,  and  ml,  and  mental  foramen;  LACM  19306,  fragmentary 
right  dentary  with  m3,  ventral  part  of  masseteric  fossa,  and  dentar>'  foramen; 
LACM  19307,  fragmentary  left  dentary  with  m2,  m3,  and  ventral  part  of 
masseteric  fossa;  LACM  19309  fragmentary  left  dentary  with  m2  and  m3; 
LACM  19515,  Lml  or  Lni2;  LACM  19516,  fragmentary  left  dentary  with  ml 
and  broken  p4  and  m2.  Locality  LACM  6552:  LACM  19308,  fragmentary 
left  dentary  with  m2  and  m3. 


White — Late  Cenozoic  Bats 


277 


Figs.  1-6.  Anzanycteiis  auzcnsis  new  genus  and  species.  1,  left  lateral  view  of 
skull  (holotype);  2,  ventral  view  of  mandibles  (holotype);  3,  ventral  view  of 
cranium  ( LACM  19301);  4,  dorsal  view  of  right  mandibular  fragment  with  m2 
(LACM  19305);  5,  dorsal  view  of  left  mandil)ular  fragment  with  m2-m3 
(LACM  19307);  6,  cf.  Anzaiu/ctcns  sp.,  ventral  \iew  of  right  maxillarv  frag- 
ment with  P4-M3   (LACM   19302).    The  short  scale  refers  to  Figs.   1-3,  the 

long  scale  to  Figs.  4-6. 


Specimens   used  in    comparisons. — Antrozoiis    paUidiis.     California:     San 


County,  KU  11373,  11374. 
County,  ISUM  7304,  7305. 
Nevada:  Churchill  County, 
Tres   Marias   Islands,   AMNH 


Bernardino    County,    KU    63546;    Contra    Costa 
Idaho:     Idaho    County,    ISUM    7370;    Bannock 
Kansas:     Barber    County,    KU    11179,    76874. 
KU   75891.    Bauerus  duhiaquercus.    Nayarit: 
180841. 

Description. — Two  incisors  (il  and  i2)  present,  i2  markedly  reduced,  having 
a  single  crown  and  appressed  into  an  indentation  on  the  medial  side  near  base 
of  the  lower  canine;  this  indentation  is  bounded  by  the  cingulum  dorsally  and 
antero-posteriorly.  The  lower  cheekteeth  are  essentially  as  in  Antrozous  and 
Bauerus.  The  coronoid  process  is  oriented  almost  xertically  as  in  Antrozous. 
The  angular  process  is  slender  and  projects  nearly  horizontally  and  posteriorly 
and  approximately  parallel  to  its  mate.  The  masseteric  fossa  is  moderates- 
deepened,  the  deepest  portion  antero-ventral.  The  shafts  of  the  upper  incisors 
are  more  than  half  as  long  as  those  of  the  canines,  and  the  roots  of  the  canines 
and  incisors  are  pressed  close  together.  The  spaces  between  the  upper  molars 
are  V-shaped,  the  apices  of  the  V's  projecti  g  laterad.  The  contact  bet\\'een 
P4  and  Ml  is  almost  parallel,  thus  leaxing  no  space  between   them,   and  the 


278 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


.F 


Fig.  7.    Lett  lateral  view  of  the  mandible  of  Batierus  dubiaquercii.s   (AMNH 

180565),  modified  from  Van  Gelder,   1959:    fig.   lA.    See  text  for  explanation 

of  symbols  pertaining  to  the  jaw  as  a  mandibular  lever. 

canine  is  closely  appressed  to  the  P4.  There  is  no  hypocone  on  Ml  and  M2, 
but  a  loph  extends  posteriad  to  the  level  of  the  metacone  and  almost  parallel 
to  the  longitudinal  axis  of  the  palate.  The  MS  is  essentially  as  in  Aiitrozoiis, 
as  is  the  infraorbital  canal.  The  upper  dentition  is  proliably  only  slighdy 
upturned  as  in  Antrozous.  The  cingula  on  the  upper  cheekteeth  are  weakly 
developed  when  compared  to  those  in  Antrozous  and  Bauerus. 

Contpahsou.s. — In  Anzanijcteris  the  number  of  teeth  is  the  same  as  in 
Antrozous  and  two  less  than  in  Bauerus  (Van  Gelder,  1959).  In  the  latter 
genus,  i3  is  a  tiny,  button-like  tooth  crowded  between  i2  and  the  canine;  it  is 
appressed  into  an  indentation  on  the  medial  side  of  the  canine.  In  Anzanijc- 
teris, i2  is  markedly  smaller  than  il   and  is  appressed  into  an  indentation  on 


Table   1. — Comparisons   of  the 

SUBFAMILY   NyCTOPHILINAE,   USING 


structures  in  the  four  genera  of  the 
characters  noted  in  miller  (1907:235). 


Character 

Nijctovhilus 

Bauerus 

Antrozous 

Anzanijcteris 

Number  and 
character  of 
lower  incisors 

3,  unreduced 

3,  i3  mark- 
edly 
reduced 

2,  unreduced 

2,  i2  mark- 
edly 
reduced 

Character  of 
lower  canine 

Probabh' 
"normal" 

With 

exca\ation 

"Normal" 

With 

excavation 

Hypocone 

Absent 

Present 

Absent 

Absent 

M3 

More  than 
half  crown 

Less  than 
half  crown 

Less  than 
half  crown 

Less  than 
half  crown 

area  of 
Ml  M2 

area  of 
Ml  M2 

area  of 
Ml  M2 

area  of 
Ml  M2 

m3 

Talonid 
unreduced 

Talonid 
reduced 

Talonid 
reduced 

Talonid 
reduced 

Rostrum 

Slighth- 
upturned 

Slightly 
upturned 

Slightly 
down- 
turned 

Probably 
slightly 
downturned 

White — Late  Cenozoic  Bats  279 

Table  2. — Cranial  measurements  (in  millimeters)  of  Anzanycteris  anzensis 
NEW  GENUS  and  SPECIES  AND  CF.  Anzanijctcris  sp.  All  numbers  rel.\te  to 
specimens   in   the    Los   Angeles    County   Museum    of    Natural    History. 


Measurement 

No.  19300 

No.  19301 

No.  19304 

No.  19306 

No.  19302 

Length  of  ma.xillary  tootlirow 
Length  of  infraorliital  canal  .. 
Depth  of  mandible  at  ni2  _— 

5.480 

.349 

1.948 

5.882 

5.315 

From  posterior  end  of 
m3  to  posterior  end  of 
angular  process   

APL  at  cingiilum  of  C      

5.707 
1.585 

1.629 
1.436 

Width  at  cingulum  of  C  

APL  P4 -__ 

1.065 

1.369 
1.633 

.985 

Width  P4  

1.824 

APL  Ml  

1.847 

1.882 

1.907 

Width   Ml 

2.108 

1.987 
1.973 
2.097 

2,294 

APL  M2  

Width   M2   

1.948 
2.539 

APL  M3  

.543 

.664 

.818 

Width    M3   .__-... 

Width  talonid  of  ml  

1.874 

1.023 

2.043 

APL  m2  _.-.. - -. 

Width  trigonid  of  m2 



1.694 
1.164 

1.288 

Width  talonid  of  m2 -..- 

APL  m3   

Width  trigonid  of  m3 

.448 

1.571 

1.032 

.447 

Width  talonid  of  m3  

the  medial  side  near  the  base  of  the  canine,  whereas  in  Anirozons  il  and  12  are 
subequal  in  size  and  the  canine  has  no  indentation. 

The  upper  toothrow  in  Anzanycteris,  as  in  Antrozous,  slopes  slightly  antero- 
dorsad,  whereas  in  Bauerus  this  upturning  of  the  tooth  row  is  more  pronounced 
or  more  bulldog-like,  and  the  sagittal  crest  is  far  more  pronounced. 

Because  the  holotype  of  Anzanycteris  has  undergone  some  distortion,  pre- 
sumably during  preservation,  the  degree  of  upturning  of  the  upper  toothrow 
was  determined  by  assimiing  the  mandibles  to  move  in  articulation  in  a  simple, 
hinge-like,  up-and-down  motion.  The  condyloid  process  has  a  configuration  not 
unlike  that  of  a  canid  and  the  postglenoid  process  is  well  de\eloped.  Assimiing 
that  the  latter  characteristics  of  the  ma  dible  would  restrict  it  to  a  minimal 
amount  of  lateral  motion,  it  follows  that  the  temporal  muscles  would  exert  the 
principal  force  in  adducting  the  mandible  while  the  masseter  and  pterygoid 
musculature  would  function  primarily  in  positioning  the  mandible  with  respect 
to  the  glenoid  fossa. 

Using  a  modification  of  a  technique  devised  by  Ostrom  (1966),  the  center 
of  the  glenoid  fossa  was  established  as  a  fulcrum  and  the  coronoid  process 
aboxe  the  level  of  the  fulcrinn  as  a  lever.  The  mandible  was  then  occluded 
and  the  center  of  origin  of  the  temporal  muscles  was  estimated.  A  line  was 
then  drawn  from  the  tip  of  the  coronoid  process  to  the  center  of  origin  of 
the  temporal  muscles.    Using  these  data  the  moment  arm  of  the  applied  force 


280  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

.10  .05  -0+  .05        ...  .10   . 


.15 


2  - 

3  - 

4  - 

5  - 

6  - 

7  - 

8  - 

9  - 

10  - 

11  ' 

12  - 

13  - 


Fig.  8.  Ratio  diagrams  modified  from  Simpson  et  al.  ( 1960),  comparing  se\eral 
cranial  dimensions  of  the  largest  and  smallest  specimens  of  Antrozotis  (open 
squares)  and  Bauems  (solid  circles)  with  Anzamjcteiis  (the  line  at  zero). 
The  logs  of  the  measrnements  of  Atizcini/ctchs  are  assumed  to  be  zero,  while 
the  differences  between  the  log  of  the  measurements  in  the  latter  genus  (stan- 
dard )  and  genera  being  compared  are  plotted  on  die  positixe  (  + )  or  negati\'e 
( — )  sides  of  the  zero  line.  The  dimensions  are  as  follows:  1  length  of 
maxillary  tooth  row;  2  posterior  end  of  m3  to  posterior  end  of  angular  process; 
3  APL  C  at  cingulum;  4  W  C  at  cingulum;  5  APL  P4;  6  APL  Ml;  7  APL  M2; 
8  W  talonid  ml;  9  APL  m2;  10  W  trigonid  m2;  11  W  talonid  m2;  12  APL  m3; 

13  W  trigonid  m3. 

(by  the  temporal  muscles)  was  calculated  for  the  holotxpe  and  for  all  com- 
parative specimens,  and  expressed  as  percentages  of  jaw  lengths  to  eliminate 
size  as  a  factor  in  comparisons.  To  quote  from  Ostrom  (1966:  302-303): 
"It  is  not  possible  to  measiue  this  moment  arm  directly  with  any  precision 
when  the  jaw  is  in  articulation  and  fully  adducted  (the  critical  position),  it 
must  be  calculated  from  other  parameters  taken  from  the  skull  and  jaws." 
See  Figure  7. 

Ostrom  continued:  "For  example:  the  moment  arm  of  the  applied  force  is 
a  function  of  1)  the  height  (h)  of  the  articulation,  2)  the  lever  distance 
( a )  betw  een  the  center  of  the  articulation  and  the  base  of  the  coronoid  process 
[determined  by  the  level  of  the  center  of  articulation,  above  or  below   the  level 


White — Late  Cenozoic  Bats  281 

of  tlie  toothrow]  and  3)  the  attitude  (angle  0)  of  the  line  of  action  of  the 
applied  force  (F)  relative  to  the  fulcrum.  .  .  .  the  moment  arm  of  the  applied 
force  is  calculated  by  m  =  ( ^  +  5 )  d  where  m  equals  the  length  of  the  moment 
arm  of  the  applied  force,  0  the  angle  between  die  applied  force  and  the  lever 
axis,  5  the  angle  between  the  diagonal  distance  ( d )  from  the  coronoid  ape\ 
to  the  center  of  the  glenoid  fossa  and  the  le\er  axis."  The  length  of  the 
moment  arm  of  the  applied  force  is  in  direct  proportion  to  the  force  applied  to 
the  mandible,  and  as  such  is  a  measure  of  the  force  itself. 

The  moment  arm  of  the  applied  force  ( m )  and  the  angle  0  are  greater  in 
Bauerus  than  in  Antrozoiis  and  Auzauyctciis.  This  is  thought  to  be  related  to 
the  greater  degree  of  the  "bulldog"  effect  in  Bauerus  than  in  the  other  two 
genera. 

Only  the  base  of  the  coronoid  process  is  kno\\'n  for  Auzanycteiis,  but  the 
slope  of  the  anterior  and  posterior  margins  of  the  process  are  visible  and  match 
favorably  the  condition  in  Antrozous,  but  not  Bauerus.  The  measurements  of 
the  moment  arms  expressed  in  percentages  of  mandible  lengths  are  as  follows: 
Antrozous— K\J  11374,  29;  KU  11373,  27;  KU  63546,  28;  KU  94363,  28; 
KU  94362,  26;  KU  75891,  27;  KU  76874,  25;  KU  11179,  24;  ISUM  7305,  26; 
AMNH  2159,  28  (calculated  from  Van  Gelder,  1959:  fig.  IB);  Bauerus— AMNH 
180841,  32;  AMNH  180565,  31  (calculated  from  Van  Gelder,  loc.  cit.:  fig. 
lA);  Anzanijcteris — LACM  19300:  22  (if  »  is  assumed  to  be  20  degrees) 
or  26  (if  ^  is  assiuned  to  be  50  degrees). 

It  can  be  inferred  from  Table  1  that  Nyetophilus  and  Bauerus  represent  an 
adaptive  type  with  a  bulldog-like  upturned  rostrum  as  opposed  to  the  condition 
in  Anzamjcteris  and  Anirozous.  The  length  of  the  upper  incisor  relative  to  the 
length  of  the  upper  canine  seems  to  vary  inversely  with  the  degree  of  uptinning 
of  the  rostrum. 

The  bulldog-like  upturning  of  the  cheekteeth  in  Bauerus  \'ersus  the  down- 
turning  in  Anirozous  and  Anzamjcteris  may  be  related  to  feeding  habits. 
Antrozous  is  known  to  feed  at  times  almost  exclusively  on  flightless  insects 
(Orr  1954).  Although  the  feeding  habits  in  Bauerus  are  unknown,  it  is  possible 
that  food  is  obtained  exclusi\ely  in  flight,  the  "bulldog"  effect  enabling  a 
stronger  hold  on  larger  insects  than  Antrozous  and  presumably  Anzanycteris. 

The  three  genera  are  almost  equally  morphologically  distinct  from  one 
another  (Fig.  8).  Study  of  the  reproductive  systems  in  the  living  forms  of 
the  subfamily  Nyctophilinae  probably  would  shed  further  light  on  their  rela- 
tionships. 

The  number  and  character  of  the  lower  incisors  and  the  related  character 
of  the  lower  canine  seem  to  pro\'ide  clues  to  the  diversification  within  the 
Nyctophilinae.  Nyetophilus  seems  to  be  the  most  primitive  genus  in  that  it 
has  three  fidly  developed  lower  incisors,  probably  no  indentation  on  the  lower 
canine,  relatively  large  M3,  and  an  unreduced  talonid  on  m3.  From  this  primi- 
tive condition  a  morphological  series  develops,  proceeding  from  Nyetophilus 
to  Bauerus  to  Antrozous  to  Anzanycteris.  Since  Anzanycteris  is  known  to  occur 
only  in  the  early  Pleistocene,  it  is  probable  this  diversification  occurred  some- 
time in  the  early  or  middle  Tertiary. 

Iguana,  Hypolagus  regaUs,  Perognathus,  Geomys,  Neototna,  and  a  small 
Sigitiodon  were  found  at  approximately  the  same  le\'el  and  associated  with 
Anzanycteris.  A  tropical  or  subtropical  savannah  is  suggested  by  this  faunal 
assemblage. 


282  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

cf.  Anzanycteris  sp. 

figure  6 

LACM  19302  is  a  right  maxillary  fragment  with  P4-M3,  and  is  not  readily 
referable  to  any  known  genus.  Here  it  is  tentatively  referred  to  Anzanycteris 
for  the  following  reasons:  (1)  anterior  to  P4,  the  posterior  and  medial  portion 
of  the  alveolus  of  the  canine  is  visible  and  there  is  no  exidence  for  the  presence 
of  a  small  premolar;  (2)  in  all  cheekteeth  except  P4,  the  dimensions  are  essen- 
tially as  in  Anzanycteris,  P4  being  markedly  shorter  antero-posteriorly  (Table 
2).  Also  the  cinguliun  in  Ml  and  M2  forms  a  prominence  seen  from  an 
occlusal  \iew,  which  projects  from  the  center  of  each  tooth,  mediad. 

Literature  Cited 

DiBBLEE,  T.  W.,  Jr. 

1954.  Geology  of  the  Imperial  Valley  Region,  California.  Bull.  California 
Div.  Mines,  170  (chap.  II):    21-28,  3  figs.,  1  map. 

HiBBARD,  C.  W. 

1949.  Techniques  of  collecting  micro-vertebrate  fossils.  Contrib.  Mus. 
Paleo.,  Univ.  Michigan,  8(2):  7-19. 

McKexxa,  M.  C. 

1962.  Collecting  small  fossils  by  washing  and  screening.  Curator,  3:221- 
235. 

Miller,  G.  S.,  Jr. 

1907.     The  families  and  genera  of  bats.    Bull.  U.  S.  Nat.  Mus.,  57:xvii  + 
1-282,  14  pis.,  49  figs. 
Orr,  R.  T. 

1954.     Natural  history   of   the   pallid   bat,   Antrozous   pallidiis    (LeConte), 
Proc.  California  Acad.  Sci.,  fourth  series,  28:   165-246,  28  figs. 
Orr,  R.  T.,  and  G.  Silva  Taboada 

1960.  A  new  species  of  bat  of  the  genus  Antrozous  from  Cuba.  Proc.  Biol. 
Soc.  Washington,  73:83-86. 

OSTROM,   J.    H. 

1966.     Fimctional  morphology  and  evolution  of  the  ceratopsian  dinosaurs. 
Evolution:  20:  290-308,  12  figs. 
Slmpson,  G.  G.,  a.  Roe,  and  R.  C.  Lewontix 

1960.     Quantitati\'e  zoology.    Harcout,  Brace  and  Co.,  New  York,  revised 
ed.,  vii  +  440  pp.,  64  figs. 
Van  Gelder,  R.  G. 

1959.     Results  of  the  Puritan-American  Museum  of  Natural  History  Expedi- 
tion to  western  Mexico.    8.    A  new  Antrozous  ( Mammalia,  Vesperti- 
lionidae)    from   the   Tres   Marias   Islands,   Nayarit,    Mexico.     Amer. 
Mus.  Novit.,  1973:1-14,  6  figs. 
Woodard,  G.  D. 

1963.  The  Cenozoic  stratigraphy  of  the  western  Colorado  Desert,  San 
Diego  and  Imperial  counties.  Southern  California.  Ph.  D.  thesis, 
Univ.  California,  Berkeley. 

WOODRIXG,   W.   P. 

1931.  Distribution  and  age  of  the  Tertiarv  deposits  of  the  Colorado  Desert. 
Publ.  Carnegie  Inst.  Washington,  148:1-25. 


A  REVIEW  OF  THE  AFRICAN  MICE  OF  THE 
GENUS  DESMODILLISCUS  WETTSTEIN,  1916 

BY 

Henry  W.  Setzer 

Wettstein  proposed  the  generic  name  Desmodilliscus  in  1916, 
with  the  type  species,  D.  braueri,  based  on  a  specimen  from  the 
road  between  Um  Ramad  and  Nubbaka,  south  of  El  Obeid,  in  the 
Sudan.  In  1920,  Thomas  and  Hinton  described  a  second  species, 
D.  Inichanani,  from  near  Kano,  Nigeria.  Ellerman  (1941)  regarded 
braueri  and  buchanani  as  synonymous  at  the  specific  level,  but 
retained  buchanani  as  a  distinct  subspecies.  Few  specimens  have 
been  recorded  since  the  original  descriptions  of  braueri  and  buchan- 
ani, although  Dckeyser  (1955:220)  reported  material  from  Niger, 
Mali,  and  western  Chad;  these  specimens  have  not  been  seen  by 
me.  Also,  a  recent  paper  by  Heim  de  Balsac  ( 1967 )  recorded  some 
additional  interesting  localities  based  on  specimens  now  housed  at 
the  Museum  National  de  Histoire  Naturelle  in  Paris. 

The  majority  of  the  specimens  mentioned  abo\'e  are  from  owl 
pellets.  It  should  be  noted  further  that  not  a  single  individual  here 
reported  was  taken  in  a  trap;  all  were  obtained  at  night  by  hand  or 
in  an  insect  sweeping  net.  The  fact  that  no  specimens  were  trapped 
is  rather  significant  in  that  animals  of  even  smaller  size  (Mus 
minutoides)  were  taken  in  Museum  Special  traps  in  the  same  areas 
where  Desmodilliscus  was  caught  by  hand. 

Other  than  the  above-mentioned  new  cranial  material,  which  I 
have  not  seen,  the  specimens  of  Desmodilliscus  braueri  available 
for  study  were  pitifully  few  until  1966,  when  field  teams  of  the 
Smithsonian  Institution  African  Mammal  Project  obtained  material 
from  Senegal  and  Nigeria.  Additional  specimens  were  obtained  in 
1967  from  Mauritania  and  Nigeria.  As  a  result  of  the  acquisition  of 
this  new  material,  it  is  felt  that  enough  specimens  now  exist  to 
warrant  an  attempt  to  define  sexual  as  well  as  geographic  variation 
in  this  genus.  All  measurements  are  in  millimeters  and  color  terms 
are  from  Ridgway  ( 1912 ) . 

Tooth  wear  seems  to  be  rather  consistent  in  the  populations 
studied,  and  it  has  been  possible  to  sort  out  five  age  classes  based 
on  this  criterion.  Age  class  I  consists  of  animals  with  unworn  or 
relatively  unworn  teeth,  but  with  M3  in  place;  indixiduals  younger 

(283) 


284 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


than  this  were  not  considered.  The  remaining  four  age  classes  were 
arbitrarily  selected  based  on  the  degree  of  closure  of  various  enamel 
lakes  with  each  other.  Age  classes  II  and  III  are  relatively  close, 
the  last  loph  in  Ml  being  contiguous  in  class  III,  but  separate  in 
class  II.  Class  IV  was  determined  by  the  closure  of  the  enamel 
lakes  in  M2,  whereas  age  class  V  was  characterized  by  the  absence 
of  cusps  but  with  all  enamel  lakes  complete.  No  individuals  showing 
extreme  old  age  ha\'e  been  seen.  The  degree  of  wear  for  the  various 
age  classes  is  illustrated  in  Figure  1. 


Fig.  1.    Crown  patterns  of  upper  and  lower  molar  teeth  of  five  age  classes  of 
Desmodilliscus  braueri.    Top  figures  are  right  upper  molars;  lower  figures  are 

right  lower  molars. 


A  small  sample  of  comparably  aged  males  and  females  from  the 
vicinity  of  Aleg,  Mauritania,  was  tested  for  sexual  variation.  In  only 
two  characters,  greatest  breadth  of  rostrum  and  greatest  breadth 
across  zygomatic  arches,  of  the  16  tested  was  any  significance  noted. 
Thus,  in  future  studies  it  may  be  possible  to  pool  males  and  females 
of  comparable  ages  to  enlarge  samples  for  statistical  purposes. 

Further  statistical  testing  between  populations  was  attempted 
but  sample  sizes  were  so  small  that  realistic  interpretation  of  the 


Setzer — Review  of  Desmodilliscus  285 

results  was  not  reasonable.  However,  certain  morphological  features 
of  the  Senegalese  and  Mauritanian  specimens  indicate  the  presence 
of  an  undescribed  subspecies  of  this  small  gerbil,  which  is  named 
beyond,  followed  by  a  synopsis  of  the  other  two  sulispecies  of 
D.  hraueri. 

Desmodilliscus  braueri  fuscus,  new  subspecies 
HoJotype. — Adult  male,  skin  and  skull,  U.S.  National  Museum  no.  378291, 
from  Richard  Toll,   River  Region,  Senegal;  obtained  on   March   18,   1966,   by 
Richard  M.  Davis,  original  no.  2752. 

Diagnosis. — Upper  parts  Olive-Brown  in  general  appearance  but  most  hairs 
tipped  with  Tawny-Olive;  pure  color  (Tawny-Olive)  edging  dorsal  Olive-Brown 
pattern;  dorsal  pattern  generally  strip-shaped,  but  extending  from  between  the 
eyes  over  head  and  neck,  dropping  o\er  shoulders  then  oxer  back,  tapering  to  a 
point  at  the  base  of  the  tail  (if  the  skin  were  to  be  laid  flat  a  cross-shaped 
pattern  would  be  visible);  underparts,  dorsal  surfaces  of  front  and  hind  feet, 
minute  supraorbital  and  postauricular  spots  pure  white;  tail  sparsely  covered 
with  short,  white  hairs;  skull  small  and  delicate;  auditory  bullae  extremely 
inflated;  braincase  broad  and  somewhat  inflated;  rostrum  relatively  short;  and 
anterior  palatine  foramina  long  and  wide  open. 

Comparisons. — When  compared  with  a  specimen  of  Desmodilliscus  braueri 
])uchanam  of  comparable  age  from  Panisau  (=Farniso),  Northern  Region, 
Nigeria,  individuals  of  D.  h.  fuscus  are  generally  darker  and  slightly  smaller 
in  external  measurements.  The  rostrum  is  generally  broader,  the  breadth  of 
the  braincase  is  less,  the  zygomatic  arches  are  more  nearly  parallel  sided,  the 
anterior  palatine  foramina  are  longer,  and  the  auditory  bullae  are  slightly 
less  inflated. 

Specimens  of  Desmodilliscus  braueri  hraueri  have  been  studied  at  the 
British  Museum  (Natural  History)  but  unfortunately  these  have  not  been  com- 
pared directly  with  specimens  of  D.  b.  fuscus.  However,  because  the  range  of 
D.  b.  buchanani  lies  between  the  ranges  of  D.  b.  braueri  and  D.  b.  fuscus  it  may 
be  assumed  that  these  latter  two  kinds  difter  from  each  other. 

Measurements. — External  and  cranial  measurements  of  the  holotype  are  as 
follows:  total  length,  97;  length  of  tafl,  43;  length  of  hind  foot,  15;  length  of 
ear,  8;  greatest  length  of  skull,  21.2;  least  interorbital  breadth,  3.8;  condylo- 
incisive  length,  18.8;  breadth  across  zygomatic  arches,  12.2;  greatest  breadth  of 
l)raincase,  10.5;  greatest  length  of  nasals,  6.0;  greatest  breadth  of  rostrum,  2.5; 
greatest  length  of  audital  portion  of  auditory  bulla,  9.4;  greatest  breadth  across 
auditory  bullae,  12.6;  length  of  anterior  palatine  foramina,  3.5;  length  of 
posterior  palatine  foramina,  2.5;  crown  length  of  maxillary  toothrow,  2.9. 

Average  and  extreme  measurements  for  six  males  from  Richard  Toll,  River 
Region,  Senegal,  and  six  females  from  Ranerou,  River  Region,  Senegal,  are, 
respectively:  total  length,  99.5  (95-111),  102.8  (96-111);  length  of  tail,  41.3 
(39-4.3),  40.0  (38-45);  length  of  hind  foot,  15.0  (15),  14.6  (14-15);  length  of  ear 
from  notch,  8.7  (8-9),  8.3  (8-9);  greatest  length  of  skull,  21.5  (21.2-21.9),  20.9 
(20.0-22.3);  least  interorbital  breadth,  3.8  (3.5-4.0),  3.7  (3.5-3.9);  condyloincisive 
length,  18.9  (18.5-19.3),  18.4  (17.6-19.1);  greatest  breadth  across  zygomatic 
arches,  12.5  (12.0-13.7),   12.3  (11.6-13.0);   greatest  breadth  of  braincase,   10.5 


286  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

(10.1-10.7),  10.4  (9.9-10.7);  greate.st  length  of  nasals,  6.9  (6.0-7.6),  7.0  (6.5-8.1); 
greatest  breadth  of  rostrum,  2.6  (2.3-3.0),  2.4  (2.2-2.6);  greatest  length  of  audital 
portion  of  auditory  bulla,  9.2  (8.7-9. .5),  8.9  (8.4-9.6);  greatest  breadth  across 
auditory  bullae,  12.5  (12.1-12.8),  12.3  (11.7-12.8);  length  of  anterior  palatine 
foiamina,  3.6  (.3.2-3.8),  3.4  (3.0-3.7);  length  of  posterior  palatine  foramina, 
2.4  (2.3-2.5),  2.4  (2.2-2.6);  crown  length  of  maxillary  toothrow,  3.0  (2.9-3.3), 
2.9  (2.7-3.0). 

Remarks. — The  influence  of  the  Senegal  and  the  Niger  rivers  on  the  dis- 
tribution of  these  small  rodents  is  not  at  all  clear.  It  appears  that  the  Mavui- 
tanian  and  Senegalese  populations  of  Destnodilliscus  have  not  been  long 
separated  by  the  Senegal  River.  Assuming  a  center  of  dispersal  in  northern 
Nigeria,  it  is  possible  that  the  westward  dispersion  was  north  of  the  Niger  and 
then  southwestward  and  westward  around  the  headwaters  of  the  Senegal 
River.  If  this  assumption  is  accepted,  then  we  would  e.xpect  to  find  markedly 
closer  relationship  between  populations  on  either  side  of  the  Senegal  River 
than  would  be  expected  between  these  populations  and  a  population  nearer 
the  center  of  di.spersal.    This  is  actually  what  has  been  observed. 

Heim  de  Balsac  (1967:162)  cited  DesmodiUiscus  from  Dori,  Upper  Volta, 
which  lies  south  of  the  great  bend  of  the  Niger  River.  If  this  is  a  valid  locality 
record,  it  is  of  extreme  interest  inasmuch  as  all  other  distributional  records  for 
this  genus  lie  to  the  east,  north,  and  west  of  the  Niger.  The  zoogeographic 
implications  of  this  distribution  are  difficult  to  explain.  It  can  only  be  hoped 
that  additional  material  can  be  obtained  from  the  region  of  the  great  bend  of 
the  Niger  to  enable  us  to  furnish  some  explanation  of  this  rather  odd  pattern. 

The  habitat  of  Desmodilliscus  appears  to  be  rather  level  hard  clay  with 
small  bushes  and  trees  scattered  throughout  on  small  tussocks.  Burrows,  with- 
out any  appreciable  amount  of  dirt  around  or  in  front  of  them,  were  foimd 
under  the  scattered  bushes  (Fig.  2). 

Other  small  mammals  taken  in  this  same  habitat  were  Jacuhts,  Taterillus, 
and  Geihdhis  (subgenus  Dipoddlus). 

Specimens  examined,  72. — Senegal:  Ranerou,  River  Region,  15°  18'  N, 
13°  58'  W,  13;  5  km.  S  Bakel,  River  Region,  14°  51'  N,  12°  28'  W,  1;  Linguere, 
Diourbel  Region,  15°  24'  N,  15°  07'  W,  5;  Ogo,  13  km.  SW  Matam,  River 
Region,  15°  33'  N,  13°  17'  W,  5;  Podor,  River  Region,  16°  40'  N,  14°  57'  W, 
1;  Richard  Toll,  River  Region,  16°  28'  N,  15°  41'  W,  12.  Mauritania:  3  km.  S 
Aleg,  17°  02'  N,  13°  55'  W,  32;  5  km.  S  Aleg,  17°  02'  N,  13°  55'  W,  1;  6.2  km. 
S  Aleg,  17°  02'  N,  13°  55'  W,  1;  26.7  km.  S  Aleg,  16°  48'  N,  13°  53'  W,  1. 

Desmodilliscus  braueri  braueri  Wettstein,   1916 
Desmodilliscus  lyraueri  Wettstein,  Anz.  k.  Akad.  Wiss.,  Wien,  53   (14):    153, 

1916;  type  locality,  on  the  road  between  Um  Ramad  and  Nubbaka,  S  of 

El  Obeid,  Sudan. 

Measurements. — An  adult  male  from  75  mi.  \V  El  Obeid  measures  as 
follows:  total  length,  102;  length  of  tail,  45;  length  of  hind  foot,  15;  greatest 
length  of  skull,  21.8;  least  interorbital  breadth,  4.0;  condyloincisive  length, 
18.5;  greatest  length  of  nasals,  7.7;  greatest  breadth  of  rostrum,  2.6;  greatest 
length  of  audital  portion  of  auditory  bulla,  9.3;  crown  length  of  maxillary 
toothrow,  2.9. 


Setzer — Review  of  Desmodilliscus  287 


Fk;.    2.    Typical   habitat   of   Desmodilliscus   hraucri   at   a   site   3   km.    S    Aleg, 

Mauritania. 

Remarks. — It  must  be  assumed  that  the  two  specimens  examined  represent 
the  nominate  subspecies  inasmuch  as  they  come  from  reUitively  near  the  t\'pe 
locality.   The  type  specimen  of  D.  hraucri  has  not  been  examined. 

Specimens  examined,  2,  both  in  British  Museum  (Natural  History). — Sudan: 
75  mi.  W  El  Obeid,  1;  140  mi.  E  El  Fasher,  1. 

Desmodilliscus  braueri  buchanani  Thomas  and  Hinton,   1920 
Desmodilliscus  buchanani  Thomas  and  Hinton,  Novitates  Zool.,  27:317,  15  June 

1920;  t>'pe  locality,  Farniso  (:=Panisau),  near  Kano,  Nigeria. 

Measurements. — Averages  and  extremes  for  seven  males  from  Karaduwa, 
Northern  Region,  Nigeria,  and  measurements  for  a  single  female  from  12  mi. 
N  Sokoto,  Northern  Region,  Nigeria,  are,  respecti\ely:  total  length,  97.4  (95- 
101),  100;  length  of  tail,  37.4  (36-39),  40;  length  of  hind  foot,  15.6  (15-16),  15: 
length  of  ear,  9.3  (9-10),  8;  greatest  length  of  skull,  21.1  (20.6-21.9),  21.0; 
least  interorbital  breadth,  3.7  (3.6-3.8),  3.4;  condyloincisive  length,  18.5  (IS.l- 
19.0),  18.4;  greatest  breadth  across  zygomatic  arches,  12.1  (12.0-12.2),  11.8; 
greatest  breadth  of  braincase,  10.3  (10.1-10.5),  10.4;  greatest  length  of  nasals, 
6.4  (6.0-6.9),  7.1;  greatest  breadth  of  rostrum,  2.4  (2.2-2.5),  2.5;  greatest  length 
of  audital  portion  of  auditory  bulla,  9.2  (8.9-9.6),  9.1;  greatest  breadth  across 
auditory  bullae,  12.3  (12.0-12.7),  12.1;  length  of  anterior  palatine  foramina,  3.2 
(3.1-3.5),  3.0;  length  of  posterior  palatine  foramina,  2.5  (2.2-2.7),  2.5;  crown 
length  of  maxillary  toothrow,  3.1  (3.0-3.2),  2.9. 

Remarks.- — Specimens  from  Sokoto  and  Panisau  have  been  compared  with 
the  type  specimen  of  D.  h.  huchanani  in  the  British  Museum  (Natural  History) 
and  found  to  agree  in  detail  both  in  color  and  cranial  features.    It  is  apparent, 


288  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

from  the  few  specimens  available,  that  D.  h.  huchanani  is  darker  in  coloration 
and  somewhat  larger  in  most  cranial  measurements  than  other  subspecies. 

Specimens  examined,  23. — Nigeria:  Panisau,  Northern  Region,  12°  05' 
N,  8°  32'  E,  3  (1  British  Museum);  Karaduwa,  Northern  Region,  12°  19'  N, 
7°  41'  E,  11;  Tangaza,  Northern  Region,  13°  08'  N,  5°  09'  E,  5.  Nicer:  Teg- 
guida,  N'tisem,  W  of  Air,  1. 

Literature  Cited 

Dekeyser,  p.  L. 

1955.     Les    niammiferes    de    I'Afrique    Noire    Francaise.     Inst.    Francaise 
d'Afrique  Noire,  2nd  ed.,  1:1-426,  illustrated 

Heim  de  Balsac,  H. 

1967.     La  distribution  reelle  de  Desmodilliscus  (Gerbillinae).    Mammalia, 
31:160-164. 
RincwAY,  R. 

1912.     Color  standards   and   color  nomenclature.     Washington,    D.C.,   pri- 
vately published,  iii  +  43  pp.,  53  colored  pis. 

Thomas,  O.,  and  M.  A.  C.  Hintox 

1920.     Captain  Angus  Buchanan's  Air  Expedition.    1.  On  a  series  of  small 
mammals  from  Kano.    Novit.  Zool.,  27:315-320. 
Wettsteix,  O.  v. 

1916.     Neue   Gerbillinae   aus   Nordostafrika.     Anz.   k.    Akad.    Wiss.,   Wien, 
53:151-154. 


AN  ANALYSIS  OF  PATTERNS  OF  VARIATION  IN 

SOME  REPRESENTATIVE  MAMMALIA.    PART  II. 

STUDIES  ON  THE  NATURE  AND  CORRELATION 

OF  MEASURES  OF  VARIATION 


BY 


Charles  A.  Long 

The  coefficient  of  variation  (CV)  has  been  calculated  for  many 
mammalian  measurements,  and  is  usually  used  as  a  measure  of  in- 
dixidual  variability.  Understanding  of  variability  is  of  importance 
in  understanding  evolution  and  in  establishing  classifications.  Terms 
such  as  "conservatism,"  "genetic  load,"  "evolutionary  plasticity,"  and 
others  have  limited  \'alue  until  the  complex  and  subtle  contributions 
to  variability  are  better  understood.  Therefore,  a  sample  of  96  fairly 
representative  species  of  mammals  was  obtained  and  analyzed  to 
determine  any  apparent  patterns  of  \ariability.  Coefficients  of  vari- 
ability were  obtained  for  the  available  measurements  of  total  length, 
length  of  cranium,  length  of  maxillary  toothrow,  and  cranial,  zygo- 
matic, and  interorbital  breadths.  The  raw  CV's,  mean  dimensions, 
and  standard  deviations  of  the  mean  were  tabulated  in  a  preliminary 
paper  (Long,  1968).  The  selection  of  data  was  discussed  in  that 
paper,  and  it  is  here  sufficient  to  note  that  CV's  were  not  taken  in 
any  order,  but  generally  include  ^'alues  available  from  my  own 
work,  easily  derived  from  published  raw  data  or  statistics,  and 
known  to  me  in  the  literature.  A  few  values  were  obtained  delib- 
erately to  make  the  sample  as  truly  representative  of  mammalian 
orders  as  possible  (e.g.,  whales  and  duck-billed  platypus).  I  doubt 
that  such  choosing  effected  significant  bias.  The  measurements 
likewise  were  chosen  generally  on  the  basis  of  availability,  and  it  is 
noted  that  important  measurements  on  teeth  and  other  structures 
related  to  fitness  and  niche  exploitation  are  not  so  abundant  as  the 
measurements  used  standardly  by  taxonomists,  and  utilized  herein. 

Methods  and  acknowledgments  are  listed  in  Long  ( 1968).  Some 
computations  were  performed  by  the  University  of  Illinois  SSUPAC 
computer.  Some  of  the  patterns  of  variation  observed  are  statisti- 
cally significant,  in  spite  of  numerous  problems  involved  in  sampling, 
whereas  other  results  are  inconclusive.  The  latter,  however,  have 
value  as  hypotheses  that  warrant  further  testing. 

(289) 


290  Misc.  Publ.  51,  Uxr'.  Kansas  Mus.  Nat.  Hist. 

The  problems  discussed  here  generally  are  correlations  of  taxo- 
nomic  position  and  amount  of  \ariation,  the  effect  of  size  on  varia- 
tion, the  intercorrelation  of  the  variations  of  several  measurements, 
the  apparent  influence  of  ecological  niche,  or  function,  on  some 
variations,  the  degree  of  preponderance  of  high  variation  in  either 
sex  or  in  wild  or  domestic  species,  and  the  nature  of  frequency 
distributions  of  CV  values  for  taxonomic  categories.  It  is  empha- 
sized that  the  data  utilized  are  measurements  of  variation  that 
estimate  amounts  of  variability,  and  are  not  morphometric  values. 

Results  and  Conclusions 
Taxonomy,  Variation,  and  Their  Correlation 

The  class  Mammalia  has  a  good  fossil  record.  Some  orders 
studied,  however,  have  problematical  affinities.  Nevertheless,  ar- 
rangement of  numerous  rodents  and  carnivores  juxtaposed  between 
primitive  monotremes,  marsupials,  insectivores,  and  bats  on  the  one 
hand  and  advanced  ungulates  on  the  other,  provides  a  meaningful 
analysis  of  "phylogenetic"  variation  in  contemporaneous  (Recent) 
taxa.  The  horizontal  arrangement  of  Recent  taxa  (see  Long,  1968) 
permits  general  comparison  of  taxa  of  different  times  of  origin.  For 
example,  the  rodents,  having  more  recently  radiated  into  various 
ecological  niches  than  ha\'e  the  insectivores,  may  show  different 
characters  of  variability  from  those  of  the  insectivores.  Biological 
factors,  such  as  adaptation  to  flight  in  bats,  may  influence  some 
examples  of  variation  reported  herein,  but  these  may  be  indirectly 
related  to  phylogeny.  Other  causes  of  variation  may  cloud  the 
comparisons,  but  significant  correlations  between  taxonomic  posi- 
tion and  amount  of  variation  were  obtained  ( see  below ) .  They  may 
result  mainly  from  size  or  jointly  from  size  and  taxonomic  position. 

In  animals  of  increased  size,  the  CV  measure  of  variability  would 
be  lowered  unless  the  standard  deviation  had  increased  in  step  with 
the  mean  (Long  and  Frank,  1968,  and  others).  A  general  rule  of 
higher  variation  in  larger  mammals,  discussed  beyond,  is  opposite 
that  which  would  be  predicted.  Large  size  tends  to  appear  in  more 
modern  and  more  variable  groups.  It  is  well  known  that  mammals 
generally  have  increased  in  size  throughout  their  history,  although 
small  species  are  numerous. 

Where  several  samples  (e.g.,  sexes,  subspecies)  in  a  wild  (non- 
domestic)  species  were  available,  mean  CV's  were  determined;  the 
available  CV's  of  the  six  measurements  were  then  correlated  with 
taxonomic  position  scaled  from  one  (the  platypus)  to  60  (Rang,ifer), 
and  from  four  to  60  for  wild  placentals  ( see  Long,  1968 ) . 


Long — Patterns  of  Variation  291 

Rather  high  CV's  were  obtained  in  the  primitive  Monotremata 
and  Marsupiaha,  although  more  data  are  necessary  for  firm  conclu- 
sions. Even  so,  in  all  wild  species  some  positive  correlations  be- 
tween taxonomic  position  and  the  amount  of  the  CV  were  obtained, 
represented  with  significance  levels  showing  the  r's  different  from 
zero;  position  and  skull  length,  0.245,  0.1  level;  maxillary  toothrow, 
0.394,  0.01;  total  length,  0.532,  0.01;  cranial  breadth,  0.186;  zygo- 
matic breadth,  0.041;  interorbital  breadth,  0.239. 

When  only  wild  placentals  are  considered,  there  are  stronger 
correlations:  position  and  skull  length,  0.430,  0.001  level;  maxillary 
toothrow,  0.396,  0.02;  total  length,  0.371,  0.05;  cranial  breadth,  0.186; 
zygomatic  breadth,  0.107;  interorbital  breadth,  0.290,  0.01. 

The  significant  correlations  certainly  reveal  that  variation  is 
generally  increased  in  some  measurements  in  advanced  placentals. 
A  similar  phenomenon  was  observed  in  xariation  of  eggs  in  passerine 
and  non-passerine  birds  by  Fisher  (1937).  However,  in  the  few 
primitive  non-placentals  sampled,  \ariation  appears  higher  than 
average  placental  variation.  Also  exceptional  is  the  low  variation 
seen  in  some  of  the  higher  placentals  (e.g.,  Peromyscus  truei,  En- 
hijdra,  Ovibos). 

One  correlation  that  was  surprisingly  low  is  between  taxonomic 
position  and  variation  of  zygomatic  breadth  of  placentals  or  of  all 
wild  species.  In  this  case  phylogenetic  effect  seems  unlikely.  There 
is  increased  complexity  of  the  masticatory  apparatus  in  many  ad- 
vanced mammals  (e.g.,  microtines,  badger,  horse),  and  the  zygo- 
matic arches  are  used  more  or  less  for  muscle  attachment.  It  is 
pertinent  that  in  some  primitive  mammals  the  zygomata  are  incom- 
plete (e.g.,  Sorex),  and  in  fact  appear  in  mammals  as  a  consequence 
of  related  enlargement  of  ancestral  reptilian  fenestrae  of  the  skull 
(see  Romer,  1962). 

In  a  different  analysis  of  the  data  where  Eumetopios  and  the 
wild  horses  were  not  included,  and  the  whales  were  placed  between 
Pinnipedia  and  Proboscidia,  the  r  of  cranial  breadth  CV's  and 
taxonomic  position  was  also  significant,  at  the  0.1  level  only.  In  yet 
another  analysis  where  Eumetopios  and  all  horses  (even  the  domes- 
tic horse  and  ass)  were  included  and  the  whales  were  placed 
between  rodents  and  carnivores,  the  r's  for  taxonomic  position  and 
maxillary  breadth  (0.459)  or  cranial  breadth  (0.369)  were  increased. 

Four  analyses  were  made  to  determine  the  relati\'e  importance 
of  phylogeny  or  size.  In  the  first,  the  sexually  dimorphic  kinds  are 
analyzed.    The  sexes  have  endured  the  same  environments  for  the 


292 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Table  1. — Comparisons  of  significant  sexual  variations  in  relation  to 

MEAN    SIZE    in    SOME    WILD    AND    DOMESTICATED    MAMMALS.      VaLUES    REPRESENT 
THE  FREQUENCIES  OF  MAMMALS   FOR  WHICH  THE   SEXES  DIFFER  SIGNIFICANILY  IN 

CV  VALUE. 


Wild  mammals 

Domestic  : 

mammals 

Measurement 

Males  equal 
or  larger 

Females  larger 

Males  equal 
or  larger 

Females  larger 

SkuU 
length 

Males  more 
variable 
5 

Males  more 
variable 
0 

Males  more 
variable 
4 

Males  more 
variable 
1 

Females  more 
variable 

2 

Females  more 
varialsle 

1 

Females  more 
variable 
0 

Females  more 

\ariable 

0 

Maxillary 
toothrow 

Males  more 
varia]:)le 

2 

Males  more 
variable 
1 

Males  more 
variable 
0 

Males  more 
\ariable 
0 

Females  more 
variable 

1 

Females  more 

variable 

1 

Females  more 
variable 
0 

Females  more 

variable 

0 

Total 
length 

Males  more 
variable 
0 

Males  more 
varialile 
0 

Males  more 
variable 

1 

Males  more 
variable 
1 

Females  more 

variable 

1 

Females  more 
varialile 

1 

Females  more 
variable 
0 

Females  more 
variable 
0 

Cranial 
breadth 

Males  more 
variable 
1 

Males  more 
variable 

1 

Males  more 
variable 
2 

Males  more 
variable 
0 

Females  more 

variable 

0 

Females  more 

variable 

0 

Females  more 
variable 
0 

Females  more 
variable 
0 

Zygomatic 
breadth 

Males  more 
varial^le 
0 

Males  more 
varial^le 
0 

Males  more 
variable 

1 

Males  more 
variable 
0 

Females  more 

variable 

0 

Females  more 
variable 

1 

Females  more 
variable 
0 

Females  more 
varialale 
0 

Interorbital 
breadth 

Males  more 
variable 
1 

Males  more 
varial:)le 
1 

Males  more 
variable 
0 

Males  more 
variable 
0 

Females  more 

varia]:)le 

1 

Females  more 
variable 
0 

Females  more 

varial)le 

0 

Females  more 
variable 
0 

same  time.  Probably  the  genomes  of  the  sexes  do  not  differ  greatly, 
such  as  those,  say,  between  families.  No  evidence  was  found  in  this 
analysis  that  mean  size  is  an  important  factor  in  effecting  the  gen- 
eral magnitude  of  CV  values  (see  Table  1). 

In  the  second  analysis,  comparisons  of  close  relatives  were  made. 
Of  Sorex,  Myotis,  Peromijscus  tniei,  PJujUotis,  Clethriono7nys, 
Martes,  and  Rangifer  (all  of  which  had  comparable  intergeneric 
homogeneous  samples),  only  Rangifer  and  Peromyscus  truei  showed 
any  indication  of  trends  in  size.  Rangifer  tended  to  vary  more  when 


Long — Patterns  of  Variation  293 

larger,  and  P.  truei  tended  to  the  opposite  ( small  samples,  less  than 
16  specmiens,  were  not  used ) . 

The  third  analysis  involved  matching  some  similarly  sized  mam- 
mals of  different  taxonomic  advancement  to  see  if  the  CV's  varied 
independent  of  mean  size.  Sorex  and  Myotis  hardly  differed;  P. 
truei,  Clethrionomys,  and  Blarina  brevicauda  were  about  the  same 
in  CV  amounts;  MepJutis  and  Lynx  were  similar,  but  they  exceeded 
Taxidea.  Both  whales  had  high  variation.  Leptis  flavigularis  showed 
less  variation  than  Ateles,  which  is  probably  more  advanced. 

The  final  analysis  inxolved  the  most  mammals.  By  classifying 
them  as  minute,  small,  medium,  medium-large,  and  large,  it  is 
apparent  that  mean  variabilities  increase  with  size.  In  Table  2,  the 
minute  mammals  all  show  low  variation.  None  of  their  orders,  how- 
ever, is  far  advanced.  In  the  small  category,  some  variations  were 
significantly  higher.  Medium-sized  and  larger  mammals  show  in- 
creasingly higher  mean  variations,  and  reveal  some  extremely  high 
\'alues  in  the  largest  wild  mammals.  Apparently  the  mean  increase 
of  variation  with  increased  size  is  paralleled  by  a  slight  increase  of 
variance  of  the  CV's  in  each  size  group.  Some  of  these  means  dif- 
fered significantly  (0.05  level)  from  one  another,  especially  where 
the  size  discrepancies  were  greatest.  The  variances  were  seldom 
significantly  different  (using  a  one-sided  F-test  to  show  that  the 
variance  was  larger  in  large  mammals ) .  ( The  ratio  of  xariances  of 
medium  to  minute  was  significant;  the  ratio  of  variances  of  small 
to  minute  was  perhaps  significant,  at  0.1  le\'el  only,  as  was  large 
to  small.) 

Scrutiny  of  Table  2  reveals  three  problems  in  determining  the 
effect  of  phylogeny  on  variability.  First,  the  samples  are  small  for 
several  size  classes.  Second,  the  primitive,  small  species  are  of  only 
three  genera,  in  two  closely  related  orders,  and  the  adxanced,  larger 
species  are  likewise  closely  related.  Third,  phylogeny  is  expressed 
at  least  as  well  from  left  to  right  as  vertically.  These  facts  defeat  the 
purpose  of  a  two-factor  analysis  of  variance. 

In  Table  3,  the  species  in  Table  2  are  arbitrarily  lumped  where 
necessary  to  form  a  table  of  five  by  five  values  (mean  CV's)  for  a 
one  factor  analysis  of  variance,  to  determine  any  effect  of  size.  None 
of  the  column  variances  is  significantly  different  from  another  (but, 
of  course,  the  degrees  of  freedom  are  only  four ) .  The  mean  CV  for 
each  column  can  be  seen  to  vary  directly  with  size.  The  ratio  of 
mean  square  among  size-classes  to  residual  mean  square  is  3.53  with 
four  and  20  degrees  of  freedom.   The  ratio  is  significant  at  the  0.05 


294 


Misc.  PuBL.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Table  2. — List  of  cv  values  for  cranial  length  of  adequate  samples  of 

WILD  SPECIES  SELECTED  FROM  LoNg's  (1968)  COMPILATION.  ThESE  ARE  CATE- 
GORIZED ACCORDING  TO  SIZE  AS  FOLLOWS:  MINUTE  (<  20  Mm),  SMALL  (< 
30    mm),    MEDIUM     (<     100    mm),    MEDIUM-LARGE     (<    200    Mm),     LARGE     (> 

200  MM,  EXCEPT  whales),  AND  HUGE  (PJiijseter  AND  Balacnoptera) .    In  the 

VERTICAL  COLUMNS,  PRIMITIVE  TAXA  ARE  ARRANGED  ABOVE  ADVANCED  TAXA. 


Minute 

Small 

Medium 

Medium-large 

Larjre 

Huge 

Sorex 

Sorex 

Tiipaia 

Canis 

Urstis 

Pliyseter 

1.98 

2.48 

2.24 

5.80 

4.39 

18.97 

1.35 

Blarina 

Alouatta 

Taxidea 

Hydrurga 

Balaenoptera 

2.64 

2.30 

2.68 

2.49 

2.88 

7.99 

Tad  (11  id  a 

Macrotm 

Atcdes 

Liitra 

Odoeoileiis 

X=  13.48 

N=  2 

1.70 

1.72 

3.50 

4.83 

4.55 

Myotis 

Dipodomijs 

Ochotona 

Enhydra 

Rangifer 

1.68 

1.59 

4.12 

2.45 

4.38 

2.30 

1.76 

Sylvdagiis 

Lynx 

Ovibos 

X=   1.94 

Pcromt/sctis 

3.24 

3.45 

2.80 

N=  6 

2.05 

Lcpua 

X=   3.80 

X=  3.80 

S-=  0.22 

PJiyUotis 

2.21 

N=  5 

N=  5 

4.05 

Spcnnopliilus 

S^=  2.18 

S^=  0.77 

3.22 

3.01 

4.02 

Thomomys 

3.43 

4.17 

Microtiis 

Ondatra 

4.82 

3.92 

2.58 

Ratius 

Clethrionom 

ys         5.91 

2.50 

Maries 

2.95 

2.76 

Zapus 

Mephitis 

2.66 

3.66 

X=  2.81 

X=   3.45 

N=15 

N=12 

S-^=  0.89 

S'=  1.05 

level,  showing  a  definite  effect  of  size.  But  an  effect  of  phylogeny  is 
not  precluded.  In  Table  3,  as  in  Table  2,  the  rows  express  phylogeny 
at  least  as  well  as  the  columns.  For  example,  in  the  first  column, 
compare  three  species  of  Sorex  and  two  bats  with  a  bat,  rodents, 
carnivores,  and  the  musk  ox  in  the  lowest  row. 


Intercor  relation 

Intercorrelation  of  the  measures  of  \'ariation  of  the  morphometric 
values  yielded  some  positive  correlations.  Including  the  non-placen- 
tals  with  the  placentals,  the  levels  of  significance  for  the  r's  that 
significantly  differ  from  zero  are  as  follows:    skull  length  and  total 


Long — Patterns  of  Variation  295 

length,  0.01  level,  and  interorbital  breadth,  0.01;  maxillary  toothrow 
and  zygomatic  breadth,  0.1;  total  length  and  interorbital  breadth, 
0.01;  cranial  breadth  and  maxillary  toothrow,  0.01,  and  zygomatic 
breadth,  0.02;  zygomatic  breadth  and  interorbital  breadth,  0.05. 

Considering  only  the  placentals,  the  significant  correlations  are 
as  follows:  skull  length  and  total  length,  0.01  le\el,  and  interorbital 
breadth,  0.001;  maxillary  toothrow  and  cranial  breadth,  0.01,  and 
zygomatic  breadth,  0.05;  total  length  and  interorbital  breadth,  0.02; 
cranial  breadth  and  zygomatic  breadth,  0.02;  zygomatic  breadth 
and  interorbital  breadth,  0.05.  Placentals  in  general  show  a  high 
amount  of  correlation  of  measures  of  amounts  of  variation. 

Adaptive  Variability 

Sympatric  differentiation  of  phenotypes  in  a  population  theo- 
retically permits  superior  exploitation  of  the  ecological  niches  and 
is  considered  adaptive  ( Ludwig,  1950;  Dobzhansky  ei  ah,  1950; 
Dobzhansky,  1951;  da  Cunha  and  Dobzhansky,  1954;  Dobzhansky 
and  Pavlosky,  1961;  Van  Valen,  1965;  and  others).  Probably  related 
is  increased  abundance  of  individuals  in  variable  species  (Darwin, 
1859;  Fisher  and  Ford,  1928;  Dobzhansky  et  al,  1950;  and  others). 
Cain  and  Sheppard  ( 1954 )  questioned  whether  adaptiveness  re- 
sulted from  polymorphism  and  several  authors  ( Simpson,  1953; 
Bader  and  Hall,  1960;  and  others)  have  pointed  out  that  vestigial 
( functionless )  structures  are  highly  variable.  Considering  samples 
from  nature,  not  only  are  the  causes  of  high  and  low  variation 
numerous  and  complex,  but  even  the  contributions  of  adaptive  and 
inadaptive  variability  are  greatly  complicated  (see  Mayr,  1963). 
Van  Valen  ( 1965 )  used  bill  width  ( often  considered  closely  asso- 
ciated with  the  mechanics  of  food  utilization  and  "obserx'ational" 
niche  size)  to  compare  variations  of  several  birds  having  wide 
niches  with  those  of  several  others  having  narrow  niches.  His 
results  were  consistent  with  the  hypothesis  that  high  variation  is 
adaptive. 

Efficient  support  of  weight  requires  some  expenditure  of  energy 
and  is,  therefore,  more  or  less  adaptive.  Coefficients  of  variation  in 
three  groups  of  taxa  with  different  problems  of  weight  are  com- 
pared: bats,  which  fly,  have  a  tangible  weight  problem;  rabbits, 
rodents,  and  carnivores  apparently  have  a  lesser  problem;  some 
aquatic  mammals  probably  have  the  least  weight  problem,  owing 
perhaps  to  buoyancy  (data  in  Long,  196S).  Bats  show  the  lowest 
overall  variation  of  the  mammals  studied   (Long,  1968).    The  In- 


296  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

sectivora,  including  or  not  including  the  subprimate  Tupaia,  are 
also  low  in  variation  but  usually  exceed  the  variation  found  in  bats. 
Rodents,  carnivores,  and  lagomorphs  show  a  wide  range  in  varia- 
tion, low  to  moderately  high. 

The  aquatic  mammals  sampled  are  diverse  in  phylogeny,  mor- 
phology, size,  diet,  and  even  in  their  aquatic  adaptations.  Hence, 
buoyancy  is  hardly  a  sole  factor  involved  with  variation.  The  ani- 
mals included  are  the  platypus  (Ornithorhijnchus),  water  shrew 
(Sorex  pahistris),  muskrat  (Ondatra),  river  otter  (Lutra),  sea  otter 
(Enliydra),  leopard  seal  (Hydrurga),  sea  lion  (Einnetopias),  walrus 
(Odohenus),  and  two  cetaceans  (Physeter,  Balaenoptera).  Vari- 
ability in  some  of  these  mammals  is  high — Ornithorhynchus  (unfor- 
tunately a  small  sample),  Lutra  (as  compared  to  other  mustelids), 
and  Odohenus,  PJnjseter,  and  Balaenoptera  (which  by  virtue  of 
huge  body  size  rely  primarily  on  buoyancy).  These  sketchy  obser- 
vations are  consistent  with  the  hypothesis  that  xariability  generally 
varies  inversely  with  these  weight  limitations,  perhaps  directly  with 
niche  size. 

Table  3. — A  five-by-five  matrix  of  exact  or  mean  cv  values  that  are 

DERIVED  FROM  TaBLE  2.  ThE  COLUMNS  ARE  SIZE  CATEGORIES.  ThE  MEANS  IN 
THE  TABLE  ARE  ACCOMPANIED  BY  LETTERS,  WHICH  DENOTE  THE  GENERA  THAT 
THE  MEANS  REPRESENT  ( SEE  TaBLE  2).  ThE  ANALYSIS  IS  EXPLAINED  IN  THE 
TEXT.    The  GRAND  MEAN  AND  TOTAL  VARIANCE  ARE  IN  THE  LOWER  RIGHT  BLOCKS. 


Minute 

Small 

Medium 

Medium-large 

Large 

X 

s- 

1.98 

2.17 
SBM 

2.81 
TAA 

5.80 

4.39 

3.43 

2.66 

1.35 

1.80 
DP 

3.19 
OSL 

2.49 

2.88 

2.34 

0.58 

2.64 

3.68 
P 

3.59 
ST 

4.83 

4.55 

3.86 

0.75 

1.70 

3.70 
M 

4.91 
OR 

2.45 

4.38 

3.43 

1.78 

1.99 

2.70 

3.21 

3.45 

2.80 

2.83 

0.32 

M 

cz 

MM 

X 

1.93 

2.81 

3.,54 

3.80 

3.80 

3.18 

s^ 

0.22 

0.75 

0.66 

2.18 

0.77 

1.31 

Considering  the  problems  of  diet  broadly,  insectivorous  feeding 
is  probably  related  to  narrow  niche  and  omni\'orous  habits  to  a 
niche  that  is  relatively  broad.  Herbivorous  and  carnivorous  modes 
of  feeding  are  difficult  to  classify,  involving  diverse  and  often  con- 
flicting physiological  functions.  To  illustrate,  herbivorous  rodents 
usually  have  at  least  four  pairs  of  highly  functional  chewing  molars 
and  two  pairs  of  highly  functional  gnawing  incisors,  but  the  exten- 


Long — Patterns  of  Variation  297 

sive  loss  of  teeth  from  the  primitive  dental  formula  emphasizes  the 
shifting  in  niches  and  adaptive  zones  that  has  occurred.  \^ariations 
in  teeth  are  excluded  from  consideration  here,  although  they  are 
intimately  and  dii-ectly  related  to  mode  of  feeding  (the  actual 
mechanics  of  feeding  are  largely  unknown  in  mammals).  Most 
Recent  animals  remain  essentially  unstudied  with  respect  to  tooth 
\'ariation. 

Insectivorous  mammals,  by  and  large,  have  low  overall  variation 
(e.g.,  shrews,  bats,  Tupaia).  Definitely  omnivorous  are  the  opossum 
(Didelphis),  monkeys  and  man,  the  rat  (Rattus),  the  skunk  (Mephi- 
tis), and  the  bear  (Ursus).  In  their  groups  these  mammals  are  mod- 
erately to  highly  variable,  if  Didelphis  is  considered  variable  (in 
some  places  it  appears  to  be  highly  ^'ariable — Bader,  1955).  Gen- 
erally herbivorous  mammals  ( such  as  rodents,  rabbits,  artiodactyls ) 
vary  from  low  to  high  in  CV  \'alue  but  are  frequently  highly  vari- 
able in  cranial  and  body  dimensions.  Seemingly,  the  Carnivora  are 
lower  on  the  average  in  cranial  \ariation  than  the  herbivores.  The 
data  appear  generally  consistent  with  the  hypothesis  that  width  of 
niche  varies  with  CV  value;  several  related  problems  warrant  addi- 
tional study. 

Domestication,  Inbreeding,  and  Sexual  Preponderance 

of  High  Variation 

Generally,  domestic  mammals  sampled  are  more  variable  than 
wild  mammals,  a  phenomenon  sexeral  authors  have  previously  ob- 
served (Lee  and  Pearson,  1(S97;  Alpatov  and  Boschko-Stepanenko, 
1928;  Latimer,  1936;  Stockhaus,  1965;  and  others).  In  mammals 
considered  domestic,  or  somewhat  independent  of  natural  ecologi- 
cal factors,  it  is  interesting  that  in  Homo,  Cavia,  Fells,  and  Eqinis 
cranial  breadth  is  especially  high  in  variations.  As  might  be  ex- 
pected, variability  in  uncategorized  dogs  is  extremely  high  ( Stock- 
haus, 1965),  but  markedly  lower  in  inbred  strains  (boxer,  for 
example).  Bader  (1956)  found  that  inbreeding  of  Mus  did  not 
markedly  lower  the  variability  of  his  samples  in  comparison  to  wild 
animals.  Probably  neither  of  these  examples  adequately  reveals 
effects  of  natural  mammalian  inbreeding.  For  one  thing,  the  effects 
of  domestication,  although  apparently  accelerant  to  increased  vari- 
ability, are  poorly  understood  in  relation  to  the  theoretically  inhibi- 
tory effect  on  xariation  of  inbreeding.  Also,  Mus  is  apparently  not 
so  variable  as  Canis  (wild  or  domestic)  and  may  have  been  long 
stabilized  in  numerous  characters  prior  to  its  use  in  experimentation. 


298  Misc.  Publ.  51,  Uxiv.  Kansas  Mus.  Nat.  Hist. 

Generally,  those  who  have  studied  variability  in  mammalian 
sexes  state  that  a  preponderance  in  variability  in  one  sex  does  not 
occur  consistently  (Pearson  and  Davin,  1924;  Schultz,  1926;  and 
numerous  others).  Table  1  categorizes  mammals,  listed  in  Long 
(1968),  as  wild  or  domestic,  and  compares  variation  between  the 
sexes  of  mammals  in  which  the  male  averages  larger,  or  smaller. 
This  is  done  for  all  available  samples  including  two  of  Canis  lupus 
and  several  subspecies  of  Rangifer  tarandus  and  Maries  americana. 
In  Rangifer,  data  involving  only  two  measurements  are  available 
and  in  Maries  only  one.  The  standard  errors  of  the  CV's  of  males 
and  females  for  six  characters  were  used  to  determine  significant 
differences. 

In  the  wild  mammals,  whether  males  or  females  exceed  the 
opposite  sex  in  size,  there  is  as  yet  no  significant  basis  for  attributing 
greater  variability  to  one  sex  in  mammals  as  a  group.  In  domestic 
mammals  males  definitely  tend  to  vary  more  than  females,  irrespec- 
tive of  whether  males  are  larger  or  smaller.  The  samples  in  Table  1 
are  somewhat  biased  in  including  among  the  wild  mammals  two 
samples  of  Canis  lupus  and  several  subspecies  of  Maries  americana 
(only  skull  lengths),  samples  of  Homo,  and  two  of  Oryciolagus 
(not  "dwarfs")  (see  Long,  1968).  That  the  variabilities  of  the  six 
characters  are  more  or  less  correlated  is  another  source  of  bias.  Even 
so,  that  males  differing  significantly  in  CV  values  between  sexes  are 
more  variable  in  10  tests,  whereas  females  are  more  variable  in  none, 
points  to  the  conclusion  that  domestic  mammals  generally  vary 
more  in  the  male  sex,  and  that  those  characters  of  the  female  sex 
are  less  mutable  (x"  table  yields  P  <  one  per  cent  that  the  ratio 
10:0  is  attributable  to  chance). 

Frequency  Distributions  for  Coefficients  of  Variation 

For  a  given  measurement,  frequency  distributions  of  mammalian 
CV's  generally  show  skewed  cur\'es,  which  rise  abruptly  in  the 
range  of  low  xariation  ( Fig.  1 ) .  Lack  of  correlation  of  subsidiary 
lengths  and  mathematical  definitions  of  CV  (see  Long  and  Frank, 
1968)  might  effect  lower  mean  variation,  or  especially  modal  varia- 
tion, in  measurements  of  complex  structures  such  as  length  of  skull. 
Where  correlation  of  component  bones  is  positive  and  significant 
( such  as  interorbital  breadth ) ,  the  variation  of  the  sum  is  somewhat 
increased.  Curves  of  most  of  the  six  measurements  show  sharp,  high 
peaks.  Maxillar\'  toothrow,  with  a  high  modal  frequency,  is  least 
skewed  of  the  six  measurements.   Zygomatic  breadth  is  also  high  in 


Long — Patterns  of  Variation 


299 


Inltrorbital 
Breadth 


8     9    10     II 


Fig.  1.    Frequency  distributions  for  coefficients  of  \ariation  of  wild  nianimals. 
Abcissas   are  CV   values,   ordinates   are   frequencies.     See   text   for   fiuther  ex- 
planation. 


modal  frequency,  and  also  does  not  show  a  conspicuously  skewed 
curve.  The  curxe  for  total  length  is  not  conspicuously  peaked,  but 
is  skewed  strongly  nexertheless. 

It  should  be  noted  that  the  curves  are  not  drawn  from  the  same 
nor  entirely  different  animals,  availability  of  the  data  of  wild  ani- 
mals being  the  only  criterion  of  usage.  The  uppermost  curve  of  each 
graph  in  Figure  1  represents  all  possible  samples,  within  kinds  and 
of  each  sex  or  of  both  sexes.  The  next  lower  curve  represents  taxa  or 
kinds,  the  next  species,  and  then  genera.  The  lo\\'ermost  curve 
represents  families.  The  lower  curves  are  made  up  from  the  data 
yielding  the  upper  curves,  and  where  possible,  of  means  of  the  more 
subdi\ided  data.  Thus,  the  CV  for  total  length  in  Ormthorhijnchus 
is  utilized  in  all  curxes  for  total  length,  but  the  curve  for  cranial 
breadth  based  on  data  of  families  incorporates  a  mean  of  the  mean 
CV  for  '^orex  (derived  from  the  mean  of  the  specific  values)  and 
the  CV  for  Blarina  brevicauda  (mean  of  both  sexes).  The  use  of 
means  permits  a  more  reliable  interpretation  of  the  lower  curves, 
whereas  the  biasing  effected  by  "stacking"  a  higher  taxonomic 
category  with  a  poor  sample  of  its  true  variability  may  lower  the 


300  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

reliability  of  that  interpretation.  The  upper  and  lower  curves  agree 
essentially  in  showing  skewedness  and  low  modal  values.  The 
curves  reveal  that,  in  relation  to  the  mode,  high  values  for  CV  are 
more  often  found  than  low  values. 

Concerning  total  length  of  skull,  few  CV's  were  less  than  1.5,  and 
the  mode  is  low,  not  exceeding  2.5.  Few  values  are  greater  than  6.5 
(the  abundance  compares  with  those  considered  less  than  1.5). 
Values  exceeding  5.5  are  uncommon  ( see  Fig.  1 ) . 

Concerning  maxillary  toothrow,  few  \'alues  are  less  than  2.5, 
and  the  mode  is  high,  about  4.0.  Few  values  are  greater  than  6  5 
(Fig.  1). 

Concerning  total  length  of  head,  bod>',  and  tail  in  representative 
mammals,  one  value  was  obtained  that  was  less  than  2.5,  and  values 
less  than  3.5  were  not  uncommon.  The  mode  is  high,  about  4.0  or 
5.0,  but  few  values  exceed  8.5  (Fig.  1). 

Concerning  cranial  breadth,  the  curve  falls  abruptly,  tailing  off 
to  high  values.  Values  below  1.5  are  infrequent,  but  those  approxi- 
mately 2.0  are  common.  The  mode  is  about  3.0.  Few  values  exceed 
4.5  ( Fig.  1). 

Concerning  zygomatic  breadth,  few  values  are  less  than  2.5. 
The  mode  is  high,  about  4.0,  but  variation  exceeding  6.5  is  uncom- 
mon (Fig.  1). 

Concerning  interorbital  breadth,  few  values  are  less  than  2.5  and 
few  exceed  7.5,  but  some  were  surprisingly  high.  Most  values  are 
about  3.0  or  4.0.  Interorbital  breadth  approaches  length  of  toothrow 
and  zygomatic  breadth  in  high  variability. 

Considering  wild  placental  species  (not  utilizing  all  the  horses 
and  Eiimetopias) ,  the  mean  CV's  with  rather  high  standard  devia- 
tions for  all  axailable  species  for  each  measurement  are  as  follows: 
cranial  length,  3.21  ±  1.40  (N,  51);  maxillary  toothrow,  3.99  it  1.09 
(N,  37);  total  length,  5.31  ±  1.96  (N,  31);  cranial  breadth,  3.05  ± 
1.24  (N,  27);  zygomatic  breadth,  3.95  ±  1.34  (N,  30);  interorbital 
breadth,  4.36  ±  1.51  (N,  32). 

The  means  slightly  exceed  the  modes,  as  would  be  expected  of 
the  skewed  distributions. 

Summary 

Patterns  of  variation  observed  in  96  mammals  indicate  that 
larger  mammals  are  generally  more  variable  among  wild  placentals. 
There  are  significant  correlations  of  taxonomic  position  and  amount 
of  variation  for  several  linear  measurements  of  mammals,  especially 


Long — Patterns  of  Variation  301 

of  placentals.  A  high  amount  of  intercorrelation  of  variations  of  the 
six  measurements  is  noted.  Concerning  \ariation  and  problems  of 
weight,  bats  are  low  and  some  marine  mammals  are  high  in  varia- 
tion. Insectivorous  mammals  are  generally  low  in  variation,  and 
omnivores  are  often  high.  Herbivores  vary  greatly  in  CV  value,  and 
tend  to  exceed  carnivores.  In  wild  species,  neither  sex  shows  a 
significant  preponderance  of  high  CV's,  but  in  domestic  mammals 
the  males  were  more  variable  in  10  tests.  Domestic  mammals  show 
moderate  to  high  xariations.  Coefficient  of  variability  values  of 
wild  taxa  are  skewed  right,  with  mean  CV's  and  modes  varying 
from  approximately  2.5  to  5.3. 

Literature  Cited 

Alpatov,  W.  W.,  and  a.  M.  Boschko-Stepanenko 

1928.  Variation  and  correlation  in  serially  situated  organs  in  insects,  fishes, 
and  birds.    Amer.  Nat.,  62:409-424. 

Bader,  R.  S. 

1955.  Variability  and  evolutionary  rate  in  the  Oreodonts.  Evolution,  9: 
119-140. 

1956.  Variability  in  wild  and  inbred  mammalian  populations.  Jour. 
Florida  Acad.  Sci.,  19:14-34. 

Bader,  R.  S.,  and  J.  S.  Hall 

1960.  Osteometric  variation  and  function  in  bats.  Evolution,  14:8-17, 
3  figs. 

Calv,  a.  J.,  and  p.  M.  Sheppard 

1954.     The  theory  of  adaptive  polymorphism.    Amer.  Nat.,  88:321-326. 
da  Cunha,  a.  B.,  and  T.  Dobzhansky 

1954.  A  further  study  on  chromosomal  polymorphism  in  Drosophila 
willistotii  in  its  relation  to  the  environment.    Evolution,  8:119-134. 

Darwin,  C. 

1859.  On  the  origin  of  species  by  means  of  natural  selection.  J.  Murray, 
London. 

Dobzhansky,  T. 

1951.     Genetics  and  the  origin  of  species.    Columbia  Press,  New  York,  3rd 
ed.,  xi  +  364  pp.,  illustrated. 
Dobzhansky,  T.,  H.  Burla,  and  A.  B.  da  Cunha 

1950.     A    comparative    study    of    chromosomal    polymorphism    in    sibling 
species  of  the  ivillistojii  group  of  Drosophila.    Amer.  Nat.,  84:229- 
246. 
Dobhansky,  T.,  and  O.  Pavlosky 

1961.  A  further  study  of  fitness  of  chromosomally  polymorphic  and  mono- 
moiphic  populations  of  Drosophila  pseudoobscura.  Heredity,  16: 
169-179. 

Fisher,  R.  A. 

1937.     The    relation    between    variability    and    abundance    shown    by    the 
measurements  of  the  eggs  of  British  nesting  birds.    Proc.  Royal  Soc. 
London,  ser.  B,  122:1-26. 
Fisher,  R.  A.,  and  E.  B.  Ford 

1928.     The   xariability   of   species   in   the   Lepidoptera,    with    reference   to 
abundance  and  sex.    Trans.  Ent.  Soc.  London,  76:367-384. 
Latimer,  H.  B. 

1936.  Weights  and  linear  measurements  of  the  adult  cat.  Amer.  Jour. 
Anat.,  58:329-347. 


302  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Lee,  a.,  and  K.  Pearson 

1897.  On  the  relative  variation  and  correlation  in  civilized  and  uncivilized 
races.   Science,  n.s.,  6:49-50. 

LoxG,  C.  A. 

1968.  An  analysis  of  patterns  of  variation  in  some  representative  Mam- 
malia. Part  I.  A  review  of  estimates  of  variability  in  selected 
measurements.    Trans.  Kansas  Acad.  Sci.,  71:201-227. 

Long,  C.  A.,  and  T.  Frank 

1968.  Morphometric  variation  and  function  in  the  baculum,  with  com- 
ments on  correlation  of  parts  and  the  condition  of  vestigiality.  Jour. 
Mamm.,  49:32-43,  2  figs. 

LUDWIG,   W. 

1950.  Zur  Theorie  der  Konkurrenz.  Neue  Ergeb.  Prob.  Zool.,  Klatt-Fest., 
pp.  516-537. 

Mayr,  E. 

1963.  Animal  species  and  evolution.  Harvard  Press,  Cambridge,  Mass., 
xvi  +  797  pp.,  illustrated. 

Pearson,  K.,  and  A.  G.  Davin 

1924.  On  the  biometric  constants  in  the  human  skull.  Biometrika,  16: 
328-363. 

Romer,  a.  S. 

1962.  The  vertebrate  body.  W.  B.  Saunders  Co.,  Philadelphia,  3rd  ed., 
viii  +  627  pp.,  illustrated. 

SCHULTZ,  A.  H. 

1926.  Studies  on  the  variability  of  platyrrhine  monkeys.  Jour.  Mamm., 
7:286-305,  9  figs. 

Simpson,  G.  G. 

1953.  The  major  features  of  evolution.  Columbia  Univ.  Press,  New  York, 
XX  +  434  pp.,  illustrated. 

Stockhaus,  K. 

1965.  Metrische  Untersuchungen  an  Schadeln  von  Wolfen  und  Hunden. 
Zeit.  Zool.  Syst.  Evolutionsforschung,  3:157-258. 

Van  Valen,  L. 

1965.  Morphological  variation  and  width  of  ecological  niche.  Amer.  Nat., 
99:377-390. 


MIGRATION  IN  THE  GUANO  BAT, 
TADARIDA  BRASILIENSIS 

BY 

E.  Lexdell  Cockrum 

Between  September,  1952,  and  September,  1967,  a  total  of 
162,892  guano  bats,  Tadarida  hrasiliensis  mexicana  (Saussure),  were 
banded  in  Arizona  and  adjacent  regions  by  persons  associated  with 
the  Department  of  Biological  Sciences  at  the  University  of  Arizona. 
Most  of  these  were  taken  in  a  few  caves,  but  smaller  numbers  were 
captured  in  a  variety  of  situations  including  roosts  in  buildings, 
under  bridges,  and  in  mine  tunnels,  as  well  as  in  mist  nets  set  over 
water.  Appendices  1  and  2  and  Table  1  summarize  the  majority  of 
these  banding  activities  and  Figure  1  shows  the  geographic  distribu- 
tion of  the  banding  sites. 

Materials  and  Methods 

Al]  of  the  bands  used  in  this  study  were  furnished  by  the  Bat  Banding 
Office,  Fish  and  Wildlife  Service,  U.S.  National  Museum,  Washington,  D.C. 
Various  sizes  and  styles  of  bands  were  used;  mostly  we  employed  the  size  0 
standard  bird  band,  but  others  used  included  sizes  1,  IB,  and  2  of  the  standard 
bird  band,  a  few  of  the  rounded-end  style,  and  a  few  lipped  bands.  See 
Herreid  et  al  (1960)  for  a  discus.sion  of  these  band  styles  and  their  relative 
merits.  All  were  applied  to  the  forearm.  Some  bands  were  color-coded  for 
use  at  specific  localities.  Anodized  bands  provided  \arious  colors  that  would 
remain  evident  for  several  weeks  to  years.  Just  how  long  the  color  remained 
e\ident  depended  upon  how  much  the  bat  chewed  on  the  band.  This  in  turn 
seems  to  inxohe  two  factors,  band  application  technique  and  variations  in  the 
reaction  of  individual  bats  to  a  band.  Poorly  applied  bands  (too  tight,  too 
loose,  crooked)  caused  irritation,  infections,  and  swellings  and  growths  that 
sometimes  covered  the  whole  band. 

Even  when  bands  were  apparently  well  applied,  some  individuals  reacted 
l)y  chewing  at  the  band.  On  many  occasions,  after  a  large  banding  operation, 
bats  would  be  observed  landing  in  the  cave,  making  vocal  noises  and  chewing 
on  the  newly  applied  band.  Especially  with  the  size  0  bands,  such  chewing 
soon  obliterated  one  or  more  numbers  on  the  band.  For  example,  a  female 
banded  at  Silverbell  (locality  28)  on  May  10,  1963,  and  recovered  at  Eagle 
Creek  (locality  24)  on  June  5,  1963,  had  already  so  chewed  the  band  that 
one  number  was  not  legible.  Bats  recovered  one  or  more  years  after  banding 
often  had  so  chewed  the  band  that  three  or  more  numbers  were  illegible,  and, 
in  some  cases,  holes  were  worn  completely  through  the  band. 

This  condition  plus  the  appearance  of  the  forearm  of  certain  unhanded  bats 
leads  to  the  conclusion  that  many  bands  are  actually  removed  by  the  bats.  In 
other  cases,  .swelling,  irritation,  and  infection  residting  from  such  chewing  may 

(303) 


304 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


no 


0-7   .4-6 
'{^■8      -9-11 

<    .12 
J 


J ^2. 


13 


.61 


14-21-  .23-25 

26-27.  22- 


62-64 


Fig.  1.    Map  showing  geographic  distribution  of  the  l^anding  sites  of  Tadarida 
hrasiliensis.    Numbers  correspond  to  those  listed  in  Appendices  1   and  2,  and 

used  in  text. 


have  caused  the  death  of  certain  individuals.  If  these  conjectures  are  true,  then 
recovery  data  in  this  study  cannot  be  used  to  construct  meaningful  mortality 
tables.  Guano  bats  appear  to  be  much  more  aggressi\-e  and  persistent  in  their 
attempts  to  remove  bands  than  are  other  species  banded  in  southern  Arizona, 
although  some  Eptesicus  fuscus  and  Mijotis  velifcr  and  a  few  Macwtus  watcr- 
housii  also  chewed  bands  extensively. 

Many  different  techniques  of  capture  were  used  in  this  study.  Most  bats 
were  captmed  with  \arious  modifications  of  the  Constantine  bat  trap  (Con- 
stantine,  1958).  Light-weight  versions,  using  vertical  monofilament  nylon 
strands  held  in  place  by  brads  spaced  at  intervals  of  three-quarters  of  an  inch 


CocKRUM — Migration  of  Tad  arid  a  305 

Table  1. — Summary  of  numbers  of  Tadarida  hrasiliensis  banded,  arranged 

BY   the   major  types   OF    HABITATS    FROM    WHICH    TAKEN. 

Water  Mine  or  Cave  Building  or  bridge 

Locality      Males      Females  Locality      Males      Females  Locality      Males      Females 

10  1  2  0             8  13  3  680 

4  8  2  3  11  14  1  4 

5  4  0  7  4             7  15  11  1 

6  0  2  8  820      1,502  16  56  30 

11  6         169  9         662         347  17  5  4 

12  10  0  10  1  0  18  10  15 
23             2             2               24    10,902    77,324                19         465         350 

29  3  0  25  12         305  20  0  41 

30  20  1  26  98         292  21  10  12 
32           19             8                27           29           67                22              1  0 

35  52  51  28         533      4,003  31  7  36 

36  2  3  37         222         554  33  13  12 

43  23  2  39  39  8  34  16  60 

44  1  6  40  21  92  38  0  1 

45  1  2  41  0  1  42  221  1,041 

46  3  8  61  1,714  1,137  49  27  22 

47  0  2  65  868  1,305  50  17  29 

48  1  1  66  19,155  21,639  52  10  100 

49  0  2  70  7,700  3,900  53  269  551 

55  569         317  71  7  0  54  2  0 

56  3  0  -..         58         215      1,101 

57  9  3  -..-         . . 59         131  3 

60  1  0  ....  .-...         ..._  62  59  91 

67  45  25  -.         - -—  63  1  0 

68  0         205  .-.         --         ..--  64         147         102 

69        1        _i  --         --  --       _^^         -.-.. 

783         816  42,788  112,492  1,697      4,316 

on  wooden  frames  made  of  pine  boards  ( two  by  two  inches )  of  appropriate 
lengths,  were  assembled  in  the  field  on  many  occasions.  Plastic  collecting 
baskets  were  positioned  at  the  bottom  of  the  trap  or,  when  the  trap  had  to  be 
snspended  above  the  gronnd,  long  funnels  were  made  of  four  mil  sheet  plastic. 
Experience  showed  that  funnels  with  a  small  throat  size  became  clogged  when 
the  trap  was  placed  in  a  dense  flight.  No  such  troubles  were  encountered  when 
utilizing  steep-sided  funnels  having  throat  diameters  in  e.xcess  of  24  inches. 
Retaining  cages  of  sufficient  number  and  size  to  avoid  suffocation  of  the  bats 
are  necessary.  Various  kinds  were  used,  all  with  open  tops.  Some  were  made 
at  the  trapping  site,  utilizing  lumber  and  hardware  cloth.  Typically  these 
included  a  plastic  sheet  along  the  upper  edge  of  the  inside  of  the  cage  to 
prevent  bats  from  crawling  out  or,  in  large  cages,  baffles  of  plastic  hanging 
down  into  the  cage  to  prevent  bats  from  flying  out;  others  were  long  narrow 
tubs  of  sheet  plastic  supported  by  side  poles  and  end  frames.  For  smaller 
operations,  20  gallon  plastic  garbage  cans,  with  a  number  of  large  holes  cut 
in  the  sides,  lined  with  a  hardware  cloth  insert,  and  with  a  hole  cut  in  the  lid 
leaving  about  four  inches  of  rim,  were  found  to  be  useful. 

Description  of  Roosts  and  Populations 
Based  upon  such  factors  as  season  of  the  year  and  composition 
of  the  population  present,  various  types  of  roosts  are  recognized: 


306 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


(1)  Maternity  Colonies,  containing  primarily  adult  females  and 
young  of  the  year  that  appeared  to  have  been  born  there;  (2)  Sum- 
mer Male  Roosts;  (3)  Transient  Roosts;  and  (4)  Winter  Roosts 
(Villa-R.  and  Cockrum,  1962:58).  A  given  locality  may  house  suc- 
cessively in  the  same  year  different  types  of  roosts  that  are  made  up 
of  different  individuals.  Specific  examples  of  this  were  seen  at 
Carbo,  Sonora  ( locality  66 ) ,  and  at  the  railroad  bridge  near  Conti- 
nental, Arizona  (locality  42). 


Fig.  2.    A  modified  Constantine  bat  trap  at  tlie  mouth  of  Eagle  Creek  Cave, 
Greenlee  Co.,  Arizona  (locality  24).    Photograph  by  Bruce  Hayward. 


CocKRUM — Migration  of  Tadarida  307 

Maternity  Colonies 

Maternity  colonies  exist  in  a  number  of  different  situations. 
Small  colonies  are  known  from  crevices  in  bridges,  in  mine  tunnels, 
in  attics  and  crevices  in  buildings,  and  in  natural  caves.  Large 
maternity  colonies  exist  in  Eagle  Creek  Cave  (locality  24),  Silver- 
bell  Cave  ( locality  23 ) ,  and  Cueva  del  Tigre  ( locality  66 ) . 

Eagle  Creek  Cave  (Figs.  2  and  3)  is  in  the  eastern  canyon  wall 
of  Eagle  Creek.  The  entrance  is  approximately  100  feet  above  the 
present  stream  level  and  is  a  vertical  slit  roughly  24  feet  wide  and 
65  feet  high.  Inside,  the  cave  opens  into  a  single,  large  football- 
shaped  chamber  that  is  approximately  288  feet  long,  65  feet  wide, 
and  100  feet  tall.  Some  wide,  deep,  chimney-like  crevices  in  the 
ceiling  provide  additional  roosting  space  for  the  bats.  This  cave  has 
been  known  as  a  "guano  cave"  for  many  years  and  formerly  biennial 
accumulations  of  guano  were  routinely  removed  and  sold. 

The  cave  is  not  readily  accessible,  even  by  vehicles  with  four- 
wheel  drive — especially  during  the  seasons  of  high  water.  It  is  4.2 
miles  down  the  narrow  valley  of  Eagle  Creek.  The  walls  of  the 
canyon  are  400  to  500  feet  high  and  are  sheer  or  nearly  so.  The 
stream  meanders  from  one  canyon  wall  to  the  other  and  thus  must 
be  forded  many  times.  The  course  of  the  stream  varies  from  year  to 
year.  In  the  dry  season  (usually  March  to  June)  it  has  little  running 
water,  but  in  the  rainy  season  (July  to  September)  is  a  raging  tor- 
rent. During  the  times  of  run-off  of  rain  and  melting  snow  in  the 
headwaters  region  (winter  and  early  spring),  flooding  conditions 
exist,  washing  out  most  of  the  "road."  The  relative  inaccessibility  of 
this  cave  results  in  a  minimal  disturbance  to  the  population  by 
vandals  but  does  impose  difficulties  in  studying  the  bats. 

This  cave  is  occupied  by  bats  only  during  the  warmer  parts  of 
the  year.  A  few  are  present  by  early  April.  The  population  increases 
during  April,  May,  and  early  June,  until,  by  late  June,  25  to  50  mil- 
lion bats  are  present.  Estimates  are  based  on  computation  of  area 
covered  by  roosting  bats  and  counts  of  numbers  hanging  in  a  num- 
ber of  sample  areas.  After  the  young  reach  adult  size,  the  population 
decreases  so  that  only  a  few  thousand  are  present  in  mid-September, 
and  by  mid-October  all  are  gone.   Pertinent  field  notes  follow. 

On  April  3,  1961,  Russell  Davis  and  Bob  Neal  visited  this  cave. 
They  reported  ( field  notes ) :  "The  only  bats  present  [estimate,  500] 
occupied  a  fissure  at  the  highest  point  in  the  ceiling." 

On  April  13,  1963,  William  J.  McCauley  visited  the  cave.  He 
reported  (field  notes):    "The  large  crack  in  the  ceiling  of  the  Eagle 


308 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Fig.   3.    Entrance  to  Eagle  Creek  Cave,  Greenlee  Co.,  Arizona   ( localit>-  24). 

Photograph  by  Bruce  Hayward. 

Creek  Cave  appeared  to  be  filled  with  Tadarida  and  the  population 
was  beginning  to  overflow  onto  the  horizontal  ceiling  of  the  main 
cave  room.  A  number  of  bands  could  be  seen  but  colors  were  not 
clearly  discernible  at  that  distance  (75  feet  and  greater).  Due  to 
the  sun  being  more  directly  in  line  with  the  cave  entrance  than  at 
other  times  when  the  cave  has  been  visited,  the  lighting  in  the  cave 
was  better  and  it  was  possible  to  see  that  the  crack  in  the  ceiling 
continues  upward  at  least  60-70  feet  above  the  level  of  the  ceiling. 


CocKRUM — Migration  of  Tad  arid  a 


309 


< 
o 

N 


> 

<: 
U 

U 

u 
« 
U 

M 

J 
O 

< 

W 

H 

< 


i» 


CO 

2 
-a 

■^ 
5 

o 

w 

u 
> 
o 
u 

u 

P5 
W 

u 
< 


Q 
Z 
< 

a 
u 

Q 

«: 
m 

K 
U 

2 

z 

b 
O 

>- 

pa 


p 


CM 

U 

n 


1) 


« 


f-H 

o 
1> 

CO 

CO 

o 
in 

1-H 

CO 

o 

o 
CO 

00 

00 

CO 

o 

lO 

CO 

CO 


o 
o 


CD 


O      "^ 


CO     a: 

CO 


fM 


CO 


CO      o 
CO     -^ 


C-1 


00 


O        -T 


G5 


O  O  -t 

r-l  IC  CO 

t--  ^__  O 

^  cm'  ct 


o 
o 

co" 

CO 


o 
o 


>0       O       CO 

in     OD     00 
.— I     CO  CO 


o    o 


in 


o 
o 
o 


OO        ^ 


00     O  00  o  o  o 

(M     o  CJ5  o  in  in 

C5  ■-<  00_^  l>  ^ 

of  co"  co' 


o 


00 

CO 


o 
o 

o 
o 

1-H 

05 

1 — 1 

05 

05 
00 

1 — 1 
05 

. — 1 
05 

o 

o 

in 

05 

T-H 

05 

05 

C5 


o>     ^ 

CO    I  C5 


o 

o 

i-H 

in 

00 

fM 

05 

in 

00 

T-H 

00 

00 

00 

CO 
00 

o 

in 

00 

in 

05 

I— H 

<M 

i-H 

I-H 

in 
o 


05 

in 


05 

00 
05 

1-H 

00 

CO 
00 

o 

05 

O-l 

05 

5 

GO 

00 

fM 

fM 

05 

1-H 

^H 

O 

^H 

00 

CO 

fM 

1-H 

fM 

co_^ 


fM 

in 

00 


00 

in 

m 

r7> 

f— ( 

CI 

-r 

^H 

>n 

CO 

CO 

C^ 

-r 

CO 

C75 

05 

C5 

CO 

CD 

05 

(15 

I-H 

I-H 

1-H 

Ci 

05 

1 — 1 

^ 

1-H 

, 

T— 1 

»-H 

^ 

4-* 

00 

in 

CO 

^ 

^ 

t^ 

c/: 

CM 

fM 

fM 

m 

fM 

CN 

o 

M 

OJ 

Oj 

0; 

CD 

<o 

0) 

H 

C 

C 

C 

C 

r^ 

< 

-J 

-J 

X. 

hI:; 

<— ' 

1— ' 

310  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

The  highest  part  of  the  crack  visible  is  still  fairly  wide  and  it  appears 
that  the  crack  continues  considerably  beyond  that  point.  The  areas 
of  the  ceiling  and  walls  of  the  main  ca\'e  room  observed  to  be  solidly 
occupied  by  bats  last  June  were  stained  so  that  sharp  lines  of 
demarcation  were  clearly  seen.  It  was  clear  that  the  crack  in  the 
ceiling  could  hold  as  large  a  population  as  could  all  of  the  area  of 
ceiling  wall  ordinarily  occupied  in  June.  There  was  no  evidence  of 
young  bats  already  born  and  it  was  conjectured  that  the  cave  might 
still  be  serving  as  a  Spring  Transient  Locality  with  the  animals 
presently  occupying  it  destined  to  move  to  caves  still  further  north 
before  parturition.  Bats  were  in  motion  flying  in  the  usual  direction; 
clockwise  as  seen  from  below.  Ammonia  odor  was  lighter  than 
that  obser\'ed  in  June,  1962,  at  the  extreme  south  end  of  the  main 
room.  Dermestids  were  absent,  or  at  least  not  obviously  present  in 
the  guano." 

Table  3. — Analysis,  by  sex  and  time  lapse,  between  banding  and  in-place 
RECOVERIES  OF  Tadaridu  brasiliensis  at  Eagle  Creek  Cave,  Arizona. 

Time  lapse  at        Females Males 

time  of  recovery  No.  No.  banded        Per  cent  No.      No.  banded      Per  cent 

Same  year  208  8,400  0.0247  16  750  0.0213 

One  year 537  66,114  0.0081  56  9,898  0.0056 

Two  years  179  34,124  0.0053  14  6,126  0.0023 

Three  years  20  14,260  0.0001  5  2,954  0.0018 

Apparently  most  of  the  gravid  females  arrive  at  this  colony 
within  a  few  days  before  their  young  are  born.  For  example,  on 
June  1,  1964,  the  population  in  the  Eagle  Creek  Cave  was  estimated 
to  be  less  than  100,000.  During  the  early  morning  return  flight, 
many  more  bats  came  in  than  had  left  so  that  the  population  was 
estimated  at  about  250,000.  Many  bats  were  hanging  in  a  part  of 
the  cave  not  occupied  the  previous  day.  Again  on  the  night  of 
June  2,  even  more  bats  came  in  so  that  about  1,000,000  were  present 
on  June  3. 

A  total  of  88,176  guano  bats  have  been  banded  at  this  cave 
(Table  2).  Because  most  were  banded  in  June,  they  were  adults  at 
least  one  year  old  that  had  spent  the  winter  elsewhere  and  had 
mo\'ed  into  the  ca\'e  during  the  weeks  preceding  banding.  No  spe- 
cial efforts  were  made  to  band  indi\'iduals  of  a  gi\'en  sex  so  that  the 
sex  ratios  of  bats  banded  probably  represents  that  of  the  whole 
population.  Note  that  the  ratio  of  those  banded,  12  males  to  88 
females,  compares  favorably  to  the  ratio  in  total  recoveries,  14  males 
to  86  females.    However,  adult  males  banded  at  the  cave  do  not 


CocKRUM — Migration  of  Tadarida 


311 


s 


o  s 

i-i  z 

w  o 

C   N 

g  2 
< 

a    . 
<^ 

i  w 
U 

ca  CO 

«    § 

n  •!- 
>^    Co 

.^  a 

^-^ 

5  S 
&° 

*^    CO 


C/3 


U 
1-1 

m 
<: 
H 


> 

o 
o 


O 
CD 


CO 


CO 


o 

o 


o 
o 


o    o 
o    o 


o    o 


o 

o 


Ol 


O      ^      ^      CO      CO      o 


O      O      O      C-1     o     o 


CO 


CO 
CO 


fM 


CO 
CO 


lo    o    CO    in    00    ^ 
CO    lo     in    CO 


— I    in    t-    X    -^    o 


CO 
c^ 


CO 
00 


o 

CD 


Tt*      o 


CO      o      ^ 


o    m    -H     -f     00 
o<i    in    in    CO 


— I     ^ 


o 


o— iint-OTPoooo 


ooooocococDfint^oo^ 
cocqcocoTfooasoa;  •-< 

Cv|  >— I    ■— I    CO   --H 


in  ^  m  CO 

CD 


in  00  CO  CO  o  ■—I 

CO     CO  -* 


in 


vn 

I— I 

CO 


CD 

05 

<— 1 

,-H 

t-H 

i—H 

^H 

^H 

I— f 

CO 

o' 

CD 

CD 

CD 

CD 

CD 

05 

»— ( 

05 

05 

Oi 

C5 

O) 

t-H 

, 

1 — 1 

i-H 

f— ( 

1— t 

^H 

^ 

53 

t- 

o" 

1-H 

co' 

-r' 

t-' 

^^ 

p 

CO 

CO 

CO 

C^l 

CO 

•J^ 

5 

>> 

>. 

>s 

>. 

>> 

M 

■*-* 

CS 

C3 

C3 

rt 

ca 

3 

O4 

S 

:s 

S 

S 

S 

0* 
C/2 

Oi 


*;;   5   aj 


o 
o 

o 


Ol 

CO 

CD 

CD 

C^l 

05 

05 

r^-\ 

CD 
0:1 

^^ 

f-H 

CD 

C5 

1— 1 

in 

O^" 

^H 

y^ 

1 — 1 

i-H 

t- 

4-' 

•t^ 

■^'' 

r— 1 

ir. 

in 

CO 

0 

5J 

^ 
M 

>. 

H 

rt 

^4 

< 

< 

312  Misc.  Publ.  51,  Univ.  Kansas  Mus,  Nat.  Hist. 

return  in  the  same  ratio  as  do  the  adult  females.  The  in-place  re- 
covery ratio  is  nine  males  to  91  females,  whereas  the  foreign  recov- 
ery ratio  is  39  males  to  61  females.  A  possible  explanation  is  that  a 
large  percentage  of  the  males  in  the  colony  are  yearlings,  born  in 
the  colony  the  preceding  summer  and  thus  not  subject  to  banding  at 
Eagle  Creek.  Howexer  these  males  would  have  been  subject  to 
banding  at  other  locations  during  the  preceding  autumn,  winter, 
and  spring  months. 

An  analysis  of  in-place  recoveries  by  time  lapsed  before  recovery 
is  given  in  Table  3.  Movements  shown  by  the  foreign  recoveries  is 
given  in  a  later  section  of  this  paper. 

Other  smaller  maternity  colonies  have  been  found  in  caves 
(localities  8,  61),  bridge  crevices  (locality  13),  and  a  mine  tunnel 
(locality  28). 

Summer  Male  Roosts 

Few  roosts  of  this  type  were  studied.  Many,  perhaps  most,  of 
these  roosts  consist  of  small  groups  of  10  to  300  individuals.  The 
roost  examined  on  July  23,  1963,  in  the  attic  of  St.  Patrick  Church, 
Bisbee,  Arizona  (locality  59),  is  the  largest  such  colony  observed. 
Approximately  1300  guano  bats  were  present  of  which  134  ( 131 
males,  three  females)  were  captured.  The  females  showed  no 
obvious  signs  of  having  borne  young  that  year. 

Transient  Roosts 
In  the  early  part  of  the  year  and  again  in  late  summer  and  in 
autumn,  guano  bats  occupy  a  variety  of  roosts  on  a  temporary  basis. 
Some  appear  to  be  occupied  by  a  given  group  of  bats  for  only  a 
few  (possibly  only  one)  days,  with  another  group  being  present  a 
few  days  later.  Examples  of  such  roosts  include:  crevices  in  bridges 
(localities  14-21,  31,  42,  49,  50,  52,  53,  54,  58);  mine  tunnels  (locali- 
ties 26,  27,  37,  40,  41);  buildings  (localities  33,  34,  38,  62,  63.  64), 
and  natural  caves  (locality  39).  Populations  in  such  roosts  vary 
widely  in  numbers  and  sex  ratios  (Table  4). 

Winter  Roosts 
No  major  winter  roosts  were  found  in  the  course  of  this  study. 
The  volcanic  cave,  Cueva  Montelarga,  17  mi.  W  Pericos,  Sinaloa, 
was  thought  to  be  a  winter  roost  when  first  discovered  in  late 
January,  1963.  Visits  to  this  cave  in  December  of  1963  and  early 
January  of  1964,  however,  revealed  that  no  guano  bats  were  present. 


CocKRUM — Migration  of  Tadarida 


313 


< 
o 

o 

6 

a 
a 
< 
U 


"3 

b 
O 

a 

5 
a 
> 
o 
u 
u 

u 
u 
<: 

>-i 


Q 
< 

w 

Q 
Z 

<< 

n 

in 

q: 
u 
ca 

z 

b 
O 


u 


£ 

"a 


> 

o 


T3 

0) 

c 


O 


oo    i 

1    ;    lino    :>*    : 

I      1      IClTjl      ICO      1 

ICOW     lOcO     ' 

lint-    I-*-* 

o 

lolO 

I05^ 

oo 

t— (  rH 

'    '    'irao     CO    ' 

I>«2       CD 

r-in     ot> 

TfM     coin 

'ooo 

oo    : 

1     1 0  CO  O  O  c-l  O 

irocDOcico 

l<M-H 

lOroioOTfo    ixc-ioco-* 


CO  cl  O      00  o 
<M  O  ^  O  O  CO 


■*  00  O  O  CI  O 
t^C:5  050550 


C!0     com 

CO  CO  O  lO  CO 

-^dooio 

cooo-rco 


oDc-ixooO'-i    locooinoo 
CO    lift  coi> 


CO  rt  00  C-l  Cl  'i*     I  -<  Tt  rt  — I  ci 

c^-Hco     Oi'^    :05i-i     co-"^ 


ICO      00  o 

I  M  Tf  ^  O  03  CO 


OCO       Cl^ 

-i<-^OlCCD 


'■^COCIO^O       Ot-OMCE 

i>roxooM^     ccioofco 


OOlOXOOi-H       COCIOO-H 

i-(         rHCO         COi-H         cox 


Oi— iffi 


OlOO 

OOCO 

OlOO) 


OCOOI 


lOcOX 
<M 


CO  10 
Otj<0 

6i>io 

010  05 


OCOcq 


lOTfO 

CO 


ICO 

lO 


t> 

05 


:o 

'lO 
ICO 


O     "^       ^05COCO^lOX050X10-HOt^OCO'^     1 1-     I -H     :rt 

CI    ;o     coco-fci'i-^iO'rio-rooxxo5C350>    io5    ico    lO 


CO  O  CO  CI  CO  -tf 
CI  CI  X       05  Tf 


O5CO^-t<C0 
lOSr-l         CO-^ 


IrtCOOOOO 


oooo 

I  CI  O  O  O  O  O  lO 
^|>05rt'y<-H'^O 

'xcoooodio 

CI  CI 


rt  ■^  CO  CO  I— I  rt  CO 
10 -H  10  CO 


OCOOOOO^ 


OOOO 
llOt-OOOOcO 

I  O  05  »— '  I— I  ^H  ^H  X 

iocodddoi> 

CI  1-1 


CO  lO  CO  CO -H  rt  lO 
CO^IOCO 


■ICOOOOO-H 


CI 


COt-i-H 

I^O 

coio 

■OCO 
lot- 

ICOO     1 
I05O5     1 

ICl 
lO 

lO 

l«!5 

IM 

ICO 

ICO 
lO 

-tCOOJ 
10I>10 

'0)0 
COIO 

ot 

lOCl 

oo    ' 

'X 

'CO 

'X 
CO 

't 
cq 

coo 
inoo 

LO^O 

lOO 

coo 
xo 

CO 

C05 
COM 

050      1 
CDC     I 
Icoct:     ' 

O 

c 
c 

CO 

o 

05 

IC 

c 

X 

lO 
lO 
IX 

Tf^Cl 

i-H 

1—1 

CI 

1—1 

CI 

xoo 

CSOIO 
COOX 

CO 

to 

lOO 

oo 

OC1 

cot 

-iO     1 
XO     1 
1—1  f^ 

Ico 

X 

It 

105 

lO 

o 

lO 

Ico 
ICO 

ei 

10  CO  CI 

1—1 

O       CD       05-HttC510Cl^Ccll0050COCOtCD 
X       05       50CDlOOlOXtlOlOlOC1ClC1i-l 


CO       05 
CO 


I.-IO 


CI 


ox 
oo 


to 


lO 


CO 


lO 

CO 
o 


o 

t 

CO 
lO 


IOC     I     li-i     I       'H 


X 

o 


o 

lO 


CO 

t 
o 

lO 


c;5 

CI 
lO 


o 

05 

CD 


0510--I     ;C0 
COO-H      ,  t 


^10  05 
COCIX 


o 

lO 


COOOOCCOOCCCC  coo  OCO  c 

ooocococooccc  coc  ccc  c 

CCOC]_0_0  0_0_^0_0_^CCO_  co_c  coo_  o 

cfioio'cfLo'cio'c'occ'ic'c  coc  ccc  c 

Ci-hcocooc  coc  ccc  o 

rt         r— -Hrt  CI  lOlClO  Cl'v'^  '^' 

r4  COcf 


oocooooooo 

occ  ooooco 

o  o  c_  o_o_^ci  o_o_o_ 

o'o'-H  -ho'    i-i'io'io' 

CO  -H 
lO 

c{ 


S;cO  — CCCDG5  — 


o       coxg 

CI  CO  com  5^ 

CDCD05C:2 

o 

C^  t  CI  CO 

O 

ft 

OiGi'-*-^ 

CI  O  CD  CD  CD  CD 

OC  '^^C]  ro 

f^l 

^H^H 

^  CD  05  05  05  -. 

O'OO^CCCC 

cr 

>—' 

lo'o"^ 

C5o;XH^rt-^ 

cc  :7^C^^G^c^ 

io^ 

CO"— ' 

-^ 

OdrH^Cl 
COCO    .      .      ^ 

-  .lo'^'t'c 

2^' 

■— i 

oa> 

^ 

CO  o  -H  cl  CI  CO 

,-H-X^CC 

*■  .^ 

r^^ 

/ 

r^  r"  f-  r-  rt 

X  1— '.-H^HOlCI 

CI 

X.X.X.AX.X. 

Ci'.'t 

-HC) 

— 

i^      —      C      ^      fcH 

o   CJ   o   O   O   CJ 

1l 

c  C-S-S-? 

1_     J«     b.     t_     ^     Im 

^   '"   I--   ^  ^  ^ 

^.  -^ 

l' 

^^^^^ 

S22SS2 

««« 

ss 

1— i 

^.^ 

^-^ 

CO  C-'  Xr~^^^ 

COCO        O    r^C5mCDIOCD 

O    05<^IOC1^100500505 

Cl^'^C0CC(-,C5rtrt^i— I 

05CD     ^     ,0505^;^     _     ^     _ 

lO0;^C0'-''^      ,     .CKOCc-i 

ClCl      ,     -COOi-IClCOC) 

"     -  .     ,_.— iCO 

,<<cXcJ5ooz;2:zzzQ 


1        CI      1 
lO-H       1 

O     1 
1—1 

t 
t 

'     'OX     ' 

oo 

1—1 

'O 

CO 
lO 

1     1 0  CI     1 

1 1— <     1 

o 

X 
lO 

o 

CI 

o 

lOO     1 

;o    1 

o 

t 
t 

1—1 

s 
o 
U 

d 


1     1     1     lO 

o    : 

05 

1     1     1     lO 

10    : 

CO 

1    1    ;    iio 

1 1—1 

CD 

f-H 

1— 1 
CI 

1     1     1     lO 

ICO     1 

IC 

1     1     1     IIO 

ICI      1 

lO 

1     1     1     Ico 

It    1 

1—1 

t 

05 

o 

o' 
t 


c 
& 

U 

IP 
bC 
o 
PC 


P 

2 


CO 
CO 
05 


a; 


c 


H     O 


314  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

As  bats  were  being  banded  at  the  Pericos  locality  on  March  2, 
1963,  an  unusual  behavior  of  captives  was  noted.  Males  were  much 
more  active  than  the  females;  thus  bats  taken  from  a  freshly  filled 
retaining  cage  tended  to  be  primarily  males,  whereas  primarily 
females  remained  in  the  cage.  A  gross  examination  revealed  that 
males  were  reproductively  active  and  that  \'aginal  plugs  were 
present  in  many  females.  In  contrast  to  their  beha\ior  during  most 
other  banding  operations,  females  reacted  to  the  touch  of  the 
banders  by  crouching  down,  not  by  aggressive  biting  and  attempts 
to  escape. 

Multi-use  Roosts 

The  localities  listed  above  fall  rather  clearly  into  a  given  type  of 
roost.  A  few  localities  serve,  in  season,  as  transient  roosts  and  as  a 
maternity  colony.  One  such  locality  that  was  visited  many  times 
during  this  study  is  a  volcanic  tube  cave,  Cueva  del  Tigre,  14.9  mi. 
SSE  Carbo,  Sonora.  It  was  in  this  cave  that  Henry  Mitchell  carried 
out  a  year-round  survey  of  the  cave  environment  as  part  of  his  study 
of  the  respiratory  physiology  of  certain  molossid  bats  (Mitchell, 
1963). 

Table  5  summarizes  visits  to  this  cave,  estimates  of  population 
size,  and  information  on  bandings,  recoveries,  and  sex  ratios.  Note 
the  low  population  in  the  winter  months  with,  in  general,  more  males 
present  than  females.  Males  predominate  through  mid-March.  By 
late  June  and  July,  when  the  population  reaches  a  maximal  size, 
females  make  up  most  of  the  population. 

In  all,  3240  guano  bats  were  recovered  in  place.  Most  of  these 
involved  a  single  recapture  of  a  given  individual,  often  a  year  or 
more  after  the  time  of  banding.  Such  recoveries  give  little  insight 
into  the  question  as  to  how  "loyal"  a  bat  is  to  a  given  transient  roost 
or  maternity  colony.  However,  a  few  individuals  were  recaptured  in 
place  three  or  more  times  (Table  6).  This  demonstrates  that  a 
given  individual  will  utilize  a  gi\'en  temporary  roost  both  in  the 
spring  and  again  in  the  autumn  ( Silver  Creek,  St.  David )  as  well  as 
in  successive  years  in  the  same  season.  This  "loyalty"  to  the  use  of 
transient  roosts  persists  in  spite  of  the  bats  being  physically  removed 
and  released  at  another  place.  The  Silver  Creek,  Arizona,  popula- 
tions were  utilized  along  with  Myotis  velifer  and  Antrozous  paUidus 
in  a  series  of  homing  experiments  (Davis,  1966).  For  example,  65 
female  Tadarida  taken  at  Siher  Creek  on  May  20,  1961,  all  were 
released  five  miles  southwest  of  the  site.  None  returned  the  same  or 
the  following  night,  although  one  returned  the  third  night  (May  22), 


CocKRUM — Migration  of  Tadarida 


315 


Table    6. — Partial   list   of   ix-place   recoveries   of    Tadarida    brasiliensis 

TAKEN    more    THAN    ONCE. 


Original  capture 


1st  recapture 


2nd  recapture 


3rd  recapture 


Carbo,  Sonora  (males) 


April  18,  1959 
November  6,  1959 
March  7,  1960 
March  7,  1960 
March  7,  1960 
March  7,  1960 
April  9,  1960 
April  9,  1960 
April  9,  1960 
April  9,  1960 
February  15,  1963 


April  9,  1960 


May  20,  1961 
Mav20,  1961 
September  10,  1961 


Tune  22, 1959 
June  1, 1959 
July  24,  1959 


April  9,  1960 
March  7,  1960 
April  21,  1962 
Febniarv  13,  1963 
April  9,  1960 
February  15,  1963 
February  15,  1963 
March  30,  1963 
March  20,  1963 
March  15,  1963 
March  15,  1963 


July  2,  1961 
March  15,  1963 
April  11,  1963 
April  11,  1963 
May  20,  1963 
March  30,  1963 
March  21,  1964 
April  11,  1963 

March  21,  1964 
March  30,  1963 


Carbo,  Sonora  (females) 
July  18,  1960  July  21,  1961 

Silver  Creek,  Arizona  (females) 
October  10,  1961  June  17,  1962 


September  10,  1961 
June  17, 1962 


June  17, 1962 
June  4,  1963 


June  1,  1959 
June  1,  1959 
June  22,  1959 
July  24,  1959 


June  5, 1963 


June  24, 1951 
June  23,  1962 


St.  David,  Arizona  (males) 

March  27,  1960  May  28,  1960 

July  24,  1959  May  28,  1960 

March  27,  1960  May  28,  1960 

St.  David,  Arizona  ( females ) 

June  22,  1959  October  1,  1960 

June  24,  1959  July  24,  1959 

May  28,  1960  September  3,  1960 

May  28,  1960  August  22,  1960 

Eagle  Creek,  Arizona  (males) 
June  2, 1964  June  27, 1964 

Eagle  Creek,  Arizona  (females) 

June  23,  1962  June  4, 1963 

June  4, 1963  June  27,  1964 


April  11,  1963 
March  21,  1964 

April  11,  1963 


August  5,  1962 


August  27,  1960 
August  27,  1960 
September  23,  1961 


April  15,  1961 
May  28,  1960 


three  by  May  27,  and  14  others  returned  one  or  more  times  in  the 
following  18  months,  even  though  each  recapture  was  followed  by 
a  release  some  distance  from  the  roost.  For  example,  a  female 
released  5  mi.  SW  Silver  Creek  on  May  20,  1961,  was  recaptured  on 
May  27  and  released  at  Maricopa,  Pinal  County.  It  was  again 
recaptured  at  Silver  Creek  on  September  10,  1961,  and  released  at 
Tucson  (100  miles  northwest),  was  captured  again  on  October  1,  1961, 
and  again  released  at  Tucson,  and  finally  was  recaptured  at  Silver 
Creek  on  June  17,  1962.  Six  others  of  this  group  showed  three  or 
more  such  recaptures.  Twelve  of  65  males  reacted  in  a  similar 
manner.  One  male,  banded  and  released  5  mi.  SW  Silver  Creek  on 
May  20,  1961,  was  recovered  on  August  7,  1961,  and  released  10  mi. 
SW  Silver  Creek;  it  was  recovered  on  September  10  and  again 
released  10  mi.  SW  Silver  Creek;  recovered  again  on  October  1, 
1961,  and  released  at  Benson  (60  miles  northwest);  recovered  again 
on  June  12,  1962,  and  released  at  Tucson  ( 100  miles  northwest ) ; 
and  finally  recovered  on  August  5,  1962. 


316  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Foreign  Recoveries 

In  all,  539  guano  bats  have  been  recovered  at  points  other  than 
the  site  of  banding.  Most  (all  but  22)  of  these  were  recovered  by 
persons  actively  collecting  bats  or  involved  in  bat  banding  activities. 
Thus,  in  spite  of  an  intensive  publicity  campaign  designed  to  enlist 
the  cooperation  of  the  general  public  in  reporting  captured  banded 
bats,  less  than  four  per  cent  of  the  foreign  returns  came  from  that 
source.  Foreign  returns  are  summarized  in  Tables  7  and  8.  A  com- 
plete breakdown  of  these  recoveries,  giving  time  and  date  of 
banding  and  recovery,  is  available  from  the  author  on  request. 

Discussion 

Most  of  the  foreign  recoveries  represent  indixiduals  retaken  only 
a  single  time  ( Table  7 ) .  However,  fourteen  indi\'iduals  were  retaken 
twice  ( Table  8 )  and  some  of  these  clearly  indicate  movement  back 
and  forth  between  transient  roosts.  Added  to  information  gained 
from  an  analysis  of  movements  (see  maps,  Figs.  4-7)  and  of  roost 
types,  the  following  sequence  of  movements  appears  to  characterize 
the  migration  of  guano  bat  populations  in  Arizona.  Most  spend  the 
winter  months  ( December  and  January )  in  Mexico,  presumably  at 
locations  south  of  Pericos,  Sinaloa.  Some,  mainly  males,  winter  as 
far  north  as  Carbo,  Sonora,  and  a  few  probably  winter  in  the  lower 
valleys  of  Arizona.  In  late  February  and  early  March,  breeding 
occurs  in  Sinaloa  and  Sonora.  In  March,  April,  May,  and  June  sub- 
populations  move  progressively  northward,  occupying  successively 
a  series  of  transient  roosts.  Males  appear  to  move  more  rapidly  than 
females  ( see  sex  ratios  of  bats  taken  at  Carbo,  Sonora,  during  these 
months.  Appendix  1).  During  May  and  early  June  a  large  number 
of  females  arrive  in  northern  Sonora  and  southern  Arizona.  By  mid- 
June  to  early  July,  when  gestation  occurs,  most  females  have  con- 
gregated in  a  relatively  few  maternity  colonies  (e.g..  Eagle  Creek, 
25  million;  Carbo,  three  million;  Powerline  Cave,  8,000  to  10,000; 
Perkinsville,  3,000;  Gaits  Cave,  5,000 ) .  During  this  time  males  gen- 
erally occur  in  smaller  colonies,  consisting  primarily  of  males. 

Probably  not  all  males  migrate  northward,  for  Alfred  Gardner 
(field  notes)  observed  some  40,000  guano  bats  in  a  cave  near 
Comitan,  Chiapas,  on  June  23,  1964.  All  examined  were  male. 

After  the  young  reach  adult  size,  the  maternity  colonies  rapidly 
break  up  (but  are  not  abandoned  by  all  bats)  and,  in  late  August, 
September,  and  early  October,  a  number  of  transient  roosts  are 
occupied  in  Arizona  and  northern  Sonora.   During  this  time,  popula- 


CocKRUM — Migration  of  Tadarida 


317 


H   '-r. 

<  a 

cs 

^  w 

r^  K* 

S  0 

2^ 

„   S   W 

"   W    (2 

5  to  w 

<  "  > 

"is 

J  z  2 

fO     ,   K 

<  w  '^ 

H   ffi  en 

t-   W 

fc.        J 

0  u  «; 

fc  ^  § 

2  ^  " 

^  Q  ^ 

zSg 

0     _ 

TED 
IBERS 
AND 

M      <      W 

►J  ;=  ^ 

z  < 

rri         "^ 

W   u  '^ 

H   S   W 

n 

U    >,     73 

q  ffi  z 

hJ  w  «: 

>  w 

c^S- 

2  So 

^       ^ 

,:s  u< 

•E?  O  w" 

2  w  iJ 

-§1 

a  J  -<: 

"t;  t,  X 

•=  a  w 

a  ^  O 

h,  w  i^ 

K"-^ 

RIES 

LEF 

CALI 

"  Z  0 

l^^ 

U  ="  z 

W   W   H 

«H   > 

Z  J  u 

OREIG 
LOCA 
AT   A 

fe    • - 

O   X 

fc  w  S 

0  o  " 

z 

>;  <;  K 

ffi  o  u 

"<    n    < 

v<    W    u 

-=^  «  ^ 

S  w  tx. 

5  ^  o 

^  73 

^ 

'•  < 

t-  ffl 

w  ffi 

►J  u 

pa  a 

H  ^ 

SlUUAEnQ 

oqjB3 
soijfBX 

OAi'o  s;[BO 

3{aaJ3  JaA[!S 
aaqsig 

SUMS 

piAT^a  IS 

uosusg 

Il!}uaU]lUO[) 

aAiqaag 

uosonx 

nsqj8Ai!S 

>iaaj3  ai3i3a 
UJBQ  sSpqoo^ 

9UIASUI3|J3J 

aAB3  3ut{japA\oj 


03  ^ 

CC  O  C^  CC  CI  O  CO 

1,111 I    )     I     I     I     I     I     I     I     I     t    I     I     I    I     I    I     I     I    1     I     I     I 

rHOrt-*0-^'-^OXC1^0clCOCc'3CIC-^iOOOC'-^OC>C"*rtOOcl 
;0  1-1  CD 


<M 


ID 


CI 

10 


ot-o 


I 
CI 


CO 
I 

o 


-HOC  CI 

I     t      I      I 
OCICIO 


o 

I 


1> 

c'l 


o 

CO 


ei 


o 


<M 


I 


lo 


cq 


t 
o 


--(        CO 

O         I 
CI 


cq 


O       --I 


CO 


CD       CI,— ICD       O'-H       I— iCO-Hr-<rt  ffiO^-^'^       O-HrtTft       •-H'-j< 

1     OrHO     '^''H     oc'iooo        o-^ooo     '^oo'T'^     oo 


oo 


cq 


t.SS  0^5  g^  c-S  §.£;  ■.-  o- 
HPL,UWe,Qc/3pM:«HpatfUapa 


2 

'> 

Dp; 

C/3C/3 


U 

.  ,   CJ 

'Air. 


c 

3 
en 

s 


V 


■So  c  o  P  S.-3  g 


ti-D 


318 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Table  8. — Foreign  recoveries  of  Tadarida   hrasiliensis  that  were  taken 

MORE    THAN    ONCE. 


Banded 


1st  recovery 


1 


Carbo 

(April  18,  1958) 
1  Carbo 

(February  15,  1963) 
1  Carbo 

(February  15,  1963) 
1  Eagle  Creek 

(June  5,  1963) 


1 


1 


1 


1 


Carbo 

(March?,  1960) 
Silver  Creek 

(March  21,  1961) 
Eagle  Creek 

(June  23,  1962) 
Continental 

(May  3,  1963) 
1  Continental 

(May  3,  1963) 
4  Silverbell 

(May  10,  1963) 
Silverbell 

(May  10,  1963) 


1 


Males 
SWRS 

(July  16,  1959) 
Eagle  Creek 

(June  4,  1963) 
Eagle  Creek 

(June  4,  1963) 
Gaits  Cave 

(June  21,  1963) 

Females 
Silverbell 

(May  10,  1963) 
Eagle  Creek 

(June  24,  1961) 
Silverbell 

(May  10,  1963) 
Silverbell 

(May  10,  1963) 
Silverbell 

(May  10,  1963) 
Continental 

(May  16,  1963) 
Continental 

(May  16,  1963) 


2nd  recovery 


Carbo 

(November  6,  1959) 
Carbo 

(March  21,  1964) 
Carbo 

(March  21,  1964) 
Eagle  Creek 

(June  27,  1964) 


Carbo 

(March  21,  1964) 
Silver  Creek 

(October  1,  1961) 
Eagle  Creek 

(June  27,  1964) 
Continental 

(June  11,  1963) 
Continental 

(May  16,  1963) 
Silverbell 

(June  10,  1963) 
Tajitos 

(September  6,  1963) 


tion  samples  reveal  sex  ratios  approaching  50:50  in  most  roosts.  By 
mid-October,  most  bats  have  left  Arizona  and  are  occupying  tran- 
sient colonies  in  Sonora. 

The  details  of  flight  compositions  and  patterns  as  well  as  dis- 
tances covered  in  a  given  flight  are  not  known.  The  spectacular 
increase  in  the  population  at  the  maternity  colony  at  Eagle  Creek 
in  early  June  of  1964  already  has  been  discussed.  A  similar  spec- 
tacular increase  in  the  population  at  the  cave  near  Carbo,  Sonora, 
was  observed  by  Henry  Mitchell  (1963:38)  and  William  J.  McCau- 
ley  (field  notes).  No  bats  were  present  in  the  cave  on  March  24, 
1962,  but  by  the  morning  of  March  25  approximately  100,000  were 
present  ( 85  per  cent  of  those  examined  were  males ) . 

Some  recoveries  indicate  that  rapid  flights  o\'er  relatively  long 
distances  are  possible  (Table  9).  One  female  moved  175  miles  in 
no  more  than  four  nights  (averaging  43.8  miles  per  night);  two 
others  covered  85  miles  in  no  more  than  two  nights  (42.5  miles  per 
night).  Two  females  and  one  male  traveled  765  miles  in  no  more 
than  39  nights  (averaging  19.6  miles  per  night).  Since  exact  times 
of  departure  and  arrival  are  not  known,  these  probably  represent 
minimal  values.   Glass  (1959:544)  reported  movements  of  approxi- 


CocKRUM — Migration  of  Tadarida 


319 


6t 


a 
u 

a 
c 


CD     C<j     oq     ■*     — ;     CO 

ai   ai    ^    00   t^    n 


05     ■*     CD     O     t-     CO 


lo    in    o    i/^    o    lo 
CD     in     CO    CO     oi    i>- 

t^     (M     •-H     CO     t-^ 


00100t:^CD^<MlOC<JI>C<IincOCp 

coc^icDcooiod^'ooxcDiO'^co 


iTiioommiooominoinoo 

t-^OOCOOJCDinoO^HlTit^i— lOi— <CD 

>— I  r-H  r^       (M       ^H       Ol       C<I       ^       C-1       ^H       fM       T-^ 


t- 

t> 

lO 

»n 

05 

CO 

1— 1 

1— 1 

CO 

CD 

CO 
CD 
CJ5 

I-H 

CO 

CD 

1 — ( 

CO 
CD 
Gi 

1 — 1 

CO 
CD 

CO 
CD 
05 

ro 
CD 

Oi 

CO 
CD 

I-H 

o> 
in 

05 

ll 

i-H 

^ 

r 

^H 

z 

CO" 

1— ( 

X 

CD 

CO' 

o 

OJ 

dj 

Oj 

TS 

T^ 

0; 

>, 

Cj 

Oj 

0; 

> 
c 

c 

c 

o 

r- 

J^ 

z 

IX 

ijl 

< 

< 

1 — . 

^ 

fH.^ 

*D^, 

1 — 1 

z 

CO  CO      CO      CO 

^     CD     CO     CO     CO     ^      " 


05 


05     ^     ^       •-     ^     -H     05 


CO      Ol 


CD     CO  05  05  05  CD 

05     05  ■— I  •— I  ■— '  C35 

^H      I— (  ^  ^  ,.  ^H 

„  CO  CO  CO 

^H       ^H  ^  <M  Ol  CO 


>.    >. 


CLi       Oj 


-^    6 

^  ^  C- 


CO 


C     CO     CO 


CM 


O         lA 


> 


o 
c 


0) 

U 


rt    CD 


c 
c 


^  U  u  - 

S  ^  =/=  S  S  ^ 

•r  M  —  i:  'n  M 

n,  rt  :^  rt  rt  rt 

fi,  W  G  U  U  W 


U     i    U 


CM 


a; 


r^         ^        ^'       '^^  C        '"' 


<^1      ^ 


^  w>  -^ 


O 

d 
c 


a* 
U 


CO 

d 
c 


> 

a 

U 


CO  CD 

CO  CO 

d  d 

c  c 


-     o     >r-     a>     h 

u  a  c;5  M  &:; 


c     c 


CO     ^ 

d  ? 


c      C 


a    U   0   U   U   U   H 


CD  CO 

CO  CO 

d  d 

c  c 


U   U 


in 


CO 


S    CO 


^   a   ^ 


CO 

CD 
CX> 


cm' 


05 

in 

05 


CO  ^ 

;S  05 

1 — I 

CO  ^ 


CO 

CD 

05 


CO 
CD 
05 


in 


in 

05 


in 

05  CO 

r-H  ;(~ 

,  c:5 


CO 


CO 

CD 
05 


:::;  05  ;::;  05  ^ 


CO  CO  CO  CO  t— I 

CO  CD  CO  CO  CD  CO 

O  05  CJ5  05  c:5  05 
05 


in 


■5rtCc8Q,C  rtCG 


CM  ^  (M  1^  -^ 

O;  O  >.  (U  <U  OJ  OJ 

c  o  «^  S  ^  S  £ 

^  O  S  ►H,  x.  ^H,  ^. 


o   o   o   o 

I-H       1—1       ^H       CM 


c 


03      rt 


K^     ^ 


^ 

'-^ 

-f 

t 

^ 

- — - 

00 

CM 

^__ 

CM 

CM 

'-^ 

CM 

.-^ 

CM 

CM 

._^ 

^-^ 

^^ 

in 

CO 

CO 

d 
c 

o 

CO 

in 

in 

CM 

in 

d 

CO 
in 

co 

CD 

r* 

00 

CM 

d 
c 

d 

00 

CM 

00 
(M 

00 

CM 

d 

d 

CO 

d 

^^ 

S 

d 

CO 

5 

-^ 

u 

03 

3 

c 

s 

^-' 

f— 

a; 

c 

y^ 

c 

■w 

, 

I' 

3 

^-' 

c 

<D 

-- 

4H 

^ 

^ 

^ 

^ 

u 
> 

f^ 

V5 

o 
o 

o; 

C 

■£ 

c 

U 

> 
Q 

> 

o 

c 

■J3 

>- 
a; 

u 

a; 

'a3 

t^ 

'bb 

•r; 

rt 

^ 

c 

C 

"m 

% 

^ 

M 

> 

"m 

C 

^ 

> 

> 

> 

a; 

05 

03 

11 

<— ' 

c 

di 

rt 

■4— ' 

QJ 

c« 

03 

r^ 

c 

03 

U 

W 

Oh 

u 

C/2 

U 

02 

M 

CAl 

CQ 

u 

w 

c75 

w 

U 

CO 

CO 

CO 

c^ 

1 — I       I — ^       CO       ' — ^       ^H 


^     CI     CD     ^     CM     CM     1-H     CI 


CO       >— I       i— I      '-H      ■— I 


320 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Fig.  4.    Map  showing  northward  movements  of  banded  female  Tadarida  hra- 

siliensis. 


mately  500  miles  (from  Selman's  Cave,  Oklahoma,  to  Carlsbad 
Caverns,  New  Mexico).  Four  bats  made  the  trip  in  12  nights  (an 
average  of  about  42  miles  per  night)  and  one  in  13  nights  (an 
average  of  about  3(S  miles  per  night).  Again,  these  represent  mini- 
mal values. 

Some  females,  but  no  males,  have  been  reco\'ered  moving  south- 
ward in  the  late  spring  (Table  9).  For  example,  three  females 
banded  at  Continental,  Arizona,  on  June  11,  1963,  were  retaken  in 
the  maternity  colony  at  Carbo  on  July  1,  and  a  female  banded  at 


CocKRUM — Migration  of  Tadarida 


321 


Fig.  5.    Map  showing  northward  movements  of  banded  male  Tadarida  hra- 

siliensis. 


Silverbell,  Arizona,  on  June  10,  1963,  was  taken  at  Tajitos,  Sonora, 
on  June  28.  The  significance  of  these  movements  is  not  known. 
However,  at  least  some  were  not  gravid  and  showed  no  obvious 
signs  of  recent  parturition  at  the  time  of  recapture.  Perhaps  the 
stresses  of  the  original  capture  and  banding  had  caused  a  miscar- 
riage (a  situation  observed  at  times  in  guano  bats  and  more  com- 
monly in  other  species)  and  the  migration  to  a  northern  maternity 
colony  was  not  completed.  Glass  (1959:544-545)  has  reported  a 
similar  southward  movement  of  a  female,  of  some  400  miles,  when 


322 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Fig.  6.    Map  showing  southward  movements  of  banded  female  Tadarida  hra- 

siliensis. 


the  time  of  the  year  coincided  "with  the  peak  of  reproductive  and 
nursing  activity,  when  females  might  be  expected  to  be  most 
sedentary." 

The  guano  bat,  Tadarida  hrasiliensis  Geoffroy,  has  been  recorded 
from  locahties  as  far  north  as  Medford,  Oregon,  and  Lincoln, 
Nebraska  (Hall  and  Kelson,  1959:206),  southward  through  the 
southern  United  States,  Mexico,  Central  America  and,  in  South 
America,  at  least  to  Chimpay,  Rio  Negro,  Argentina  (Schwartz, 
1955:108).    It  also  occurs  on  most  of  the  islands  in  the  Caribbean 


CocKRUM — Migration  of  Tad  arid  a 


323 


Fig.   7.     Map   showing   the   southward   movements   of   banded   male    Tadarida 

brasiliensis. 


(Hall  and  Kelson,  Joe.  cit.).  As  currently  understood  the  species  is 
di\'isible  into  a  number  of  races  ( Schwartz,  op.  cit. ) ,  but  the  inter- 
relationships of  the  various  populations  within  the  species  complex 
are  poorly  known.  The  discovery  that  at  least  some  of  the  popula- 
tions of  the  north  temperate  portion  of  this  range  are  migratory  has 
led  to  even  further  confusion  concerning  these  relationships. 

In  recent  years  all  guano  bats  in  the  southwestern  United  States 
and  Mexico  have  been  referred  to  a  single  subspecies,  Tadarida 
brasiliensis  mexicana  (Saussure).  Large  numbers  of  these  bats  have 


324 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


been  banded  in  Oklahoma  (Bryan  Glass  and  associates),  Texas 
(Davis  et  al;  Eads  et  ah),  New  Mexico  (see  Constantine,  1967:54, 
for  list),  and  California  (Albert  J.  Beck,  Philip  Leitner,  and  others). 
One  of  the  surprising  aspects  of  this  study  was  that  none  of  the  bats 
banded  in  Arizona  was  recovered  in  California  or  in  the  Oklahoma- 
Texas  region.  Further,  none  of  the  bats  banded  in  western  Arizona 
were  recovered  in  eastern  Arizona  ( or  vice  versa ) .  In  fact,  only  one 
of  the  bats  banded  in  western  Arizona  was  retaken  at  other  than  the 
point  of  banding  and  it  demonstrated  only  a  local  movement.  These 
facts,  coupled  with  an  analysis  of  the  Hterature  reports  of  the 
seasonal  distribution  of  populations  (Villa-R.  and  Cockrum,  1962), 
and  the  banding  results  reported  by  others  leads  to  the  conclusion 
that  four  or  more  behaviorally  (and  possibly  genetically)  separate 
populations  of  Tadarida  hrasiliensis  mexicana  occur  in  the  western 
United  States  during  the  summer  months  (see  map.  Fig.  8). 


Fig.  8.  Map  showing  approximate  limits  of  behaviorally  (and  probably 
genetically)  separate  groups  of  Tadarida  hrasiliensis  mexicana  (Saussure)  in 
the  southwestern  United  States.  Group  A  performs  only  local  movement; 
group  B  migrates  only  a  short  distance;  groups  C  and  D  migrate,  in  separate 
flyways,  distances  in  excess  of  1000  miles.    See  text  for  discussion. 


CocKRUM — Migration  of  Tadarida  325 

Group  A,  inhabiting  southern  Oregon  and  CaHfornia,  probably 
is  composed  of  resident  populations.  They  appear  to  perform  only 
local  seasonal  movements  as  argued  by  Benson  ( 1947 ) .  At  that  time, 
Christensen  ( 1947 )  had  concluded,  on  the  basis  of  observations  at 
Carlsbad  Ca\'erns,  New  Mexico,  that  T.  h.  mexicana  was  migratory. 
Benson  argued,  on  the  basis  of  obser\'ations  in  California,  that 
mexicana  probably  was  not  migratory  and  stressed  the  necessity 
for  banding  studies  to  clarify  the  situation.  At  this  point  it  appears 
that  the  two  groups  of  Tadarida  beha\'e  differently.  California 
populations  appear  to  make  only  local  seasonal  movements;  Carls- 
bad populations  do  appear  to  perform  long  movements  (Constan- 
tine,  1967). 

Group  B,  inhabiting  western  Arizona,  southern  Nevada,  and  at 
least  southeastern  California  along  the  Colorado  River  probably 
does  migrate,  but  not  southward  into  Sonora  and  Sinaloa.  Perhaps 
these  bats  mo\e  southward  into  Baja  California,  or  perhaps  west- 
ward into  the  low  interior  \'alleys  of  southern  California.  In  any 
case,  they  appear  to  be  absent  in  Ne\'ada,  Arizona,  and  extreme 
southeastern  California  during  the  winter  months  and  none  has  been 
recovered  in  Sonora  or  Sinaloa.  Extensive  field  work,  especially  in 
the  winter  months,  is  needed  to  clarify  the  status  of  these  popula- 
tions. 

Group  C,  inhabiting  central  and  eastern  Arizona  as  well  as 
western  New  Mexico,  appears  to  have  a  well-developed  fly- way 
through  Sonora  and  Sinaloa,  west  of  the  Sierra  Madres.  A  map  ( Fig. 
8 )  combining  all  of  the  moxements  found  in  this  study  demonstrates 
many  of  the  features  of  this  fly-way. 

Group  D  involves  the  populations  of  central  and  eastern  New 
Mexico,  western  Kansas,  Nebraska,  Texas,  and  adjacent  portions  of 
Mexico.  Constantine  (1967)  reported  movements  to  and  from 
Carlsbad  Caverns,  New  Mexico.  With  few  exceptions,  these  in- 
volved movements  within  the  area  ascribed  here  to  Group  D.  None 
of  the  exceptions  ( Eagle  Creek  Cave,  Arizona,  and  Carbo,  Sonora, 
as  also  reported  above  in  this  paper)  in\'olved  movements  in  the 
same  season.  The  findings  of  Da\'is  et  ah  ( 1962:320)  further  support 
this  concept  as  do  the  findings  of  Glass  ( 1958,  1959)  for  bats  banded 
in  Oklahoma. 

Literature  Cited 

Bexsox,  S.  B. 

1947.     Comments  on  migration   and   hibernation   in  Tadarida   brasiliensis. 
Jour.  Mamm.,  28:407-408. 
Christexsex,  E. 

1947.     Migration    or    hibernation    of    Tadarida    mexicana.     Jour.    Mamm., 
28:59-60. 


326 


Misc.  PuBL.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


COXSTAXTIXE,    D.    G. 

1958.     An    automatic   bat-collecting  device.     J.    Wildlife    Mgt.,    22:17-22, 

4  figs. 
1967.     Acti\ity  patterns  of  the  Mexican  free-tailed  bat.    Univ.  New  Mexico 

Publ.  Biol.,  7:1-79,  11  figs. 
D.wis,  R. 

1966.     Homing  performance  and  homing  ability  in  bats.    Ecol.   Monogr., 

36:201-237,  4  figs. 

Davis,  R.  B.,  C.  F.  Herreid,  II,  and  H.  L.  Short 

1962.  Mexican  free-tailed  bats  in  Texas.  Ecol.  Monogr.,  32:311-346,  11 
figs. 

Eads,  R.  B.,  J.  S.  WisEMAX,  AXD  G.  C.  Mexzies 

1957.  Observations  concerning  the  Mexican  free-tailed  bat,  Tadarida 
mexicana,  in  Texas.    Texas  Jour.  Sci.,  9:227-242,  4  figs. 

Glass,  B.  P. 

1958.  Returns  of  Mexican  free-tailed  bats  banded  in  Oklahoma.  Jour. 
Mamm.,  .39:4.35-4.37,  1  fig. 

1959.  Additional  returns  from  free-tailed  bats  banded  in  Oklahoma.  Jour. 
Mamm.,  40:542-545,  4  figs. 

Hall,  E.  R.,  axd  K.  R.  Kelsox 

1959.  The  mammals  of  North  America.  Ronald  Press  Co.,  New  York, 
l:x.xx  +  1-546  +  7.9,  illustrated. 

Herreid,  C.  F.,  II,  R.  B.  Davis,  axd  H.  L.  Short 

1960.  Injuries  due  to  l^at  banding.    Jour.  Mamm.,  41:398-400. 
Mitchell,  H.  A. 

1963.  Respiratory  physiology  and  anatomy  in  certain  members  of  the 
Chiropteran  family  Molossidae.  Ph.D.  dissertation.  University  of 
Arizona,  vi  +  124  pp. 

Schwartz,  A. 

1955.  The  status  of  the  species  of  the  hrasiliensis  group  of  the  genus 
Tadarida.    Jour.  Mamm.,  36:106-109. 

Villa  R.,  B.,  axd  E.  L.  Cockrum 

1962.  Migration  in  the  guano  bat,  Tadarida  hrasiliensis  mexicana  (Saus- 
sure).   Jour.  Mamm.,  43:43-64,  1  fig. 


Appendix  1 

Summary   of   Tadarida   hrasiliensis  b.a.xded,    1952-1965.     Locality   numbers 
correspoxd  to  the  numbers  on  figure   1  AXD  to  the  detailed  locality 

data  given  in  Appendix  2. 


Locality 


Type  of  habitat 


Date 


Males 

Females 

0 

1 

0 

8 

1 

1 

0 

2 

1 

0 

7 

0 

4 

0 

0 

1 

0 

1 

4 

2 

80 

1.57 

0 

119 

740 

826 

0 

400 

Arizona:    Mohave  County 

1.  Red  Lake  Water  hole 

2.  5  mi.  NE  Topock  Sink  hole 

3.  Jim  Kane  Mine  Mine 

4.  1  mi.  N  Kingman  Water  tank 


5.  SE  Kingman  Water  tank 

6.  SSE  Kingman  Water  tank 

7.  Da\is  Dam  Crevices 

8.  Powerline  Ca\e  Ca\"e 


July  8,  1963 
May  13,  1961 
Septeml^er  9.  1961 
lulv  15,  19.59 
"October  17,  1961 
April  18,  1962 
September  26,  1961 
lulv  16,  1960 
August  8,  1961 
September  17,  1960 
August  16,  1960 
lune  23,  1961 
lulv  25,  1962 
July  2,  1963 


CocKRUM — Migration  of  Tad  arid  a 


327 


Locality 

Type  of  habitat 

Date 

Males 

Females 

9.  Burro  Creek 

Mine  tunnel 

March  27,  1960 

0 

3 

April  19,  1960 

657 

344 

Tune  6,  1961 

2 

0 

April  25,  1962 

3 

0 

10.  Lower  Burro  Creek 

Mine  tunnel 

April  28,  1962 

1 

0 

11.  Alamo  Crossing 

Water  tank 

July  30,  1963 

6 

169 

Arizona:    Yuma  County 

12.  Kofa  Came  Refuge 

Water  tank 

April  23,  1959 

10 

0 

Arizona:    Ya\apai  County 

13.  Perkinsville 

Railroad  bridge 

July  13,  1963 

3 

680 

Arizona:    Graham  County 

14.  Bridge  no.  1 

Highway  bridge 

March  11,  1961 

1 

4 

15.  Bridge  no.  2 

High\va>'  bridge 

March  11,  1961 

1 

1 

April  20,  1963 

10 

0 

16.  Railroad  bridge 

Railroad  bridge 

March  11,  1961 

56 

30 

17.  Bridge  no.  5 

Highway  bridge 

March  11,  1961 

4 

3 

March  17,  1961 

1 

1 

18.  Bridge  no.  7 

Highway  bridge 

March  11,  1961 

9 

14 

March  18,  1961 

1 

1 

February  25,  1961 

2 

1 

19.  Bridge  no.  9 

Highway  bridge 

March  11,  1961 

242 

161 

March  17,  1961 

83 

13 

September  10,  1961 

0 

62 

April  20,  1963 

138 

113 

20.   Bridge  no.  10 

Highway  bridge 

August  9,  1961 

0 

41 

21.  Bridge  no.  22 

Highway  bridge 

April  20,  1963 

10 

10 

22.  Safford 

Building 

June  21,  1959 

1 

0 

Arizona:    Creenlee  County 

23.  Pump  station 

Eagle  Creek 

August  18,  1958 

2 

2 

24.  Eagle  Creek 

Cave 

August  19,  1958 

26 

100 

[une  28,  1959 

28 

1710 

Tune  24,  1961 

2900 

12450 

Tune  23,  1962 

3198 

19964 

Tune  5,  1963 

3800 

33700 

June  2,  1964 

750 

8400 

June  27,  1964 

200 

1000 

25.  Plantsite,  Morenci 

Railroad  tunnel 

August  20,  1958 

7 

210 

June  13,  1959 

5 

95 

Arizona:    Pinal  County 

26.  Picacho  3-hole 

Mine  tunnel 

August  25,  1955 

98 

284 

May  6,  1958 

0 

7 

May  21,  1960 

0 

1 

27.  Picacho  drive-in 

Mine  tunnel 

April  29,  1955 

0 

1 

August  25,  1955 

3 

36 

October  9,  1955 

9 

13 

October  16,  1955 

0 

1 

May  6,  1958 

1 

1 

Mav  3,  1959 

0 

6 

October  3,  1959 

3 

6 

August  28,  1960 

0 

1 

September  18,  1960 

0 

2 

Arizona:    Pima  Counts' 

28.  SiKerbell 

Mine  tunnel 

May  10,  1963 

380 

2903 

June  10,  1963 

133 

1000 

May  1,  1964 

20 

100 

328 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Locality 

Type  of  habitat 

Date 

Males 

Females 

29. 

Rillito 

Irrigation  ditch 

July  14,  1958 

3 

0 

30. 

Sabino  Canyon 

Stream 

Julv  21,  1958 

9 

0 

June  2,  1961 

1 

0 

June  4,  1961 

3 

0 

June  8,  1961 

1 

1 

June  30,  1961 

5 

0 

July  12,  1961 

1 

0 

31. 

Caiion  del  Oro 

Highway  bridge 

May  3,  1959 

1 

9 

May  30,  1959 

3 

22 

April  9,  1960 

1 

0 

June  18,  1960 

2 

June  19,  1960 

0 

September  24,  1960 

0 

October  1,  1960 

0 

October  14,  1960 

0 

32. 

NE  Tucson 

Pond 

April  17,  19.54 

10 

0 

April  20,  1954 

0 

8 

April  22,  1954 

9 

0 

33. 

Tucson 

Building 

November  4,  1953 

13 

12 

34. 

University  of 

Building 

June  29,  1955 

6 

39 

Arizona  campus 

July  1,  1955 

6 

21 

July  24,  1955 

3 

0 

March  11,  1960 

1 

0 

35. 

Gravel  pit,  Tucson 

Water  hole 

Tune  10,  1958 

0 

6 

June  12,  1958 

15 

8 

June  13,  1958 

22 

25 

June  16,  1958 

11 

4 

"Tune  24,  1958 

1 

4 

Tune  27,  1958 

2 

3 

July  10,  1958 

1 

1 

.36. 

Valencia  Road, 

Water  hole 

May  12,  1958 

2 

2 

Tucson 

May  21,  19.58 

0 

1 

37. 

Beehive  Mountain 

Mine  tunnel 

April  11,  19.55 

26 

0 

September  6,  1955 

38 

140 

May  12,  1956 

17 

89 

May  13,  1956 

8 

21 

May  19,  1956 

0 

September  27,  1956 

8 

13 

Octolier  2,  1956 

4 

2 

September  24,  1957 

3 

8 

October  6,  1957 

91 

99 

April  24,  1958 

6 

22 

April  27,  1958 

1 

4 

April  29,  1958 

4 

7 

May  2,  1958 

7 

15 

May  4,  1958 

0 

6 

May  7,  1958 

0 

3 

May  11,  1958 

0 

1 

May  14,  1958 

0 

3 

August  16,  19.58 

2 

3 

August  24,  1958 

1 

38 

September  20,  1958 

0 

2 

October  5,  1958 

0 

1 

May  3,  1958 

0 

4 

May  30,  19,58 

1 

32 

September  10,  1959 

0 

2 

September  20,  1959 

5 

36 

38. 

San  Xavier  Mission 

Building 

May  2,  1954 

0 

1 

39. 

Saguaro  Nat'l 
Monument 

Cave 

March  7,  1960 

39 

0 

CocKRUM — Migration  of  Tad  arid  a 


329 


Locality 


Type  of  habitat 


Date 


Males    Females 


40.  Helmet  Peak 


Mine 


41.  Twin  Bnttes 

42,  Continental 


43.  Madera  Canvon 


Mine 

Railroad  bridge 


Water  hole 


44.  Fresnal  Canyon  Water  hole 

45.  NW  Sasabe  Water  hole 

46.  Garcia's  Represso       Water  hole 


Arizona:    Santa  Cruz  County 

47.  Lochiel  Water  hole 

48.  Patagonia  Water  hole 

49.  Hayward's  bridge      Bridge 


50.  Davis'  bridge 


Biidge 


Arizona:    Cochise  County 

51.  Canello  Water  hole 

52.  Benson  Highway  bridge 


53.  St.  Da\id 


Highway  bridge 


October  1,  1952 
March  28,  1953 
April  9,  1953 
April  17,  1953 
April  24,  1953 
Ma>  10,  1953 
August  3,  1953 
April  24,  1954 
June  26,  1954 
April  11,  1955 
April  30,  1955 
May  1,  1955 
Mav  2,  1955 
April  17,  1953 
May  3,  1963 
Mav  16,  1963 
Tune  5,  1963 
Tune  11,  1963 
May  23,  1958 
Mav  26,  1958 
April  2,  1959 
August  1,  1961 
August  2,  1961 
June  10,  1962 
Julv  15,  1958 
July  24,  1958 


June  6,  1958 
Tune  8,  1961 
"March  27,  1960 
September  6,  1960 
September  11,  1960 
August  28,  1961 
March  27,  1960 
April  3,  1960 
October  15,  1960 
November  6,  1960 
April  15,  1961 
April  3,  1963 


June  17,  1960 
September  25,  1961 
Tune  1,  1963 
tune  1,  1959 
Tune  22,  1959 
June  24,  1959 
July  8,  1959 
Tulv  24,  1959 
March  27,  1960 
April  3,  1960 
May  28,  1960 
Tuly  3,  1960 
July  4,  1960 
Tuly  17,  1960 
Tune  23,  1960 
August  6.  1960 
August  18,  1960 
August  22,  1960 


1 

1 

0 

1 

1 

0 

1 

18 

10 

9 

0 

31 

0 

1 

3 

4 

1 

9 

1 

9 

0 

1 

1 

3 

2 

5 

0 

1 

78 

441 

43 

476 

100 

5 

0 

124 

6 

0 

0 

1 

17 

1 

0 

5 

1 

1 

1 

2 

3 

6 

0 

2 

0 

2 

1 

1 

3 

3 

22 

16 

0 

1 

2 

2 

3 

3 

7 

18 

1 

0 

0 

1 

4 

6 

2 

1 

0 

2 

0 

2 

10 

98 

41 

146 

160 

160 

14 

17 

3 

4 

21 

74 

6 

8 

2 

0 

12 

70 

3 

4 

1 

1 

0 

2 

0 

4 

0 

1 

1 

3 

2 

1 

330 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Locality 

Type  of  habitat 

Date 

Males 

Females 

August  27,  1960 

2 

32 

September  3,  1960 

1 

10 

October  8,  1960 

0 

1 

April  15,  1961 

0 

1 

May  7,  1961 

0 

3 

September  10,  1961 

0 

1 

June  13,  1962 

0 

8 

54. 

St.  David 

Highway  bridge 

March  27,  1960 

1 

0 

May  28,  1960 

1 

0 

55. 

Southwestern 

Swimming  pool 

May,  1955  (2) 

13 

3 

Research  Station 

Tune,  1955  (56) 

56 

28 

July,  1955  ( 1 ) 

1 

0 

August,  1955  (6) 

34 

7 

September,  1955  (9) 

86 

86 

October,  1955  (23) 

101 

78 

November,  1955  (2) 

2 

3 

March,  1956  ( 1 ) 

1 

1 

April,  1956  (4) 

4 

0 

May,  1956  (12) 

21 

5 

Tune,  1956  (13) 

33 

14 

Tuly,  1956  (10) 

15 

1 

August,  1956  (3) 

5 

0 

September,  1956  (19)  35 

23 

October,  1956  (12) 

25 

32 

November,  1956  (4) 

7 

5 

February,  1957  (3) 

4 

4 

March,  1957  (6) 

2 

7 

Tune,  1958  (6) 

12 

4 

Tuly.  1958  (9) 

39 

3 

September,  1958  (3) 

9 

7 

October  15,  1958  (3 

2 

0 

November  30,  1958 

1 

0 

Tune  27,  1959 

12 

3 

Tuly  30,  1959 

47 

0 

May  24,  1963 

2 

0 

March  28,  1964 

0 

1 

56. 

Portal 

Water  hole 

Tuly  2,  1958 

1 

0 

July  3,  1958 

2 

0 

57. 

South  Ca\'e  Creek 

Stream 

Tune  2,  1955 

1 

0 

Tuly  11,  1958 

3 

1 

julv  21,  1958 

5 

1 

August  21,  1958 

0 

1 

58. 

Silver  Creek 

Highway  bridge 

May  20,  1961 

65 

238 

May  21,  1961 

1 

20 

May  22,  1961 

5 

38 

Mav  24,  1961 

2 

22 

Mav  27,  1961 

11 

43 

August  7,  1961 

35 

186 

September  10,  1961 

8 

194 

October  1,  1961 

33 

205 

Tune  17,  1962 

43 

197 

August  19,  1962 

1 

0 

May  24,  1963 

11 

11 

59. 

Bisbee 

Building 

Tuly  23,  1963 

131 

0 

60. 

10  mi.  E  Douglas 

Water  tank 

April  1,  1961 

1 

0 

New 

Mexico:    Valencia  County 

61. 

Calts  Ca\e 

Cave 

July  21,  1963              1714 

1137 

CocKRUM — Migration  of  Tad  arid  a 


331 


Locality 

Type  of  habitat 

Date 

Males 

Females 

New  Mexico:    Grant  County 

62.  Silver  City, 

no. 

1 

Building 

September  1,  1962 
September  5,  1962 
September  17,  1962 
September  19,  1962 

0 

15 
36 

8 

2 
12 
71 

6 

63.  Silver  City, 

no. 

2 

Building 

Septeml^er  8,  1961 

1 

0 

64.  Silver  City, 

no. 

3 

Building 

October  29,  1962 

64 

46 

October  30,  1962 
October  31,  1962 
November  6,  1962 

60 
5 

18 

38 
11 

7 

Mexico:    Sonora 

65.  Tajitos 

Mine 

June  28,  1963 
September  5,  1963 
October  5,  1963 

3 

863 
2 

4 

1300 

1 

66.  Carbo  (see  ' 

Table  5)  Cave 

Various 

19155 

21639 

67.  Gil  ay  mas 

Water  hole 

March  28,  1959 
March  31,  1959 
April  1,  1959 
April  19,  1960 
April  20,  1960 
Tuly  22,  1961 

2 
26 

2 

1 
13 

1 

0 
0 
4 
0 
21 
0 

68.  Matorrena 

Water  hole 

tuly  26,  1960 
Tulv  27,  1960 

0 
0 

105 
100 

69.   Mocteziima 

Water  hole 

August  5,  1961 

1 

4 

Mexico:    Sinaloa 

70.  Pericos 

Cave 

March  2,  1963 

7700 

3900 

Me.xico:    Chiapas 

71.  Comitan 

Cave 

Appendix 

June  23,  1964 
:  2 

7 

0 

Detailed    locality    data,    associated    species,    and    names    of    principal 
BANDERS  OF  Tadoiida  brasiliensis  reported  in  Appe.ndix  1. 

Arizona:    Molia\e  County 

1.  Mist  net  over  water  hole  in  Red  Lake  area,  about  35  mi.  N  Kingman; 
netted  July  8,  1963,  1  Tadarida  brasiliensis  and  1  PipistrcUus  hcspcrus 
(Bill  Musgrove). 

2.  Sink  hole,  5  mi.  NE  Topock,  near  old  railroad  bed;  sink  hole  opening 
about  30  feet  in  diameter,  40  feet  deep,  larger  at  bottom  than  top,  and 
with  15  foot  o\erhang;  May  13,  1961,  about  300  Tadarida  brasiliensis 
in  cre\ices;  July  31,  1961,  maternity  colony  of  400  to  500  T.  brasiliensis 
(Bill  Musgrove). 

3.  Jim  Kane  mine,  17  mi.  NW  Kingman,  short  tunnel  about  100  feet  deep; 
SeptemlDer  24,  1961,  about  40  Antrozous  pallidus  and  3  Tadarida 
brasiliensis  present  (Bill  Musgrove). 

4.  Mist  net  over  water  hole  near  rock  cliffs,  1  mi.  N  Kingman,  3400  ft.; 
July  15,  1959,  31  Pipistrellus  hesperus,  13  Eptesietis  fiiscus,  2  Antrozous 
pallidns,  and  2  T.  brasiliensis-,  October  17,  1961,  3  T.  brasiliensis,  7 
Myotis  californicus,  1  Antrozous  pallidus;  April  18,  1962,  7  T.  brasilien- 
sis, 1  Lasiurus  cinereus,  1  Pipistrellus  Iiesperus,  1  Antrozous  pallidus, 
1  Eptesicus  fuscus  (Bill  Musgrove). 

5.  Mist  net  o\er  water  hole,  1.5  mi.  SE  Kingman  about  3300  ft.;  Septem- 
ber 26,  1961,  2  Pipistrellus  hesperus,  4  T.  brasiliensis,  1  Myotis  cali- 
fornicus, 1  Myotis  thysanodes,  6  Antrozous  pallidus,  1  Tadarida  macro- 
tis  (Bill  Musgrove). 


332  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

6.  Mist  net  over  water  tank  at  windmill,  4.5  mi.  SSE  Kingman,  about  4000 
ft.;  nets  set  for  28  nights  between  July  11,  1959  and  July  6,  1962; 
only  3  T.  brasiliensis  taken  during  this  time,  but  1  Plecotus  iownsendii, 
13  Mijotis  suhulatus,  14  Mijotis  volans,  5  Mijotis  thijsanodes,  9  Mijotis 
evotis,  31  Myotis  californicus,  39  Anirozous  pallichis,  253  Eptesicus 
fusctis,  and  486  PipstreUus  hespenis  were  also  taken   (Bill  Musgrove). 

7.  Slits  and  crevices  in  Davis  Dam  across  Colorado  Ri\'er;  September  17, 
1960,  estimated  10,000  Myotis  yumanensis  and  10,000  T.  brasiliensis; 
April  9,  1961,  nothing;  April  15,  1962,  estimated  500  T.  brasiliensis 
and  3500  Myotis  yumanensis  (Bill  Musgrove). 

8.  Volcanic  tube  on  W  slope  Chemehue\'i  Mtns.,  8  mi.  E  Site  6  and 
62  mi.  SW  Kingman;  cave  75  by  100  feet,  ceiling  25  feet  high,  opening 
20  by  8  feet;  outside  temperatiue  86°F  on  August  16,  1960,  88°F  to 
108°F  on  July  2,  1963;  August  16,  1960,  estimated  2000  T.  brasiliensis; 
July  10,  1961,  about  8000  T.  brasiliensis;  April  15,  1961,  nothing; 
July  25,  1962,  about  10,000  T.  brasiliensis;  July  2,  1963,  about  8000 
T.   brasiliensis. 

9.  Mine  timnel,  about  3/2  mi.  E  Burro  Creek  bridge  on  highway  93;  March 
27,  1960,  about  500  T.  brasiliensis,  1  Macrotis  waterhousii;  April  19, 
1960,  about  3000  T.  brasiliensis,  25  Myotis  velifer,  1  Macrotus  ivater- 
Iwusii;  July  16,  1960,  793  Myotis  velifer,  7  Myotis  californicus;  August 
7,  1960,  1  Myotis  velifer;  May  17,  1961,  102  Myotis  velifer,  1  Plecotus 
toivnsendii;  April  28,  1962,  30  Plecotus  townsendii,  22  Myotis  velifer, 
3  Tadarida  brasiliensis  (Bill  Musgrove,  E.  L.  Cockrum). 

10.  Mine  timnel  near  Burro  Creek,  2/4  mi.  below  bridge  over  Kaiser  Spring 
Wash  on  highway  93;  this  tunnel  visited  10  times  between  May  17, 
1961  and  April  7,  1963;  Macrotus  waterhousii  and  Myotis  velifer  pres- 
ent in  numbers  but  only  1  T.  brasiliensis  was  seen  (April  28,  1962) 
(Bill  Musgrove). 

11.  Mist  net  over  water  hole  in  Bill  Williams  Ri\'er  at  Alamo  Cros.sing; 
July  30,  1963,  178  T.  brasiliensis,  6  T.  femorosacca,  1  Myotis  velifer, 
1  Myotis  yumanensis,  1  Antrozotis  pallidus,  1  Macrotus  waterhousii 
(Bill  Musgrove). 

Arizona:    Yuma  County 

12.  Mist  net  o\er  Horse  Tank,  Kofa  Game  Refuge;  April  28,  1959,  10  T. 
brasiliensis,  2  Eptesicus  fuscus,  1  Antrozous  pallidus  (E.  L.  Cockrum). 

Arizona:    Yavapai  County 

13.  Railroad  bridge,  '2  mi.  N  Perkinsville;  June  30,  1960,  maternity  colony 
estimated  at  3000  T.  brasiliensis  (Bruce  Hayward);  July  13,  1963, 
maternity  colony  estimated  at  3000  T.  brasiliensis  ( Robert  Schwab  and 
Russell  Davis). 

Arizona:    Craham  County 

14.-21.  Crevices  in  series  of  highway  bridges  on  old  U.S.  highway  80  east  of 
Coolidge  Dam;  (locality  14)  26.4  mi.  E  Coolidge  Dam,  (15)  4  mi.  E 
Coolidge  Dam,  (16)  just  50  yards  S  of  no.  15,  ( 17)  22.7  mi.  E  Coolidge 
Dam,  (18)  21.1  mi.  E  Coolidge  Dam,  (19)  18.1  mi.  E  Coolidge  Dam, 
(20)  16.8  mi.  E  Coolidge  Dam,  (21)  2.6  mi.  E  Coolidge  Dam;  these 
bridges  \isited  five  times  between  February  25,  1961  and  July  20,  1965. 
Antrozous  pallidus  present  in  warm  months,  T.  brasiliensis  present  in 
spring  and  autumn  (Russell  Da\'is,  Bruce  Hayward,  E.  L.  Cockrum). 

22.  Attic  abandoned  church,  3.7  mi.  S  Safford;  June  21,  1959,  1  Myotis 
velifer,  1  T.  brasiliensis  (G.  Bradshaw,  Bruce  Hayward,  E.  L.  Cockrmn). 

23.  Mist  net  over  Eagle  Creek,  1  mi.  N  Piunp  Station  and  ca.  5  mi.  W 
Morenci,  3600  ft.;  August  18,  1958,  4  T.  brasiliensis  (Alfred  Gardner, 
E.  L.  Cockrum). 

24.  Cave,  E  wall  of  Eagle  Creek  Canvon,  4.2  mi.  S  Pmnp  Station  and  4  mi. 
S  and  2.3  mi.  W  Morenci,  3500  ft.;  (E.  L.  Cockrum,  Wm.  J.  McCauley, 
Russell  Da\ds,  Bill  Musgro\e,  Bruce  Hayward,  Gordon  Bradshaw, 
Alfred  Gardner,  and  about  30  others — see  Table  5). 


CocKRUM — Migration  of  Tadarida  333 

25.  Abandoned  railroad  tunnel,  '2  mi.  S  of  East  Plant  Site,  Morenci;  August 
20,  1958,  about  500  T.  hra.siliensis-  June  13,  1959,  about  1000  T. 
brasiliensis;  June  21,   1959,  1   T.  brasiliensis,  5  Myotis  ijuinanemis. 

Arizona:    Pinal  County 

26.  Mine  tunnel,  SE  corner  Picacho  Mtn.  about  3  mi.  SE  Picacho,  ca.  2200 
ft.;  visited  33  times  bet\veen  April  16,  1955  and  Octolier  15,  1960;  up 
to  4000  Myotis  vclifer  in  summer,  up  to  500  T.  brasiliensis  in  August, 
up  to  100  Macrotus  waterhoiisii  (E.  L.  Cockrum,  Gordon  Bradshaw, 
Bruce  Hayward,  Alfred  Gardner,  Anthony  Ross,  and  others). 

27.  Group  of  three  mine  tunnels,  SW  corner  Picacho  Mtn.  about  2/2  mi. 
SSE  Picacho,  ca.  2200  ft.;  \isited  22  times  l:)etween  July  28,  1955  and 
July  31,  1960;  up  to  42  Antruzous  pallidus,  3000  Myotis  velifer,  15 
Macrotus  waterhoiisii,  400  Tadarida  hrasiliensis  (same  investigators  as 
under  no.  26). 

Arizona:    Pima  County 

28.  Mine  tunnel  and  associated  natural  cave  in  "Old  Silverbell"  mine; 
May  27,  1958,  300  Tadarida  brasiliensis;  luly  10,  1962,  se\eral  hundred 
T.  brasiliensis;  May  10,  1963,  20,000  T.  brasiliensis,  300  Myotis  velifer; 
June  10,  1963,  10,000  T.  brasiliensis,  3000  Myotis  velifer;  September  13, 
1963,  nothing;  May  1,  1964,  1000  T.  brasiliensis  (Gordon  Bradshaw, 
Russell  Davis,  Robert  Schwab). 

29.  Mist  net  over  irrigation  ditch,  1  mi.  S  and  1  mi.  W  Rillito;  July  14, 
1958,  1  Pipistrellus  hesperus,  3  Eptesicus  fiiscus,  3  Tadarida  brasiliensis, 
14  Tadarida  femorosacca,  2  Antrozous  pallidus  (E.  L.  Cockrum). 

30.  Mist  nets  over  various  water  holes  in  Sabino  Canyon;  netted  40  nights 
between  July  21,  1958,  and  August  24,  1962;  totals — 1  Choeronycteris 
mexicana,  1  Leptomjcteris  sanborni,  3  Myotis  californicus,  294  Pipi- 
strellus hesperus,  138  Eptesicus  fuscus,  1  Lasiurus  cinereiis,  1  Plecotus 
townsendii,  7  Antrozous  pallidus,  22  Tadarida  brasiliensis,  30  T.  femo- 
rosacca, 6  T.  macrotis,  8  Eumops  perotis  (Alfred  Gardner,  E.  L.  Cock- 
rum, Bruce  Hayward,  Thomas  J.  Co.x,  Stephen  Cross,  Bill  Musgrove). 

31.  Crevices  under  highway  bridge  over  Canon  del  Oro,  10  mi.  NNW 
Tucson;  visited  26  times  between  May  3,  1959,  and  August  5,  1962; 
in  May  to  September  up  to  50  Antrozous  pallidus;  in  May  and  June 
and  September  and  October  up  to  50  T.  brasiliensis  (E.  L.  Cockrum, 
Bruce  Hayward,  Gordon  Bradshaw,  Russell  Davis,  and  others). 

32.  Mist  net  o\'er  pond  2  mi.  E,  7  mi.  N  Tucson;  only  T.  brasiliensis  taken 
on  dates  indicated  (E.  L.  Cockrum). 

33.  Crevice  between  two  buildings,  downtown  Tucson;  November  4,  1953, 
about  50  T.  brasiliensis  (  Lee  Beatty ) . 

34.  In  crevices  in  Library  Building  and  Old  Main  Building  on  University 
of  Arizona  campus,  Tucson   (E.  L.  Cockrum). 

35.  Mist  net  over  water  hole  in  gravel  pit,  S  end  Cottonwood  Lane,  Tucson; 
netted  eight  nights  bet\veen  June  10,  1958,  and  July  10,  1958;  totals — 
1  Myotis  yumanensis,  8  Myotis  velifer,  57  Eptesicus  fuscus,  17  Antro- 
zous pallidus,  106  T.  brasiliensis,  37  T.  femorosacca  (Alfred  Gardner). 

36.  Mist  net  over  water  hole,  S  side  of  Valencia  Road,  just  west  of  Santa 
Cruz  River,  Tucson;  netted  si.x  nights  between  April  29,  1958,  and 
June  9,  1958;  totals — 14  Myotis  velifer,  15  Eptesicus  fuscus,  4  Lasiurus 
cinereiis,  1  Plecotus  townsendii,  5  Tadarida  brasiliensis  (Alfred  Gard- 
ner ) . 

37.  Mine  tunnel,  W  slope  Beehi\e  Mtn.,  Tucson  Mtns.;  ^•isited  69  times 
between  April  11,  1955,  and  September  23,  1961;  T.  brasiliensis  present 
April  and  May  and  August  to  October,  usually  less  than  100  but 
April  29,  1955,  about  10,000  were  present;  Myotis  velifer  present  about 
same  time  (E.  L.  Cockrum,  Alfred  Gardner,  Gordon  Bradshaw,  Bruce 
Hayward,  and  others). 


334  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

38.  Under  eaves  of  buildings  at  San  Xavier  Mission,  SSW  Tucson;  about 
50  r.  bicmliensis  present  (E.  L.  Cockrum). 

39.  Cave  in  Box  Canvon,  Saguaro  National  Monument;  August  5,  1958, 
2000  Myotis  velifer,  August  15,  1959,  2000  Myotis  velifer,  500  T. 
hrasilien.sis;  March  7,  1960,  50  T.  hiasiliensis;  May  11,  1960,  150 
Leptonycieris  sanhorni,  35  Myotis  velifer,  15  Plecotus  townsendii; 
August  24,  1960,  2  Leptonycieris  sanhorni,  3000  Myotis  velifer,  15 
Plecotus  townsendii,  500  T.  hrasiliensis  (Alfred  Gardner,  E.  L.  Cock- 
rum,  Bruce  Hayward). 

40.  Mine  tunnel,  S  side  Helmet  Peak,  18  mi.  S  and  7  mi.  W  Tucson, 
3500  ft.;  visited  97  times  bet\veen  October  1,  1952,  and  May  30,  1962; 
T.  hrasiliensis  (up  to  50)  March  to  June  and  August  to  October, 
Myotis  velifer  (up  to  3000)  April  to  October,  Macrotus  waterhoiisii 
(up  to  50),  Antrozous  pallidus  (night  roost)  April  to  October  (same 
investigators  as  listed  in  no.  37). 

41.  Mine  tunnel,  S  side  W  Butte  of  Twin  Buttes,  21  mi.  S  and  5/2  mi.  W 
Tucson,  3500  ft.;  visited  32  times  l^etween  April  17,  1963,  and  May  20, 
1962;  usually  nothing  or  only  10  to  20  Antrozous  pallidus  and  10  to 
20  Myotis  velifer  present;  only  1  T.  hrasiliensis  seen  (same  investigators 
as  no.  37). 

42.  Railroad  bridge,  4.5  mi.  S  Continental;  maternity  colony  for  T.  hrasili- 
ensis and  Eptesicus  fuscus;  July  8,  1962,  1000  T.  hrasiliensis,  60 
Eptesicus  fuscus;  August  8,  1962,  3000  T.  hrasiliensis,  40  Eptesicus 
fuscus;  August  15,  1962,  5000  T.  hrasiliensis;  March  31,  1963,  100 
T.  hrasiliensis,  1  Eptesicus  fuscus;  April  3,  1963,  15  T.  hrasiliensis; 
May  3,  1963,  530  T.  hrasiliensis,  16  Eptesicus  fuscus;  May  8,  1963, 
nothing;  June  5,  1963,  200  T.  hrasiliensis;  June  11,  1963,  500  T. 
hrasiliensis;  April  24,  1964,  none;  June  25,  1965,  300  T.  hrasiliensis, 
30  Eptesicus  fuscus  ( Russell  Davis,  Stephen  Cross,  Robert  Schwab, 
E.  L.  Cockrum). 

43.  Mist  net  over  pond,  mouth  Madera  Canyon,  4400  ft.;  May  23,  1958, 
1  Myotis  velifer,  4  Pipistrellus  hesperus,  6  Eptesicus  fuscus,  2  Lasiurus 
cinereus,  8  T.  hrasiliensis;  May  26,  1958,  4  Pipistrellus  hesperus,  12 
Eptesicus  fuscus,  1  Lasiurus  cinereus,  1  T.  hrasiliensis;  April  2,  1959,  1 
Eptesicus  fuscus,  18  T.  hrasiliensis;  April  10,  1959,  2  Myotis  velifer, 
1  Myotis  volans,  1  Pipistrellus  hesperus,  1  Eptesicus  fuscus,  3  Lasiurus 
cinereus,  1  Antrozous  pallidus,  6  Tadarida  hrasiliensis;  April  15,  1959, 
1  Eptesicus  fuscus,  1  Tadarida  hrasiliensis  (Alfred  Gardner,  E.  L. 
Cockrum ) . 

44.  Mist  net  over  pool  in  Fresnal  Canyon,  W  slope  Baboquivari  Mtns.; 
August  1,  1961,  7  Myotis  velifer,  8  Eptesicus  fuscus,  1  Antrozous 
pallidus,  5  T.  hrasiliensis;  August  2,  1961,  3  Myotis  velifer,  1  Myotis 
californicus,  6  Pipistrellus  hesperus,  12  Eptesicus  fuscus,  2  Tadarida 
hrasiliensis  (Thomas  Cox,  Jaime  Maya). 

45.  Mist  net  over  pond  in  Los  Encinas  Wash,  8  mi.  NW  Sasabe,  3950  ft.; 
5  T.  hrasiliensis,  8  T.  femorosacca,  2  Eumops  underwoodi. 

46.  Mist  net  over  Garcia's  Represso,  2  mi.  E  Sasabe;  July  15,  1958,  13 
Eptesicus  fuscus,  6  Antrozous  pallidus,  9  T.  hrasiliensis,  11  Eumops 
underwoodi;  July  24,  1958,  5  Myotis  velifer,  1  Pipistrellus  hesperus, 
1  Eptesicus  fuscus,  4  Antrozous  pallidus,  2  T.  hrasiliensis,  4  T.  femo- 
rosacca, 7  Eumops  underwoodi  (A.  L.  Gardner). 

Arizona:    Santa  Cruz  County 

47.  Mist  net  over  water  hole,  3  mi.  N  Lochiel,  June  6,  1958,  2  Myotis 
velifer,  4  Eptesicus  fuscus,  1  Lasiurus  cinereus,  2  T.  hrasiliensis  ( Alfred 
Gardner). 

48.  Mist  net  over  water  hole,  6.8  mi.  N  Patagonia;  June  7,  1961,  1  Myotis 
californicus,  6  Myotis  keenii,  1  Pipistrellus  hesperus,  4  Eptesicus  fuscus, 
1  Lasiurus  horealis,  1  Antrozous  pallidus,  2  T.  hrasiliensis  (E.  L. 
Cockrum  and  others). 


CocKRUM — Migration  of  Tadarida  335 

49.  "Hayward's  Bridge,"  crevices  in  bridge  on  old  U.S.  highway  89,  near 
U.S. -Mexico  border  at  Nogales;  visited  39  times  between  March  27, 
1960,  and  April  3,  1963;  Myotis  velifei;  Eptesicus  ftiscus,  and  T. 
hrasiliensis  (Bruce  Hayward,  Russell  Davis,  Cordon  Bradshaw,  E.  L. 
Cockruni  and  others ) . 

50.  "Davis  Bridge,"  crevices  in  old  highway  82  bridge  across  Santa  Cruz 
R,  NE  Nogales;  visited  21  times  between  March  20,  1960,  and  April 
29,  1964;  Myotis  velifer,  Eptesicus  ftiscus,  Antrozous  pallidus,  and 
T.  hrasiliensis  (same  investigators  as  no.  49). 

Arizona:    Cochise  County 

51.  Mist  net  over  stock  pond,  2.2  mi.  SE  Canello,  5000  ft.;  3  Myotis 
velifer,  1  Myotis  keenii,  3  Eptesicus  ftiscus,  3  Lasiurtis  cinereiis,  1  An- 
trozous pallidus,  2  T.  hrasiliensis  (Bruce  Hayward). 

52.  Whetstone  Overpass  of  U.S.  80  over  railroad,  about  1.6  mi.  W  Benson, 
4000  ft.;  luly  18,  1961,  about  500  T.  hrasiliensis;  September  25,  1961, 
200  T.  hrasiliensis,  10  Antrozotis  pallidus;  June  1,  1963,  125  T.  hrasili- 
ensis, 10  Antrozous  pallidus  (Russell  Davis,  Robert  Schwal>). 

53.  Crevices  in  highway  bridge,  S  St.  David,  3800  ft.;  visited  56  times 
between  June  1,  1959,  and  May  24,  1963;  Eptesicus  ftiscus  (up  to 
110),  Antrozous  pallidus  (up  to  150)  and  T.  hrasiliensis  (up  to  360) 
(mainly  Russell  Davis,  Init  also  E.  L.  Cockrum,  Bill  Musgrove,  Bruce 
Hayward,  Gordon  Bradshaw  and  others). 

54.  Crevices  in  highway  bridge,  S  St.  David,  4000  ft.;  visited  33  times  be- 
hveen  March  27,  1960,  and  May  24,  1963;  Mtjotis  velifer  ( 1 ),  Eptesictis 
ftiscus  (up  to  75),  Antrozous  pallidus  (up  to  45),  T.  hrasiliensis  (up 
to  3)   (same  investigators  as  in  no  53). 

55.  Mist  net  over  swimming  pool,  5360  ft.;  Southwestern  Research  Station, 
Cave  Creek  Canyon,  Chiricahua  Mtns.;  netted  315  nights  between 
May  29.  1955,  and  March  28,  1964  (mainly  by  Ellen  Ordway,  but 
many  others  involved). 

56.  Mist  net  over  Represso,  1  mi.  WNW  Portal,  4900  ft.;  netted  three 
nights,  total — 1  Myotis  californictis,  2  Myotis  volans,  2  Myotis  keenii, 
32  Pipistrellus  hesperus,  13  Eptesicus  ftiscus,  2  Lasiurus  cinereus,  3 
Plecotus  phtjllotis,  1  Plecotus  townsendii,  3  T.  hrasiliensis  (Larry  Com- 
missaris ) . 

57.  Mist  net  over  pools  in  S  fork  Cave  Creek  5400  ft.;  Chiricahua  Mtns.; 
netted  28  nights  between  June  2,  1955,  and  August  21,  1958;  totals — 
16  Choeromjcteris  mexicana,  34  Mtjotis  volans,  3  Myotis  thysanodes, 
50  Myotis  evotis,  39  Myotis  calif  amicus,  3  Myotis  suhulatus,  117  Pipi- 
strellus hesperus,  81  Eptesicus  fusctis,  2  Lasionycteris  noctivagans,  17 
Lasiurus  cineretis,  2  Lasiurus  horealis,  18  Plecotus  townsendii,  12  Pleco- 
tus phtjllotis,  46  Atitrozous  pallidus,  12  T.  hrasiliensis  (Larry  Com- 
misaris,  Ellen  Ordway,  E.  L.  Cockrum,  and  others). 

58.  Crevices  in  highway  bridge  over  Silver  Creek,  on  U.S.  80,  NE  Douglas, 
4500  ft.;  visited  27  times  between  April  1,  1961,  and  May  24,  1963; 
up  to  60  Myotis  velifer,  1  Eptesicus  ftisctis  (1  time  only),  30  Antrozous 
pallidus,  350  T.  hrasiliensis  (Russell  Davis). 

59.  Attic  of  St.  Patrick  Church,  Bisbee,  5300  ft.;  July  23,  1963,  about  200 
Eptesicus  fuscus,  1300  T.  hrasiliensis  (E.  L.  Cockrum). 

60.  Mist  net  over  concrete  cattle  tank,  10  mi.  E  Douglas,  on  Cuadalupe 
Canyon  road,  4500  ft.;  April  1,  1961,  1  Plecotus  townsendii,  1  T.  hra- 
siliensis (  Russell  Davis ) . 

New  Mexico:    Valencia  Co. 

61.  Calts  Cave,  about  7200  ft.;  about  15  mi.  S  San  Rafael;  June  24,  1963, 
5000  to  10,000  r.  hrasiliensis;  July  21,  1963,  5000  T.  hrasiliensis 
(  Robert  Schwali ) . 


336  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

New  Mexico:    Grant  County 

62.-64.   In  crevices  in  attic  of  buildings  in  Silver  City,  5900  ft.;  autumn  tran- 
sient colonies  of  T.  hrasiliensis  (  Bruce  Hayward ) . 

Mexico:    Sonora 

65.  Mine  tunnels  ( Mina  de  la  Virgen),  Tajitos;  June  13,  1963,  "several 
million"  T.  hrasiliensis;  June  28,  1963,  5000  Leptonycteris  sanborni, 
500  Macrotiis  waterhousii,  40  Antrozous  pallidus,  .500  Tadarida  hrasili- 
ensis; August  30,  1963,  20,000  Leptonycteris  sanhorni,  50,000  T.  hrasili- 
ensis; September  5,  1963,  50  T.  hrasiliensis;  October  .5,  1965,  10  Macro- 
tits  waterhousii,  50  Myotis  velifer,  5  Antrozous  pallidus,  1  Tadarida 
hrasiliensis  (Robert  Schwab,  E.  L.  Cockrum,  Russell  Davis,  Robert 
Baker ) . 

66.  Volcanic  tube,  la  Cueva  del  Tigre,  14.9  mi.  SE  Carbo;  visited  32 
times  between  November  12,  1967,  and  October  21,  1965  by  many 
different  persons  ( see  Table  7  for  results ) . 

67.  Mist  net  over  fresh  water  ponds  on  N  shore,  Bocachimpampa  Bay,  NW 
Guaymas,  20  ft.;  March  28,  1959,  1  Leptonycteris  sanhorni,  12  Macro- 
tus  waterhousii,  5  Pipistrellus  hesperus,  2  T.  hrasiliensis;  March  31, 
1959,  1  Lasiurus  ega,  26  T.  hrasiliensis,  1  T.  macrotis;  April  1,  1959, 
6  T.  hrasiliensis;  April  19,  1960,  10  Macrotus  waterhousii,  10  Pipistrel- 
lus hesperus,  5  Lasiurus  ega,  1  T.  hrasiliensis;  April  20,  1960,  1  Lasiurus 
ega,  24  T.  hrasiliensis  (Alfred  Gardner,  Bruce  Hayward,  E.  L.  Cock- 
rum,  Anthony  Ross,  and  others). 

68.  Mist  net  over  water  hole,  2  km.  SW  Matorrena  ( NE  Guaymas);  July 
26,  1960,  4  Antrozous  pallidus,  108  T.  hrasiliensis,  1  T.  femorosacca, 
3  Eumops  underwoodi;  July  27,  1960,  5  Myotis  velifer,  1  Pipistrellus 
hesperus,  1  Antrozous  pallidus,  100  T.  hrasiliensis,  1  T.  femorosacca, 
3  Eumops  underwoodi  (Alfred  Gardner). 

69.  Mist  net  over  stock  tank,  11  mi.  W  Moctezuma,  5  T.  hrasiliensis 
(Thomas  Cox). 

Mexico:    Sinaloa 

70.  La  Chinacatua,  Cueva  Montelargo,  17  mi.  W  Pericos;  March  2,  1963, 
5000  Chilonycteris  psilotis,  10,000  Pteronotus  damji,  100,000  Tada- 
rida hrasiliensis  (E.  L.  Cockrum,  Wm.  J.  McCauley,  Alfred  Gardner, 
Russell  Davis,  and  others). 

Mexico:    Chiapas 

71.  Cueva  la  Trinitaria,  19  km.  SE  Comitan;  June  23,  1964,  40,000  T. 
hrasiliensis  (Alfred  Gardner). 


INTRASPECIFIC  POPULATION  STRUCTURE  OF 
THE  SPECIES  PAPPOGEOMYS  CASTANOPS 

BY 

Robert  J.  Russell 

Comparati\'ely  little  is  known  about  species  in  a  multidimen- 
sional system.  Complex  intraspecific  population  structure  and  rela- 
tionships have  not,  so  far  as  I  know,  been  described  in  detail  for 
any  mammal.  In  the  course  of  recent  taxonomic  studies  of  the 
pocket  gopher,  Pappogeomys  ccistanops,  I  became  interested  in  the 
intraspecific  relationship  of  populations  consisting  of  groups  of  sub- 
species. These  populations,  here  called  subspecies  clusters,  seen  to 
l)ehave  as  units,  and,  thereby,  play  an  important  role  in  the  evolu- 
tion and  difi^^erentiation  of  the  species.  I  was  especially  interested 
in  the  way  in  which  these  populations  were  distinguished  and  the 
way  in  which  they  originated.  The  complex  relationships  between 
them  in\'olve  the  essential  feature  of  populations  of  a  species  in  a 
multidimensional  system — interbreeding  and,  therefore,  gene  flow 
between  populations.  Restriction  of  gene  flow  results  in  divergence, 
and,  ultimately,  in  the  formation  of  separate  species.  A  range  of 
intraspecific  compatability  was  found  to  exist  between  the  popula- 
tions of  Pappogeomys  castanops. 

This  paper  is  dedicated  to  Professor  E.  Raymond  Hall,  who  I 
would  especially  like  to  thank  for  his  many  personal  favors  to  me 
over  the  past  years.  The  illustrations  were  made  by  Donna  Lynn 
Helbing,  under  my  direction,  with  financial  assistance  from  the 
Department  of  Biology,  Uni\ersity  of  Missouri-Kansas  City.  Special 
thanks  is  due  Maureen  Arnold,  who  assisted  with  clerical  duties. 

Analysis  of  Variation 

Intraspecific  variation  in  Pappogeomys  castanops  occurs  mainly 
in  external  and  cranial  dimensions.  Color  distinctions  have  devel- 
oped in  some  populations,  but  most  of  these  pocket  gophers  are 
characterized  by  approximately  the  same  range  of  variation  in 
pigmentation.  Differences  in  qualitative  morphological  traits  be- 
tween the  populations  were  not  obserxed.  On  the  basis  of  geo- 
graphic \ariation  in  quantitative  characters,  Russell  ( 1968:62L691) 
recognized  25  subspecies.  At  the  time  that  geographic  variation  was 
analyzed,  it  was  noted  that  the  subspecies  did  not  differ  from  each 
other  either  in  the  same  combination  of  characters  or  to  the  same 

(337) 


338  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

degree.  Thus,  the  variation  characterizing  the  species  was  found 
to  be  heterogeneously  distributed  throughout  the  range  of  P.  casta- 
nops.  Moreover,  critical  comparison  of  the  subspecies,  disclosed,  as 
expected,  that  geographical  adjacent  subspecies  closely  resembled 
one  another  morphologically,  and,  therefore,  could  be  grouped  into 
geographic  clusters.  Each  subspecies  cluster  differed  from  other 
such  clusters  in  the  same  traits  and  usually  to  the  same  degree  ( see 
Figs.  2-13). 

The  subspecies  cluster  is  here  recognized  as  a  secondary  division 
on  the  population  level  within  the  species.  Evidence  obtained  in 
this  study  suggests  that  the  subspecies  cluster  tends  to  be  perma- 
nent over  relatively  long  periods  of  time  and  that  it  evolves  as  a 
population  unit.  Therefore,  the  subspecies  cluster  has  great  sig- 
nificance as  a  unit  of  evolution,  especially  in  relation  to  divergent 
trends  within  the  species.  The  recognition  of  subspecies  clusters 
and  the  analysis  of  the  complex  relationships  between  them  provide 
a  better  understanding  of  intraspecific  trends  in  evolution. 

In  Pappogeomys  castanops,  six  subspecies  clusters  and  two  geo- 
graphic isolates  are  recognized  on  the  basis  of  distribution  of 
geographic  variation.  These  populations  have  been  given  letter 
designations  A  through  H.  The  geographic  distribution  of  the 
clusters  and  isolates  are  depicted  in  Figure  1,  and  include  the 
following  subspecies:  (A)  geographic  isolate,  P.  c.  parviceps; 
( B )  subspecies  cluster  including  P.  c.  consitus,  P.  c.  perexiguus,  and 
P.  c.  surculus;  (C)  subspecies  cluster  including  P.  c.  subnuhilus, 
P.  c.  elihatus,  P.  c.  planifrons,  and  P.  c.  peridoneus;  (D)  subspecies 
cluster  including  P.  c.  goldmani  and  P.  c.  rubellus;  (E)  subspecies 
cluster  including  P.  c.  excelsus,  P.  c.  subsimus,  P.  c.  jucundus,  P.  c. 
sordididus,  and  P.  c.  ustulatiis;  (F)  subspecies  cluster  including 
P.  c.  clarkii,  P.  c.  protensis,  P.  c.  perplanus,  P.  c.  simulans,  and  P.  c. 
castanops;  (G)  geographic  isolate,  P.  c.  hirtus;  (H)  subspecies  clus- 
ter including  P.  c.  buUatus,  P.  c.  tamaulipensis,  P.  c.  angusticeps, 
and  P.  c.  torridus. 

Several  of  the  subspecies  included  in  subspecies  clusters  may 
actually  be  geographic  isolates.  If  so,  they  evidently  have  only 
recently  become  isolated,  and  have  not  developed  significant  dis- 
tinctions from  the  other  taxa  in  the  subspecies  cluster  to  which  they 
are  assigned.  For  instance,  P.  c.  torridus  of  population  H  is  probably 
now  isolated  from  P.  c.  angusticeps  to  the  east,  and  angusticeps  is 
separated  from  P.  c.  bullatus  to  the  south  by  the  Rio  Grande.  The 
Rio  Grande  probably  serves  only  as  a  filter  barrier  for  these  pocket 


Russell — Population  Structure  of  Pappogeomys         339 


Fig.  1.  Distribution  of  intraspecific  populations  of  Pappogeomys  castanops 
depicting  subspecies  clusters  and  geographic  isolates  (A-H)  and  the  areas 
where   either   sympatry    (1-6)    or   secondary   intergradation    (7)    occurs.     For 

detailed  explanation,  see  text. 


340  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

gophers;  therefore,  gene  flow  is  only  reduced,  not  totally  restricted. 
In  population  F,  records  of  occurrence  indicate  that  P.  c.  clarkii 
probably  no  longer  has  contact  with  either  P.  c.  pratensis  to  the 
north  or  with  P.  c.  sorclidiihis  to  the  south  (a  subspecies  of  popula- 
tion E).  Also  in  population  E,  P.  c.  usftdatus  evidently  has  become 
isolated  recently  from  P.  c.  jiicundus  to  the  west. 

The  subspecies  that  comprise  any  one  of  the  population  clusters 
cannot  be  reduced  to  one  taxon  owing  to  the  wide  range  of  variation 
that  is  heterogeneously  distributed  among  the  included  subspecies. 
The  range  of  variation  characterizing  a  subspecies  cluster  yields 
high  coefficients  of  variation,  suggesting  that  the  sample  is  hetero- 
geneous. Since  variation  due  to  sex  and  age  has  been  eliminated, 
the  significantly  high  coefficiency  must  be  ascribed  to  taxonomic 
variation  of  the  geographic  races  that  make  up  the  population 
cluster.  It  should  be  pointed  out,  however,  that  neither  the  popula- 
tion clusters  nor  the  geographic  isolates  are  at  once  obvious.  Only 
after  careful  analysis  of  subspecific  variation  was  I  able  to  define 
these  population  units. 

Twelve  metric  characters  were  analyzed.  The  result  is  repre- 
sented by  histograms  in  Figures  2-13.  Ten  of  the  characters  are 
cranial  dimensions  and  two  are  external  dimensions. 

Condylohosal  length  (see  Figure  2). — Of  the  characters  studied, 
condylobasal  length  was  found  to  have  the  greatest  range  of  varia- 
tion. On  the  basis  of  this  character,  populations  of  Pappogeomys 
castanops  fall  into  two  distinct  and  sharply  defined  groups,  one 
characterized  by  short  skulls  including  populations  A,  B,  and  C,  and 
the  other  characterized  by  long  skulls  including  populations  E,  F, 
G,  and  H.  Population  D  is  intermediate,  both  morphologically  and 
geographically  (see  map,  Fig.  1),  between  the  two  divergent 
groups. 

Of  populations  A,  B,  and  C,  population  C,  although  it  occurs  in 
a  relatively  small  geographical  area  compared  with  B,  has  a  much 
greater  range  of  variation  in  condylobasal  length.  Also  population 
B,  a  subspecies  cluster,  yields  slightly  more  variation  than  popula- 
tion A,  an  isolated  subspecies. 

Comparison  of  populations  E,  F,  G,  and  H  reveals  that  both 
E  and  F  (both  subspecies  clusters)  are  about  equally  variable  and 
that  G,  a  single  isolated  species,  shows,  as  expected,  considerably 
less  variation  than  the  subspecies  clusters.  Of  these  populations,  the 
skull  of  H  is  significantly  shorter,  overlapping  the  range  of  variation 
developed  in  the  group  with  shorter  skulls  ( A,  B,  and  C ) . 


Russell — Population  Structure  of  Pappogeomys 


341 


I   I  I   I  I   I   I   M   I  I    I  I  I  I   I   I   I  I   I   I    I  I  I  I  I   I   I   I  I   I  I   I  I   I  I  I  I   I   I   I  I   I   I   I  I  I   I   I  I   I  I   I  I   I  I  I  I  I  I   I    I  I  I   I   I  ll   I  I    I  I  I  I  I 


_A 


_B 


CO 

o 

< 


Oh 


Dd    c£]  Dd 


D 
E 


DC2£]Unc5D 

ddUdcdd    eCis0 


F 

U6 


E^ 


o 


zLJa         a 


39     10     11     42    43     41     45     46    47     4S     49     50     51     52     53 

J  I  Ll  I  ij  lij.ij  Li.ij  liuj  1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1  Li^-ij_i-i-j_i  1 1 1 1 1 1 1 1 


Fig.  2.    Histograms  of  condylobasal  length   (mm.)   in  intraspecific  populations 

of  P.  castanops. 

The  highest  degree  of  distmction  in  condylobasal  length  of  E, 
F,  G,  and  H,  as  compared  with  A,  B,  and  C,  is  developed  in  popula- 
tion E.  Populations  F,  G,  and  H  show  progressively  less  distinction, 
although  all  but  H  fail  to  o\erlap  the  maximum  range  of  variation 
developed  in  A,  B,  and  C. 

LengtJ}  of  palate  (see  Figure  3). — Length  of  palate  is  separated 
into  two  distinct  groups — one  characterized  by  a  short  palate,  in- 
cluding populations  A,  B,  and  C,  and  one  characterized  by  a  long 
palate,  including  populations  E,  F,  G,  and  H.  Population  D  is 
characterized  by  a  range  of  variation  that  overlaps  both  the  maxi- 
mum \ariation  of  C  and  the  minimum  xariation  of  E,  and,  as  in 
condylobasal  length,  this  character  provides  dimensions  inter- 
mediate between  the  shortest  measurements  of  the  palate  in  popu- 
lation C  and  the  longest  in  E. 

Compared  with  populations  A,  B,  and  C,  population  E  is  the 
most  distinct.  Indeed,  there  is  no  o\  erlap  in  the  ranges  of  minimum- 
maximum  variation,  a  significant  fact  considering  that  the  geo- 
graphic ranges  of  B  and  C  are  contiguous  with  that  of  E.  Only 
slight    o\erlap    obtains    between    populations    A,    B,    and    C    and 


342  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

I  I  MM  I  I  I  I  I  I  I  I  I  I  I  I  I  M  I  I  I  I  I  I  i  II  I  |l  I  II  I  II  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I 


C/2 

Z 

o 

< 

o 

PL, 


B 

c  D 

D 


£ 


r-OrnJ 


Zl       D 


d 


I 1     I T 


23 


0 


2ZI  0    0 


e2 


1771  rr/i    n 


2B  27  28  29  30  31  32  33  34  35  35  37 

I  I  I  I  i  I  I  I  I  I  I  I  I  I  I  I  I  '  M  I  ll  I  I  I  I  II  I  I  I  II  I  I  I  I  I  I  ll  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I 


Fig.  3.    Histograms  of  length  of  palate   ( mm. )   in  intraspecific  populations  of 

P.  castanops. 

populations  F  and  G;  only  populations  A  and  B  are  in  geographic 
contact  with  F  and  G.  The  least  distinction  is  developed  in  popu- 
lation H  in  comparison  with  A,  B,  and  C.  In  this  case,  a  significant 
proportion  of  the  smaller  individuals  of  population  H  have  palates 
no  longer  than  those  at  the  maximum  range  of  ^'ariation  in  A,  B, 
and  C. 

Of  the  group  with  short  palates,  population  C  has  a  greater 
range  of  variation  than  does  either  A  or  B,  and,  of  the  group  with 
long  palates,  population  F  pro\'ides  the  greatest  range  of  variation. 
The  range  of  \'ariation  in  population  H  is  less  than  that  of  E  and  F, 
and  although  H  broadly  oxerlaps  F,  there  is  only  slight  overlap 
with  E. 

Palatofrontal  deptJi  of  skull  (see  Figure  4). — The  populations 
show  less  distinctiveness  in  palatofrontal  depth  than  they   do  in 


Russell — Population  Structure  of  Pappogeomys         343 

condylobasal  length  and  palatal  length;  the  same  pattern  of  varia- 
tion is  evident  but  with  less  of  a  degree  of  difference  and  with 
decidedly  more  overlap  between  populations.  Major  population 
groups  within  the  species  can  not  be  separated  readily  on  the  basis 
of  this  character. 


1 1 1 1  j  1 1 1 1 1 1 1 1 1 1 1  M  1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1  r 


Z 

o 
< 

D 

Ph 

O 

PL, 


czi  CD     n  n  a 


_=    D 


pJLnJ 


D     □ 


n=3 


D  D       O 


^ 


run  nz 


0 


0  EZ 


0E^ 


21 


0 


\Z7\\A     U\     \77\ 


IZl 


160        165       no 

I  I   I  M  I  I   I   I   I  I  I  I 


17  5  18  0  18  5  19  0  195  20  0  205  210 

I   I  I    M   I   I  II   I   1  I   I  I   I  I    I  II    II   I  II    I  I  I   I  I   I   I   I  I  I   I  I  I    III  I 


Fig.  4.    Histograms  of  palatofrontal  depth  of  skull  (mm.)  in  intraspecific  popu- 
lations of  F.  castanops. 

Even  so,  the  shallowest  skulls  are  found  in  populations  A,  B,  and 
especially  C,  and  the  deepest  skulls  in  populations  G  and  H,  and 
especially  E  and  F.  The  variations  of  populations  E  and  F  differ 
little  from  each  other  in  this  character;  both  are  significantly  larger 
than  population  H.  As  in  other  features,  the  greatest  degree  of 
difference  is  found  between  populations  C  and  E,  but,  in  this  case, 
populations  C  and  F  are  equally  distinct.  Geographically  and 
morphologically,  populations  E  and  C  are  linked  by  population  D, 
which  links  the  two  extremes.    The  greatest  range  of  variation  of 


344 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


1  I  I  I  I  I  I  I  I  I  M  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I   I  I  I  I  I  I  I  I  I  I  I   I  I  I 


B 
D 


1=]       CZl 


D 


dBD     D&3 


0  0 


o 

I— I 

D 
O 


24 


25 

,1 


0 


35 


J_LLLJJ  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I 


Fig.  5.    Histograms  of  greatest  zygomatic  breadth  (mm.)  in  intraspecific  popu- 
lations of  P.  castanops. 

any  single  population  is  developed  in  population  C.  Population  H 
more  broadly  o\  erlaps  A,  B,  and  C  than  any  of  the  other  populations 
with  deep  skulls. 

Zygomatic  breadth  (see  Figure  5). — Zygomatic  breadth,  meas- 
ured across  its  greatest  distance,  depicts  the  same  pattern  as  ob- 
served in  other  characters;  however,  the  degree  of  o\'erlap  between 
population  H  and  populations  A,  B,  and  C  is  greater  than  in  either 
condylobasal  length,  length  of  palate,  or  palatof rental  depth.  Re- 
gardless, the  narrowest  skulls  are  developed  in  populations  A,  B, 
and  especially  C,  and  the  broadest  in  populations  E,  F,  and  G,  and 
to  a  lesser  degree  H.  Population  D  provides  an  intergrading  series 
between  E  and  C,  both  of  which  overlap  slightly  in  their  ranges  of 
variation  in  this  trait. 

Zygomatic  breadth  is  especially  useful  in  separating  popidations 
A  and  B  from  populations  F  and  G.  All  foiu-  occur  in  geographically 
adjacent  regions  ( see  map.  Fig.  1 ) .  Population  B  also  occurs  adja- 
cent to  the  range  of  population  C,  and  both  are  clearly  distinguished 


Russell — Population  Structure  of  Pappogeomys 


345 


in  this  character.  The  greatest  range  of  variation  was  observed  in 
population  C,  in  which  the  range  of  variation  exceeds  both  the 
minimum  and  maximum  \ariation  of  populations  A  and  B. 

Squamosal  Jjreadth  (see  Figure  6). — The  distinctiveness  of  the 
populations  is  not  so  clearly  defined  in  squamosal  breadth  as  it  is 
by  zygomatic  breadth,  but  the  same  pattern  of  population  variation 
emerges.  Indeed,  the  patterns  between  these  t\\'o  measurements  of 
cranial  breadth  are  quite  similar.  The  narrowest  skulls  are  charac- 
teristic of  populations  A,  B,  and  C,  and  the  broadest  skulls  are  found 
in  populations  E,  F,  G,  and  H.  The  greatest  extremes  in  minimum- 
maximum  breadth,  as  in  zygomatic  breadth,  are  developed  in 
populations  C,  E,  and  F,  respectively.  As  in  other  dimensions,  the 
two  groups  are  connected  by  the  intermediate  and  overlapping 
range  of  variation  found  in  population  D. 

The  overall  trend  in  dixergcnce  is  apparent  in  Figure  6;  how- 
ever, squamosal  breadth  is  useful  taxonomically  only  in  separating 


en 
Z 

o 

< 

Ph 

C 
P-( 


I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  M  I  I  I  I  I  I  I  I  M  I  I  I 


_A 

_B  : 

.C  DD     [HH 

.D 

_E 

.F 
.G 

_H 

22  23  24 


^ 


!£Lo 


D    n 


D  Em 


D  tba 


0  0     Z3 


^^^ 


J_L 


I  I  1  I  I  I 


25  26  27  28  29  30  31  32 

I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  1  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I 


Fig.   6.    Histograms  ot  squamosal  breadth    ( mm. )    in  intraspecific  populations 

of  P.  castanops. 


346 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


population  E  from  A,  B,  and  C;  only  a  slight  overlap  in  minimum- 
maximum  dimensions  occurs  (with  C)  between  these  particular 
populations.  The  distinctions  between  populations  F  and  H  are 
less  developed  in  this  character  and  the  minimum  of  the  ranges  of 
variation  in  both  populations  broadly  overlap  the  maximum  ranges 
of  variation  of  populations  A,  B,  and  C. 

The  highest  degree  of  variation  obtained  in  any  one  population, 
once  again,  was  found  in  population  C,  and,  as  in  zygomatic 
breadth,  the  range  of  variation  in  C  exceeds  both  the  minimum 
and  maximum  variation  in  A  and  B. 


1 1  I  I  I  1 1  1 1 1  I  I  I  I  I  I  I  1 1  I  1 1  I  1 1  1 1  I 


n  czi     U 


TT 


TT 


c/3 
Z 

o 

H 
< 

o 

(In 


cD 


n 


D 


en 


I 1 


I 1 


UQ 


0  0    0  0 


I 


jm^Mt^ 


m 


0 


0  0 


J 


0  dJ  a 


^Mn 


12  13  14  15  II 

I   M  M   I  I  I   I  I  I   I  I  I   I  I  I  I   I  I  I 


i6  17  1 

I  I  I  M  I  I  I 


8  1|9  20  2,1 

I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I 


Fig.  7.    Histograms  of  length  of  nasals   ( mm. )   in  intraspecific  populations  of 

P.  castanops. 


Russell — Population  Structure  of  Pappogeomys         347 

LengtJi  of  nasals  (see  Figure  7). — The  pattern  of  population 
xariation  noted  in  the  preceding  characters  is  weakly  developed  in 
length  of  nasals.  As  usual,  the  shortest  nasals  are  developed  in 
populations  A,  B,  and  C,  and  the  longest  nasals  are  in  E,  F,  G,  and 
H,  but  the  degree  of  overlap  is  more  extensively  developed  than 
usual.  Population  D  provides  an  intermediate  range  of  variation 
between  the  two  groups. 

The  distinction  between  the  two  groups  in  length  of  nasals  is 
more  highly  developed  between  population  C  and  population  E 
than  between  other  populations  of  the  two  groups.  The  lack  of 
overlap  in  the  ranges  of  \ariation  of  this  character  in  C  and  E  is 
significant  in  view  of  the  fact  that  they  occupy  adjacent  geographic 
ranges.  Only  slight  o\'erlap  occurs  between  B  and  E,  which  also 
occur  in  adjacent  geographic  areas.  The  distinction  between  these 
adjacent  populations  also  is  reflected  in  population  D,  where  the 
intergrading  series  includes  samples  referable  to  the  small  group  of 
subspecies  on  the  lower  end  of  the  gradient  and  some  referable  to 
the  large  group  of  subspecies  on  the  upper  end  of  the  gradient, 
without  overlapping.  This  is  the  only  example  where  the  range  of 
\ariation  of  population  D  exceeds  that  of  population  C. 

Population  F  and  especially  H  are  less  clearly  distinguished  from 
A,  B,  and  C,  and  all  five  populations  broadly  overlap  in  variation. 
Distinctions  between  A,  B,  and  C,  and  F,  G,  and  H  are  poorly 
defined. 

Length  of  rostrum  (see  Figure  8). — The  populations  have  de- 
\'eloped  the  same  pattern  of  variation  in  length  of  rostrum  as 
developed  in  other  cranial  dimensions.  The  shortest  rostra  are 
characteristic  of  populations  A,  B,  and  C  and  the  longest  rostra 
in  E,  F,  G,  and  H.  As  in  other  characters,  population  D  indicates 
intergradation  between  the  two  groups. 

Differences  among  the  populations  with  short  rostra  (A,  B,  and 
C )  is  not  great,  with  the  exception  of  A,  which  has  developed  longer 
rostra  than  usual  for  the  group.  Also,  in  the  populations  with  long 
rostra  (E,  F,  G,  and  H),  the  differences  are  not  significant,  except  in 
population  H,  which  has  decidedly  shorter  rostra.  Even  so,  the 
range  of  variation  in  H  broadly  overlaps  that  of  E,  F,  and  G  as  well 
as  A,  B,  and  C. 

Difference  in  length  of  rostrum  between  the  two  groups,  there- 
fore, is  especially  significant  between  populations  B  and  C  of  the 
small  group  and  E  and  F  of  the  large  group.  In  fact,  there  is  no 
overlap  between  B-C  and  E  in  the  samples  available  to  me,  and 


348 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


in 
Z 

o 

H 

< 

o 


I  I  M  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  j  I  I  I  I  I  I  I  I  I  M 

□  D  □ 


rr 


^ 


u     nd]E\hi    0 


21 


Fig.  8.    Histograms  of  length  of  rostrum  ( mm. )  in  intraspecific  populations  of 

P.  castanops. 

only  slight  ox'erlap  between  B-C  and  F.  The  overlap  between  A 
and  F-G  is  noteworthy  because  these  three  populations  have  adja- 
cent ranges.  Of  course,  population  H  is  not  in  geographic  contact 
with  any  of  the  small  populations  ( A,  B,  and  C ) . 

The  minimum-maximum  range  of  variation  between  populations 
C,  D,  F,  and  H  is  comparable,  being  only  slightly  greater  in  popula- 
tion H. 

Breadth  of  rostrum  (see  Figure  9). — The  narrowest  rostra  are 
characterized  by  populations  A,  B,  and  C,  and  the  broadest  by 
E,  F,  G,  and  H,  especially  E  and  F.  As  in  other  dimensions,  popula- 
tion D  yields  a  range  of  \ariation  intermediate  between  the  small 
and  large  groups,  and,  unlike  some  of  the  other  characters,  the 
\ariation  of  the  representatives  of  the  two  groups  that  comprise 
population  D  broadly  overlap. 


Russell — Population  Structure  of  Pappogeomys 


349 


Although  the  two  groups  show  a  tendency  toward  divergence  in 
breadth  of  rostrum,  the  trend  is  not  so  strongly  developed  in  this 
character  as  in  other  measurements.  Therefore,  broad  overlap  is 
observed  in  the  minimum-maximum  ranges  of  variation  of  the  two 
groups.  The  greatest  degree  of  difference  between  the  two  groups 
occurs  between  populations  C  and  E.  Less  distinction  and  more 
overlap  than  usual  is  found  between  the  contiguous  populations 


I  I  I  I  I  I  I  I  M  I  I  I  M  I  I  II  I  I  I  I  I  I  [  I  I  I  I  I  I  I  I  I  [  I  I  I  I  I  I  I  I  I [  I  I  ' 

:   Cb 


C/3 

z 

o 

H 
< 

O 

Oh 


.c  DDd 


F 
G 


CZ] 


c^t^t. 


D 


D     ED     UZD 


& 


Q 


□ 


0 


8.0 


8.5 


9.0         9.5        10.0        10.5        11.0         11.5        12.0        12.5        13.0 
i  I  I  I  I  I  I  I  I  II  I  II  II  I  I  I  II  I  I  I  II  I  I  I  I  1  I  I  I  I  I  I  I  I  I  I  I 


Fig.  9.    Hi.stograni.s  of  l:)ieadth  of  rostrum   (mm.)    in  intra,specific  populations 

of  P.  castanops. 

A-B  and  E-F,  although  those  populations  of  F  that  have  actual 
geographic  contact  with  populations  of  both  A  and  B  are  composed 
of  individuals  that  provide  the  upper  range  of  the  variation  ascribed 
to  F.  In  the  usual  pattern,  population  H  is  decidedly  smaller  than 
F  and  especially  E,  but,  in  breadth  of  rostrum,  H  completely  over- 


350 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


I  M  I  M  I  I  [  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I 

a      D  n  n  D 


C3 


o 

< 
D 

Ph 

O 


n 


0 


D 


UtMM^ 


ZZZA    23 


(za 


zi    la 


7.5 


0 

10  5 


I   I  I  I  I   I  I  I   I  I  I   I  I   I  I  M  I  I   I  I   I  I  I  I   I   I  I  I  I 


110  11.5 

I  I  I  11  I  I 


Fig.  10.    Histograms  of  alveolar  length  of  maxillary  toothrow   (mm.)   in  intra- 

specific  populations  of  P.  castanops. 

laps  F  and  broadly  overlaps  E.  However,  individuals  of  population 
H  do  not  obtain  so  great  a  breadth  as  the  largest  in  E,  F,  or  G. 

Population  B  of  the  small  group  has  a  greater  maximum  variation 
than  either  A  or  C.  Unusually  broad  rostra  are  developed  in  local 
populations  in  the  northern  ( Chihuahua )  part  of  the  range  of 
population  B. 

Alveolar  len^ith  of  maxillary  toothrow  (see  Figure  10). — The 
populations  studied  differ  less  in  length  of  maxillary  toothrow  than 
any  other  cranial  feature,  except  for  breadth  of  braincase.  Even 
though  the  distinctiveness  of  each  population  is  minimal,  the  same 
pattern  of  variation  developed  in  other  cranial  features  emerges. 
Although  broadly  overlapping  in  range  of  variation,  the  populations 
with  the  shortest  toothrows  tend  to  be  A  and  especially  B  and  C. 


Russell — Population  Structure  of  Pappogeomys         351 


en 
Z 

o 

H 
< 

D 

O 

0H 


I  I  I  I  I  I  I  I  I  I  I  I  M  I  I  I  I  I  I  I  I  I  M  I  I  I  M  I  I  I  I  I  1 


D     C 


[L 


UIJI-, 


^   si: 


IZZl 


EZI 


23  24 


18  19  20  21  22  23  24  25 

I  I  I  I  I  I  I  I  I  I  I  I  1  1  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  ll   I  I  I  I 


Fig.   11.    Histograms  of  breadth  of  braincase    (mm.)    in  intraspecific  popula- 
tions of  P.  castanops. 

The  greatest  degree  of  overlap  of  populations  A,  B,  and  C  is  with 
populations  G  and  H.  Those  with  the  longest  toothrows  are  E  and 
F.  In  population  D,  representatives  of  the  two  subspecies  groups 
intergrade  completely,  with  no  indication  of  divergence. 

As  is  usually  the  case,  the  smallest  individuals  are  found  in 
population  C  and  the  largest  in  population  E.  Also,  the  highest 
degree  of  distinction  occurs  between  populations  B-C  and  E-F.  In 
this  feature,  population  H,  rather  than  C,  yields  the  highest  degree 
of  variation. 

BreadtJi  of  braincase  (see  Figure  11). — Significant  overlap  is 
developed  in  breadth  across  the  braincase  between  all  the  popula- 


352  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

tions  except  A,  a  geographic  isolate,  and  E.  The  congruity  of  the 
populations  in  this  feature  largely  obscures  the  characteristic  pat- 
tern of  variation  noted  in  most  other  cranial  characters  analyzed; 
however,  a  tendency  toward  divergence  still  can  be  ascertained  by 
comparing  population  E  with  populations  A,  B,  and  C.  The  brain- 
case  in  populations  A,  B,  and  C  is  not  significantly  narrower  than 
other  populations  ( F,  G,  and  H ) ;  rather,  there  is  a  weakly  defined 
trend  toward  a  broader  than  usual  braincase  in  population  E.  Yet 
broad  overlap  occurs  with  other  populations.  Aside  from  popula- 
tion E,  breadth  of  braincase  a\'erages  slightly  greater  in  population 
F  as  compared  with  the  other  populations. 

Although  some  difference  is  developed  between  the  geographi- 
cally adjacent  populations  C  and  E,  the  intergrading  series  D  shows 
no  tendency  toward  divergence.  Both  populations  C  and  H  show 
a  greater  range  of  variation  than  the  other  populations;  but  that  of 
H  is  due  entirely  to  one  indi\'idual,  which  has  a  considerable 
broader  braincase  than  is  usual  for  that  population. 

Length  of  head  and  body  (see  Figure  12). — Two  external  di- 
mensions were  analyzed,  length  of  head  and  body  and  length  of 
hind  foot.  Both  show  a  considerable  range  of  variation  when 
lumped  in  subspecies  clusters,  although  length  of  hind  foot  some- 
times distinguishes  two  or  more  of  subspecies  included  within  the 
subspecies  clusters.  Total  length  is  especially  variable  owing  to  the 
high  degree  of  individual  variation  of  the  length  of  tail  in  these 
pocket  gophers.  The  length  of  head  and  body  (total  length  less 
length  of  tail)  is  a  much  more  rehable  character  in  the  study  of 
geographic  variation,  and  the  population  relationships  of  subspecies 
clusters  are  defined  more  clearly  in  length  of  head  and  body  than 
in  other  external  features  in  this  species. 

The  pattern  observed  in  most  of  the  cranial  dimensions  is  evident 
in  length  of  head  and  body;  but,  the  range  of  overlap  between  most 
populations  is  great,  especially  observed  in  populations  A,  B,  and  C 
when  compared  with  F,  G,  and  H.  There  is  no  significant  difference 
among  populations  A,  B,  and  C  of  the  small  group  or  F,  G,  and  H 
of  the  large  group.  However,  there  is  a  weakly  developed,  but 
nonetheless  clearly  defined,  trend  toward  larger  size  in  population 
E  of  the  large  group. 

The  tendency  toward  large  size  in  population  E  is  noteworthy 
because  this  population  has  evidently  had  longer  contact  with  the 
small  group  (C,  D,  and  especially  B)  than  have  other  populations 
of  the  large  group  ( F,  G,  and  H ) .   This  difference  suggests  a  trend 


Russell — Population  Structure  of  Pappogeomys        353 


z 

o 

H 
< 

P^ 
O 
Pi 


I  I  M  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  M  I  I  I  I  I  I  I  I  I  I  M  I 


B 


D 


zDDoDt^^i 


D     D     C 


^^^ 


D 


^Et 


ZZI 


II 


0 


/ 


^mm 


\m 


(Z2 


0 


IZI 


'  ^IJ         '50  160  170  leo  190         200         2  10         220 

I  I  I  I  I   I  I   I  I  I   I  I  M  I  I   I  1  I  I   I  I  I  I  I  I  I  II  II  I  I  I  I  I  I  I  1  I  I  I 


Fig.  12.    Histograms  of  length  of  head  and  body  (mm.)  in  intraspecific  popu- 
lations of  P.  castanop-s. 

toward  larger  size  in  population  E.  Moreover  population  D,  which 
includes  an  intergrading  series  between  population  E  and  C,  clearly 
is  intermediate  between  these  two  populations.  For  instance, 
gophers  in  population  D  are  greater  in  ma.xinium  size  than  those  of 
population  H,  which  is  the  population  least  distinguished  from  A, 
B,  and  C. 

The  smallest  individuals,  as  in  most  cranial  characters,  are  found 
in  population  C  and  the  largest  in  E.  The  greatest  range  of  varia- 
tion was  observed  in  population  F. 

Length  of  hind  foot  (see  Figure  13). — Length  of  hind  foot  was 
recorded  in  whole  numbers;  hence,  this  accounts  for  the  even  spac- 
ing between  bars  in  the  histograms.  However,  \\'hen  a  particular 
variable  occurred  at  a  higher  frequency  than  could  be  conveniently 
recorded  in  the  diagram,  the  bar  was  doubled,  or  in  one  case  (F) 
tripled,  by  duplicating  the  bar  behind  the  initial  one  with  as  many 


354  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

I  I  I  I  I  I  I  I  I  I  I  I  M  I  I  I  M  I  I  I  I  I  I  I  III  I  I  I  I  I  I  I  I  I  I  I 


TTT 


Z 

o 

p— ( 

< 

O 


D 


DaD 


B  n 


D  D  LLD  D  Ud 


D 


n  a 


0 


□  [k[feD 


0 


ii 


0 


_G 


_H 


0 


iZl 


la 


00a 


27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42 

I  I  I  I  I  i  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  II  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I 


Fig.  13.    Histograms  of  length  of  hind  food  (mm.)  in  intraspecific  populations 

of  F.  castanops. 

squares  as  necessary.  The  same  system  was  used  in  population  D, 
where  two  subspecies,  one  from  each  of  the  major  groups,  are  in- 
cluded in  the  same  subspecies  cluster;  however,  in  this  case  each  is 
distinguished  in  the  diagram  by  difference  in  shading. 

Variation  in  length  of  hind  foot  shows  an  overall  pattern  that 
resembles  in  general  that  of  length  of  head  and  body  excepting  that 
it  is  not  so  strongly  developed.  No  significant  difference  could  be 
ascertained  between  populations  A,  B,  and  C  or  populations  F,  G, 
and  H.  Moreover,  the  distinction  between  the  small  group  (A-C) 
and  the  large  group  (E-H)  is  weakly  developed.  Only  in  popula- 
tion E,  as  in  length  of  head  and  body,  is  there  a  tendency  toward 
deviation  from  the  other  populations  in  that  there  is  a  poorly  defined 
trend  toward  greater  length  of  the  hind  foot,  emphasized  in  com- 
parison with  other  populations  of  the  large  group  by  the  lack  of 


Russell — Population  Structure  of  Pappogeomys        355 

extreme  measurements  at  the  minimal  range  of  variation.  This  sug- 
gests that  selection  is  favoring  divergence  in  population  E  by 
elimination  of  smaller-sized  individuals.  Unlike  length  of  head  and 
body,  population  H  shows  no  tendency  toward  distinctiveness. 

There  is  a  tendency  in  all  populations  for  development  of  a 
relatively  wide  range  of  variation  in  length  of  hind  foot.  In  most 
populations,  individuals  are  recorded  that  are  characterized  by  an 
extreme  measurement  separated  by  a  significant  hiatus  from  the 
main  body  of  the  population.  This  is  especially  noted  in  populations 
A,  B,  E,  G,  and  H. 

Population  Structure 

Analysis  of  geographic  variation  in  Pappogeomys  castanops  dis- 
closes a  complex  relationship  between  populations  and  groups  of 
populations.  One  of  the  fundamental  disclosures  is  that  subspecies 
may  be  grouped  into  intraspecific  population  units,  here  called  sub- 
species clusters.  Each  subspecies  cluster  has  a  well-defined  geo- 
graphic range,  and,  of  course,  all  of  the  individual  and  mutually 
exclusive  ranges  of  the  subspecies  of  each  cluster  are  contiguous. 
The  analysis  of  variation  in  the  preceding  section  shows  that  all  of 
the  subspecies  that  comprise  a  cluster  are  characterized  by  a  similar 
pattern  of  variation.  This  suggests  that  the  subspecies  of  a  particu- 
lar cluster  occur  under  fairly  uniform  environmental  conditions 
throughout  the  geographic  range  occupied  by  that  cluster.  There- 
fore, the  populations  of  the  cluster  are  subjected  to  the  same  or 
similar  selection  pressures  resulting  in  the  strong  resemblance  of 
patterns  of  variation  among  the  member  subspecies  of  a  cluster. 

The  fact  that  the  subspecies  of  a  cluster  do  difl^er  significantly 
from  each  other,  usually  in  combinations  of  several  features  (see 
Russell,  1968:621-691),  offers  evidence  suggesting  that  the  direction 
and  intensity  of  selection  is  not  precisely  the  same  everywhere 
within  the  range  of  the  cluster.  Hence,  divergence  on  a  minor  scale 
is  occurring  among  the  subspecies  of  a  cluster,  allowing  perhaps 
for  a  more  precise  adaptation  to  environmental  optima  from  place 
to  place  over  the  geographic  range  of  the  cluster.  Evidently,  the 
environmental  conditions  and,  concomitantly,  the  selective  pressures 
vary  between  the  different  subspecies  clusters.  The  interaction  of 
these  evolutionary  agents  have  produced  significant  divergence 
between  the  subspecies  clusters.  As  to  be  expected,  the  degree  of 
divergence  between  particular  subspecies  clusters  varies,  being 
greater  in  some  cases  than  in  others.    It  is  of  interest  to  note  that 


356  Misc.  Publ.  51,  Unr'.  Kansas  Mus.  Nat.  Hist. 

the  differences  between  subspecies  clusters  are  greater  than  the 
differences  between  the  subspecies  of  a  cluster. 

Therefore,  intraspecific  population  structure  in  Pappogeomys 
castanops  is  organized  into  population  units  greater  in  scope  than 
the  subspecies.  These  population  units  consist  of  two  or  more  sub- 
species that  are  characterized  by  a  unique  pattern  of  \'ariation, 
occupy  a  specific  geographical  range,  and  evidently  evolved  as  a 
unit.  No  doubt,  the  subspecies  clusters  of  Pappogeomys  castanops 
have  some  degree  of  constancy  in  time  because  they  have  developed 
distinctive  characteristics. 

Population  units  below  the  level  of  subspecies  include  local 
populations  and  denies.  The  variation  and  relationships  of  these 
small  units  are  not  analyzed  in  this  report  for  lack  of  appropriate 
data.  Huge  collections,  including  samples  from  all  places  of  occur- 
rence, would  be  necessary  for  a  detailed  study  of  this  sort;  to  my 
knowledge,  none  has  been  made. 

Not  all  subspecies  are  united  into  clusters.  In  Pappogeomys 
castanops,  two  are  clearly  geographic  isolates.  In  analyzing  the 
geographic  variation  of  the  species,  the  isolated  subspecies  were 
treated  along  with  the  subspecies  clusters;  however,  their  range  of 
variation  is  not  so  great  as  that  of  the  clusters  and  their  distinctive- 
ness is  not  so  sharply  defined.  In  both  cases,  the  geographically 
isolated  subspecies  could  ha\  e  been  assigned  to  one  of  the  neighbor- 
ing subspecies  clusters  (geographic  isolate  A  to  cluster  B  and  geo- 
graphic isolate  G  to  cluster  F)  without  difficulty.  Probably,  each 
originated  in  the  not-too-distant  past  from  (or  along  with)  the 
subspecies  group  with  which  the  closest  relationship  exists.  How- 
ever, the  geographically  isolated  subspecies  are  discrete  population 
units  owing  to  the  lack  of  gene  flow  with  neighboring  populations, 
and,  therefore,  deserve  to  be  analyzed  separately.  Considering  long 
range  evolution,  however,  I  do  not  regard  them  to  be  as  important 
as  the  subspecies  cluster  because  they  have  neither  the  store  of 
variability  nor  the  potential  for  mutation  of  the  larger  population 
units. 

The  four  subspecies  of  one  cluster,  population  H  ( see  preceding 
section),  actually  may  be  a  collocation  of  geographic  isolates.  Cer- 
tainly the  Rio  Grande  restricts  gene  flow  between  P.  c.  hullatus  and 
P.  c.  angusticeps,  and  mountains  at  least  restrict,  if  not  prevent, 
gene  flow  between  P.  c.  angusticeps  and  P.  c.  torridus.  Geographi- 
cally intermediate  populations  between  P.  c.  hullatus  and  P.  c. 
tamaulipensis  of  the  lower  Rio  Grande  Valley  have  not  been  se- 


Russell — Population  Structure  of  Pappogeomys         357 

cured.  However,  all  four  subspecies  are  closely  related,  and,  there- 
fore, are  treated  as  a  subspecies  cluster.  Isolation,  if  it  exists,  is 
probably  of  recent  origin. 

The  sharp  discontinuity  between  subspecies  clusters,  relative  to 
that  between  subspecies  of  a  cluster,  is  especially  noteworthy.  The 
subspecies  clusters  meet  along  narrow  zones  or  belts  of  intergrada- 
tion  as  indicated  in  Figure  1.  The  abrupt  shift  from  one  character- 
istic range  of  variation  to  a  different  and  equally  characteristic 
range  of  variation  indicate  secondary  zones  of  intergradation  where 
the  ranges  of  subspecies  clusters  abut.  Secondary  zones  of  inter- 
gradation suggest  secondary  contact  after  prior  geographic  isolation. 
On  the  other  hand,  zones  of  contact  between  subspecies  of  the  same 
cluster  are  characteristic  of  primary  zones  of  intergradation  that 
have  developed  in  the  absence  of  geographic  isolation  (not  shown 
in  Fig.  1,  but  see  Russell,  196(3:622).  Evidently,  the  member  sub- 
species of  a  cluster  have  developed  while  in  geographic  contact  and 
not  while  isolated,  at  least  not  while  isolated  for  significant  periods 
of  time. 

There  seems  to  be  no  restriction  to  gene  flow  between  contigu- 
ous subspecies  of  the  same  cluster,  thus  contributing  to  the  cohesive- 
ness  of  the  population  unit.  To  the  contrary,  evidence  suggests  that 
there  is  some  restriction  to  gene  flow  between  subspecies  clusters, 
more  in  some  cases  than  in  others.  For  instance,  no  evidence  of 
intergradation  could  be  found  between  subspecies  clusters  E  and  H 
where  their  ranges  meet  in  north-central  Coahuila;  however,  sec- 
ondary intergradation  occurs  between  clusters  H  and  F  in  the 
vicinities  of  Sanderson  and  Dryden,  southwestern  Texas  (see 
Russell,  1968:632).  Reduction  of  gene  exchange  between  adjacent 
subspecies  clusters  would  be  advantageous  in  promoting  local 
adaptation  by  preventing  the  destruction  of  advantageous  gene 
combinations  through  the  introgression  of  foreign  genes.  The 
unique  gene  combinations  of  each  subspecies  cluster  are  maintained, 
therefore,  by  the  low  incidence  of  gene  introgression.  Whether 
there  is  selection  against  intergrades  in  the  secondary  zones  of 
intergradation  or  some  restriction  to  interbreeding  between  in- 
di\  iduals  of  different  subspecies  clusters  is  not  known;  however,  I 
would  suppose  that  the  former  is  more  likely  the  case. 

Even  so,  owing  to  the  intraspecific  population  structure,  a  situa- 
tion exists  that  lends  itself  to  the  ready  establishment  of  reproduc- 
ti\e  isolation  between  subspecies  clusters.  If  lengthy  geographic 
isolation  occurs,   the  already   divergent  populations   may   develop 


35(S  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

biological  isolating  mechanisms  that  will  severely  restrict  or  prevent 
interbreeding  once  the  isolated  populations  have  re-established 
contact.  E\'idently,  such  is  the  case  between  populations  A,  B,  and 
C  and  populations  E,  F,  G,  and  H  in  Pappogeomys  castanops.  As 
pointed  out  in  a  preceding  publication  (Russell,  1968:623-627  and 
769-771),  there  are  a  number  of  places  (points  1,  2,  3,  4,  5,  and  6 
on  Fig.  1 )  where  the  ranges  of  these  two  groups  are  sympatric  with 
no  evidence  of  interbreeding,  and  in  these  areas  they  behave  as  full 
species. 

That  populations  A,  B,  and  C  are  referable  to  the  same  species 
as  populations  E,  F,  G,  and  H  is  established  by  a  chain  of  inter- 
breeding populations  between  population  C  and  E.  However,  this 
can  be  demonstrated  in  only  one  geographic  area — southwestern 
San  Luis  Potosi,  northern  Zacatecas,  and  eastern  Durango.  The 
intergrading  populations  consisting  of  two  subspecies,  P.  c.  nihellus 
and  P.  c.  goldmani,  are  included  here  in  population  D,  which  in 
most  characters  (see  Figs.  2-13)  clearly  depicts  the  intermediate 
relationship  between  the  two  distinctly  different  subspecies  clusters 
( populations  C  and  E )  that  it  connects.  Save  for  gene  flow  between 
the  populations  in  this  area,  the  two  groups  of  subspecies  could  be 
recognized  as  two  separate,  albeit  closely  related,  species.  Russell 
(loc.  cit.),  recognizing  these  distinctions,  assigned  the  populations  to 
one  of  two  subspecies-groups:  populations  A-C  to  the  subnubilus- 
group  (unshaded  histograms)  and  E-H  to  the  excelsus-group 
(histograms  with  diagonal  lines  in  Figs.  2-13).  Population  D,  the 
intergrading  series,  includes  two  subspecies,  one  (P.  c.  goldmani) 
referable  to  the  excelsus-gioup  and  the  other  (P.  c.  ruheUus)  refer- 
able to  the  suhmihihis-growg. 

Examination  of  Figures  2-13  reveal  that  those  populations  of 
the  excehus-gxoxxp  that  are  in  contact  with  populations  of  the 
subnubilus-group  have  developed  the  greatest  degree  of  difference. 
In  contrast,  subspecies  cluster  H,  which  has  no  contact  with  the 
populations  of  the  sii])nubihis-gYouv),  has  developed  the  least  degree 
of  differentiation.  Moreover,  subspecies  cluster  E,  which  has  prob- 
ably been  in  contact  with  the  suljmibilus-group  longer  than  any 
other  populations  of  the  excelstis-group  (see  discussion  beyond),  is 
characterized,  considering  the  entire  cluster,  by  the  highest  degree 
of  differentiation.  This  example  of  character  displacement  (see 
Brown  and  Wilson,  1956:49-64)  or  character  di\'ergence  (Mayr, 
1963:82-86),  suggests  that  the  excelsus-grou^  has  responded  to 
sympatry  with  the  subnubilus-giou-p  by  divergent  evolution,  result- 


Russell — Population  Structure  of  Pappogeomys         359 

ing  in  an  increase  in  cranial  dimensions.  Therefore,  most  cranial 
dimensions  of  populations  G,  F,  and  especially  E  are  significantly 
larger  than  those  of  the  subniil)ilus-group  (see  Figs.  2-9),  usually 
with  only  slight  overlap,  or  no  overlap,  in  the  respective  ranges  of 
variation. 

On  the  other  hand,  most  cranial  distinctions  between  population 
H  and  the  subniibiJus  populations  are  not  great  and  usually  are 
characterized  by  broad  overlap  in  their  ranges  of  variation.  Evi- 
dently, since  population  H  has  had  no  contact  with  the  subnubilus- 
group  there  has  also  been  no  selection  pressure  toward  character 
divergence.  Therefore,  population  H  of  the  eAce/s  us -group  shows 
closer  resemblance  to  the  subnubilus-gioup  than  do  adjacent  popu- 
lations of  the  excehus-group. 

Also,  there  is  less  resemblance  between  population  H  and  popu- 
lations E,  F,  and  G  of  the  excelsus-grou-p  than  there  is  among  E,  F, 
and  G.  Indeed,  as  mentioned  above,  field  data  indicate  that  popula- 
tions H  and  E  may  not  interbreed  where  they  are  in  contact  in 
north-central  Coahuila.  Russell  (1968:6.34)  could  find  no  evidence 
of  intergradation  between  P.  c.  btiUatus  (subspecies  cluster  H)  and 
P.  c.  ustulatus  and  P.  c.  jucimdus  (subspecies  cluster  E)  in  this 
area.  However,  population  H  does  intergrade  with  population  F 
both  in  northern  Coahuila  ( Russell,  loc.  cit. )  and  in  eastern  part  of 
the  Trans-Pecos  region  of  Texas  (Russell,  op.  cit.:  632).  Moreover, 
the  subspecies  P.  c.  pratensis  of  population  F,  occurring  in  the  cen- 
tral part  of  the  Trans-Pecos  of  Texas  (point  7,  Fig.  1),  seems  to 
ha\e  differentiated,  judging  from  its  characters,  as  the  result  of 
unrestricted  gene  exchange  between  ancestral  stock  of  the  small 
animals  of  population  H  and  the  large  animals  of  population  F  that 
became  sympatric  in  this  area  ( see  discussion  beyond ) . 

Populations  A,  B,  and  C  of  the  subnubilus-group  are  less  sharply 
differentiated  in  most  features  than  those  of  the  excelsiis- group. 
Most  of  the  contact  with  the  excelsus-group  occurs  with  populations 
A  and  B;  population  C,  which  occurs  south  of  the  mountainous 
barrier  formed  by  the  Sierra  Guadalupe-Sierra  Parras  ranges  in 
southwestern  Coahuila,  has  only  limited  contact  with  population  E 
of  the  excehus-group.  For  example,  contact  occurs  only  in  a  few 
high  passes  in  the  mountains,  as  in  Santo  Domingo  Caiion  ( point  6, 
Fig.  1).  Hence,  it  is  unlikely  that  there  would  be  selection  for 
character  divergence  in  this  population,  and  none  is  evident.  There- 
fore, population  C  probably  has  remained  little  changed  from  its 
earlier  Pleistocene  ancestor;  except,  of  course,  for  subspecific  differ- 
entiation within  the  cluster. 


360  Misc.  Publ.  51,  Uxiv.  Kansas  Mus.  Nat.  Hist. 

It  is  at  first  surprising  to  find  that  the  differences  between  popu- 
lations A,  B,  and  C  are  so  weakly  developed,  especially  in  view  of 
the  broad  contact  of  A  and  B  with  populations  of  the  excelsus- 
group.  One  would  expect  population  A  and  B  to  have  developed 
some  degree  of  character  dixergence,  and,  therefore,  to  have  de- 
veloped distinctive  differences  from  population  C.  Such  is  not  the 
case.  Instead  of  in  the  suhnubilus-grouTp,  character  divergence  has 
occurred  in  the  excelsus-group,  particularly  in  populations  in  the 
zone  of  sympatry. 

x\nalysis  of  the  species  Pappogeomijs  castanops  as  a  multidimen- 
sional system,  reveals  a  complex  population  structure  with  a  \xdde 
range  of  interpopulation  relationships.  At  one  extreme  are  the  col- 
lection of  denies  and  local  populations  that  compose  the  subspecies. 
Such  groupings  are  characterized  by  unlimited  gene  exchange, 
especially  on  the  interdemic  level.  Next,  subspecies  are  usually 
grouped  into  larger  intraspecific  population  units,  the  subspecies 
cluster.  Unrestricted  gene  flo\\'  through  primary  intergradation 
within  the  subspecies  cluster  results  in  a  characteristic  pattern  of 
variation  and  degree  of  adaptation  for  the  entire  population  unit. 
The  relationship  between  subspecies  clusters  seems  to  vary  in  this 
species  between  nearly  maximum  rates  of  gene  exchange  (which 
evidently  occurred  in  the  differentiation  of  P.  c.  pratensis)  down  to 
what  seems  to  be  severely  restricted  gene  flow  ( as  between  popula- 
tions H  and  E ) .  At  any  rate,  the  zones  of  contact  between  the  sub- 
species clusters  are  evidently  zones  of  secondary  intergradation,  and 
were  established  after  periods  of  geographic  isolation.  At  the  other 
extreme,  are  the  sympatric  occurrence  of  populations  of  different 
subspecies  clusters.  There  is  no  indication  of  interbreeding  between 
these  populations  at  most  localities.  Reproductive  isolation  in  this 
non-dimensional  system  is  probably  due  to  the  formation  of  bio- 
logical isolating  mechanisms.  The  nature  of  these  is  unknown,  but 
both  ecological  and  ethological  mechanisms  are  indicated.  Regard- 
less, the  sympatric  populations  behave  as  species  in  these  areas. 
Correlated  with  their  sympatric  occurrence  has  been  the  develop- 
ment of  considerable  character  divergence.  Breakdown  of  the  iso- 
lating mechanisms  that  prevent  interbreeding  between  these  sym- 
patric populations  occurs  only  in  one  area.  There,  two  of  the  sub- 
species of  the  diverging  groups  do  interbreed  where  their  ranges 
are  contiguous,  resulting  in  gene  introgression  into  adjacent  popula- 
tions of  both  subspecies-groups.  At  least  two  geographic  isolates, 
with  the  possibility  of  several  others   (as  mentioned  above),   are 


Russell — Population  SxRucxuiyi  of  Pappogeomys        361 

known.  Taxonomically,  both  have  been  assigned  subspecific  status. 
On  the  intraspecific  lexel,  the  subspecies  cluster  in  Pappog,eomys 
castanops  seems  to  be  the  most  important  population  unit  of  evolu- 
tion. Intraspecific  divergence  seems  to  occur  on  this  level,  including 
divergent  trends  that  could  lead  in  the  future  to  the  separation 
of  the  complex  into  t\\'o  or  more  species  if  geographic  isolation 
develops. 

Evolution  of  Intraspecific  Population  Structure 

In  order  to  reconstruct  the  evolutionary  history  of  the  complex 
population  structure  of  Pappogeomys  castanops,  two  sorts  of  data 
are  vital.  First,  it  is  necessary  to  ha\e  a  knowledge  of  current 
population  structure,  relationships,  and  geographic  variation.  Sec- 
ond, at  least  a  general  knowledge  of  the  immediately  past  climatic 
changes  in  the  area  of  distribution  also  is  pertinent.  One  could 
argue  that  actual  fossil  e\idence  would  be  ideal  for  reconstructing 
the  past  history  of  the  species.  I  disagree  with  this  opinion,  for  two 
reasons :  ( 1 )  we  are  concerned  with  the  formation  of  intraspecific 
structure  of  a  modern  species,  and  thus  we  are  dealing  only  with 
evolutionary  changes  in  the  terminal  stage  of  the  Pleistocene  (Wis- 
consin and  post- Wisconsin  periods)  and  in  the  Recent,  too  short 
a  time  span,  therefore,  for  fossil  evidence  to  be  of  much  value; 
(2)  fragmentary  fossils  from  widely  scattered  sites,  even  if  avail- 
able, would  offer  little  evidence  for  solving  the  details  of  intra- 
specific population  \'ariation.  Therefore,  the  dexelopment  of  the 
complex  must  be  based  on  the  interpretation  of  other  data,  and  it 
is  primarily  a  problem  of  neontology.  Some  idea  of  the  formation 
of  intraspecific  population  structure  helps  to  appreciate  the  complex 
relationships  of  the  species  and  also  provides  some  notion  as  to  how 
they  may  de\'elop. 

Pleistocene  Changes  in  CJiniate  and  Environment 
Late  Pleistocene  climatic  changes  and  concomitant  changes  in 
the  environment  have  had  a  direct  effect  upon  the  process  of 
speciation  and  the  patterns  of  distribution  of  living  species.  These 
climatic  changes  took  place  during  the  time  of  the  last  glacial 
advance  (Wisconsin)  in  the  northern  part  of  the  continent,  and, 
of  course,  the  climatic  changes  continued  into  the  post- Wisconsin 
period.  E\idence  supporting  trenchant  en\'ironmental  changes  in 
the  southwestern  United  States  and  northern  Mexico  is  from  three 
sources:    analysis  of  pollen  profiles;  analysis  of  contemporary  dis- 


362  Misc.  Publ.  51,  Univ.  Kansas  Mus,  Nat.  Hist. 

tributions  and  ecologies;  and,  to  a  lesser  extent,  the  distribution  of 
late  Pleistocene  fossils. 

During  much  of  the  Wisconsin  glaciation  of  North  America,  and 
probably  as  late  as  the  Cochran  Readvance  (see  Dorf,  1960:342; 
and  Hibbard,  1955:82-84),  northern  Mexico  and  the  southwestern 
United  States  experienced  a  cool,  moist  chmate.  Jaeger  ( 1926 )  sug- 
gested that  these  pluvial  cycles  were  characterized  by  the  develop- 
ment of  numerous  lakes,  many  of  large  size,  throughout  the  northern 
part  of  the  Mexican  Plateau  (see  Flint,  1947:476,  for  a  brief  dis- 
cussion). Old  stream  courses,  now  dry,  and  the  large  number  of 
old  lake  beds  bear  out  Jaegers  opinion. 

Before  the  Wisconsin  pluvial  cycles,  in  the  terminal  stages  of 
the  Sangamon  interglacial  period,  this  region  was  evidently  hot  and 
dry,  characterized  in  the  main  by  desert  environments.  Hibbard 
(1960:22,  25)  pointed  out  that  the  Sangamon  interglacial  stage 
terminated  with  a  particularly  arid  interval  just  preceding  the 
Wisconsin.  Extensive  deposits  of  caliche  formed  on  the  Great  Plains 
(southwestern  Kansas)  at  the  close  of  the  Sangamon,  indicating  a 
hot,  dry  climate.  It  is  not  unreasonable  to  suppose  that  aridity  also 
developed  over  the  major  part  of  the  region  to  the  south  and  west 
at  the  same  time.  An  abrupt  shift  to  cool,  moist  climates  of  the 
Wisconsin  stage  followed  the  arid  interval. 

The  most  convincing  evidence  of  the  shift  to  cool,  moist  climates 
in  northern  Mexico  and  the  southwestern  United  States  is  afforded 
by  palynological  evidence.  Wisconsin  and  post- Wisconsin  fossil 
pollen  sequences  have  been  studied  from  many  sites,  especially  from 
the  northern  part  of  this  region  by  Martin  (1963),  Martin  and 
Mehringer  (1965),  Hafsten  (1961),  and  more  recently  by  Wendorf 
(Symposium  on  Pleistocene  and  Recent  Environments  on  the  Cen- 
tral Great  Plains,  University  of  Kansas,  October  26,  1968).  Analysis 
of  these  fossil  pollen  records  indicates  that  the  region  inhabited  by 
Pappogeomys  castanops  was  characterized  by  a  cool,  moist  climate 
during  the  Wisconsin  pluvial.  The  dominant  vegetation  was  a 
boreal  forest  of  pine,  spruce,  and  fir.  Evidently,  during  somewhat 
dryer  subintervals,  oak  increased  in  abundance  resulting  in  mixed 
pine-oak  forests  on  dryer  sites.  The  pluvial  interval  and  its  boreal 
environment  lasted  to  the  end  of  the  Wisconsin,  approximately 
10,000-11,000  years  BP.  During  the  height  of  pluviation  the  evi- 
dence suggests  that  boreal  forests  were  more  or  less  continuous  in 
both  lowlands  and  highlands.    Toward  the  close  of  the  Wisconsin 


Russell — Population  Structure  of  Pappogeomys         363 

( 14,000  to  10.000  BP ) ,  the  continuous  boreal  forests  of  the  lowlands 
gave  way  to  grasslands  mixed  with  boreal  woodlands. 

In  the  post-Wisconsin,  conditions  became  progressively  more 
arid.  Probably  the  trend  toward  aridity  began  in  the  southern  part 
of  the  region  and  gradually  advanced  northward.  As  the  climate 
became  hot  and  dry,  desert  shrubs  and  forbs  once  again  became 
dominant.  On  some  sites,  probably  at  higher  elevations  in  the  foot- 
hills of  the  mountains,  grassland  savannas  of  oaks  and  junipers 
replaced  forests. 

During  the  Wisconsin  pluvial  cycle,  many  northern  species  ex- 
tended their  ranges  southward  into  northern  Mexico.  For  example, 
the  remains  of  Sorex  cinereus  (Findley,  1953:635-636),  Sijnaptomiis 
cooperi,  and  an  unidentified  species  of  Marmota  (see  Jakway,  1958: 
319,  321;  Gushing,  1945:185)  were  recovered  from  Wisconsin  de- 
posits from  San  Josecito  Cave  in  Nuevo  Leon.  At  the  present  time, 
none  of  these  mammals  occur  nearer  the  site  of  the  cave  than  800 
miles  to  the  north  in  the  mountains  of  central  New  Mexico  (Sorex 
and  Marmota).  Furthermore,  these  terrestrial  species,  restricted  to 
relatively  small  home  ranges,  must  have  reached  the  area  of  the 
ca^'e  when  the  intervening  area  was  favorable  to  their  occurrence. 
In  the  case  of  the  shrew,  this  would  have  necessitated  a  relatively 
cool  and  mesic  environment  according  to  Findley  {loc.  cit.:  636), 
although  less  mesic  environments  could  have  favored  the  two 
rodents.  The  postulation  that  Sorex  cinereus  enjoyed  a  continuous 
distribution  in  the  pine-spruce  forests  throughout  this  region  is 
substantiated  by  the  occurrence  of  Sorex  miJIeri,  a  living  monotypic 
species  endemic  to  higher  elevations  in  the  Sierra  del  Carmen  of 
north-central  Coahuila  and  the  Sierra  Madre  Oriental  in  extreme 
southeastern  Coahuila  (see  Baker,  1956:167-168).  Findley  (1955: 
617)  found  Sorex  milleri  to  be  a  close  relative  of  Sorex  cinereus, 
and  he  postulated  that  the  southern  segment  of  S.  cinereus,  that 
which  expanded  southward  during  the  Wisconsin  pluvial  stage,  be- 
came isolated  in  post-Wisconsin  time  and  gave  rise  to  S.  milleri. 
The  post- Wisconsin  shift  to  aridity  contributed,  as  Findley  (loc.  cit.) 
suggested,  to  the  present  disjunct  distribution  of  milleri.  The  two 
small  populations  of  S.  milleri,  therefore,  are  relics  that  have  found 
refuge  in  isolated  mountains. 

Subsidiary,  but  important,  evidence  is  provided  by  two  other 
relics  of  this  region,  Scalopus  montanus  and  Cynonnjs  mexicana. 
Both  species  have  restricted  ranges  in  northeastern  Mexico  (see 
Baker,  op.  cit.:  170,  202),  and  both  species  are  geographic  isolates, 


364  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

separated  from  the  main  range  of  their  nearest  relatives  to  the  north 
and  northeast.  We  may  assume  that  both  of  these  isolated  popula- 
tions reached  the  sites  of  their  present  occurrence  by  range  exten- 
sions of  the  main  body  of  their  respective  genera  when  environ- 
mental conditions  were  more  mesic  than  now  in  the  intervening 
areas.  Both  populations  were  left  in  isolation  by  the  extirpation  of 
the  connecting  populations  with  the  post-pluvial  trend  to  aridity. 
Scalopiis  is  of  particular  interest  in  view  of  the  restrictions  of  move- 
ment imposed  by  its  fossorial  adaptations.  Moist,  friable  soils  with 
suital:)le  insect  populations  are  necessary  for  its  occurrence. 

Also,  the  disjunct  distributional  patterns  in  northern  Mexico  of 
Eiitamias  dorsalis,  Neotoma  mexicana,  Sigmodon  ochrognathus, 
Microtus  montaniis,  M.  longicaudus,  M.  mexicanus,  and  M.  pennsijl- 
vanicus  ( see  reports  of  Baker,  1956;  Findley  and  Jones,  1960;  Ander- 
son, 1959,  1961;  Findley  and  Jones,  1962)  furnish  additional  in- 
formation as  to  the  effects  of  the  post- Wisconsin  shift  to  xeric  con- 
ditions in  the  lowlands.  All  the  aforementioned  species  are  presently 
restricted  to  pockets  of  favorable  habitat  found  at  higher  elevations 
in  the  isolated  mountains  of  the  region.  No  doubt  the  currently 
disjunct  populations  of  each  of  these  species  enjoyed  a  more  wide- 
spread and  continuous  distribution  across  the  intervening  lowlands 
during  the  more  mesic  plu\'ial  stage. 

The  detailed  investigation  of  Sigmodon  ochrognathus  by  Findley 
and  Jones  (1960:462-469)  revealed  a  distributional  history  that 
probably  is  characteristic  of  the  other  species  listed  above.  These 
authors  could  find  no  significant  differences  among  the  many  iso- 
lated populations  of  S.  ochrognathus  that  they  sampled,  indicating 
to  them  that  all  of  these  isolated  populations  were  recently  con- 
nected by  continuous  distribution.  Moreover,  this  species  of  Signio- 
don  is  restricted  ecologically  to  the  oak  forest  zone.  These  data 
suggest  that  oak  forest,  and  a  more  mesic  environment,  were  wide- 
spread throughout  the  lowlands  between  the  mountain  ranges  of 
this  region  synchronously  with  the  uninterrupted  distribution  of 
the  cotton  rats. 

In  summary,  the  pre- Wisconsin  climate  of  northern  Mexico  and 
the  southwestern  United  States  was  hot  and  dry  with  decidedly  arid 
environments  prevailing  especially  in  the  lowlands.  The  Wisconsin 
glacial  period  that  followed  was  marked  by  a  shift  to  cool,  moist 
climates  that  supported  more  mesic  environments.  During  the 
plu\  ial  maxima,  boreal  forests  of  pine,  spruce,  and  fir  became  the 
dominant  vegetation  in  the  lowlands.    Fluctuating  subcycles  prob- 


Russell — Population  Structure  of  Pappogeomys         365 


Fk;.  14.    Pnssilile  pattern  of  distriljiition  of  P.  castanops  in  the  early  Wisconsin. 


ably  occurred  during  the  pluvial  period  until  the  Wisconsin  termi- 
nated. The  post-Wisconsin  was  characterized  by  the  progressive 
development  of  hot,  dry  climates,  accompanied  by  a  trend  toward 
xeric  environments.  The  return  to  arid  conditions  was  thus  marked 
by  the  rise  to  dominance  of  the  desert  flora  and  fauna  that  occurs 
in  this  region  at  the  present  time.  This  general  history  of  northern 
Mexico  provides  a  model  for  analysis  of  microevolution  in  Pappoge- 
omys castanops. 


366 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Fig.  15.   Possible  disjunct  pattern  of  distribution  of  major  populations  (A-C)  of 
P.  castanops  during  the  time  of  pknial  maximum  in  the  Wisconsin. 


Microevolution  of  Pappogeomys  castanops 
Remains  of  Pappogeomys  castanops  from  the  Wisconsin  deposits 
of  San  Josecito  Cave  in  Nuevo  Leon,  Mexico  (Russell,  1960:543) 
demonstrates  the  occurrence  of  the  species  in  northeastern  Mexico  in 
the  late  Pleistocene.  In  pre-Wisconsin  times,  this  species  probably 
enjoyed  an  extensive  range  in  the  lowlands  of  northeastern  Mexico 
as  suggested  in  Figure  14.  The  extent  of  its  distribution  outside  of 
northeastern  Mexico,  if  indeed  it  occurred  elsewhere,  is  not  known; 


Russell — Population  Structure  of  Pappogeomys        367 


10-1 


102 


100 


98 


106 


Fig.   16.    Possible  directions  of  emigration  from  the  three  major  refugia  in  the 
early  stages  of  post-Wisconsin  time.    Note  initial  points  of  contact  between  the 

populations  A-B  and  B-C. 


there  is  no  evidence  at  this  time  of  occurrence  beyond  the  region 
indicated.  The  hot,  dry  cHmates  at  the  close  of  the  Sangamon 
would  have  favored  the  development  of  xerophytic  vegetation,  thus, 
proxiding  adequate  habitat  for  these  pocket  gophers. 

The  shift  to  cool,  moist  climates  in  the  early  stages  of  the  Wis- 
consin would  have  been  unfavorable  to  Pappogeoi7iys  castanops; 
and,  as  boreal  vegetation  from  the  north  progressively  replaced  the 


36S 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


106 


104 


102 


100 


98 


Fig.  17.  Possil^le  patterns  of  gene  flow  with  continued  emigration  in  post- 
Wisconsin  time  and  the  establishment  of  zones  of  sympatry  in  areas  where 
intraspecific  inter-breeding  was  restricted.  Moxements  of  populations  in  this 
final  phase  of  post- Wisconsin  readjustment  has  resulted  in  the  superspecies 
complex  that  occurs  at  the  present  time  (see  Fig.  1). 


xerophytic  species  in  the  lowlands,  Pappogeomys  castanops  would 
have  been  forced  out  of  most  of  the  areas  where  it  previously  oc- 
curred. As  a  result  of  adverse  conditions  the  range  of  this  species 
decreased,  finally  fragmenting  into  isolated  populations,  restricted 
to  small  areas  of  preclimax  conditions,  at  least  during  the  height  of 
the  pluvial  cycles.  These  refugia  may  have  been  located  in  the  rain 


Russell — Population  Structure  of  Pappogeomys        369 

shadow  of  the  higher  mountains  where  drier  environments  pre- 
vailed. Although  these  areas  probably  did  not  offer  optimum  habi- 
tat, they  did  allow  survival  of  the  species.  There  is  no  way  of 
knowing  how  many  isolated  populations  existed;  probably  they 
varied  in  number  and  size,  becoming  less  numerous  as  the  Wiscon- 
sin pluviation  continued.  Current  patterns  of  geographic  variation 
and  population  structure  suggest  that  there  were  at  least  three  main 
refugia,  which  were  located,  in  general,  in  the  areas  indicated  in 
Figure  15.  f-lefugia  A  and  B  were  in  the  rain  shadow  of  the  Sierra 
Madre  Oriental;  refugium  A  was  located  to  the  south  of  the  east- 
west  oriented  Sierra  Guadalupe-Sierra  Parras  ranges,  and  refugium 
B  was  located  to  the  north  of  the  same  escarpment.  A  third  re- 
fugium ( C )  must  have  occurred  on  the  low  coastal  plain  somewhere 
to  east  of  the  sierras.  Its  exact  location  is  not  clearly  discernible, 
and  perhaps  several  refugia  existed  on  the  coastal  plain  rather  than 
one.  During  the  period  of  isolation,  the  disjunct  populations  evolved 
as  separate  systems.  The  small  populations  would  have  been 
favored  by  maximum  rates  of  evolution,  and,  therefore,  divergent 
evolution  took  place  in  a  relatively  short  time.  However,  the  period 
of  isolation  was  not  long,  and  reproductive  isolation  was  not  fully 
developed  in  all  the  isolated  populations.  Thus,  the  trend  toward 
speciation  was  incomplete. 

At  the  end  of  the  Wisconsin  with  the  return  to  arid  conditions, 
the  isolated  populations  of  Pappogeomys  castanops  expanded  their 
ranges  out  from  the  Wisconsin  refugia.  Contact  was  re-established 
between  the  isolated  populations  as  theii'  ranges  became  more  wide- 
spread, as  suggested  in  Figure  16.  Wisconsin  populations  A  and  B 
probably  first  established  contact  on  the  elevated  plains  in  north- 
eastern Durango  and  southwestern  Coahuila  west  of  the  Sierra 
Parras.  Populations  B  and  C  evidently  came  into  contact  along  the 
western  limits  of  the  coastal  plain,  probably  in  the  area  presently 
inhabited  by  P.  c.  huUatus  and  P.  c.  ustulatus  (see  Russell,  1968: 
622).  There  is  no  evidence  the  Wisconsin  population  C  moved 
westward  in  Mexico.  Also  there  is  no  evidence  that  Wisconsin 
populations  B  and  C  made  contact  during  the  time  of  range  expan- 
sion; probably  the  high  sierras  of  this  area  still  functioned  as  a 
barrier  to  their  dispersal. 

Relationships  between  the  descendants  of  these  Pleistocene  iso- 
lates indicates,  that  interbreeding  was  restricted  when  they  made 
contact  at  the  close  of  the  Wisconsin  and  the  incidence  of  cross- 
breeding was  small.   Reproductive  isolation  was  especially  well  de- 


370  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

veloped  between  Wisconsin  populations  A  and  B.  Probably  bio- 
logical isolating  mechanisms,  developed  while  in  isolation,  func- 
tioned to  severely  reduce  the  incidence  of  cross-mating  between 
these  populations.  As  a  response  to  intergroup  competition  in  the 
areas  of  sympatry,  especially  between  populations  A  and  B,  charac- 
ter divergence  evolved,  resulting  in  the  distinctions  between  living 
descendants  of  these  populations.  As  a  result  of  reduced  gene  flow 
(or  lack  of  it),  each  of  the  populations  was  able  to  maintain  its 
unique  features. 

As  arid  conditions  of  post- Wisconsin  time  progressively  moved 
northward,  replacing  the  more  mesic  environments  of  the  pluvial 
stage,  the  populations  of  Pappogeomys  castanops  expanded  their 
ranges  northward  as  suggested  in  Figure  17.  At  places  where  the 
ranges  of  populations  A  and  B  became  contiguous  they  maintained 
reproductive  isolation  and  did  not  interbreed.  Indeed,  a  broad  area 
of  sympatry  was  developed  between  A  and  B  in  northeastern  Du- 
rango  and  northern  Zacatecas  by  the  southwestward  expansion  of 
population  B  from  Coahuila.  Population  B  also  extended  its  range 
northward  into  the  Trans-Pecos  of  Texas,  there  making  contact  with 
northwestward-expanding  segments  of  population  C.  In  this  area, 
(point  7,  Fig.  1),  potential  isolating  mechanisms  between  popula- 
tions B  and  C  failed  to  function,  and  the  resultant  interbreeding 
between  the  two  units  produced  a  population  with  intermediate 
features  (P.  c.  pratensis).  Further  range  extension  to  the  north 
resulted  in  the  present  pattern  of  distribution  depicted  in  Figure  1. 

Wisconsin  population  A,  therefore,  gave  rise  to  living  popula- 
tions A,  B,  and  C  depicted  in  Figure  1.  Wisconsin  population  B 
gave  rise  to  the  modern  populations  E,  F,  and  G,  and  Wisconsin 
population  C  to  the  modern  population  H.  The  characteristics  and 
relationships  of  these  populations  are  described  in  the  aforegoing 
accounts. 

Literature  Cited 

Anderson,  S. 

1959.     Distribution,    variation,    and    relationships    of    the    montane    vole, 

Microtus   montanus.    Univ.   Kansas   Publ.,   Mus.   Nat.   Hist.,   9:415- 

511,  12  figs. 
1961.     A    relict   population    of   Microtus    pennsylvanicus   in    Southwestern 

New  Mexico.    Amer.  Mus.  Novit,  2034:1-3,  1  fig. 

Baker,  R.  H. 

1956.     Mammals  of  Coahuila,  Mexico.   Univ.  Kansas  Pulil.,  Mus.  Nat.  Hist., 
9:125-335,  75  figs. 

Brown,  W.  L.,  Jr.,  and  E.  O.  Wilson 

1956.     Character  displacement.    Syst.  Zool.,  5:49-64,  6  figs. 


Russell — Population  Structure  of  Pappogeomys         371 

CusHixG,  J.  E.,  Jr. 

1945.  Quaternary  rodents  and  lagoniorphs  of  San  Josecito  Cave,  Nuevo 
Leon,  Mexico.   Jour.  Mamm.,  26:182-185. 

DORF,  E. 

1960.     Climatic  changes  of  the  past  and  j)resent.    Amer.  Scientist,  48:341- 

364,  3  figs.,  1  pi. 
Flint,  R.  F. 

1947.     Glacial   geology  of  the  Pleistocene  Epoch.     John  Wiley   and   Sons, 

New  York,  fourth  printing  (1953),  xviii  +  589  pp.,  6  pis.,  88  figs. 

FiXDLEY,  J.    S. 

1953.  Pleistocene  Soricidae  from  San  Josecito  Cave,  Nuevo  Leon,  Mex- 
ico.   Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.,  5:633-639. 

1955.  Taxonomy  and  distribution  of  some  American  shrews.  Univ.  Kansas 
Publ.,  Mus.  Nat.  Hist.,  7:613-618. 

FiXDLEY,   J.    S.,   AXD   C.   J.   JOXES 

1960.  Geographic  variation  in  the  yellow-nosed  cotton  rat.  lour.  Mamm., 
41:462-469,  1  fig. 

1962.  Distribution  and  variation  of  voles  of  die  genus  Miciotiis  in  New 
Mexico  and  adjacent  areas.    Jour.  Mamm.,  43:154-166,  5  figs. 

Hafstex,  U. 

1961.  Pleistocene  de\elopment  of  \egetation  and  climate  as  evidenced  by 
pollen  analysis.  Pp.  59-91,  figs.  19-34,  in  Paleoecology  of  the  Llano 
Estacado,  Mus.  New  Mexico  Press,  Santa  Fe,  Publ.  no.  1,  Fort 
Burgwin  Research  Center. 

HiBBARD,  C.  W. 

1955.     Pleistocene  vertebrates  from  the  Upper  Becerra   ( Becerra  Superior) 

Formation,   Valley   of   Texquixquiac.    Mexico,   with   notes   on   other 

Pleistocene  forms.    Contrib.  Mus.  Paleo.,  Univ.  Michigan,  12:47-96, 

9  pis.,  5  figs. 
1960.     An    interpretation   of   Pliocene    and    Pleistocene    climates    in    North 

America.     62nd    Annual    Report,    Michigan    Acad.    Sci.,    Arts,    and 

Letters,  30  pp.,  1  pi.,  2  figs. 

Jaeger,  F. 

1926.     Forschungen   uber  das   diluviale   Klima  in    Mexiko.    Peterm.    Mitt., 
Erganzungsh.,  190,  64  pp. 
Jakway,  G.  E. 

1958.     Pleistocene  Lagomorpha  and  Rodentia  from  the  San  Josecito  Cave, 
Nuevo  Leon,  Mexico.    Trans.  Kansas  Acad.  Sci.,  61:.313-327,  1  fig. 
Martix,  p.  N. 

1963.  The  last  10,000  years,  a  fossil  pollen  record  of  the  American  South- 
west.   Univ.  Arizona  Press,  Tucson,  iv  -)-  87  pp.,  14  pis.,  37  figs. 

Martix,  P.  N.,  and  P.  J.  Mehrixger,  Jr. 

1965.     Pleistocene  pollen  analysis  and  biogeography  of  the  Southwest.    Pp. 
433-451,   in  Quaternary  of  the   United  States    (H.   E.   Wright  and 
D.  G.  Frey,  Eds.).    Princeton  Univ.  Press,  x  -f  922  pp. 
Mayr,  E. 

1963.     Animal    species   and   evolution.     Harvard    Univ.    Press,    Cambridge, 
Massachusetts,  xiv  +  795  pp.,  65  figs. 
Russell,  R.  J. 

1960.     Pleistocene  pocket  gophers  from  San  Josecito  Ca^•e,   Nuevo   Leon, 

Mexico.    Univ.  Kansas  Publ.,  Mus.  Nat.  Hist.,  9:539-548. 
1968.     Revision  of  pocket  gophers  of  the  genus  Pappogeomys.    Univ.  Kan- 
sas Publ.,  Mus.  Nat.  Hist.,  16:581-776,  10  figs. 


TAXONOMIC  REVIEW  OF  THE  GOLDEN  MOUSE, 
OCHROTOMYS  NUTTALLI 

BY 

Robert  L.  Packard 

Golden  mice  (genus  Ochrotomys)  are  rather  unique,  semi- 
arboreal  cricetines  in  the  neotomyine-peromyscine  phyletic  line. 
The  single  species,  Ochrotomys  nuttaUi,  occurs  in  the  southeastern 
United  States  (see  Fig.  1) — westward  to  eastern  Texas,  north  as 
far  as  northern  Kentucky  and  central  Virginia,  and  south  to  central 
Florida.  Ecologically,  the  distribution  of  these  mice  is  closely  cor- 
related with  that  of  the  southeastern  deciduous  ( oak-hickory )  hard- 
woods and  pine  stands,  or  the  Lower  Austral  and  portions  of  the 
Upper  Austral  life-zones  ( Austroriparian  and  part  of  the  Carolinian 
biotic  provinces).  Golden  mice  are  found  from  lowland,  heavily 
forested  floodplains,  to  pine  uplands  where  there  is  considerable 
underbrush,  but  seem  most  numerous  in  the  former  type  of  habitat. 
Considerable  information  has  appeared  recently  on  the  biology  of 
these  mice  (see  McCarley,  1958;  Layne,  1960;  Rippy  and  Harvey, 
1963;  Packard  and  Garner,  1964;  and  Linzey  and  Linzey,  1967a, 
1967b),  but  there  has  been  only  a  single  systematic  revision  (Os- 
good, 1909).  This  was  based  on  250  specimens  from  segments  of 
the  currently  known  geographic  range,  and  two  subspecies  were 
recognized. 

Golden  mice  were  described  by  Harlan  in  1832  as  Arvicoh 
miitaUi,  with  the  type  locality,  Norfolk,  Norfolk  Co.,  Virginia.  In 
1841,  Audubon  and  Bachman  described  Mus  (Calomys)  atireolus 
from  oak  forests  of  South  Carolina.  Baird,  in  1858,  followed  Wagner 
(1843:51)  in  considering  golden  mice  as  belonging  to  the  genus 
Hesperomys  and  stated  (p.  468):  "...  I  am  impelled,  by  a  strict 
regard  for  the  law  of  priority,  to  change  the  expressive  name  of 
aureolus,  hitherto  applied  exclusively  to  this  species,  for  the  less 
meaning  one  of  mittaUi."  Bangs  (1898),  presumably  following 
Trouessart  ( 1897),  placed  the  golden  mice  in  the  genus  Peromysctis. 
When  Osgood  ( 1909)  revised  Peromyscus,  he  accorded  golden  mice 
subgeneric  standing  with  the  description  of  Ochrotomys,  based  on 
the  numerous  characters  by  which  these  mice  differ  from  other 
peromyscoids. 

Aside  from  the  description  of  two  new  subspecies  (P.  n.  leivisi 
Howell,  1939,  and  P.  n.  fammeus  Goldman,  1941),  no  changes  were 

(373) 


374 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


made  in  the  taxonomy  of  golden  mice  until  Blair  (1942)  suggested 
that  they  might  represent  a  distinct  genus.  However,  Miller  ( 1912, 
1924),  Miller  and  Kellogg  (1955),  and  Hall  and  Kelson  (1959)  all 
followed  Osgood's  arrangement.  Hooper  ( 1958 )  concluded  from 
his  study  of  the  male  phalli  in  peromyscoids  that  golden  mice  should 
be  accorded  generic  rank.  More  recently  (Manville,  1961;  Rinker, 
1963;  Hooper  and  Musser,  1964;  Patton  and  Hsu,  1967),  using  dif- 
ferent approaches,  have  strengthened  the  position  that  golden  mice 
warrant  generic  recognition.   There  has  been  no  recent  analyses  of 


Fig.  1.    Geographical  distribution  of  sulxspecies  of  Ochiotomys  nuttaUi:    O.  n. 

aiircolus  (1);  O.  n.  flammcti.s  (2);  O.  ».  floiidauiis-  (3);  O.  n.  lisae  (4);  O.  n. 

nuttaUi  (5).    Cross-hatching  indicates  areas  of  intergradation. 


geographic  variation  and  synthesis  of  the  available  data  bearing  on 
the  phylogeny  of  these  mice.  The  objectives  of  this  report  are: 
(1)  analyze  variation  in  a  population  resulting  from  age,  season,  and 
sex;  (2)  record  the  amount  of  variation  within  and  between  popula- 
tions; (3)  comment  on  the  systematic  status  and  phylogeny  of  the 
golden  mice. 


Packard — Review  of  Ochkotomys  375 

Acknowledgments 

This  study  was  conducted  with  financial  aid  from  the  National  Academy  of 
Sciences  ( Bache  Fund),  and  in  part  from  National  Science  Foundation  Grant 
G- 12059.  Students  who  aided  me  at  \'arious  times  in  the  course  of  this  work 
were  H.  Rex  Arnold,  Herschel  W.  Garner,  and  David  J.  Schmidly.  I  am 
indebted  to  the  following  named  institutions  and  persons  for  making  specimens 
a\ailable  for  study:  American  Museum  of  Natural  History  (AMNH),  R.  G. 
Van  Gelder;  Auburn  University  (AU),  Julian  Dusi;  Duke  University  (DU), 
Department  of  Zoology;  Field  Museum  of  Natural  History  (FM),  Joseph  C. 
Moore  and  Pliillip  H.  Hershkovitz;  Louisiana  State  University,  Museum  of 
Natiual  Science  (LSU),  George  H.  Lowery,  Jr.;  Museum  of  Comparative 
Zoology  (MCZ),  Barbara  Lawrence;  Mississippi  State  Game  and  Fish  Com- 
mission (MGF);  North  Carolina  State  University  (NCS),  F.  S.  Barkalow; 
Stephen  F.  Austin  State  College  (SFA),  Edwin  D.  Michael;  U.S.  National 
Museum  (USNM),  Charles  O.  Handley,  Jr.  and  Henry  W.  Setzer;  University 
of  Arkansas  (UA),  John  A.  Sealander;  University  of  Florida  (UF),  J.  H.  Kauf- 
man; University  of  Georgia  (UG),  Eugene  P.  Odum;  Southern  Illinois  Uni- 
versity (SIU),  W.  D.  Klimstra;  Texas  A  and  M  University',  Cooperative  Wild- 
life Research  Collection  (TAM),  W.  B.  Davis;  University  of  Kansas,  Museum 
of  Natural  History  (UK),  E.  Raymond  Hall  and  J.  Knox  Jones,  Jr.;  University 
of  Illinois,  Museum  of  Natural  History  (UI),  Donald  F.  Hoflmeister;  LTniver- 
sity  of  Michigan,  Museum  of  Zoology  (MZ),  W.  H.  Burt  and  E.  T.  Hooper; 
University  of  Missouri  (UM),  William  Elder;  University  of  Texas  (UT), 
Gerald  G.  Raun;  Virginia  Polytechnic  Institute   (VPI),  Henry  S.  Moseby. 

Materials  and  Methods 

This  report  is  based  on  the  study  of  approximately  825  museum  specimens 
(skins,  skulls,  complete  skeletons,  and  entire  animals  preserved  in  liquid,  ap- 
proximately t\vo-thirds  more  specimens  than  were  available  to  Osgood,  1909). 
Most  specimens  were  accompanied  with  data  relative  to  locality,  date  of 
capture,  sex,  and  standard  external  measurements.  In  addition,  live  golden 
mice  were  captured  and  retained  in  the  laboratory  where  breeding  tests  were 
conducted  and  beha\'ioral  data  gathered. 

Specimens  were  grouped  for  study  by  sex,  age,  and  season  of  capture 
( \\  hen  feasible ) .  Because  of  the  limitations  placed  on  statistical  analyses  by 
extremely  small  sample  sizes,  specimens  were  regrouped  into  units  for  study  of 
geographic  variation.  These  units  were  selected  on  the  basis  of  knowledge 
of  habitat  (the  degree  of  imiformity),  altitude,  and  general  geographic  rela- 
tionship. As  a  result,  composition  of  the  groups  were:  (1)  Florida  Peninsula; 
(2)  coastal  plains  of  Louisiana,  Mississippi,  Alabama,  and  Florida  (north  of 
ca.  30°  N  latitude);  (3)  coastal  plain,  piedmont,  and  mountains  of  Georgia; 
(4)  South  Carolina;  (5)  coastal  plain,  piedmont,  and  moimtains  of  North 
Carohna;  (6)  coastal  plain,  piedmont,  and  area  of  Amelia  Court  House, 
Virginia;  (7)  mountains  of  western  Virginia,  southeastern  Kentucky,  and 
eastern  Tennessee;  (8)  Mississippi  River  Valley  of  southern  Illinois  and  Mis- 
souri; (9)  Arkansas  (principally  mountainous  area);  (10)  eastern  Texas. 
Measurements  from  classes  3  to  5  (see  section  on  age),  adults  of  both  sexes, 
were  programmed   (according  to  the  above  listed  locaHties)   in  a  7040  IBM 


376 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


PARACONE 
MESOLOPH 


METACONE 
MESOSTYLE 


ANTERIOR    CINGULUM 


PROTOCONE 
MURE 


HYPOCONE 


ENTOSTYLE 


Fig.  2.  Right  upper  molar  row  of  Ochrotomijs  nuitaUi,  age  three  male,  no.  1072 
Stephen  F.  Austin  State  College,  2  mi.  SW  Bienxille,  Bienville  Parish,  Louisiana. 

computer  at  this  institution  using  the  biomed  simple  data  description,  version 
of  February  6,  1964,  Health  Sciences  Computing  Facility,  University  of  Cali- 
fornia (Los  Angeles). 

External  measurements  used  were  those  recorded  by  collectors  plus  body 
length,  which  was  obtained  by  subtracting  the  length  of  the  tail  from  the  total 
length.  Fifteen  cranial  measurements  were  analyzed  in  the  study  of  non- 
geographic  variation  of  a  Texan  population.  Coefficient  of  variation  was  calcu- 
lated for  all  measurements  in  the  population  studied  and  those  that  varied  least 
were  selected  for  comparing  samples  from  different  geographic  areas.  As  a 
result,  the  following  cranial  measurements  (see  Packard,  1960:585,  for  de- 
scription) were  selected:  condylobasal  length;  zygomatic  breadth;  depth  of 
cranium;  length  of  rostrum;  alveolar  length  of  maxillary  toothrow;  post-palatal 
length;  least  interorbital  breadth;  alveolar  length  of  mandibular  toothrow. 

Some  of  the  color  notations  refer  to  the  Munsell  Soil  Color  Charts  (1954), 
whereas  other  terms  used  do  not  refer  to  any  one  standard.  The  names  of 
cusps  and  ridges  on  the  teeth  (see  Fig.  2)  are  those  of  Wood  and  Wilson 
(1936)  and  the  enamel  grooves  and  folds  are  those  of  Hershkovdtz  ( 1944)  and 
Hooper  (1952). 

Secondary  sexual  variation  was  not  found  significant  (see  analyses  of  age 
and  sexual  \ariation  and  Fig.  3)  and  adult  males  and  females  were  considered 
together  in  the  analysis  of  geographic  patterns. 

Non-geographic  Variation 
Non-geographic  variation  in  Ochrotomijs  has  been  little  studied. 
Aside  from  Osgood's  (1909)  revision  and  Hooper's  (1957)  study  of 
variation  in  dental  patterns,  only  cursory  comments  on  this  subject 
are  in  the  literature.  Important  contributions  in  the  dexelopment, 
growtli,  and  molt  of  golden  mice  have  been  reported  by  Linzey 


Packard — Review  of  Ochrotomys  377 

and  Linzey  (1967a,  1967b)  and  Layne  (1960).    Blus  (1966)  con- 
tributed information  on  variation  in  litter  size. 

The  largest  sample  available  to  me  for  study  w^as  79  specimens  of 
O.  n.  fatnmetis  from  Nacogdoches  County,  Texas.  These  were  sep- 
arated by  sex,  thence  into  age  categories,  and  measured.  Most  speci- 
mens were  captured  between  the  months  of  February  to  May.  All 
measurements  were  analyzed  by  the  7040  computer  using  the  simple 
data  description  program  previously  mentioned. 

Variation  with  Age 

Specimens  of  both  sexes  were  grouped  into  five  age  categories 
as  follows:  (1)  juvenile;  (2)  young  or  subadult;  (3)  young  adult; 
(4)  adult;  (5)  old  adult  (hereafter  age  categories  will  be  referred 
to  by  number ) .  Linzey  and  Linzey  ( 1967b )  pointed  out  that  golden 
mice  have  a  prolonged,  but  slow,  growth  rate.  Thus,  mice  beyond 
10  to  12  weeks  of  age  do  not  change  appreciably  in  size.  My  age 
categories  were  based  first  on  morphological  criteria  and  secondly 
were  compared  with  the  linear  data  on  growth  of  Linzey  and  Linzey 
(1967b)  and  Layne  (1960).  External  measurements  (particularly 
length  of  tail)  of  mice  of  age  group  five  compare  favorably  with 
those  of  individuals  48  weeks  of  age,  age  group  four  with  mice  28 
to  44  weeks  old,  age  group  three  with  those  16  to  28  weeks  old,  age 
group  two  with  mice  that  were  seven  to  12  weeks  old,  and  age  group 
one  with  individuals  less  than  seven  weeks  old.  Linzey  and  Linzey 's 
(loc.  cit.)  and  Layne's  (loc.  cit.)  growth  data  for  external  measure- 
ments (total  length,  body  length,  length  of  tail,  length  of  hind  foot) 
show  a  significant  leveling  in  rate  of  increase  between  three  to  six 
weeks.  Golden  mice  eight  weeks  old  or  older,  while  still  growing 
at  a  rather  slow  but  constant  rate,  changed  little  in  size.  Layne 
(1960)  stated  that  cheekteeth  are  fully  developed  between  four  to 
five  weeks.  On  this  basis,  my  age  one  mice  would  be  less  than  four 
to  five  weeks  of  age,  whereas  my  age  two  mice  would  seemingly 
be  five  weeks  old  or  older. 

For  purpose  of  comparisons  of  mice  from  different  geographic 
areas,  only  individuals  of  ages  three  through  five  were  used.  In  most 
cases,  specimens  in  comparable  pelages  ( having  completed  the  post- 
juvenile  molt,  see  Linzey  and  Linzey,  1967a)  were  used  for  geo- 
graphic analyses. 

Age  1. — Mice  with  M3  and  m3  erupted  tn  alveolar  surface;  cusps,  styles, 
and  lophs  of  other  molars  all  distinct  with  little  or  no  evidence  or  wear;  all 
cranial  sutures  generally  distinct  with  presphenoid-basisphenoid  suture  widely 
open;  tympanic  bullae  rugose  with  mottled  appearance;  dentary  bone  rugose; 


378 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Table  1. — Analysis  of  variation  with  age  and  secondary  sexual  vaiuation 
OF  Ochrotomys  nuttaUi  from  eastern  Texas. 


Variant 

Sex 

N 

Ranife 

Mean 

SE 

SD 

Age  1 

Total  length 

M 
F 

M 
F 

M 
F 

M 
F 

M 
F 

M 

F 

1 
1 

1 
1 

1 
1 

1 

1 

1 

2 

125.0 
109.0 

65.0 
60.0 

60.0 
49.0 

16.0 
17.0 

8.0 
12.0 

18.6 

.40 

Length  of  body 

Length  of  tail 

Length  of  hind  foot 

Length  of  ear 

Condylobasal  length 

18.2-19.0 

.57 

Zygomatic  breadth 

M 
F 

1 
2 

11.9 
11.6 

740 

11.2-12.0 

.57 

Depth  of  braincase 

M 
F 

1 

2 

8.6 

8.7 

7l5 

8.6-8.9 

.21 

Length  of  rostrum 

M 

F 

M 
F 

1 

1 

2 

776 

3.7 
3.3 

'.2'5 

Alveolar  length  of  upper  molar  row 

3.0-3.5 

.35 

Post-palatal  length 

M 

F 

1 

2 

8.1 

7.2 

.15 

7.0-7.3 

.21 

Length  of  mandil^ular  molar  row 

M 
F 

M 

F 

1 

1 

2 

3^3 

4.20 

3.75 

.55 

Least  interorbital  Ijreadth 

3.2-4.3 

.78 

Age  2 

Total  length 

M 
F 

15 

12 

130-154 
135-167 

142.3 
146.4 

1.6 

2.3 

6.1 
8.1 

Length  of  body 

M 
F 

15 
12 

70-103 
70-97 

81.1 
79.5 

1.8 
2.0 

6.9 
7.0 

Length  of  tail 

M 

F 

15 
12 

52-73 
62-77 

63.1 
66.9 

1.3 
1.2 

5.1 
4.3 

Length  of  hind  foot 

F 

15 
12 

13-19 
15-19 

17.1 
17.5 

.4 
.3 

1.7 
1.1 

Length  of  ear 

M 

F 

15 
12 

10-17 
10-18 

14.9 
14.6 

.4 
.6 

1.8 

2.2 

Condylobasal  length 

M 
F 

14 
9 

19.9-21.9 
21.0-22.4 

21.1 
21.8 

.2 
.2 

.6 
.5 

Zygomatic  breadth 

M 

F 

15 
10 

11.5-13.0 
12.0-13.4 

12.4 
12.6 

.1 
.1 

.4 

.4 

Depth  of  braincase 

M 
F 

14 
9 

8.4-9.8 
9.1-9.5 

9.1 
9.3 

.1 
.04 

.3 
.1 

Packard — Review  of  Ochrotomys 


379 


Table  1. — Continued. 


Variant 


Sex 


N 


Ran};e 


Mean 


SE 


SD 


Length  of  rostrum 


M      15 
F       10 


Alveolar  length  of  upper  molar  row    M      14 

F         9 


Post-palatal  length 

Length  of  niandiliular  molar  row 

Least  interorbital  Ijreadth 


M  14 

F  7 

M 
F 

M  15 

F  10 


7.5-10.0 
8.5-10.0 

3.4-3.9 
3.6-3.9 

7.5-8.9 
8.4-9.0 


3.5-4.5 
4.1-4.5 


8.9 
9.3 

3.7 
3.7 

8.4 

8.8 


4.2 
4.4 


2 
A 

.03 
.02 

.1 
.1 


.1 
.03 


.7 
.5 

.1 
.1 

.4 

2 


.3 
.1 


Total  length 
Length  of  body 
Length  of  tail 
Length  of  hind  foot 
Length  of  ear 
Condylobasal  length 
Zygomatic  breadth 
Depth  of  iMaincase 
Length  of  rostrum 


Age  3 

M  11 

F  12 

M  11 

F  12 

M  11 

F  12 

M  11 

F  12 

M  10 

F  12 

M  11 

F  8 

M  11 

F  8 

M  11 

F  8 

M  11 

F  8 


Alveolar  length  of  upper  molar  row    M      11 

F         8 


Post-palatal  length 

Length  of  mandibular  molar  row 

Least  interorjjital  breadth 


M      11 
F         8 


M 
F 


1 


146-160 
139-174 

74-90 
76-106 

62-77 
55-80 

15-21 
16-20 

11-16 
13-17 

21.5-23.2 
21.4-23.2 

12.3-13.3 
12.3-13.6 

8.6-9.7 
9.0-9.7 

8.7-10.1 
9.1-10.1 

3.5-3.8 
3.4-3.9 

8.0-9.6 
8.6-9.8 


M      11 
F         8 


3.8-4.8 
3.7-4.6 


1.53.9 
158.1 

83.8 
88.6 

70.1 
69.5 

17.1 
17.8 

13.4 
15.5 

22.4 
22.3 

12.9 
12.9 

9.3 
9.3 

9.4 
9.6 

3.7 
3.7 

8.9 
9.2 

3.6 


1.5 
3.0 

1.6 
2.3 

1.2 
2.2 

.5 

.4 

.5 
.3 

.1 

2 

.1 

.2 

.1 
.1 


4.4 
4.3 


.1 
.1 

.02 
.1 

.1 

.2 


5.0 
10.5 

5.2 
8.1 

3.9 

7.8 

1.5 
1.3 

1.6 
1.1 

.5 

.7 

.3 
.4 

.3 
.3 

.4 
.4 

.1 

.1 

.4 

.4 


.2 


Total  length 
Length  of  body 
Length  of  tail 


Age  4 

M  10 

F  10 

M  10 

F  10 

M  10 

F  10 


153-171 
140-170 

80-96 
80-103 

68-79 
63-85 


160.9 
159.7 

87.6 
89.5 

73.1 
73.2 


1.8 
3.3 

1.6 

2.5 

1.1 
2.4 


5.9 
10.6 

5.0 
7.9 

3.4 
7.5 


380  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Table  1. — Continued. 

Variant                                                 Sex  N             Range  Mean 

Length  of  hind  foot                               M  10          17-19  17.6 

F  10          17-19  17.9 

Length  of  ear                                         M  10          11-17  15.6 

F  10          1.3-17  16.1 

Condylobasal  length                               M  10  22.5-23.4  22.9 

F  9  22.3-24.0  23.1 

Zygomatic  breadth                                M  10  12.8-13.6  13.3 

F  9  13.4-13.9  13.6 

Depth  of  braincase                                M  10  9.3-9.7  9.5 

F  9  8.7-9.7  9.3 

Length  of  rostrum                                 M  9  9.6-10.8  9.9 

F  9  9.5-10.5  9.9 

Alveolar  length  of  upper  molar  row    M  10  3.6-3.9  3.8 

F  9  3.5-3.8  3.6 

Post-palatal  length                                 M  10  9.1-9.7  9.4 

F  8  8.2-9.7  9.3 

Length  of  mandibular  molar  row        M  

F  2  3..3-3.4  3.4 

Least  interorbital  breadth                    M  10  4.2-4.6  4.4 

F  9  3.6-4.6  4.3 

Age  5 

Total  length                                            M  5  167-174  170.0 

F  1  131.0 

Length  of  body                                      M  5          84-100  93.2 

F  1  68.0 

Length  of  tail                                         M  5          70-85  76.9 

F  1  63.0 

Length  of  hind  foot                               M  5          18-19  18.2 

F  1 19.0 

Length  of  ear                                           M  5           16-17  16.6 

F  1  11.0 

Condylobasal  length                               M  5  23.0-23.8  23.4 

F  -  

Zygomatic  breadth                                 M  5  13.5-13.8  13.7 

F 

Depth  of  braincase                                M  5  9.3-9.7  9.5 

F  -  

Length  of  rostrum                                  M  5  9.4-10.4  10.1 

F  _.  

Alveolar  length  of  upper  molar  row    M  5  3.7-3.8  3.8 

F  

Post-palatal  length                                 M  5  9.5-9.9  9.7 

F  - .- 

Length  of  mandibular  molar  row        M  -  

F  ._  - 

Least  interorbital  breadth                     M  5  4.2-4.6  4.4 

F  - - 


SE 


SD 


.2 

.2 

.7 

.7 

.6 
.5 

1.8 
1.4 

.1 

.2 

.O 

.6 

.] 
.1 

.3 
.2 

.03 

.1 

.1 
.3 

.1 
.1 

.4 
.3 

.03 
.04 

.1 
.1 

.1 
.2 

2 
.5 

"T 

.l" 

.04 
.1 

.1 

.4 

1.1 

2.5 

2.7 

6.1 

2.8 

6.3 

.2 

.4 

.2 

.5 

.1 

.3 

.1 

.1 

.1 

.1 

.2 

.4 

.02 

.1 

.1 

.2 

.1 

.2 

Packard — Review  of  Ochrotomys  381 

epiphyses  of  all  long  bones  distinct  from  diaphyses;  under  fur  of  pelage 
grayish-lilack,  guard  hairs  dusky-blackish,  remaining  hair  on  dorsiun  golden- 
brown  (agouti  tips  short)  and  near  7. SYR  5/6  ( Mimseli,  1954);  hairs  on  belly 
white  at  tip  and  gray  at  base,  giving  dusky  white  appearance;  tail  bicolored, 
ears  golden,  upper  parts  of  hind  feet  white  (see  Table  1). 

Age  2. — Mice  with  M3  and  ni3  erupted;  small  degree  of  wear  evident  on 
cusps,  styles,  and  lophs  of  all  molars;  basisphenoid-basioccipital  and  basisphen- 
oid-presphenoid  sutures  evident;  tympanic  liullae  and  dentary  bone  smooth 
( also  in  all  older  age  groups ) ;  epiphyseal-diaphyseal  sutures  not  distinct  in 
long  bones  save  the  tibia;  guard  hairs  with  brownish  tones  (less  blackish  than 
in  age  one),  agouti  ])and  broader,  resulting  in  a  more  golden  brown  color  on 
dorsum  than  in  age  one,  but  still  near  7. SYR  5/6  on  Munsell  scheme;  belly 
with  slight  suffusion  of  yellow  with  individual  hairs  pale  yellow-gray  at  base, 
whitish  near  tips;  tail  slightly  bicolored,  more  fully  haired  than  in  age  one; 
hair  on  hind  feet  white;  ears  golden  brown  (see  Table  1). 

Age  3. — Wear  on  enamel  of  major  cusps  sufficient  to  reveal  dentine  bands; 
lophs  and  styles  evident  (see  Fig.  2),  l)ut  with  wear  (particularly  on  M3  and 
m3 ) ;  sutures  in  cranium  not  evident;  distal  diaphyseal-epiphyseal  suture  of 
femur  and  proximal  diaphyseal-epiphyseal  sutine  of  tibia  faintly  visible;  guard 
hairs  on  dorsum  brow  nish-black  ( less  profuse  than  in  age  two ) ;  golden-reddish 
agouti  band  broadened  at  expense  of  grayish  base,  resulting  in  overall  golden- 
red  tone  (color  SYR  5/8  middorsally );  belly  whitish  with  yellow-orange  over- 
tones, basal  parts  of  hair  gray;  tail  indistinctly  bicolored;  hind  feet  whitish 
above;  ears  orange-red  (see  Table  1). 

Age  4. — All  teeth  show  considerable  wear  with  dentine  lake  in  center  of 
M3  and  m3;  mesolophs  and  mesolophids  of  first  and  second  molars  well  worn, 
residting  in  confluence  of  dentine  with  nuire  between  paracone  (protoconid) 
externally  and  hypocone  (entoconid)  internally;  distal  suture  bet\veen  epiphy- 
sis and  diaphysis  of  tibia  faintly  visible;  little  change  in  color  of  pelage  from 
age  three  except  overall  tones  of  upper  parts  more  golden  brown  than  in  age 
three  ( see  Table  1 ) . 

Age  5. — All  teeth  well  worn,  essentially  without  pattern  of  cusps,  lophs, 
and  styles,  and  with  only  a  thin  peripheral  band  of  enamel  surrounding  lake 
of  dentine;  sutures  on  long  bones  ankylosed;  pelage  somewhat  ragged,  but  not 
differing  appreciably  from  pelage  of  ages  three  and  four,  color  near  SYR  5/8 
middorsally  ( see  Table  1 ) . 

Secondary  Sexual  Variation 

Variation  resulting  from  differences  between  sexes  in  eacli  of  the 
age  classes  was  studied  following  the  method  of  Hubbs  and  Hubbs 
( 1953 ) .  Females  of  adult  age  groups  tended  to  have  slightly  longer 
and  broader  skulls  and  may  be  somewhat  more  variable  than  males, 
but  the  differences  were  not  significant  ( see  Fig.  3 ) . 

Because  no  significant  secondary  sexual  variation  was  revealed 
in  this  study,  the  sexes  have  been  considered  together  in  the 
analysis  of  geographic  variation.  All  specimens  used  in  analysis  of 
secondary  sexual  variation  were  collected  in  a  period  from  January 


382 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


II  16 


^p- 


111       XT 


23 

22 

■J] 
~U 

-13 

-12 

11 

10 


123  456789 

Fig.  3.  Analyses  of  secondary  sexual  variation.  Numbers  at  base  of  abscissa 
refer  to  variants  analyzed  as  follows:  (1)  alveolar  length  of  maxillary  tooth- 
row;  (2)  depth  of  cranium;  (3)  breadth  of  braincase;  (4)  length  of  rostrum; 
(5)  length  of  incisive  foramina;  (6)  least  interorbital  breadth;  (7)  postpalatal 
length;  (8)  condylobasal  length;  (9)  zygomatic  breadth.  Vertical  lines  repre- 
sent range,  solid  black  bar  equals  one  standard  error  of  the  mean,  open  bar 
equals  one  standard  deviation;  numerals  directly  above  each  diagram  represent 
size  of  sample;  in  each  matching  set  of  diagrams,  samples  of  males  are  on  the 

left  and  females  on  the  right. 

to  May  from  the  Stephen  F.  Austin  Experimental  Forest  (a  part  of 
the  Angehna  National  Forest),  located  about  10  mi.  SW  Nacog- 
doches, Texas. 

Indwidual  Variation 

Golden  mice  from  the  same  population,  when  grouped  into 
comparable  age  and  sex  categories,  vary  little  from  each  other. 
Hooper  ( 1957 )  reported  little  variation  in  the  dental  patterns  of 
these  mice.  External  measurements  were  more  variable  than  those 
of  the  cranium  but  this  may  reflect  the  technique  of  measuring  by 
the  different  collectors.  Linzey  and  Linzey  (1967a)  found  some 
variation  in  detail  of  the  pattern  of  molt,  but  the  basic  pattern 
varied  little.  My  study  of  molt  patterns  are  essentially  in  agreement 
with  theirs.  A  series  of  54  bacula  of  golden  mice  were  measured 
and  studied  and  only  minor  variations  were  recorded.  Measure- 
ments and  ratios  of  basal  width  to  total  length  and  bacular  length 


Packard — Review  of  Ochrotomys  383 

to  body  length  do  not  differ  appreciably  from  those  reported  by 
Blair  (1942).  The  hyoid  apparatus  of  38  golden  mice  from  the 
Nacogdoches  County,  Texas,  sample  did  not  vary  from  the  pattern 
reported  by  Sprague  (1941).  Illustrations  of  the  phallus  by  Hooper 
(1958)  and  Hooper  and  Musser  (1964)  suggest  little  variation 
present  in  that  organ.  Rinker  (1960)  and  Manville  (1961)  reported 
the  absence  of  the  entepicondylar  foramen  in  the  humerus  of 
Ochrotomys  (Blair,  in  Blair  et  oL,  1968:518,  erroneously  reports  its 
presence).  Rinker  had  four  specimens  and  Manville  five.  I  exam- 
ined 92  golden  mice  humeri  from  various  parts  of  the  geographic 
range  and  found  no  entepicondylar  foramen.  Patton  and  Hsu 
( 1967 )  reported  a  diploid  number  of  52  chromosomes  in  the  golden 
mouse.  No  variation  from  this  number  was  recorded  although 
sample  size  ( four )  was  exceedingly  small.  Peterson  ( 1968 )  reported 
a  stable  densito-metric  curve,  on  the  basis  of  two  specimens,  in  the 
blood  serum  pattern  of  nuttalli.  Rinker  (1963)  reported  a  consistent 
myological  plan  on  the  basis  of  three  specimens  dissected.  Examina- 
tion and  dissection  of  11  additional  specimens  in  this  study  (particu- 
larly the  M.  pronator  quodratus  and  M.  pronator  teres)  suggests 
little  variation. 

The  most  impressive  aspect  of  the  study  of  individual  variation 
of  O.  nuttalli  is  the  small  degree  of  it  when  compared  to  other 
cricetines.  The  relative  lack  of  much  individual  variation  in  golden 
mice  suggests  the  species  to  be  extremely  genetically  homogenous. 

Genus  Ochrotomys  Osgood 

1909.    Ochrotomys  Osgood,  N.  Amer.  Fauna,  28:222,  April  17.    Type,  Arvicola 
nuttalli  Harlan. 

Diagnosis. — Size  medium  (total  length  in  adults  from  140-190);  tail  longer, 
or  about  equal  to,  head  and  body;  hind  foot  in  adults  15-21;  ears  medium 
(10-18)  and  rounded;  color  of  pelage  of  young  slightly  duskier  than  of  adults, 
varying  on  dorsum  from  7. SYR  5/6  (young)  to  5YR  5/8  (adults)  on  the  Mun- 
sell  (1942)  system;  color  of  adults  tawny  ochraceous  on  upper  parts  and  ears, 
creamy  with  ochraceous  wash  on  underparts;  tail  faintly  bicolored;  feet  similar 
to  underparts  in  color;  plantar  tubercles  six  with  rudimentary  seventh  adjacent 
to  large  tubercle  at  base  of  fifth  digit;  mammae  6  (inguinal  2/2,  pectoral  1/1); 
posterior  palatine  foramina  nearer  interpterygoid  fossa  than  to  posterior  termi- 
nus of  anterior  palatine  foramen;  mental  foramen  of  mandible  situated  laterally 
and  in  medial  plane  of  ramus;  molariform  teeth  with  tendency  to  brachyodonty; 
enamel  folds  compressed  and  thick,  those  of  two  sides  of  a  molar  touching 
when  worn  and  resulting  in  five  subtriangular  islands  of  dentine  in  Ml  and  ml, 
four  in  M2  and  m2;  prominent  accessory  lophs  and  styles  on  all  molars,  full 
mesoloph  (-id),  mesostyle  (-id),  ectoloph,  and  ecotostylid,  mesoloph  (-id) 
joined  with  prominent  mesostyle  (-id)  apically;  loph  extends  lateral  from  mure; 


384  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

glans  penis  with  unusually  large  spines,  protractile  tip  and  adjoining  portions 
of  glans  without  spines,  bilobed  urethral  flap  protecting  meatus  urinarius,  distal 
margin  of  glans  body  scalloped;  baculum  capped  with  long  cartilaginous  cone, 
broad  at  base,  shaft  short;  electrophoretic  pattern  with  one  T  band,  one  po. 
A  band;  chromosome  diploid  number,  52;  dental  formula,  1/1,  0/0,  0/0,  3/3 
=  16. 

Geographic  Variation 

The  same  cranial  and  external  measurements  were  used  for  the 
analysis  of  geographic  variation  as  were  used  for  determining  the 
degree  of  variation  within  a  population.  Therefore,  13  character- 
istics were  analyzed  in  each  of  the  10  major  geographic  groupings 
( previously  described ) .  In  certain  instances,  where  relatively  large 
samples  permitted,  specimens  from  Georgia,  North  Carolina,  and 
Virginia  were  divided  each  into  coastal  plain,  piedmont,  and  moun- 
tain divisions  for  analysis.  Only  adult  specimens  (age  classes  3-5) 
were  used  in  this  study  (measurements  are  summarized  in  Table  2). 

Total  length. — Specimens  from  the  coastal  plains  of  Virginia,  North  Caro- 
lina, and  Georgia  average  larger  in  this  measurement  than  those  from  the 
Piedmont  Region  of  the  same  states.  Populations  from  Illinois,  Missouri,  Texas, 
and  the  Gulf  Coastal  plains  of  Louisiana,  Mississippi,  and  northern  Florida  are 
smallest  in  this  character  save  for  the  sample  from  Amelia  Court  House, 
Virginia  (which  merits  special  consideration,  see  summary  of  geograpliical 
variation  and  systematic  account  of  subspecies).  There  is  considerable  overlap 
in  samples  but  clinal  variation  from  Atlantic  Coastal  to  Piedmont  populations 
seems  evident.  There  is  a  rather  sharp  break  in  this  cline  of  decreasing  size  in 
that  populations  from  the  Appalachian  Mountains  average  larger  than  those 
on  the  Piedmont. 

Degree  of  variation  in  each  sample,  expressed  by  coefficient  of  variation, 
is  considerably  greater  in  external  characters  than  in  cranial  features  (see 
Table  2),  and  conclusions  concerning  trends  in  variations  may  be  less  mean- 
ingful than  those  revealed  by  cranial  measurements. 

Length  of  body. — Trends  in  variation  in  this  characteristic  are  similar  to 
those  for  total  length,  but  several  departures  of  note  exist.  Body  length  of 
specimens  from  Arkansas  ( Ozark  Mountains )  is  as  large  as  in  those  from  the 
Atlantic  Coastal  Plains.  Specimens  from  the  peninsula  of  Florida  and  the  Gulf 
Coast  region  of  Louisiana,  Mississippi,  and  Florida  are  smallest  in  body  length. 

Length  of  fail. — Pattern  of  variation  in  this  characteristic  resembles  that  in 
total  length.  A  notable  exception  is  the  Arkansas  sample  wherein  length  of  tail 
does  not  coincide  with  the  pattern  of  variation  in  length  of  body.  Thus,  mice 
from  Arkansas  have  large  bodies  with  proportionately  shorter  tails.  Samples 
from  the  mountainous  regions  of  Georgia,  North  Carolina,  Virginia,  western 
Virginia,  Tennessee,  and  Kentucky  reveal  considerable  concordance  in  having 
longer  tails  than  do  samples  from  the  Piedmont. 

Length  of  hind  foot. — This  measurement  \'aries  in  much  the  same  way  as 
do  the  foregoing  features.  Specimens  from  Texas,  Illinois,  Missouri,  and  the 
Gulf   Coastal   region   have    the    smallest   hind   feet,    whereas    specimens    from 


Packard — Review  of  Ochrotomys  385 

Arkansas  average  longer  in  hind  foot  length  (lieing  similar  to  average  size  of 
samples  from  the  Appalachian  Mountains ) .  Specimens  from  Amelia  Court 
House,  Virginia,  axerage  as  long  in  this  character  as  do  specimens  from  the 
mountains  to  the  west,  and  the  total  range  of  variation  overlaps  considerably 
that  of  the  coastal  plains  populations. 

Length  of  ear. — A  cline  of  decrease  in  length  of  ear  from  coastal  plain  to 
Piedmont  in  Georgia,  then  a  reverse  to  larger  ear  size  in  the  mountains,  is  dis- 
cernible. This  is  similar  to  the  pattern  noted  in  total  length.  However,  the 
coastal  plains  populations  of  North  Carolina  are  nearly  identical  to  those  of 
the  Piedmont  and  both  have  ears  that  average  longer  than  do  those  of  popula- 
tions from  the  mountains.  Longest  ear  length  is  in  Uie  sample  from  the  Virginia 
Piedmont,  averaging  considerably  longer  than  in  specimens  from  the  Virginia 
coastal  plains  and  western  mountainous  regions.  Seemingly  there  is  considerably 
discordance  in  trends  of  this  character  when  compared  to  other  external 
measurements. 

Conchjlobasal  length. — There  is  little  detectable  pattern  of  variation  in  this 
feature.  Eastern  coastal  plain  and  Piedmont  samples  are  nearly  identical,  and 
average  somewhat  larger  than  do  specimens  from  the  mountains  of  North 
Carolina,  but  are  similar  in  size  to  those  from  the  western  Appalachians  ( Ken- 
tucky, Tennessee,  and  Virginia).  Specimens  from  Arkansas  axerage  somewhat 
longer  than  do  those  from  smroimding  areas  of  Texas,  Illinois,  and  Missouri, 
and  approach  the  size  of  the  sample  from  the  western  Appalachian  Mountains. 

Zygomatic  breadtli. — Two  patterns  of  \ariation  in  this  character  are  e\'ident: 
( 1 )  specimens  from  the  eastern  coastal  plains  axerage  broader  with  size  de- 
creasing westward  toward  the  mountains  ( this  is  particularly  evident  in  North 
Carolina ) ;  ( 2 )  Arkansas  mice  average  larger  than  those  from  eastern  Texas, 
Missouri,  and  Illinois. 

Depth  of  braincasc. — The  trends  in  variation  of  this  character  differ  little 
from  variation  of  total  length.  Samples  from  the  Appalachian  Mountains 
average  larger  than  those  from  the  Atlantic  Piedmont  and  coastal  plains.  A 
clinal  pattern  is  evident  between  Piedmont  and  the  coastal  plain  populations, 
with  the  latter  averaging  larger.  Populations  from  eastern  Texas  and  Arkansas 
axerage  smaller  than  eastern  samples  and  oxerlap  the  range  of  variation  ob- 
served in  the  Gulf  Coast  specimens  from   Louisiana,  Mississippi   and  Florida. 

Length  of  rostrtnn. — This  character  varies  in  much  the  same  way  as  con- 
dylobasal  length.  Specimens  from  Amelia  Court  House,  Virginia,  are  signifi- 
cantly smaller  in  rostral  length  than  all  eastern  coastal  specimens  examined. 
Specimens  from  eastern  Texas,  Arkansas,  Illinois,  and  Missouri  are  similar  in 
this  character. 

Alveolar  lengtli  of  maxillary  toothrow. — Variation  in  length  of  upper  molar 
row  is  essentially  non-clinal  and  little  agreement  exists  between  the  patterns  of 
variation  of  this  character  and  others.  Variation  in  length  of  ma.xillary  toothrow 
of  specimens  from  Amelia  Court  House  is  within  the  range  of  variation  of  the 
specimens  from  the  coastal  plain  and  Piedmont  of  Virginia. 

Post-palatal  length. — This  feature  shows  no  marked  patterns  of  variation. 
There  is  considerable  overlap  among  all  populations  studied.  Specimens  from 
the  coastal  plains  of  North  Carolina  and  Georgia  are  nearly  identical  to  those 
from  the  Piedmont  of  those  states,  whereas  coastal  plains  .specimens  from  Vir- 


386 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Table  2. — Analysis  of  geographic  variatiox  in  external  and  cr,\nial 
measurements  of  adxjlt  golden  mice.  parenthetical  abbreviations  fol- 
lowing  some   localities    are:     cp    ( coastal   plain ) ;    p    ( piedmont ) ;    m 

(  mountains  )  . 


Locality 

N 

Mean  ±  1  SD 

Range 

SE 

cv 

Total  Length 

North  Carolina  (CP) 

6 

174.16±4.79 

180-167 

1.95 

2.75 

North  Carolina  ( P ) 

46 

166.89+8.54 

183-146 

1.25 

5.12 

North  CaroUna  (  M ) 

24 

170.00±12.82 

191-151 

2.61 

7.54 

Virginia  (CP) 

26 

176.80±8.52 

200-165 

1.67 

4.82 

Virginia  (P) 

18 

163.94±9.63 

180-149 

2.27 

5.87 

Virginia-Tennessee-Kentucky  (M)      11 

168.00+10.61 

183-154 

3.19 

6.32 

Amelia,  Virginia 

11 

149.63+10.55 

170-134 

3.18 

7.06 

Georgia  (CP) 

13 

164.92+10.05 

183-150 

2.78 

6.10 

Georgia  (P) 

32 

155.00+10.50 

172-127 

1.85 

6.78 

Georgia  (M) 

10 

164.50+8.05 

175-150 

2.54 

4.90 

South  Carolina 

11 

164.36+7.07 

175-152 

2.13 

4.30 

Illinois-Missouri 

28 

156.28+9.66 

174-138 

1.82 

6.18 

Arkansas 

18 

164.66±8.60 

181-146 

2.02 

5.22 

Texas 

79 

159.11+9.68 

178-139 

1.08 

6.09 

Gulf  Coast 

51 

155.78+15.88 

200-106 

2.22 

10.20 

Florida 

28 

157.60+10.65 

178-135 

2.01 

6.75 

LengtJi  of  Head  and  Body 

North  Carolina  (CP) 

6 

91.50±3.27 

96-88 

1.33 

3.57 

North  Carolina  ( P ) 

46 

91.69+6.13 

104-79 

.90 

6.69 

North  Carolina  ( M ) 

24 

90.37+9.04 

105-77 

1.84 

10.01 

Virginia  (CP) 

26 

93.46+4.76 

103-83 

.93 

5.09 

Virginia  (P) 

19 

88.05+5.95 

98-75 

1.36 

6.77 

Virginia-Tennessee-Kentucky  ( M )     11 

88.72+6.70 

98-80 

2.02 

7.56 

Amelia,  Virginia 

11 

83.72+6.58 

94-75 

1.98 

7.87 

Georgia  (CP) 

13 

89.69+5.42 

99-81 

1.50 

6.05 

Georgia  (P) 

32 

85.34+8.20 

103-68 

1.45 

9.62 

Georgia  (M) 

10 

87.00±4.89 

95-75 

1.54 

5.60 

South  Carolina 

11 

86.72+3.16 

93-82 

.95 

3.65 

Illinois-Missouri 

28 

86.42+6.99 

100-71 

1.32 

8.09 

Arkansas 

18 

92.44+7.89 

102-73 

1.86 

8.54 

Texas 

49 

86.96+7.08 

115-74 

.79 

6.09 

Gulf  Coast 

51 

82.82+9.87 

115-51 

1.38 

11.92 

Florida 

28 
Length 

85.00+7.77 
of  Tail 

100-64 

1.46 

9.15 

North  Carolina  (CP) 

6 

82.16+5.81 

91-75 

2.37 

7.08 

Packard — Review  of  Ochrotomys 


387 


Table  2. — Continued. 


Locality 

N 

Mean  ±  1  SD 

Range 

SE 

CV 

North  Carolina 

(P) 

46 

75.17+5.44 

89-61 

.80 

7.24 

North  Carolina 

(M) 

24 

79.62±6.34 

94-70 

1.29 

7.97 

Virginia  (CP) 

26 

83.34±5.42 

97-73 

1.06 

6.51 

Virginia  (P) 

18 

75.94±5.09 

89-69 

1.20 

6.71 

Virginia  ( M ) 

11 

79.27±6.16 

90-71 

1.85 

7.78 

Ameha,  Virgini; 

a 

11 

65.90±7.27 

77-55 

2.19 

11.03 

Georgia  (CP) 

13 

75.23±6.28 

86-65 

1.74 

8.36 

Georgia  (P) 

32 

69.65+8.10 

85-51 

1.43 

11.64 

Georgia  ( M ) 

10 

76.50±5.03 

83-68 

1.59 

6.59 

South  Carolina 

11 

77.63±5.85 

87-68 

1.76 

7.54 

Illinois-Missouri 

29 

69.89±4.32 

81-62 

.80 

6.19 

Arkansas 

18 

72 .22  ±7. 08 

81-59 

1.66 

9.81 

Texas 

49 

72.12+5.48 

85-60 

.61 

7.60 

Gulf  Coast 

52 

72.90+8.15 

95-50 

1.13 

11.18 

Florida 

28 

72.32+5.80 

85-60 

1.09 

8.03 

Length  of  Hind  Foot 

North  Carolina  (CP) 

6 

18.66+.51 

19-18 

.21 

2.77 

North  Carolina  (P) 

46 

17.95+1.11 

20-16 

.16 

6.21 

North  Carolina 

(M) 

24 

18.25+.84 

20-17 

.17 

4.64 

Virginia  (CP) 

25 

19.60+.57 

21-19 

.11 

2.95 

Virginia  (P) 

19 

19.05+.62 

20-18 

.14 

3.26 

Virginia  ( M ) 

9 

18.88+.60 

20-18 

.20 

3.18 

Amelia,  Virginit 

I 

6 

18.50+1.40 

20-17 

.42 

5.67 

Georgia  (CP) 

13 

18.07+.95 

20-17 

.26 

5.23 

Georgia  (P) 

32 

17.53+. 94 

19-16 

.16 

5.42 

Georgia  (M) 

10 

18.20+.91 

20-17 

.29 

5.05 

South  Carolina 

11 

19.00+.77 

20-18 

.23 

4.08 

Illinois-Missouri 

29 

17.72+1.70 

20-12 

.31 

9.64 

Arkansas 

18 

18.44+1.29 

20-14 

.30 

7.01 

Texas 

49 

17.83+1.20 

21-15 

.13 

6.74 

Gulf  Coast 

50 

18.26+2.22 

29-13 

.31 

12.21 

Florida 

28 
Length 

17.78+1.34 
of  Ear 

19-13 

.25 

7.55 

North  Carolina 

(CP) 

6 

16.50+1.04 

18-15 

.42 

6.36 

North  Carohna 

(P) 

38 

16.31  +  1.11 

20-14 

.18 

6.85 

North  Carolina  (  M ) 

12 

15.00+1.47 

17-13 

.42 

9.85 

Virginia  (CP) 

2 

17.50+2.12 

19-16 

1.50 

12.12 

Virginia  (P) 

10 

18.60+.96 

20-17 

.30 

5.19 

388 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Table  2. — Continued. 


Locality 

N 

Mean  ± 

1  SD 

Range 

SE 

cv 

Virginia 

(M) 

8 

17.37± 

1.40 

20-16 

.49 

8.10 

Amelia, 

Virginia 
(CP) 

6 

17-15 

.36 

Georgia 

16.00± 

.89 

5.59 

Georgia 

(P) 

11 

14.45± 

2.25 

18-12 

.67 

15.58 

Georgia 

(M) 

3 

15.33± 

.57 

16-15 

.33 

3.77 

South  Carolina 

4 

15.50± 

1.29 

17-14 

.64 

8.33 

Illinois-Missouri 

29 

15.41± 

1.57 

17-10 

.29 

10.19 

Arkansas 

11 

16.45± 

.68 

18-16 

.20 

4.18 

Texas 

46 

15.72± 

1.53 

18-11 

.17 

9.78 

Gulf  Coast 

41 

16.26± 

2.51 

26-12 

.39 

13.22 

Florida 

27 

15.88± 

1.36 

21-14 

.26 

8.61 

Cotuhjlobasal  Lengtl 

I 

North  Carolina  (CP) 

4 

23.22± 

1.22 

24.2-21.5 

.61 

5.29 

North  Carolina 

(P) 

32 

23.32± 

.68 

24.6-22.3 

.12 

2.92 

North  Carolina 

(M) 

26 

22.74± 

.95 

24.7-21.0 

.18 

4.18 

Virginia 

(CP) 

24 

23.94± 

.62 

25.1-22.8 

.12 

2.61 

Virginia 

(P) 

15 

22.68± 

.67 

23.7-21.8 

.17 

2.95 

Virginia 

(M) 

11 

23.58+ 

.87 

24.6-21.6 

.26 

3.71 

Amelia, 

Virginia 

5 

22.40± 

.75 

23.7-21.9 

.33 

4.70 

Georgia 

(CP) 

9 

23.26± 

.83 

24.6-22.3 

.27 

3.61 

Georgia 

(P) 

17 

22.41  ± 

1.72 

25.0-17.0 

.41 

7.69 

Georgia 

(M) 

8 

22.90± 

.71 

24.1-22.1 

.25 

3.13 

South  Carolina 

6 

23.46± 

.55 

24.1-22.7 

.22 

2.36 

Illinois-Missouri 

25 

22.70± 

.51 

23.8-21.9 

.10 

2.27 

Arkansas 

> 

23 

23.30± 

.45 

24..3-22.4 

.09 

1.96 

Texas 

43 

22.66± 

.61 

23.8-21.4 

.07 

2.70 

Gulf  Coast 

42 

22.43± 

1.48 

25.0-18.6 

.22 

6.63 

Florida 

22 

22.70± 

.97 

24.3-20.8 

.20 

4.28 

ZijgomaUc  Breadth 

North  Carolina 

(CP) 

4 

13.80± 

.14 

14.0-13.7 

.07 

1.02 

North  Carolina 

(P) 

33 

13.37± 

.37 

14.3-12.7 

.06 

2.78 

North  Carolina 

(M) 

26 

13.18± 

.54 

14.3-12.2 

.10 

4.16 

Virginia 

(CP) 

26 

13.85± 

.36 

14.5-13.0 

.07 

2.62 

Virginia 

(P) 

15 

13.19± 

38 

13.9-12.6 

.10 

2.94 

Virginia 

(M) 

11 

13.56± 

41 

14.2-12.7 

.12 

3.06 

Amelia, 

Virginia 

5 

13.10± 

27 

13.5-12.8 

.12 

2.09 

Georgia 

(CP) 

9 

13.56± 

50 

14.4-12.7 

.16 

3.69 

Packard — Review  of  Ochrotomys  389 

Table  2. — Continued. 


Locality 

N 

Mean  ±  1  SD 

Range 

SE 

cv 

Georgia  (P) 

21 

12.90  +  .75 

13.8-10.6 

.16 

5.87 

Georgia  ( M ) 

8 

I3.25±.45 

13.9-12.6 

.16 

3.45 

South  Carolina 

8 

13.56±.46 

14.3-13.0 

.16 

3.46 

Illinois-Missouri 

26 

12.99±.42 

13.8-11.8 

.08 

3.27 

Arkansas 

22 

13.45±.37 

14.2-12.7 

.08 

2.79 

Texas 

43 

13.18±.42 

13.9-12.3 

.04 

3.20 

Gulf  Coast 

46 

13.10±.72 

14.7-11.3 

.10 

5.56 

Florida 

24 
Depth  of  B 

13.17+.44 
raincase 

14.1-12.1 

.09 

3.37 

North  Carolina  (CP) 

4 

9.90±.86 

10.9-9.0 

.43 

8.69 

North  Carolina  ( P ) 

25 

9.52±.28 

10.2-9.0 

.05 

2.99 

North  Carolina  (  M  ) 

13 

9.55±.26 

10.0-9.2 

.07 

2.82 

Virginia  (CP) 

20 

9.55±.22 

10.1-9.2 

.05 

2.39 

Virginia  (P) 

15 

9.52±.42 

10.3-8.4 

.11 

4.44 

Virginia  (M) 

11 

9.87±.18 

10.1-9.6 

.05 

1.87 

Amelia,  Virginia 

4 

9.00±.40 

9.2-8.4 

.20 

4.44 

Georgia  (CP) 

5 

9.64±.38 

10.3-9.3 

.17 

3.99 

Georgia  (P) 

14 

9.35±.55 

10.7-8.3 

.15 

5.99 

Georgia  (M) 

4 

9.43±.28 

9.7-9.1 

.14 

2.92 

South  Carolina 

4 

9.80±.41 

10.1-9.2 

.20 

4.17 

Illinois-Missouri 

25 

9.29±.34 

9.9-8.7 

.07 

3.69 

Arkansas 

1 

9.30 

Texas 

43 

9.37±.19 

9.7-9.0 

.02 

2.10 

Gulf  Coast 

33 

9.28i!i.44 

10.0-8.0 

.08 

4.71 

Florida 

21 

9.67±.28 

9.8-8.5 

.06 

2.97 

Length  of  Rostrtim 

North  Carolina  (CP) 

4 

9.65±.37 

10.2-9.4 

.18 

3.83 

North  Carolina  (P) 

30 

9.72+.5I 

10.6-8.6 

.09 

5.33 

North  Carolina  (  M ) 

17 

9.34±.42 

10.2-8.7 

.10 

4.54 

Virginia  (CP) 

23 

9.61±.37 

10.2-9.0 

.08 

3.80 

Virginia  (P) 

16 

9.25±.35 

9.8-8.8 

.09 

3.79 

Virginia  (M) 

11 

9.73±.71 

10.5-8.3 

.21 

7.30 

Amelia,  Virginia 

11 

8.53±.48 

9.6-7.8 

.14 

5.63 

Georgia  (CP) 

5 

8.90+.82 

9.8-7.8 

.37 

9.27 

Georgia  (P) 

20 

8.98±.92 

10.1-5.9 

.21 

10.23 

Georgia  (M) 

6 

9.12±.13 

9.3-9.0 

.05 

1.46 

South  Carolina 

5 

9.62±.55 

10.3-8.8 

.24 

5.67 

390 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Table  2. — Coxtixued. 


Locality 

N 

Mean  ±  1  SD 

Range 

SE 

CV 

Illinois-Missouri 

26 

9.67±.55 

10.7-8.8 

.11 

5.70 

Arkansas 

1 
42 

10.00 

9.75±.49 

.06 

Texas 

10.8-8.0 

5.12 

Gulf  Coast 

38 

9.14+.70 

10.3-7.4 

.11 

7.72 

Florida 

19 

9.24±.53 

10.1-8.1 

.12 

5.79 

Length  of  Upper 

Molar  Toothrow 

North  Carolina  (CP) 

4 

3.80±.14 

3.9-3.6 

.07 

3.72 

North  Carolina  (P) 

37 

3.66±.15 

3.9-3.1 

.02 

4.09 

North  Carolina  (  M ) 

18 

3.80+.13 

4.0-3.5 

.03 

3.43 

Virginia  (CP) 

23 

3.56+.12 

3.8-3.4 

.03 

3.45 

Virginia  (P) 

16 

3.72±.13 

3.9-3.5 

.03 

3.43 

Virginia  ( M ) 

11 

3.90+.54 

4.0-3.8 

.02 

1.38 

Amelia,  Virginia 

11 

3.65±.14 

3.9-3.4 

.04 

3.76 

Georgia  (CP) 

5 

3.60±.16 

3.8-3.4 

.07 

4.39 

Georgia  (P) 

19 

3.67+.18 

3.9-3.1 

.04 

4.81 

Georgia  ( M ) 

7 

3.54±.23 

3.9-3.3 

South  Carolina 

5 

3.82±.25 

4.2-3.6 

.11 

6.52 

Illinois-Missouri 

26 

3.76±.13 

4.0-3.5 

.03 

3.52 

Arkansas 

1 
43 

3.70 
3.73+.10 

.01 

Te.xas 

3.9-3.5 

2.71 

Gulf  Coast 

39 

3.77±.10 

3.9-3.7 

.01 

2.08 

Florida 

22 

3.67+.10 

3.9-3.5 

.02 

2.81 

Post-palatal  Length 

North  Carolina  (CP) 

4 

8.98±.67 

9.5-8.1 

.33 

7.47 

North  Carolina  (  P ) 

25 

8.99±.48 

9.8-8.0 

.09 

5.28 

North  Carolina  (  M ) 

13 

9.14±.64 

10.4-8.2 

.18 

7.06 

Virginia  (CP) 

19 

9.02+.41 

9.7-8.3 

.09 

4.55 

Virginia  (P) 

14 

8.76±.36 

9.4-8.1 

.09 

4.09 

Virginia  (M) 

3 

8.80±.60 

9.4-8.2 

.35 

6.82 

Amelia,  Virginia 

5 

8.46±.39 

9.1-8.1 

.17 

4.62 

Georgia  (CP) 

5 

8.66+.15 

9.2-8.3 

.15 

3.88 

Georgia  (P) 

15 

8.72±.86 

9.7-6.2 

.22 

9.88 

Georgia  (M) 

5 

9.10±.45 

9.6-8.5 

.20 

4.98 

South  Carolina 

3 

9.60±.79 

10.5-9.0 

.46 

8.27 

Illinois-Missouri 

25 

8.67±.38 

9.8-8.3 

.08 

4.29 

Arkansas 

1 

9.20 

Texas 

42 

9.27±.39 

9.9-8.5 

.05 

4.21 

Packard — Review  of  Ochrotomys 


391 


Table  2. — Concluded. 


Locality 

N 

Mean  ±  1  SD 

Range 

SE 

cv 

Gulf  Coast 

33 

8.70+.77 

10.1-7.1 

.13 

8.89 

Florida 

21 

8.8I±.44 

9.5-8.0 

.09 

4.95 

Length  of  Lower  Molar  Toothroiv 

North  Carolina  (CP) 

4 

4.00±.16 

4.2-3.8 

.08 

4.08 

North  Carolina  (P) 

30 

3.69±.21 

4.0-3.2 

.04 

5.08 

North  Carolina  ( M ) 

16 

3.90+.17 

4.1-3.4 

.04 

4.23 

Virginia  (CP) 

23 

3.59±.12 

3.9-3.4 

.03 

3.40 

Virginia  (P) 

16 

3.84±.14 

4.0-3.5 

.04 

3.67 

Virginia  ( M ) 

3 

4.00+.20 

4.2-3.8 

.12 

5.00 

Amelia,  Virginia 

11 

3.69+.15 

4.0-3.5 

.05 

4.10 

Georgia  (CP) 

5 

.    3.60±.32 

3.9-3.2 

.14 

9.00 

Georgia  (P) 

20 

4.11  +  .17 

4.3-3.7 

.04 

4.07 

Georgia  (M) 

6 

3.48±.30 

3.9-3.2 

.12 

8.79 

South  Carolina 

4 

4.33±.24 

4.5-4.0 

.12 

5.47 

Illinois-Missouri 

26 

3.84±.19 

4.1-3.2 

.04 

4.95 

Arkansas 

1 

3.70 

i  1,1  fvaiiiimtj 

Texas 

19 

3.77±.09 

4.0-3.6 

.02 

2.48 

Gulf  Coast 

26 

3.86±.08 

4.0-3.7 

.02 

2.08 

Florida 

22 

3.80±.12 

4.5-3.7 

.04 

4.13 

Least  Interorbital  Breadth 

' 

North  Carolina  ( CP ) 

4 

4.50+.14 

4.6-4.3 

.07 

3.14 

North  Carolina  (P) 

37 

4.24+.15 

4.5-3.9 

.02 

3.53 

North  Carolina  ( M ) 

18 

4.31±.09 

4.5-4.2 

.02 

2.09 

Virginia  (CP) 

22 

4.12+.19 

4.4-3.7 

.04 

4.49 

Virginia  (P) 

16 

4.29+.16 

4.5-3.9 

.04 

3.80 

Virginia  ( M ) 

11 

4.47±.15 

4.7-4.3 

.04 

3.33 

Amelia,  Virginia 

11 

4.00±.13 

4.2-3.8 

.04 

3.24 

Georgia  (CP) 

5 

4.08±.24 

4.4-3.8 

.11 

5.85 

Georgia  (P) 

20 

4.11±.17 

4.3-3.6 

.04 

4.07 

Georgia  ( M ) 

6 

3.91±.21 

4.2-3.6 

.09 

5.46 

South  Carolina 

5 

4.36±.ll 

4.5-4.2 

.05 

2.61 

Illinois-Missouri 

26 

4.10+.20 

4.6-3.8 

.04 

4.86 

Arkansas 

1 

4.00 

Texas 

43 

4.38+.17 

4.8-4.0 

.02 

3.99 

Gulf  Coast 

40 

4.18±.14 

4.0-3.9 

.02 

3.88 

Florida 

22 

4.22±.17 

4.5-3.7 

.04 

4.13 

392  Misc.  Publ.  51,  Uxiv.  Kansas  Mus.  Nat.  Hist. 

ginia  average  shorter  than  mice  from  the  Piedmont  and  mountains  in  that  state. 
Specimens  from  the  mountains  of  Georgia  average  somewhat  longer  than  those 
from  the  Georgia  coastal  plains  and  Piedmont.  Mice  from  eastern  Texas  average 
longer  than  samples  from  Illinois  and  Missouri.  Mice  from  South  Carolina, 
mostly  from  the  Piedmont,  average  longest  of  all  specimens  examined  but 
overlap  considerably  coastal  and  Piedmont  samples  from  Georgia  and  North 
Carolina. 

Length  of  mandibular  tootlirow. — Mice  from  South  Carolina  have  a  signifi- 
cantly longer  lower  molar  row  than  all  other  samples  studied;  slight  overlap 
occurs  with  the  coastal  plains  population  from  North  Carolina  and  mice  from 
the  mountains  of  Tennessee,  western  Virginia,  and  Kentucky.  In  North  Caro- 
lina, mice  possessing  longer  molar  rows  are  on  the  coastal  plains  with  decreasing 
length  on  the  Piedmont  and  increasing  length  in  mountain  samples.  There  is 
concordance  between  all  Piedmont  samples.  Specimens  from  the  mountains  of 
Georgia  ha\'e  shortest  lower  molar  rows.  There  is  a  cline  of  decreasing  size 
from  northernmost  populations  progressing  southward  into  the  mountains  of 
Georgia.  Specimens  from  Texas,  Illinois,  Missouri,  and  Gulf  Coast  areas  of 
Louisiana,  Mississippi,  and  Florida  are  nearly  identical  in  this  feature.  Mice 
from  peninsular  Florida  are  considerably  longer  in  this  measurement  than  are 
adjoining  populations  from  the  coastal  plains  of  Georgia. 

Least  interorhital  breadth. — A  cline  of  decreasing  size  from  the  northern 
Appalachians  of  western  Virginia  south  to  Georgia  is  e\'ident.  A  cline  involving 
decrease  in  size  is  also  evident  from  the  coastal  plains  of  the  Carolinas  to  the 
Piedmont,  but  this  trend  is  not  exident  in  Georgia.  There  the  mountain  popu- 
lations are  smaller  than  Piedmont  and  coastal  plains  samples  which  resemble 
each  other  closely.  Samples  from  the  mountains  of  North  Carolina  and  Virginia 
are  larger  in  this  feature  than  Piedmont  populations.  Eastern  Texas  popula- 
tions, average  considerably  larger  than  specimens  from  Illinois  and  Missouri, 
and  those  from  the  Gulf  Coast. 

Color  of  pelage. — Specimens  from  the  Atlantic  Coastal  Plain  of  \'irginia, 
the  Carolinas,  and  Georgia  are  some\\hat  brighter  (more  reddish  yellow)  than 
those  from  the  Piedmont  and  mountainous  areas  to  the  \\est,  which  are  more 
brownish  as  a  result  of  blacker  overtones  ( greater  suffusion  of  black  guard 
hairs).  Mice  from  Texas,  northern  Louisiana,  Missouri,  and  Illinois  tend  to  a 
brighter  color  with  yellowish  o\ertones,  whereas  those  from  the  Florida  Penin- 
sula are  rich  yellowish  brown. 

Summary  of  Patterns  of  Geographic  Variation 

Several  trends  of  variation  seem  well-defined  as  follows:  (1)  a 
eline  of  deereasing  size  from  north  to  south  exists  in  the  Appalachian 
Mountains  in  all  external  features  studied  and  in  certain  cranial 
features  ( condylobasal  length,  depth  of  braincase,  length  of  rostrum, 
length  of  maxillary  toothrow,  length  of  mandibular  toothrow),  yet 
these  populations  show  more  concordance  with  each  other  than 
they  do  with  those  of  the  eastern  Piedmont  or  coastal  plains;  (2)  a 
cline  of  decreasing  size  in  breadth  of  cranium,  interorbital  breadth, 
post-palatal  length,  and  depth  of  braincase  occurs  from  the  Atlantic 


Packard — Review  of  Ociirotomys  393 

Coastal  Plain  to  the  Piedmont;  (3)  a  decreasing  cline  in  external 
size  ( total  length,  length  of  body,  length  of  tail,  and  length  of  hind 
foot)  exists  between  Atlantic  Coastal  Plains  and  Piedmont  popula- 
tions; samples  from  the  Appalachian  Mountains  mostly  average 
large  in  these  characters;  (4)  specimens  from  Arkansas  have  longer 
hind  feet,  larger  crania,  but  shorter  tails  than  surrounding  popula- 
tions from  Missouri,  Illinois,  and  eastern  Texas;  (5)  specimens  from 
Amelia  Court  House,  Virginia,  are  smaller  in  nearly  all  characters 
that  would  be  expected  to  be  small  in  a  subadult  sample,  whereas  in 
characters  such  as  length  of  toothrows  that  approach  maximal  size 
in  subadults  (class  two  mice),  they  are  not  significantly  different 
from  surrounding  populations;  (6)  Gulf  Coast  populations  from 
Louisiana,  Mississippi,  and  northwestern  Florida  are  highly  variable 
(see  coefficients  of  variation  of  measurements)  suggesting  a  secon- 
dary zone  of  intergradation;  (7)  South  Carolina  (mostly  Piedmont) 
mice  have  a  significantly  long  lower  molar  toothrow;  (8)  specimens 
from  peninsular  Florida  (Ocala  Ridge)  ha\c  the  brightest  orange- 
yellow  color  of  pelage  ( specimens  to  the  north  ha\'e  a  reddish  brown 
overtone,  and  those  to  the  west  more  yellowish  red),  with  little 
variation  in  color  of  pelage  throughout  other  parts  of  the  geographic 
range;  (9)  specimens  from  Amelia  Court  House,  Virginia,  are  a  dull 
yellow  with  blackish  overtones  similar  to  age  two  mice  ( subadults ) . 
Seemingly  there  are  five  groups  of  golden  mice  that  are  distinct 
enough  from  each  other  to  warrant  subspecific  recognition  as  fol- 
lows :  ( 1 )  Atlantic  Coastal  plains  and  Piedmont  populations  ( clinal 
variation  exists  between  these  populations  and  the  oxerlap  in  char- 
acters studied  is  of  such  magnitude  that  they  should  be  considered 
as  one  subspecies);  (2)  Appalachian  Mountains  populations  with 
specimens  from  the  mountains  in  Georgia  smaller  in  most  features 
studied  than  those  to  the  north  in  North  Carolina  and  Virginia; 
( 3 )  specimens  from  peninsular  Florida,  with  much  brighter  orange- 
colored  pelage  and  longer  molar  rows  than  in  other  races;  (4)  popu- 
lations from  eastern  Texas,  northern  Louisiana  and  Mississippi, 
western  Tennessee,  Missouri,  and  Illinois;  (5)  the  mountainous 
region  of  Arkansas  and  southern  Missouri. 

Natural  History 

Populations  and  habitat. — McCarley's   (1958)   study  in  eastern 

Texas  revealed  a  population  density,  in  a  lowland  flood  plain  with 

hardwoods  and  considerable  underbrush,  varying  from  as  few  as 

.3  per  acre  in  summer  to  as  many  as  2.2  per  acre  in  early  spring. 


394  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Home  range  in  the  same  area  was  1.46  ±  .17  acres  for  males  and 
1.40  ±  .22  acres  for  females.  Sex  ratios  in  the  study  in  eastern 
Texas  did  not  depart  significantly  from  the  expected  1:1.  McCarley 
( 1959a )  found  that  irradiation  of  500  R  on  male  testes  reduced  the 
population  density  in  similar  habitat. 

The  habitat  of  golden  mice  varies  from  flood  plain,  principally 
hardwood  forests  with  underbrush,  to  upland  pine-oak  and  pine 
stands.  Specific  notes  on  habitat  are:  Virginia,  Handley  and  Patton 
(1947)  and  Handley  (1948),  woods  and  thickets  usually  associated 
with  honeysuckle  and  occasionally  in  cane  brakes;  South  Carolina, 
Golley  (1966),  in  wooded  and  brushy  areas  particularly  in  thickets 
of  honeysuckle  and  greenbrier;  Georgia,  Golley  (1962),  lowland 
swamp  forest  to  open  drier  upland  forest;  Tennessee,  Kellogg 
( 1939 ) ,  in  hemlock  among  moss-covered  boulders,  swampy  wood- 
lands, along  borders  of  broomsedge  fields,  and  in  brier  patches; 
Kentucky,  Barbour  (1942)  and  Goodpaster  and  Hoffmeister  (1954), 
in  brushy  valleys  with  beech,  dogwood,  greenbrier,  and  honey- 
suckle; Illinois,  Hoffmeister  and  Mohr  (1957),  in  thick  timber  bor- 
dering cypress  swamps;  Missouri,  Schwartz  and  Schwartz  (1959) 
and  Easterla  (1968),  in  moist  thickets,  forests,  borders  of  broom- 
sedge  fields,  and  rocky-oak  hillsides;  Arkansas,  Sealander  (1956), 
in  oak-pine  uplands;  Louisiana,  Lowery  (1936,  1943),  in  habitat 
similar  to  that  reported  for  eastern  Texas;  Texas,  McCarley  ( 1959b), 
Davis  (1960),  Packard  and  Garner  (1964),  upland  pine-oak  wood- 
land, heavily  forested  hardwood  Hood  plain,  and  on  hillsides  with 
considerable  lianas  such  as  grapevine  and  honeysuckle.  Packard 
( 1968 ) ,  while  studying  fulvous  harvest  mice,  found  the  distribution 
of  golden  mice  on  the  same  area  of  intensive  study  (where  eight 
habitat  types  were  present)  to  be  significantly  correlated  with  that 
of  the  pine-oak  habitat. 

Behavior. — Golden  mice  are  quite  docile  in  comparison  with 
other  peromyscoid  mice,  and  are  much  less  aggressive  than  pygmy 
mice  (Baiomijs),  which  are  about  half  their  size.  Golden  mice  seem 
quite  social  (see  McCarley,  1958,  and  Dunaway,  1955).  Both  ar- 
boreal globular  nests  and  ground  nests  are  utilized  for  feeding 
(Goodpaster  and  Hoffmeister,  1954)  and  rearing  litters  (Packard 
and  Garner,  1964).  Layne  (1960)  reported  in  detail  the  behavior 
of  young  and  parental  care.  He  also  suggested  that  golden  mice 
become  increasingly  wild  after  long  periods  of  captivity.  The  tail  is 
frequently  used  prehensily  in  climbing  and  in  moving  through  xines 
(see  Packard  and  Garner,   1964,  and  Rippy  and  Harvey,   1963). 


Packard — Review  of  Ochrotomys  395 

McCarley  ( 1959b )  reported  these  mice  as  having  a  rather  distinct 
odor  different  from  that  of  other  peromyscoids.  My  studies  of 
golden  mice,  both  in  the  field  and  laboratory,  suggest  a  crepuscular 
and  nocturnal  activity  cycle. 

Reproduction. — Golden  mice  may  be  polyestrous  (see  Good- 
paster  and  Hoftmeister,  1954)  in  certain  parts  of  their  geographic 
range  and  monestrous  in  other  parts  (see  McCarley,  195<S).  Breed- 
ing occurs  chiefly  in  the  winter  to  late  winter  period  followed  by  a 
gestation  period  of  25  to  30  days  (average  of  15  litters,  27  days); 
two  to  four  young  are  born  per  litter  (average  2.65  in  85  litters — 
see  Linzey  and  Linzey,  1967b). 

Accounts  of  Subspecies 

Ochrotomys  nuttalli  aureolus  (Audubon  and  Bachman) 
Mus  (Calomys)  aureolus  Audubon  and  Bachman,  Proc.  Acad.  Nat.  Sci.  Phila- 
delphia, 1:98,  1841. 
[Peromyscus  nuttalli]  aureolus,  Elliot,  Field  Columb.   Mus.,  Zool.   Ser.,  2:140, 

1901  (part). 
Ochrotomys  nuttali  [sic]  aureolus,  Rippy  and  Harvey,  Trans.  Kentucky  Acad. 
Sci.,  24:5,  1963. 

Holotype. — Not  known  to  exist.  Because  no  holotype,  syntypes,  nor  lecto- 
type  are  known  to  be  extant,  I  designate  as  a  neotype  an  adult  female,  skin  and 
skull,  no.  104075  University  of  Michigan,  Museum  of  Zoology;  from  Marshall, 
Madison  Co.,  North  Carolina  ( a  place  approximately  45  miles  northwest  of  the 
South  Carolina  state  line). 

Range. — Appalachian  Mountains,  from  Clark  County  Virginia  south  through 
North  and  South  Carolina  into  northwestern  Georgia,  thence  northward  into 
eastern  one-half  of  Tennessee  and  Kentucky  ( see  Fig.  1 ) . 

Diagnosis. — Size  medium  to  large  for  the  species;  external  and  cranial 
characters  of  populations  in  north  axerage  larger  than  those  to  the  south  in  the 
mountains;  middorsal  region  reddish  brown  5YR  4/4,  ears  of  same  color;  belly 
dusky  cinnamon,  sides  dusky  tawn.  A  dark  reddish  brown  colored  subspecies 
resulting  from  a  profusion  of  black  guard  hairs  from  the  nape  posterior  to  the 
base  of  the  tail  (see  Table  2  for  measurements  of  samples  from  North  Carolina 
mountainous  area ) . 

Comparisons. — For  comparisons  with  O.  n.  lisae,  O.  n.  nuttalli,  O.  n.  flori- 
clanus,  see  accounts  of  those  subspecies.  From  O.  n.  flammeus,  to  the  west 
O.  n.  aureolus  differs  in:  ears  and  dorsal  color  of  body  darker  reddish  brown; 
belly  duskier,  tail  darker  above;  total  length  and  length  of  tail  averaging  longer; 
ears  average  smaller. 

Remarks. — In  most  internal  and  cranial  characters  studied,  O.  ;i.  aureolus 
shows  decrease  in  size  from  Virginia  southward  into  Georgia.  Mice  from  central 
Kentucky  and  Tennessee  show  intergradation  in  color  and  size  from  O.  n.  lisae 
to  the  west,  but  seem  best  referred  to  aureolus.  Golden  mice  in  the  Appala- 
chian  Mountains  average   larger  than  Piedmont  samples   to   the  east  in  most 


396  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

external  and  certain  cranial  featvnes  ( depth  of  braincase,  post-palatal  length, 
length  of  mandibular  toothrow).  Mice  from  the  mountains  are  duskier 
reddish-brown  than  those  from  the  Piedmont  and  coastal  plains.  This  results 
from  a  greater  profusion  of  black  guard  hairs  on  the  dorsum  in  the  samples 
from  the  mountains.  Because  of  the  sharp  break  in  the  clines  detected  from  the 
coastal  plain  to  the  Piedmont,  mice  from  the  mountains  comprise  a  distinct 
subspecies  to  which  I  restrict  the  name  O.  n.  aureolus. 

The  original  type  localit>'  for  aureolus,  "oak  forests  of  South  Carolina,"  is  of 
little  use,  because  oak  forests  are  found  in  both  upland  and  lowland  areas  there. 
I  have  chosen  to  restrict  the  type  locality  to  the  mountains  ( pine  and  oak ) ,  and 
also  select  a  place  as  close  to  the  original  generally  defined  area  as  the  avail- 
ability of  a  suitable  specimen  would  permit.  The  result  of  selecting  a  new  type 
locality  also  has  the  adxantage  of  more  accurately  correlating  the  patterns  of 
variation  detected  in  this  species. 

Specimens  examined  (169). — Georgia.  Rahien  Co.:  unspecified  locality,  1 
(UG).  Stephens  Co.:  Toccoa,  1  (USNM).  Union  Co.:  Margret,  2  (UG),  3 
(USNM).  Walker  Co.:  Pinetucky,  5  (AMXH).  Kentucky.  Barrew  Co.: 
.45  mi.  ENE  Little  Hope  Church,  1   (UI).    Carter  Co.:  10  mi.  E  Olive  Hill, 

1  (UI).  Christian  Co.:  1  mi.  NNE  Parklodge,  1  (UI).  Edmondson  Co.:  Mam- 
moth Nat'l  Park,  1  (KU).  Flemminfi  Co.:  Crancreek,  2  (UI);  8  mi.  S  Flem- 
mingsburg,  11  (UG);  Wa'lingford,  1  (KU).  Hopkins  Co.:  1.37  mi.  NW  junc- 
tion of  Ky.  109  and  US  62,  1  (UI).  Lewis  Co.:  Salt  Lick  Creek,  2M  mi.  W 
Charters,  1  (UI);  eVz  mi.  W  Vanceburg,  2  (UI);  6  mi.  E  Vanceburg,  14  (UI); 

2  mi.  S  Vanceburg,  1  (KU);  1  mi.  W  Vanceburg,  1  (UI).  Madison  Co.:  2  mi. 
S  Big  Hill,  3  (UI).  McCreahj  Co.:  4)i  mi.  NE  Greenwood,  2  (UI).  Ptdaski 
Co.:  Eubanks,  7  (USNM).  Roican  Co.:  Morehead,  1  (USNM);  2  mi.  S  Rod- 
burn,  1  (UI);  Rodburn,  21  (UI),  2  (USNM).  Trigg  Co.:  8  mi.  NNE  Golden 
Pond,  I  (MZ).  North  Carolina.  Buncombe  Co.:  Asheville,  3  (MZ),  1 
(NCS);  Weaverville,  25  (AMNH),  4  (FM).  Cherokee  Co.:  Murphy,  3  (UG). 
Macon  Co.:  Highlands,  9  (UG).  Madison  Co.:  Marshall,  10  (MZJ.  Transijl- 
vania  Co.:  White  Water  River,  1  (NCS).  Tennessee.  Carter  Co.:  3  mi.  SSW 
Roan  Mt.,  3  (MZ);  Roan  Mt.,  1  (USNM).  Jefferson  Co.:  unspecified  locality, 
2  (USNM).  Johnson  Co.:  3  mi.  NE  Holston  Mt.,  4  (USNM).  Knox  Co.:  Knox- 
ville,  1  (USNM).  Union  Co.:  Maynardville,  6  (MZ).  Virginia.  Montgomery 
Co.:  Vic.  Blacksburg,  6  (VPI),  I  (AMNH).  Scoff  Co.:  6  mi.  SE  Norton,  1 
(MZ). 

Ochrotomys  nuttalli  flammeus  (Goldman) 

Peromtjscus  nuttalli  flannncus  Goldman,  Proc.  Biol.  Soc.  Washington,  54:190, 

December  8,  1941. 

Holotype. — Adult  female,  skin  and  skull,  no.  170591  U.S.  National  Museum 
(Biological  Surveys  Collection);  from  Delight,  Pike  Co.,  Arkansas;  obtained 
on  No\ember  16,  1910,  by  Walter  G.  Savage,  original  number  8221. 

Range. — Ouachita  Mountains  of  easternmost  central  Oklahoma  and  west- 
central  Arkansas;  also  in  Arkansas  from  the  type  locality  eastward  to  the 
vicinity  Pine  Bluff,  Jefferson  County,  north  to  the  vicinity  of  Beebe,  White 
County,  thence  northwest  into  the  Boston  Mountains  in  the  \'icinity  of  Fayette- 
\ille  (see  Fig.  1 ). 

Diagnosis. — Size  medium  to  large  for  the  species;  length  of  body  larger  and 
tail  proportionately  shorter  than  in  specimens  from  Texas,  northern  Louisiana, 


Packard — Review  of  Ochrotomys  397 

southern  Missouri,  and  Illinois;  middorsal  region  an  ochraceous-tawny,  black 
guard  hair  more  profuse  on  head;  sides  cinnamon-buff,  underparts  creamy  to 
pale  cinnamon-buff";  tail  faintly  bicolored,  ochraceous  above,  cream-colored 
below  ( see  Table  2  of  measurements  for  external  and  cranial  variants ) . 

Comparisons. — O.  n.  flammeus  intergrades  with  O.  n.  lisae  to  the  southwest 
in  Texas  and  toward  the  Mississippi  Ri\er  Valley  to  the  east  of  the  Ouachita 
and  Ozark  uplifts;  to  the  north,  flammeus  intergrades  with  lisae  northeastward 
across  the  Ozark  Plateau.  Because  of  this,  flammeus  is  compared  only  with 
lisae. 

Young  adults  of  both  subspecies  in  unworn  pelage  show  best  the  colors  that 
differentiate  the  two  subspecies.  O.  n.  flammeus  dorsally  has  orange-red  over- 
tones, whereas  lisae  is  yellowish  red  (this  difference  results  mostly  from  a 
greater  profusion  of  black  guard  hairs  on  O.  n.  flammeus).  Crania  of  flammeus 
are  larger  ( averages  of  condylobasal  length  and  zygomatic  breadth )  than  those 
of  lisae,  and  flammeus  possesses  somewhat  longer  ears,  averages  larger  in  body 
length,  but  has  a  proportionately  shorter  tail. 

Remarks. — When  Goldman  (1941)  described  O.  n.  flammeus  (on  the  basis 
of  15  specimens),  he  noted  that  ".  .  .  additional  specimens  are  needed  for  the 
more  exact  delimitation  of  the  ranges  of  the  subspecies  .  .  .  ,"  suggesting  that 
the  degree  of  variation  within  v.-estern  populations  of  the  species  merited 
further  analy.sis  to  better  interpret  patterns  of  variation.  Patterns  of  variation 
rexealed  in  my  studies  suggest  that  a  combination  of  characteristics  (length  of 
tail  proportionate  to  body  length,  larger  ears  and  crania,  and  somewhat  dif- 
ferent color  patterns)  do  delimit  the  populations  from  the  Ouachita  and  Boston 
mountains  and  southwestern  part  of  the  Ozark  Plateau  from  those  to  the 
southwest,  south,  east,  and  northeast  with  which  they  intergrade.  The  habitat 
occupied  by  flammeus  in  the  uplands  is  also  slightly  different  (rocky  with  pine) 
from  those  populations  on  the  alkuial  regions  to  the  north,  east,  and  south. 

Specimens  examined  (46). — Arkansas.  Garland  Co.:  3  mi.  N  Buckville, 
1  (UA);  Sulphur  Gorge,  Hot  Springs,  2  (UI).  Jefferson  Co.:  10  mi.  N  Pine 
Bluff,  5  (UA);  Pine  Bluff,  2  (UA).  Pike  Co.:  type  locality,  7  (USNM),  4 
(MCZ).  Polk  Co.:  6.3  mi.  NW  Rich  Mountain,  2  (UA);  6  mi.  N,  3  mi.  E 
Mena,  1  (KU);  Cave,  1  (KU);  Shady  Lake  area,  1  (UA).  Scoff  Co.:  Fourch 
La  Fave  River,  1  (FM).  Washington  Co.:  Fayetteville,  1  (UA),  1%  mi.  W 
Farmington,  1  (UA);  7  mi.  NE  Winslow,  4  (KU);  Winslow  golf  course,  3 
(KU);  1  mi.  N  Winslow,  1  (KU);  Winslow,  5  (KU).  White  Co.:  Beebe,  1 
(USNM),  Big  Spring,  1  (USNM).  Oklahoma.  Redland  (in  eastern  Okla- 
homa originally  located  in  "Indian  Territory"),  2   (USNM). 

Ochrotomys  nuttalli  floridanus,  new  subspecies 
Holotijpe. — Adult  male,  skin  and  skull,  no.  119422  Museum  of  Natural  His- 
tory, The  University  of  Kansas;  from  Welaka,  Putnam  Co.,  Florida;  obtained 
on  February  18,  1948,  by  G.  H.  Pournelle,  original  number  268. 

Range. — Peninsular  Florida;  northern  limits  in  area  of  Wakulla  and  Leon 
counties  in  the  northwest,  east  to  the  Atlantic  Coast  (vicinity  of  Jacksonville); 
south  to  approximately  27°  latitude. 

Diagnosis. — Size  medium  to  small  for  the  species  (for  measurements,  see 
Table  2);  tail  short  relative  to  body  length;  hind  foot  short;  braincase  shallow; 
middorsal  region  yellowish  brown  ( 10  YR  5/8),  belly  cream  buff  with  yellowish 


398  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

overtone  extending  onto  region  of  upper  and  lower  jaws;  ventralmost  vibrissae 
white;  tail  bicolored,  pale  cinnamon  al)ove,  cream-colored  below;  ears  imi- 
formly  orange;  general  appearance,  burnished  orange. 

Comparisons. — From  O.  n.  nuttalli,  O.  n.  ftoiidainis  differs  in:  dorsal  colora- 
tion brighter  orange;  upper  parts  of  hind  feet  white  rather  than  gray;  tail 
cinnamon-orange  above  rather  than  blackish  brown;  ears  orange  rather  than 
brownish  red;  belly  with  yellowish-white  wash  rather  than  dusky  white;  averag- 
ing smaller  in  most  external  and  cranial  measurements  ( see  Table  2 ) . 

From  O.  n.  aureolus,  O.  n.  floridanus  differs  in:  dorsal  color  orange-red 
rather  than  brownish  red;  tail  cinnamon-orange  above  rather  than  blackish 
brown;  ears  orange  rather  than  blackish  red  to  brown;  belly  yellowish  white 
rather  than  dusky  white  with  cinnamon  overtones;  averaging  smaller  in  total 
length. 

From  O.  n.  lisae,  O.  n.  floridanus  differs  in:  dorsal  color  brighter  orange 
rather  than  yellowish;  tail  cinnamon-orange  rather  than  brown  above,  yellowish 
rather  than  dusky  white  below;  ears  orange  rather  yellowish  brown;  belly 
yellowish  white  rather  than  whitish  with  tones  of  cinnamon;  rostal  length  and 
post-palatal  length  averaging  shorter. 

Remarks. — The  most  striking  differences  of  floridanus  are  the  lack  of  a 
profusion  of  black  dorsal  guard  hairs  on  the  posterior  half  of  the  back  and 
upper  parts  of  the  tail,  resulting  in  an  orange  or  golden-red  pelage.  This 
distinctive  color  differentiates  floridanus  from  the  other  subspecies.  O.  n. 
floridanus  intergrades  with  li.sae  and  nuttalli  in  western  Florida,  southern  Ala- 
bama, and  Mississippi.  Specimens  from  southern  Georgia  and  northern  Florida 
are  intergrades  between  floridanus  and  nuttalli.  Specimens  from  the  Ocala 
Ridge  in  Florida  show  best  those  features  that  differentiate  this  subspecies. 

Specimens  examined  (53). — Florida.  Alachtia  Co.:  10  mi.  NW  Gaines- 
ville, 1  (UF);  8  mi.  NW  Gainesville,  2  ( MZ),  1  (UF);  7  mi.  NW  Gainesville, 
1  (UF);  San  Felasco  Hammock,  Gainesville,  4  (UI);  Gainesville,  2  (AMNH); 
Gracie's  Crossing,  1  (MZ),  2  (UF);  M  mi.  N  Paradise,  1  (UF).  Citrus  Co.: 
Homosassa  Springs,  3  (AMNH).  Clay  Co.:  3  mi.  SW  Middleburg,  1  (UF). 
Duval  Co.:  New  Berlin,  1  (AMNH).  Columbia  Co.:  Winfield,  1  (UF).  Gad- 
.son  Co.:  Chattahoochee,  6  (AMNH).  Highlands  Co.:  Hicoria  (Archbold  Bio- 
logical Station),  3  (UF).  Leon  Co.:  10  mi.  SE  Tallahassee,  3  (AMNH).  Levtj 
Co.:  Gulf  Hammock,  7  (UF).  Putnam  Co.:  type  locality,  8  (UF),  1  (UK). 
St.  John's  Co.:  2  mi.  W  Crescent  Beach,  1  (UF).  Taylor  Co.:  4  mi.  SW  Perry, 
1  (UF).  Wakulla  Co.:  Spring  Creek,  1  (UF).  Volusia  Co.:  Enterprise,  1 
(AMNH). 

Ochrotomys  nuttalli  lisae,  new  subspecies 
Holotype. — Young  adult  male,  skin,  skull,  and  body  skeleton,  no.  119421 
Museum  of  Natural  History,  The  University  of  Kansas;  from  La  Nana  Creek 
bottoms,  1  mi.  E  Stephen  F.  Austin  State  College  campus,  Nacogdoches, 
Nacogdoches  Co.,  Texas;  obtained  on  January  18,  1961,  by  Robert  L.  Packard, 
original  number  829. 

Range. — Eastern  Texas,  central  and  northern  Louisiana,  northward  into 
extreme  eastern  Arkansas  (in  the  Mississippi  Valley),  Missouri,  and  southern 
Illinois,  thence  to  western  parts  of  Kentucky,  Tennessee,  and  Mississippi. 

Diagnosis. — Size  medium  to  small  for  the  species  (for  measurements  see 
Table  2);  total  length  small,  with  tail  averaging  short  for  the  species  and  body 


Packard — Review  of  Ociirotomys  399 

size  small;  hind  foot  averaging  shorter  than  most  populations  from  other  areas 
examined;  ear  of  average  size  for  the  species;  cranial  l^readth  average  for  the 
species,  Init  condylobasal  length  averaging  shorter  and  braincase  shallower; 
length  of  toothrows  (upper  and  lower)  of  average  length  for  the  species;  dorsal 
ground  color  near  7.5  YR  5/6  (strong  brown),  some  specimens  with  more 
yellowish  wash  (particularly  in  western  Louisiana  and  eastern  Texas);  belly 
cream  colored  with  some  overtones  of  cinnamon;  tail  bicolored,  cinnamon 
above,  cream  below. 

Comparisons. — For  comparisons  with  O.  n.  jiammeus,  see  account  of  that 
subspecies.  From  O.  n.  aureolus,  with  which  O.  n.  lisae  intergrades  to  the  east, 
lisac  differs  in:  dorsal  color  paler,  having  yellowish  oxertones  than  being 
reddish  brown  with  dusky  overtones;  belly  somewhat  paler,  more  whitish  rather 
than  yellowish  to  cream-colored  (particularly  true  of  freshly  taken  specimens); 
tail  more  distinctly  bicolored  rather  than  tending  to  unicolor;  toothrows  averag- 
ing shorter  except  for  populations  from  northwestern  Georgia;  smaller  in  most 
cranial  dimensions  except  rostral  length;  total  length  and  length  of  tail  averag- 
ing less. 

Remarks. — O.  n.  lisae  is  typically  a  yellowish-washed,  small  subspecies  of  the 
alluvial  bottomlands  and  low  rolling  hills  of  the  Austroriparian  part  of  eastern 
Texas,  central  and  northern  Louisiana,  and  northward  into  the  Mississippi  River 
Valley.  Osgood  (1909)  assigned  the  few  specimens  available  to  him  from  west 
of  the  Mississippi  River  to  O.  n.  aureolus  or  to  O.  n.  nuttalli.  Goldman  ( 1941 ) 
considered  all  specimens  (15)  that  he  examined  from  west  of  the  Mississippi 
River  to  be  O.  n.  jiammeus.  Studies  of  much  larger  series  of  specimens  than 
were  available  to  either  of  the  aforementioned  workers  reveal  two  distinct  sub- 
species in  the  Trans-Mississippi  region.  Populations  in  the  mountainous  areas 
in  Arkansas  and  eastern  Oklahoma  seem  typical  of  jiammeus  and  are  different 
from  those  in  the  alluvium  and  low  rolling  hills  of  neighboring  regions.  It 
seems  best  to  restrict  the  range  of  jiammeus  to  the  mountains  of  Arkansas  and 
Oklahoma  in  order  to  accord  with  Goldman's  original  description. 

Lowery  (1943)  recognized  (based  on  communication  with  Goldman)  the 
possible  presence  of  an  undescribed  population(s)  of  pale  golden  mice  in 
Louisiana.  Examination  of  specimens  from  that  state  leads  me  to  conclude 
these  pale  (yellowish-colored)  mice  are  typical  of  lisac  as  they  compare  favor- 
ably with  specimens  from  eastern  Texas.  In  northern  Louisiana,  there  is  evi- 
dence of  intergradation  in  color  with  jiammeus  to  the  north,  but  specimens  from 
this  area  are  best  referred  to  lisae. 

Specimens  from  southern  Missouri  and  southern  Arkansas  intergrade  with 
jiammeus,  but  seem  to  have  more  concordance  with  lisae  in  characters  studied 
and  are  assigned  to  it.  Eastwardly,  lisae  intergrades  broadly  with  O.  n.  aureolus, 
and  specimens  from  eastern  Louisiana  and  Mississippi  are  difficult  to  assign  to 
either  lisae  or  aureolus,  or  in  some  cases  jioridanus.  1  assign  these  to  lisae  mostly 
because  there  is  concordance  in  color  of  pelage.  Interestingly,  these  specimens 
from  the  Gulf  Coast  Region  are  quite  variable  (see  Table  2),  suggesting  a 
secondary  zone  of  intergradation.  Perhaps  the  western  populations  (west  of 
the  general  area  of  the  Mississippi  River  Valley)  were  separated  for  a  period 
of  time  from  those  to  the  east,  although  there  is  nothing  in  the  fossil  record  to 
substantiate  this. 


400  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

This  suljspecies  is  named  in  memory  of  my  daughter,  Lisa  Ann. 
Specimens  examined  (289). — Arkansas.    Chicot  Co.:  Island  80,  1   (UA); 
vie.  El  Dorado,  1   (UA).    Illinois.    Alexander  Co.:  Olive  Branch,  20   (FM), 

1  (SIU).  Johnson  Co.:  I  mi.  S  Forman,  1  (UI).  Pope  Co.:  sec.  28,  T.  12  S, 
R.  5  E,  6  (UI).  Union  Co.:  Pine  Hills,  3  (UI);  Wolf  Lake,  1  (SIU);  3  mi. 
SW  Ware,  I   (SIU);  2  mi.  SE  Ware,  2   (SIU).    Louisiana.    Bienville  Parish: 

2  mi.  SW  Bienville,  2  (SFA).   Bossier  Parish:  }i  mi.  W  Webster,  ]i  mi.  S  Arkan- 
sas line,  1   (LSU).    Caddo  Parish:  2  mi.  E  Zylkes,  1   (LSU);  %  mi.  E  Zylkes, 
1   (LSU);  'A  mi.  E  Zylkes,  1   (LSU).    Caldwell  Parish:  6%  mi.  N  Columbia,  1 
(LSU);  Columbia,  8   (FM);   Hackley,  3    (FM).    Claiborne  Parish:   Y^  mi.   S 
Marsahs,  1   (LSU).    Desoto  Parish:  Mansfield,  1   (USNM).    East  Baton  Rouge 
Pari.sh:  7  mi.  SE  Baton  Rouge,  2   (LSU);  Baton  Rouge,   1    (LSU);  4  mi.  S 
Lindsay,   1    (LSU);  3  mi.  S  Lindsay,   1    (LSU);  2  mi.  S  Lindsay,  3   (LSU); 
Lindsay,  2  (USNM);  Plains,  5  (MZ);  )^  mi.  E,  2  6/10  mi.  S  Union,  1  (LSU); 
11-12  mi.  S  University,  2  (LSU);  6%  mi.  SE  University,  2  (LSU);  4-5  mi.  S 
University,  2  (LSU);  3  mi.  S  University,  3  (LSU),  1  (USNM);  2  mi.  S  Uni- 
versity, 1  (LSU).   East  Feliciana  Parish:  E.  Jackson,  1  (LSU).    Lincoln  Parish: 
Ruston,  1   (USNM);  2  mi.  N  Tremont,  3   (LSU).    Livingston  Parish:  2  mi.  S 
Watson  River,  1  (LSU).    Morehouse  Parish:  4  mi.  SE  Bastrop,  2  (LSU);  Che- 
minahaut  State  Park,  2  (LUS).    Ouachita  Parish:  12  mi.  W  Monroe,  19  (TU); 
4  mi.  W  Monroe,  2   (TU).    Rapides  Parish:  5  mi.  E  Lind,   1    (LSU).    Union 
Parish:  3)^  mi.  N,  M  mi.  E  Bernice,  4   ( UI ) ;  7  mi.  NE  Farmerville,  1   ( LSU ) ; 
4  mi.  NE  Farmerville,  1  (LSU);  3  mi.  NE  Farmerville,  2  (LSU).    Washington 
Parish:  8  mi.  SE  Angie,  1   (TT);  7  mi.  S,  5  mi.  W  Angie,  3   (TT);  6  mi.  S 
Angie,  1   (TT);  )2  mi.  N  Angie,  3  (TT).    Mississippi.    Adams  Co.:  unspecified 
locality,  1   (MGF).    Bolivar  Co.:  Bogue  Phalia,  1   (MGF).    Clark  Co.:  Linton 
area,  1    (MGF).    Coahoma  Co.:  Palmer  Lake  area,   1    (MGF).    Copiah  Co.: 
Crystal  Springs,  1   (MGF).    Forest  Co.:  Burkett  Creek,  1    (MGF);  McKinnon 
Springs,  2  (MGF);  Shelby  State  Park,  3  (TT).   /ones  Co.:  unspecified  locality, 
1  (MGF).   Lamar  Co.:  Lumberton  area,  1  (MGF).   Lauderdale  Co.:  Okatibbee 
area,  1  (MGF).  Pearl  River  Co.:  Poplarville,  1  (MGF).  Pike  Co.:  Percy  Quinn 
State  Park,  1  (TT).   Prentiss  Co.:  20-mile  Bottom  area,  2  (MGF).   Rankin  Co.: 
Spear  Farm,  1    (MGF).    Tippah  Co.:  Gillard  Farm,   1    (MGF).    Tishomingo 
Co.:  2)^  mi.  E,  2J^  mi.  S  Tishomingo,  2   (KU).    Wayne  Co.:  Trigg  Area,    1 
(MGF).     Wilkin.son    Co.:    Percis   Creek,    1    (MGF).     Winston   Co.:    Sulphur 
Springs,  1    (MGF).    Yazoo  Co.:  2  mi.  SE  Yazoo  City,   1    (LSU).    Missouri. 
Bollinger  Co.:  Duck  Creek  Refuge,  4   (UM);  Puxico,  2   (UM).    Camden  Co.: 
Hahatouka,  1  (MZ).   Dent  Co.:  Salem,  3  (UM).    Franklin  Co.:  Meramec  State 
Park,  3   (UM).    Gasconde  Co.:   Owensville,   1    (UM).    Girardeau   Co.:   Cape 
Girardeau,  3  (UM).    Howell  Co.:  Willovvspring,  3  (UM).   Iran  Co.:  2  mi.  SW 
Grantsville,  1   (UI).    Jefferson  Co.:  St.  Louis,  1    (USNM).    Pulaski  Co.:  Rich- 
land, 2  (UM).   Reynolds  Co.:  West  Fork,  1  (UM).   Ripley  Co.:  Pratt,  1  (UM). 
Texas  Co.:  unspecified  locality,  1   (UM).    Wayne  Co.:  Williamsville,  1   (UM). 
Tennessee.    Lake  Co.:  Real  Foot  Lake,  4  (UI);  Tiptonville,  1   (UI).    Lauder- 
dale Co.:  1.9  mi.  E  Open  Lake,  1   (UI).    Shelby  Co.:  2  mi.  S  Germantown,  1 
(LSU).    Texas.    Angelina  Co.:  Lufkin,   1    (TU).    Anderson  Co.:  20  mi.   NW 
Palestine,  6  (TU),  1  (SFA).   Cherokee  Co.:  5  mi.  E  Rusk,  1  (SFA);  3  mi.  SE 
Rusk,  2  (SFA);  2  mi.  N  Rusk,  1   (SFA).    Harrison  Co.:  15  mi.  NE  Marshall, 
1  (SFA).   Nacogdoches  Co.:  type  locality,  20  (SFA),  3  (TT),  1  (UK);  12  mi. 


Packard — Review  of  Ochrotomys  401 

N  Liifkin,  1  (TAM);  20  mi.  SW  Nacogdoches,  4  (SFA),  2  (KU);  14  mi.  SW 
Nacogdoches,  1  (SFA);  Stephen  F.  Austin  Experimental  Forest  (d'A  mi.  SW 
Nacogdoches),  36  (SFA),  2  (MZ),  4  (TU),  5  (TT);  6  mi.  SW  Nacogdoches, 
1  (SFA).  Panola  Co.:  4  mi.  SE  Long  Branch,  1  (SFA).  Rusk  Co.:  2  mi.  E 
Old  London,  2  (SFA). 

Ochrotomys  nuttalli  nuttalli  (Harlan) 

Arvicola   nuttalli  Harlan,   Monthly  Amer.   Jour.   Geol.   Nat.   Sci.,   Philadelphia, 

p.  446,  April,  1832. 
Ochrotomys  nuttalli,  Hooper,  Misc.  Publ.  Mus.  Zool.,  Univ.  Michigan,  105:23, 

December  29,  1958. 
Peromyscus  nuttalli  lewisi  A.  H.  Howell,  Jour.  Manim.,  20:498,  November  14, 

1939  (type  locality,  Amelia  Court  House,  Amelia  Co.,  Virginia). 

Holotype. — Not  known  to  exist.  Because  no  holotype,  syntypes,  nor  lecto- 
type  are  known  to  be  extant,  I  designate  as  a  neotype  an  adult  female,  skin  and 
skull,  no.  95889  U.S.  National  Museum  (U.S.  Biological  Surveys);  from  Lake 
Drummond,  Dismal  Swamp,  Nansemond  Co.,  Virginia  ( 30  miles  southwest  of 
the  original  type  locality  at  Norfolk,  Virginia). 

Range. — Coastal  plains  and  Piedmont  of  central  and  southern  Virginia, 
south  on  the  coastal  plains  and  Piedmont  of  North  and  South  Carolina,  and 
Georgia,  to  central  Alabama  and  northwestern  Florida. 

Diagnosis. — Size  large  for  the  species;  total  length,  body  length,  and  tail 
length  in  northern  populations  average  larger  than  other  populations  studied; 
skull  broad,  particularly  on  coastal  plains,  braincase  deeper  on  coastal  plain, 
becoming  shallower  on  Piedmont;  middorsal  region  strong  brown  ( 7.5  YR 
5/8),  ears  of  same  color;  hind  feet  dusky  white  above;  belly  white  with  pale 
yellowish  tint. 

Comparisons. — For  comparisons  with  O.  n.  ftoridanus,  see  account  of  that 
subspecies.  From  O.  n.  aureolus,  O.  n.  nuttalli  differs  in:  dorsum  brownish  red 
rather  than  with  black  overtones  ( resulting  from  fewer  black  guard  hairs ) ; 
upper  part  of  tail  brown  rather  than  black-brown;  facial  area  paler,  washed 
with  a  slight  yellowish  tone;  belly  paler,  having  less  cinnamon  wash;  tail 
averaging  longer  (particularly  in  coastal  plain  samples);  ear  and  hind  foot 
averaging  longer;  overall  cranial  size  somewhat  larger. 

From  O.  n.  li.sae,  O.  n.  nuttalli  differs  in:  dorsum  darker,  red-brown  rather 
than  yellow-brown;  upper  parts  of  hind  feet  dusky  white  rather  than  white, 
\entral  side  of  tail  duskier  and  less  fully  haired;  axeraging  larger  in  most 
external  and  cranial  characteristics  examined   ( see  Table  2  for  comparisons ) . 

Remarks. — There  are  two  trends  in  variation  in  O.  n.  nuttalli  as  follows: 
( 1 )  Virginian  specimens  from  both  coastal  plain  and  Piedmont  are  larger  in 
most  external  and  cranial  characters  than  are  samples  from  the  south  ( Georgia ) , 
and  the  pattern  in  variation  is  clinal;  (2)  specimens  on  the  coastal  plain  are 
larger  in  total  length,  length  of  tail,  length  of  hind  foot,  depth  of  braincase, 
and  post-palatal  length  than  samples  from  the  adjacent  Piedmont — thus,  a 
cline  of  decreasing  size  occurs  from  east  to  west.  Although  populations  on  the 
Piedmont  differ  to  varying  degrees  from  those  on  the  coastal  plains,  the  mice 
from  these  two  areas  share  enough  features  in  common  to  consider  them  as  the 
same  subspecies.  The  break  in  clinal  variation  occurs  chiefly  between  the  mice 
inhabiting  the  Appalachian  Mountains  and  those  of  the  Piedmont  Region. 


402  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Osgood  (1909:225)  noted  that  specimens  from  Virginia  were  larger  than 
those  from  more  southerly  localities,  but  because  of  the  small  sample  sizes 
available  to  him  (particularly  to  the  west),  the  trends  in  variation  were 
obscured.  It  seems  best  to  consider  the  mice  of  the  Atlantic  coastal  plains  and 
Piedmont  as  O.  n.  ntittalli,  and  specimens  from  the  mountains  as  O.  n.  oureoliis. 

O.  n.  Icivisi  is  regarded  as  representati\e  of  a  subadult  population  of  age 
two  (see  age  categories)  and  is  considered  as  synonymous  with  O.  n.  ntittalli. 

Specimens  examined  (244). — Alabama.  Escumbria  Co.:  Brewton,  1 
(USNM).  Houston  Co.:  Dothan,  1  (USNM).  Lee  Co.:  Auburn,  9  (AU),  2 
(MZ);  Beauregard,  1  (AU).  Mobile  Co.:  Mobile,  2  (USNM).  Russell  Co.: 
Scale,  1  (USNM).  Georgia.  Ben  Hill  Co.:  Fitzgerald,  1  (UG).  Bibb  Co.: 
5  mi.  NE  Macon,  1  (LSU).  Camden  Co.:  St.  Marys,  1  (UG).  Charlton  Co.: 
Okefinokee,  2  (KU).  Clark  Co.:  Athens,  25  (UG);  Sandy  Creek,  .3  (UG); 
White  Hall,  2  (UG).  Cobb  Co.:  Marietta,  2  (MZ);  Rozwell,  2  (KU),  4  (MZ); 
Vinings,  1  (UG).  Columbia  Co.:  Dickey's  Branch,  1  (UG).  DeKalh  Co.: 
Decatur,  1  (MZ);  2  mi.  N  Emory,  1  (UI).  Grady  Co.:  4  mi.  S  Beachton,  1 
(UI);  Sherwood  Plantation,  3  (UI).  Emanuel  Co.:  Adrian,  1  (UG).  Lincoln 
Co.:  unspecified  locality,  1  (UG).  Lloyd  Co.:  8  mi.  W  Rhome,  1  (UI). 
Lowndes  Co.:  Naylor,  1  (UG).  Merriivether  Co.:  Singer's  Hill,  I  (UG). 
Mitchell  Co.:  W  Side  Gamilla,  1  (UI).  Richmond  Co.:  unspecified  locality,  2 
(USNM).  Swnpter  Co.:  unspecified  locality,  3  (UG).  Thomas  Co.:  Thomas- 
ville,  3  (AMNH).  Ware  Co.:  2  mi.  W  Camp  Cornelia,  1  (KU).  White  Co.: 
unspecified  locality,  1  (UG).  Wilkes  Co.:  unspecified  locality,  1  (UG). 
Florida.  Okaloosa  Co.:  Ft.  Walton,  1  (KU);  5  mi.  SW  Laurel  Hill,  1  (UF). 
North  Carolina.  Beaufort  Co.:  Core  Point,  I  (NCS).  Currituck  Co.:  NW 
River  Marsh,  1  (NCS).  Durham  Co.:  4  mi.  W  Duke,  5  (DU);  Duke  Forest, 
4  (KU);  Duke  University,  4  (DU);  SY,  mi.  E  Durham,  5  (DU).  Mecklenburg 
Co.:  Charlotte,  6  (NCS),  1  (UI);  Davidson,  3  (MZ).  North  Hampton  Co.: 
Roanoke  Rapids,  1  (FM).  New  Hanover  Co.:  Carolina  Beach,  1  (NCS). 
Orange  Co.:  un.specified  locality,  1  (NCS).  Stanley  Co.:  xmspecified  locality, 
1  (NCS).  Wake  Co.:  Apex,  2  (FM);  Green  Laurel,  1  (NCS);  Raleigh,  11 
(AMNH),  5  (NCS),  3  (MZ),  2  ( FM),  1  (lU),  1  (KU).  South  Carolina. 
Abbeville  Co.:  Abbeville,  2  (KU);  Calhoun  Falls,  9  (FM).  Aiken  Co.:  unspeci- 
fied locality,  2  (UG);  Aiken  State  Park,  1  (UI).  Barnwell  Co.:  un.specified 
locahty,  2  (UG).  Dorchester  Co.:  St.  George,  1  (USNM).  Greenville  Co.: 
Cliff  Ridge  Road,  1  (KU);  Jones  Gap,  1  (KU).  Richland  Co.:  Columbia,  1 
(USNM).  Virginia.  Amelia  Co.:  Amelia,  14  (USNM),  1  (MZ),  1  (VPI). 
Brunswick  Co.:  Seward  Forest,  5  (USNM),  5  (MZ).  Campbell  Co.:  Lynch- 
burg, 9  (MZ).  Charlotte  Co.:  Brookneal,  1  (MZ);  5  mi.  E  Charlotte,  1  (MZ); 
1  mi.  W  Charlotte,  1  (MZ).  Halifax  Co.:  5  mi.  N  Clover,  1  (MZ).  Nansemond 
Co.:  Lake  Drummond  (Dismal  Swamp),  4  (AMNH);  Dismal  Swamp,  34 
(USNM). 

Summary 
There  is  little  fossil  evidence  that  would  account  for  the  origin 
and  relationship  of  the  golden  mouse.  A  single  Pleistocene  record 
exists  from  Missouri  (sec  Olson,  1940)  in  an  area  where  these  mice 
are  found  today.  Hibbard  (1968)  suggested  that  the  neotomyine- 
peromyscine  group  had  an  origin  in  the  Oligocene  or  early  Miocene. 


Packard — Review  of  Ochrotomys  403 

In  degree  of  specialization,  Ochroiomys  possesses  features  suggest- 
ing that  as  much  time  has  been  involved  in  their  evolution  as  in 
the  case  of  Omjchomys,  and  possibly  Baiomys  and  Scotinomys.  If 
so,  then  golden  mice  may  have  evolved  in  the  Pliocene,  though  no 
record  documents  this.  The  relationship  of  Ochrotomys  to  other 
cricetine  genera  as  proposed  by  Hooper  and  Musser  (1964)  seems 
most  cogent.  The  work  of  Arata  ( 1964 )  lends  strength  to  the  pro- 
posal that  Ochrotomys  is  one  of  the  more  primitive  members  of  the 
neotomyine-peromyscine  line.  Additional  data  bearing  on  the  de- 
gree of  gastric  specialization  in  Ochrotomys  and  other  genera  has 
been  reported  recently  by  M.  D.  Carleton  (personal  communica- 
tion). In  1964,  one  of  my  students,  William  Grabowski,  made  a 
brief  survey  of  the  gastric  anatomy  of  representatives  of  the  sub- 
genera (Osgood,  1909)  of  Peromyscus.  Relative  to  Ochrotomys, 
he  found  an  internal  fold  delimiting  the  pyloric  region  from  the 
rest  of  the  stomach  (a  similar  fold  was  detected  in  Baiomys).  In 
addition,  in  all  representatives  studied  except  Ochrotomys  and 
Baiomys,  he  found  a  region  located  on  the  ventral  floor  of  the 
stomach  that  appeared,  externally,  paler  in  color  in  comparison  to 
surrounding  areas.  Internally,  this  region  was  found  to  be  thicker 
in  comparison  with  the  rest  of  the  epithelium  of  the  stomach  and 
was  always  delimited  by  a  fold  ( incisura  anguJaris  of  Vorontsov, 
1957). 

Golden  mice  seem  to  have  undergone  little  differentiation  if 
Ochrotomys  is  as  old  as  other  seemingly  closely  related  genera.  I 
judge  that  the  semi-arboreal  habits  and  rather  uniform  habitat  oc- 
cupied may  account  for  this.  The  genus  is  limited  to  the  pine-oak 
forests  of  the  United  States  and  may  have  been  there  for  some 
considerable  time.  Blair  (195(S)  pointed  to  changes  in  the  Pleisto- 
cene in  the  area  now  occupied  by  golden  mice.  Perhaps  the  species 
was  divided  into  two  separate  entities  in  Pleistocene  times,  one  in 
Florida  and  the  other  westward  in  Texas  and  northern  Mexico. 
Sufficient  time  elapsed  for  the  differentiation  of  several  subspecies 
but  not  for  species  differentiation.  This  may  account  for  the  high 
degree  of  variation  in  the  Gulf  Coast  area  detected  in  this  study.  A 
second  possibility  is  suggested.  Golden  mice  may  have  differen- 
tiated south  and  west  ( in  Mexico )  of  the  area  now  occupied.  These 
mice  then  could  have  moved  northeastward  through  the  subhumid 
pine-oak  corridor  that  once  existed  in  the  Gulf  Goast  area  (vestiges 
of  which  are  the  lost  pines  near  San  Marcos,  Texas).  Upon  arrival 
in  the  southeastern  United  States,  their  geographic  range  may  have 


404  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

been  constricted  in  periods  of  glacial  advance  to  the  Florida  Penin- 
sula as  were  the  ranges  of  certain  other  species  of  mammals.  After 
glacial  retreat,  populations  then  dispersed  northeastward  and  north- 
westward to  occupy  the  current  geographic  range.  I  am  certain  this 
is  an  oversimplification,  but  it  is  all  I  can  propose  in  the  absence 
of  a  fossil  record. 

Literature  Cited 

Arata,  a.  a. 

1964.     The  anatomy  and  taxonomic  significance  of  the  male  accessory  re- 
productive   glands    of   muroid    rodents.     Bull.    Florida    State    Mus., 
9:1-42,  9  figs. 
Baird,  S.  F. 

1858.     Mammals,   /)i    Reports   of  explorations   and   surveys    .    .    .   from   the 
Mississippi  River  to  the  Pacific  Ocean  .  .  .  ,  8(  1 )  :xxl-xlviii  -f-  1-757, 
4.3  pis. 
Bangs,  O. 

1898.     The  land  mammals  of  peninsular  Florida  and  the  coastal  region  of 
Georgia.    Proc.  Boston  Soc.  Nat.  Hist.,  28:157-235,  7  figs. 
Barbour,  R.  W. 

1942.     Nests  and  habitat  of  the  golden  mouse  in  eastern  Kentucky.    Jour. 
Mamm.,  23:90-91. 
Blair,  W.  F. 

1942.  Systematic  relationships  of  Peromyscus  and  several  related  genera  as 
shown  by  the  baculum.   Jour.  Mamm.,  23:196-204,  2  figs. 

1958.  Distributional  patterns  of  vertebrates  in  the  southern  United  States 
in  relation  to  past  and  present  environments.  Pp.  433-468,  11  figs., 
in  Zoogeographv  (C.  L.  Hubbs,  ed.),  Publ.  Amer.  Assoc.  Adv.  Sci., 
51  :x  +  1-.509,  illustrated. 

Blair,  W.  F.,  A.  P.  Blair,  P.  Brodkorb,  F.  R.  Cagle,  and  G.  A.  Moore 

1968.     Vertebrates  of  the  United  States.    McGraw  Hill,  2nd  ed.,  ix  +  616 
pp,  illustrated. 
Blus,  L.  J. 

1966.     Relationship  between  litter  size  and  latitude  in  the  golden  mouse. 
Jour.  Mamm.,  47:546-547. 
Davis,  W.  B. 

1960.  The  mammals  of  Texas.  Bull.  Texas  Game  and  Fish  Comm.,  41:1- 
2.52,  73  figs.,  64  maps. 

DUNAWAY,  P.   B. 

1955.     Late  fall  home   ranges  of  three   golden  mice,   Peromyscus  nuttalli. 
Jour.  Mamm.,  36:297-298. 
Easterla,  D.  a. 

1968.     Terrestrial  home  site  of  golden  mouse.   Amer.  Midland  Nat.,  79:246- 
247. 
Goldman,  E.  A. 

1941.  A  new  western  subspecies  of  golden  mouse.  Proc.  Biol.  Soc.  Wash- 
ington, 54:189-191. 

GOLLEY,   F.   B. 

1962.     Mammals  of  Georgia.  .  .  .    Univ.  Georgia  Press,  Athens,  xii  +  218 

pp. 
1966.     South  CaroUna  mammals.    The  Gharleston  Museum,  xiv  -|-  181  pp. 

Goodpaster,  W.  W.,  and  D.  F.  Hoffmeister 

1954.  Life  history  of  the  golden  mouse,  Peromijscus  nuttalli,  in  Kentucky. 
Jour.  Mamm.,  35:16-27,  1  pi,  2  figs. 

Hall,  E.  R.,  and  K.  R.  Kelson 

1959.  The  mammals  of  North  America.  Ronald  Press,  New  York,  2:viii 
+  547-1083  +  7.9,  illustrated. 


Packard — Review  of  Ochrotomys  405 

Handley,  C.  O.,  Jr. 

1948.  Haliitat  of  the  golden  mouse  in  Virginia.  Jour.  Mamm.,  29:298-299. 
Handley,  C.  O.,  ]r.,  and  C.  P.  Patton 

1947.  Wild  mammals  of  \'irginia.  Virginia  Conmi.  Game  and  Inland 
Fisheries,  Richmond,  \i  +  220  pp.,  frontis  +  10.3  figs. 

Hershkovitz,  p. 

1944.  A  systematic  review  of  the  Neotropical  water  rats  of  the  genus 
Nectoini/s  ( Cricetinae ) .  Misc.  Publ.  Mus.  Zool.,  Univ.  Michigan, 
58:1-88;  4  pis.,  5  figs.,  2  maps. 

HiBBARD,  C.  W. 

1968.  Paleontology.  Pp.  6-26,  in  Biology  of  Peromyscus  (Rodentia)  (J.  A. 
King,  ed. ),  Spec.  Publ.  Amer.  Soc.  Mamm.,  xiii  +  593  pp.,  illus- 
trated. 

HOFFMEISTER,    D.    F.,  AND  C.    O.    MOHR 

1957.  Fieldbook  of  Illinois  mammals.  Manual  Illinois  Nat.  Hist.  Surv., 
4:xi  +  233,  illustrated. 

Hooper,  E.  T. 

1952.  A  systematic  review  of  the  harvest  mice  (genus  Reithwdontomys) 
of  Latin  America.  Misc.  Publ.  Mus.  Zool.,  Univ.  Michigan,  77:1-255, 
9  pis.,  24  figs.,  12  maps. 

1957.  Dental  patterns  in  mice  of  the  genus  Peromyscus.  Misc.  Publ.  Mus. 
Zool.,  Univ.  Michigan,  99:1-59,  24  figs. 

1958.  The  male  phallus  in  mice  of  the  genus  Peromyscus.  Misc.  Publ. 
Mus.  Zool.,  Univ.  Michigan,  105:1-24,  1  fig.,  14  pis. 

Hooper,  E.  T.,  and  G.  G.  Musser 

1964.  The  glans  penis  in  Neotropical  cricetines  (family  Muridae)  with 
comments  on  classification  of  muroid  rodents.  Misc.  Publ.  Mus. 
Zool.,  Univ.  Michigan,  123:1-57,  9  figs. 

Howell,  A.  H. 

1939.  Description  of  a  new  subspecies  of  the  golden  mouse.  Jour.  Mamm., 
20:498. 

HuBBS,  C.  L.,  and  C.  Hubbs 

1953.  An  impro\ed  graphical  analysis  and  comparison  of  series  of  samples. 
Syst.  Zool.,  2:49-56,  92,  4  figs. 

Kellogg,  R. 

1939.     Annotated  list  of  Tennessee  mammals.    Proc.  U.S.  Nat.  Mus.,  86: 
245-303. 
Layne,  J.  N. 

1960.  The  growth  and  development  of  young  golden  mice,  Ochrotomys 
mtttaUi.   Quart.  Jour.  Florida  Acad.  Sci.,  23:36-58,  4  figs. 

Llnzey,  D.  W.,  and  a.  V.  Linzey 

1967a.   Maturational  and  seasonal  molts  in  the  golden  mouse,  Ochrotomys 

uuttaUi.    Jour.  Mamm.,  48:2.36-241,  1  fig. 
1967b.  Growth  and  development  of  the  golden  mouse,  Ochrotomys  mittaUi 

nuttaUi.    Jour.  Mamm.,  48:445-458,  8  figs. 

LowERY,  G.  H.,  Jr. 

1936.     A  preliminary  report  on  the  distribution  of  the  mammals  of  Louisi- 
ana.  Proc.  Louisiana  Acad.  Sci.,  3:11-39,  4  pis.,  2  maps. 
1943.     Check-list  of  the  mammals  of  Louisiana  and  adjacent  waters.    Occas. 
Papers  Mus.  Zool.,  Louisiana  State  Univ.,  13:213-257,  5  figs. 

Manville,  R.  H. 

1961.  The  entepicondylar  foramen  and  Ochrotomys.  Jour.  Mamm.,  42: 
103-104. 

McCarley,  W.  H. 

1958.  Ecology,  behavior  and  population  dynamics  of  Peromyscus  mtttaUi 
in  eastern  Texas.    Texas  Jour.  Sci.,  10:147-171,  2  figs. 

1959a.  A  study  of  the  dynamics  of  a  population  of  Peromyscus  gossypinus 
and  P.  nuttalli  subjected  to  the  efiects  of  X-irradiation.  Amer.  Mid- 
land Nat.,  61:447-469,  4  figs. 

1959b.  The  mammals  of  eastern  Texas.   Texas  Jour.  Sci.,  11:385-426,  2  figs. 


406  Misc.  Publ.  51,  Unw.  Kansas  Mus.  Nat.  Hist. 

Miller,  G.  S.,  Jr. 

1912.     List  of  North  American  land  mammals  in  the  United  States  National 

Museum,  1911.    Bull.  U.S.  Nat.  Mus.,  79:  .xiv  +  1-455. 
1924.     List  of   North   American   Recent   mammals    1923.     Bull.    U.S.    Nat. 
Mus.,  128:  xvi  +  1-673. 
Miller,  G.  S.,  Jr.,  and  R.  Kellogg 

1955.  List  of  North  American  Recent  mammals.  Bull.  U.S.  Nat.  Mus.,  205: 
xiii  +  1-954. 

MuNSELL  Color  Company,  Inc. 

1954.     Munsell  soil  color  charts.   Baltimore,  3rd  ed.,  7  pi. 
Olson,  E.  C. 

1940.  A  late  Pleistocene  fauna  from  Herculaneum,  Missouri.  Jour.  Geol., 
48:32-56,  1  pi.,  9  figs. 

Osgood,  W.  H. 

1909.     Revision  of  the  mice  of  the  American  Genus  Peromyscus.    N.  Amer. 
Fauna,  28:1-285,  8  pis.,  12  figs. 
Packard,  R.  L. 

1960.     Speciation  and  evolution  of  the  pygmy  mice,  genus  Baiomys.    Univ. 

Kansas  Publ.,  Mus.  Nat.  Hist.,  9:579-670,  4  p's.,  12  figs. 
1968.     An  ecological  study  of  the  fulvous  harvest  mouse  in  eastern  Texas. 
Amer.  Midland  Nat.,  79:68-88,  9  figs. 
Packard,  R.  L.,  and  H.  Garner 

1964.  Arboreal  nests  of  the  golden  mouse  in  eastern  Texas.  Jour.  Mamm., 
45:369-374. 

Patton,  J.  L.,  AND  T.  C.  Hsu 

1967.  Chromosomes  of  the  golden  mouse,  Peromt/scus  (Ochrotomijs)  nut- 
ialli  (Harlan).   Jour.  Mamm.,  48:637-639,  i  fig. 

Petersen,  M.  K. 

1968.  Electrophoretic  blood-serum  patterns  in  selected  species  of  Pero- 
myscus.   Amer.  Midland  Nat.,  79:130-148,  5  figs. 

Rinker,  G.  C. 

1960.     The  entepicondylar  foramen  in  Peromyscus.    Jour.  Mamm.,  41:276. 
1963.     A  comparative  myological  study  of  three  subgenera  of  Peromyscus. 
Occas.  Papers  Mus.  Zool.,  Univ.  Michigan,  632:1-18,  1  fig. 
RippY,  C.  L.,  AND  M.  J.  Harvey 

1963.     Comparative  liehavioral  characteristics  of  six  genera  of  mice.   Trans. 
Kentucky  Acad.  Sci.,  24:5-8. 
Schwartz,  C.  W.,  and  E.  R.  Schwartz 

1959.     The  wild  mammals  of  Missouri.    Univ.  Missouri  Press,  viii   +  341 
pp.,  illustrated. 
Sealander,  J.  a.,  Jr. 

1956.  A  pro\'isional  check-list  and  key  to  the  mammals  of  Arkansas  (with 
annotations).   Amer.  Midland  Nat.,  56:257-296,  8  figs. 

Sprague,  J.  M. 

1941.  A  study  of  die  hyoid  apparatus  of  the  Cricetinae.  Jour.  Mamm.,  22: 
296-310,  5  figs. 

Trouessart,  E.  L. 

1897.     Catalogus  mammalium.  .  .  .    R.  Friedliinder  and  Son,  l:vi  4-  1-664. 

VORONTSOV,  N.  N. 

1957.  [The  structure  of  the  stomach  and  the  relati\e  de\elopment  of  parts 
of  the  gut  in  criutids  (Cricetinae,  Rodentia,  Mammalia)  of  the 
Palearctic  and  New  World.]  Doklady  Akad.  Nauk  SSSR,  117:526- 
529  ( in  Russian ) . 

Wagner,  J.  A. 

1843.  Die  Siiugethiere  ...  [of  Schreiber].  Leipzig,  3:,\iv  +  1-614,  illus- 
trated. 

Wood,  A.  E.,  and  R.  W.  Wilson 

1936.  A  suggested  nomenclature  for  the  cusps  of  the  cheek  teeth  of  ro- 
dents.   Jour.  Paleo.,  10:388-391,  2  figs. 


ALGUNOS  MURCIELAGOS  DEL  NORTE 
DE  ARGENTINA 

FOR 

Bernardo  Villa-R.  y  Martha  Villa  Cornejo 

La  rabia  es  una  zoonosis  que  a  pesar  de  su  antiguedad,  sigue 
siendo  un  problema  actual  que  atrae  la  atencion  de  muchos  in- 
vestigadores,  asi  como  de  instituciones  nacionales  e  internacionales. 
En  America  el  padecimiento  adquiere  un  mayor  relieve,  a  causa  de 
que  ha  demostrado  su  presencia  entre  los  mamiferos  del  orden 
Chiroptera  que  actuan  como  portadores,  destacandose  entre  ellos 
los  vampiros  o  murcielagos  chupadores  de  la  familia  Desmodonti- 
dae,  peculiares  solamente  del  Nuevo  Mundo,  que  por  sus  habitos 
alimenticios  pueden  ser  vectores  eficientes  para  infectar  a  otros 
mamiferos   (Villa-R.,   1966:469-471;   Bernstein,   1952:82-87,  92-93). 

Durante  los  meses  de  junio,  julio,  y  agosto  de  1965,  bajo  el 
patrocinio  de  la  Oficina  Sanitaria  Panamericana,  Organizacion 
Mundial  de  la  Salud  (a  promocion  del  Servicio  de  Luchas  Sani- 
tarias  (SELSA),  Direccion  General  de  Sanidad  Animal,  Secretaria 
de  Estado  de  Agricultura  y  Ganaderia  de  la  Nacion,  Republica 
Argentina),  uno  de  nosotros  (Villa-R.),  en  calidad  de  consultor  a 
corto  plazo,  de  la  Organizacion  Mundial  de  la  Salud,  llevo  a  cabo 
trabajos  de  Campo,  con  el  objeto  de  colectar  ejemplares  de  mur- 
cielagos y  otros  mamiferos  del  norte  de  la  Republica  Argentina 
para  estudiar  la  distribucion  geografica,  la  ecologia,  la  etologia  de 
los  murcielagos  y  sus  relaciones  con  el  virus  de  la  rabia. 

Agradecimientos 

Nos  es  placentero  manifestar  nuestra  gratitud  a  las  personas  siguientes,  por 
las  innumerables  muestras  de  fraternal  camaraderia  que,  en  una  o  en  otra 
forma,  hicieron  agradable  nuestra  estancia  en  aquel  pals  o  que,  con  su  com- 
pania  y  cooperacion  facilitaron  nuestros  trabajos:  Dr.  Boris  Szyfres,  Director 
del  Centro  Panamericano  de  Zoonosis:  Dr.  Cezar  A.  Mayer,  Director  Tecnico 
del  Servicio  de  Luchas  Sanitarias;  Dr.  Guillermo  Forrest;  Dr.  Abel  Retamozo 
Yepes,  del  mismo  servicio  (SELSA),  y  al  entusiasta  y  generoso  Dr.  Guztavo 
A.  Gonzalez  Blanco,  Segundo  Jefe  de  la  Campafia  contra  la  rabia  paresiante 
en  el  norte  de  Argentina,  quien  no  escatimo  ningun  esfuerzo  para  el  logro  de 
nuestros  propositos. 

Al  personal  a  sus  ordenes  patentizamos  tambien  nuestra  profunda  gratitud. 
Roberto  L.  Menini,  del  personal  del  Centro  Panamericano  de  Zoonosis 
(CEPANZO),  nos  acompano  durante  los  tres  meses,  llevando  al  cabo  con 
entusiasmo   y   dexocion   las   tareas   que   se   le   encomendaron.     Su   esposa    nos 

(407) 


408  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

acompano  durante  el  ultimo  mes  de  nuestra  estancia;  por  su  grata  compaiiia, 
conserxamos  imhorraliles  recuerdos  y  les  expresamos,  igualmente,  nuestra 
gratitud.  El  Dr.  Jorge  A.  Crespo  nos  permitio  exaniinar  los  ejemplares  a  su 
cargo,  en  el  Museo  Argentine  de  Ciencias  Naturales  "Bernardino  Rivadaxia"; 
William  Lopez-Forment,  nos  ayudo  a  elaborar,  estadisticamente,  las  medidas 
de  los  ejemplares  estudiados.  El  Biologo  Jose  Ramirez  Pulido  prepare  la 
mayoria  de  los  mapas  que  figiuan  en  el  texto. 

Seria  interminable  la  lista  de  todas  las  personas  que  nos  colmaron  de 
solicita  atencion,  pero  no  podemos  terminar  esta  parte  del  presente  trabajo, 
sin  hacer  mencion  de  la  hospitalidad  ejemplar  de  las  maestras  de  la  Escuela 
Pro\incial  Numero  138  de  Palma  Sola,  Departamento  de  Santa  Barbara,  al 
oriente  de  la  Provincia  de  Jujuy,  profesoras  Maria  Mercedes  del  Valle  Castro, 
Maria  Brigida  Irades,  Amanda  Davalos,  Martha  Matche,  Ana  Maria  Quibal, 
y  la  familia  del  enfermero,  seiior  S.  Roberto  Jimenez.  La  Seiiorita  Herminia 
Martinez  Gonzalez  escribio  el  manuscrito  en  su  forma  final. 

Material  y  Metodos 

Nuestras  observaciones  se  basan  en  mas  de  un  miliar  de  ejemplares  colec- 
tados  en  diferentes  localidades  de  las  provincias  de  Jujuy,  Salta,  Formosa, 
Chaco,  Catamarca,  Tucuman,  Cordoba,  Santa  Fe,  Corrientes  y  Misiones. 
De  todo  este  material  se  prepararon  ejemplares  convencionales  para  estudio 
cientifico  (piel  y  craneo),  tratando  de  tener  representadas,  en  esta  forma,  a  las 
especies  capturadas  durante  nuestra  visita.  La  mayoria  de  los  murcielagos 
obtenidos  correspondio  a  Dcsmodus  rotundus  rotundus  y  a  Tadarida  hrasdiensis 
hrasdiensis.  De  estas  especies,  los  animales  no  preparados  para  estudio  cien- 
tifico se  repartieron  en  partes  iguales  entre  el  Centre  Panamericano  de  Zoonosis 
y  SELSA,  para  investigaciones  virologicas.  Algunos  ejemplares  se  transportaron 
\ivos  hasta  la  Ciudad  de  Buenos  Aires,  al  Centre  Panamericano  de  Zoonosis, 
en  Azul,  o  se  exhibieron  en  las  expesiciones  ganaderas  de  Perico  del  Carmen, 
Jujuy  y  Palermo,  un  barrio  de  la  Ciudad  Capital  de  la  Argentina.  En  las 
colecciones  del  Institute  de  Biologia,  de  la  Uni\ersidad  Nacional  Autonoma 
de  Mexico,  estan  depesitades,  por  tanto,  100  ejemplares  para  estudio,  (piel 
y  craneo).  Ademas,  se  examine  el  material  existente  en  el  Museo  Argentine 
de  Histeria  Natural  (Bernardino  Rivadavia)  y  en  los  casos  pertinentes,  se 
hicieren  cemparaciones  con  material  pro\eniente  de  Mexico,  Costa  Rica, 
Panama,  Colombia,  Peru,  Brasil,  y  Chile. 

Mas  del  cincuenta  per  ciento  del  total  de  los  ejemplares  a  que  se  ha  hecho 
referencia  se  capture  usando  el  metode  de  envenenamiento  con  gas  Cyanamid, 
un  precedimiento  que  se  habia  puesto  en  use  por  el  personal  de  SELSA  para 
el  combate  de  les  murcielagos  per  las  razenes  que  en  el  lugar  aprepiade  se 
explicaran. 

Ademas,  se  celecto  haciendo  use  de  redes  entomelogicas  para  la  captura 
de  murcielagos  y,  sebre  tode,  se  usaren  redes  japonesas  de  seda,  "mist  nets," 
de  diferente  tamafie. 

Para  el  tratamiente  sistematico  seguimos  el  precedimiento  clasico  basado 
en  les  caracteres  merfologices  externos  y  craneales,  apoyandoles  en  carac- 
teristicas  numericas.  Come  no  fue  posible,  en  tedes  les  cases,  semeter  los 
ejemplares  a  tratamientes  estadisticos,  para  obtener  la  media  aritmetica  y  las 
medidas  de  variabilidad,  se  consignan  les  premedies  y  se  registran  la  minima 


ViLLA-R.  AND  Villa  Cornejo — Argentine  Bats  409 

y  la  maxima  de  las  caiacteristicas  mensii rallies;  el  mimero  de  ejemplares 
promediados  o  sometidos  a  tiataiiiientos  estadisticos  se  da  entre  parentesis, 
despues  de  las  cifras  relativas  a  los  otros  datos. 

Las  medidas  se  dan  en  milimetros.  De  entre  estas,  la  longitud  total  del 
cuerpo  que  generalmente  se  toma  desde  la  punta  de  la  nariz,  hasta  la  punta 
de  la  cola  \'ertebial,  sin  tomar  en  cuenta  los  pelos  cuando  los  hay,  en  D.  r. 
rotitndiis  se  elimino,  eonsideiando  que  anoja  cifras  muy  variables  y,  funda- 
mentalmente,  porque  carece  de  cola,  lo  mismo  que  Artiheus.  Las  medidas 
craneales  se  tomaron  con  un  "metric  dial  calipers,"  hasta  decimas  de  milimetro. 

Los  nomlires  de  los  colores  del  pelaje  escritos  con  mayuscnla  estan  de 
acuerdo  con  Ridgway  (1912).  Los  especimenes  que  forman  la  base  de  este 
estudio  se  encuentran  catalogados  en  el  Institute  de  Biologia,  Universidad 
Nacional  Autonoma  de  Mexico. 

En  el  estudio  de  las  soml^ras  del  mismo  pelaje,  sobre  todo  de  Desmochis 
rotitndus,  buscando  las  diferencias  entre  las  dos  subespecies  hasta  ahora 
reconocidas,  se  uso  la  tecnica  colorimetrica  utilizando  un  fotometro.  Weston 
Master,  que  coincide  con  el  procedimiento  descrito  por  Desha  ( 1965:  233-236). 

Las  Zonas  Fitogeograficas  del  Norte  Argentino 

Como  la  presencia  de  los  mamiferos  esta  determinada  funda- 
mentahnente  por  la  presencia  o  ausencia  de  "habitats"  adecuados, 
de  los  que  las  plantas  son  parte  importante  y,  puesto  que  las  obser- 
vaciones  se  hicieron  en  un  lapso  corto  y  durante  una  sola  estacion 
del  ano  (el  invierno),  para  dar  una  mejor  idea  de  la  ecologia  del 
norte  Argentino,  seguimos  aqui  el  trabajo  de  Cabrera  (1953). 
Veanse,  ademas,  Castellanes  y  Perez  Moreau  (1944)  y  Hauman 
( 1917).  Por  otra  parte,  se  vera  en  su  oportunidad,  que  los  vampiros 
de  esta  parte  de  las  Americas,  tienen,  como  refugios  diurnos,  prin- 
cipalmente  la  oquedad  de  los  arboles,  en  mayor  proporcion;  en 
tanto  que,  en  Mexico,  se  les  encuentra  mas  frecuentemente  en 
cuevas,  minas  abandonadas,  fisuras  de  rocas,  casas  deshabitadas  y, 
en  ultimo  lugar,  en  el  hueco  de  los  arboles. 

Segun  Cabrera  (1953),  la  parte  norte  de  Argentina  corresponde 
a  la  region  Neotropical  con  las  siguientes  subdiviciones  (vease 
Fig.  1). 

Dominio  de  la  America  Subtropical 

El  Dominio  de  la  America  Subtropical  ocupa  una  superficie  muy  reducida 
en  el  norte  Argentino  y  esta  tipificado  por  la  Provincia  Subtropical  Occidental 
y  la  Provincia  Oriental.  El  clima  es  calido  y  humedo  con  lluvias  principalmente 
en  el  Verano  y  heladas  durante  el  invierno.  La  vegetacion  predominante  es 
pedemontana. 

La  Provincia  Subtropical  Occidental  presenta  los  siguientes  distritos, 
Oranense,  Tucumanense,  y  Montano.  El  Distrito  Tucumanense  se  extiende 
por  los  contrafuertes  de  la  cordillera,  a  traves  del  sur  de  Salta  y  por  el  centro 
de  Tucuman,  penentrando  ligeramente  en  el  este  de  Catamarca.    La  composi- 


410 


Misc.  Publ.  51,  Uxiv.  Kansas  Mus.  Nat.  Hist. 


(bo)  (»•) 


Fig.  1.  Las  zonas  fitogeograficas  del  Norte  de  Argentina,  donde  se  efectuaron 
los  trabajos  de  campo  a  que  se  hace  referenda  en  el  texto,  senalandose  con 
una  linea  gruesa  el  Hmite  mas  sureiio  de  la  distribucion  del  vampiro  de  patas 
pelonas  Desmodus  rotimdus.  A  pesar  de  las  diferencias  de  la  flora,  de  la 
topografia,  de  las  condiciones  naturales  ambientales,  en  todas  se  registra  la 
presencia  de  estos  niurcielagos  heniatofagos. 


ViLLA-R.  AND  Villa  Coknejo — Argentine  Bats  411 

cion  es  parecida  a  la  del  Distrito  Oranense,  pero  con  nienor  numero  de  especies. 
Los  bosques  ribereiios  son  de  la  misma  composicion  que  en  el  Distrito 
Oranense. 

El  Distiito  Montano  ocupa  las  laderas  orientadas  al  este  de  las  montaiias 
del  noroeste  de  la  Argentina,  por  arriba  de  los  distritos  Oranense  y  Tucuma- 
nense  entre  los  1500  y  los  2500  metros  de  altina.  El  clima  es  mas  fresco  que 
en  los  anteriores.  La  vegetacion  dominante  esta  constituida  por  bosque  en  las 
laderas  enipinadas  y  en  las  quebradas  y  por  praderas  en  las  laderas  suaves. 

La  Provincia  Subtropical  Oriental  se  extiende  por  el  extremo  noroeste  de  la 
Republica,  ocupando  todo  el  Territorio  de  Misiones  y  el  noroeste  de  Corrientes. 
Adenias,  se  prolonga  en  forma  de  galeria,  a  lo  largo  de  los  rios  Parana  y 
Uruguay,  hasta  El  Plata.  Taml:)ien  asciende  por  los  afluentes  de  estos  grandes 
rios.  Al  sur  y  al  oeste  limita  con  el  dominio  Chaqueno,  al  norte  y  al  este 
penetra  en  el  Paraguay  y  en  el  Brasil  donde  alcanza  su  mayor  extension.  Estos 
corredores  destacan  claramente  en  vista  aerea  volando  desde  Iguazu  a  Posadas 
y  a  Resistencia. 

El  Clima  de  esta  provincia  es  calido  y  humedo,  con  precipitaciones  durante 
todo  el  aiio,  torrenciales  en  el  \erano.  La  precipitacion  puede  alcanzar  los 
2000  mm.  anuales  en  el  norte  de  Misiones,  descendiendo  a  1400  mm.  en  Cor- 
rientes. En  esta  Pro\'incia  ecologica,  el  suelo  es  lateritico,  rojo.  La  vegetacion 
predominante  es  de  selvas  y  sabanas.  Al  tiempo  de  nuestras  observaciones  im 
extenso  desmonte  modifical^a  el  estado  de  la  vegetacion  caracteristica. 

Esta  provincia  ecologica  esta  cercanamente  relacionada  con  la  subtropical 
occidental,  pero  son  notables  o  exclusivos  los  generos  Araucaria,  Balfouroden- 
dron,  Holocalijx,  Cahralea,  y  Mabbaeriuin.  Es  rica  en  bambusaseas  y  en 
lielechos  arborescentes. 

Segiin  el  autor  que  hemos  xenido  siguiendo,  en  esta  region  se  pueden 
distinguir  tres  distritos:  Distrito  de  los  Pinares,  por  la  presencia  y  generalmente 
abundancia  de  Araucaria  angustifolia;  Distrito  de  las  Selvas  Mixtas,  funda- 
mentalmente  con  los  mismo  elementos  del  distrito  anterior,  pero  sin  Araucaria; 
y  Distrito  de  los  Campos,  caracterizado  por  la  predominancia  de  las  sabanas. 

Dominio  Chaqueno 

El  Dominio  Chaqueno  es  el  que  ocupa  la  mayor  parte  del  territorio  del 
norte  Argentino  a  que  se  refiere  el  presente  trabajo.  Se  extiende  practicamente 
desde  el  Atlantico  hasta  la  Cordillera  de  los  Andes  y  desde  el  limite  con  el 
Paraguay  hasta  el  norte  de  Chul^ut.  La  fisonomia  de  su  xegetacion  es  poli- 
morfa — liosques  xerofilos,  caducifolios,  estepas  arbustivas,  estepas  herbaceas, 
sabanas,  praderas,  palmares,  pajonales  y  otras. 

Su  clima  es  continental,  con  lluvias  escasas  ( con  excepcion  de  la  Provincia 
Bonaerense),  esti\ales  en  la  parte  Norte  del  Dominio,  primaverales  y  otonales 
en  el  Sur.  La  temperatura  es  elevada  en  el  verano  y  templada  en  el  invierno. 
En  ciertas  zonas  la  oscilacion  diaria  es  muy  amplia  con  diferencia  hasta  de 
39°  C. 

El  Dominio  Chaqueiio  se  divide  en  cinco  provincias  que  se  pueden  recono- 
cer  por  los  siguientes  caracteres:  Pro\incia  Chaquena  o  del  Chaco,  con 
predominancia  de  Shinopsis  y  Aspidosperma;  Provincia  del  Espinal,  muy  seme- 
jante  a  la  anterior,  pero  sin  Schinopsia  y  con  predominancia  del  genero 
Prosopis;    Provincia    Prepunena,    con    especies    escasas    y    predominancia    de 


412 


Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 


Fig.    2.    Localidades  de  donde  precede  el  material  examinado. 


cactaceas  careiformes  y  cigofilaceas,  leguminosas  y  compuestas  arbiistivas; 
Provincia  del  Monte,  sin  arboles  o  con  especies  arboreas  enanas  y  predominan- 
cia  de  cigofilaceas  arbustivas  del  genero  Larrea  y  Provincia  Pampeana,  sin 
arboles,  con  predominancia  de  gramineas  xerofitas  de  los  generos  Stipa,  Pipto- 
diaetrtim,  Atuhopopoti,  Elionurtis  y  otros. 

El  Distrito  Chaqueno  oriental  ocupa  la  mitad  oriental  del  territorio  de 
Formosa,  del  de  la  Provincia  (poHtica)  del  Chaco,  del  norte  de  Santa  Fe  y 
al  oeste  de  Corrientes.  Entre  los  arboles  que  le  caracterizan,  entre  otros 
mnclios,  harenios  especial  mencion  del  "gua>acan,"  Cacsalpinia  paraguaijensis. 
En  general,  es  el  area  de  los  quebrachales,  algarrobales  y  palmares. 


ViLLA-R.  AND  Villa  Cornejo — Argentine  Bats  413 

El  Distrito  Chaqueiio  occidental  se  extiende  por  la  niitad  occidental  de 
Formosa  y  el  Chaco,  todo  el  oriente  de  Salta,  todo  el  extremo  oriental  de 
Jujuy  y  Tucuman.  Este  distrito  es  mas  seco  que  el  oriental,  con  vegetacion 
formada  por  bosques  xerofilos,  casi  sin  interrupcion,  algunos  palmares,  estepas 
halofitas  y  sabanas  originadas  por  incendios  y  desmontes.  Entre  la  variada 
composicion  de  arboles  que  se  encuentran  en  este  distrito,  tambien  haremos 
especial  mencion  del  "yuchan"  o  "palo  borraclio,"  CJiorisia  insignis. 

El  Distrito  Chaqueiio  Serrano  se  extiende  de  norte  a  sur,  a  lo  largo  de  las 
montaiias  bajas  que  forman  los  primeros  contrafuertes  de  la  Cordillera  Andina 
y  en  el  este  de  la  Provincia  de  Jujuy,  en  el  centro  de  Salta  y  Tucuman;  en  el 
extremo  oriental  de  Catamarca,  prolongandose  hasta  las  sierras  de  Cordoba, 
San  Luis,  y  la  Rioja.  Llega  aproximadamente  hasta  los  33  grados  de  latitud 
Sur;  suele  ocupar  las  laderas  bajas  de  los  cerros  y  quebradas,  formando  vm 
complicado  engranaje  con  la  provincia  Subtropical  Occidental  constituida  por 
bosques  serranos  y  estepas  serranas.  Esta  es  la  region  mejor  conocida  por 
nosotros,  porque  en  ella  nos  movimos  constantemente;  sobre  todo,  es  la  parte 
de  la  Argentina  en  donde  colectamos  con  mas  intensidad.  En  la  Figura  1,  se 
mdican  las  zonas  fitogeograficas  arriba  mencionadas,  en  las  que  se  extiende, 
precisamente;  la  distriliucion  de  los  vampiros  Dcsmodus  rotiindiis  rotundus, 
sefialada  con  una  linea  muy  negra  y  continua  en  su  parte  mas  sureiia. 

Nomenclator  Geografico 

Las  siguientes  localidades  en  donde  se  obtuvieron  los  ejemplares 
de  murcielagos  estudiados  (vease  Fig.  2)  fueron  registradas  sigui- 
endo  el  mapa  general  de  la  Republica  Argentina  (la  escala  1: 
3,750,000)  las  localidades  correspondientes  a  otros  paises  se  han 
anotado  de  acuerdo  con  el  "Hammond's  Map  of  Latin  America" 
(a  la  escala  de  1:  10,000).  Tambien  se  hizo  uso  de  la  Carta  Aero- 
nautica  Mmidial,  OACI  (a  la  escala  de  1:  1,000,000),  particular- 
mente  en  el  caso  de  las  localidades  de  la  Republica  del  Uruguay. 

Tratamiento  Sistematico 
Familia  Noctilioxidae 

Noctilio  leporinus  rufipes  D'Orbigny 
Murcielago  ictiofago 

Ejemplares  examinados  (3). — Provixcia  de  Salta,  Hickman,  Chaco  Sal- 
teno,  1.  Provixcia  de  Jujuy,  Embarcacion,  Finca  Tres  Pozos,  Departamento 
San  Martin,  247  km.  NE  San  Salvador  de  Jujuy,  2. 

Medidas. — Ejemplares  nos.  9495,  9497,  9498  consignadas  en  el  mismo 
orden:  longitud  total  del  cuerpo,  128.0,  11.0,  115.0;  cola  vertebral,  16.4,  20.0, 
20.0;  oreja  desde  la  escotadura  32.3,  26.0,  30.2;  pata  trasera,  25.4,  29.0,  27.0; 
antebrazo,  87.5,  86.4,  87.1;  tibia,  33.8,  32.2,  39.9;  tercer  dedo  metacarpiano, 
79.1,  78.6,  77.8;  primera  falange,  20.7,  20.2,  20.6;  segunda  falange,  9.4,  7.4, 
8.9;  longitud  mayor  del  craneo,  25.8,  26.0,  27.3;  longitud  condilobasal,  21.5, 
21.5,  21.6;  longitud  palatal,  13.1,  12.9,  13.5;  anchura  bicigomatica,  19.9,  19.1, 
19.5;  anchura  interorbitaria,  7.3,  7.1,  7.1;  anchura  del  rostro,  12.0,  12.0,  12.1; 


414  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

anchiira  de  la  caja  craneal,  15.4,  13.8,  13.8;  hilera  superior  de  dientes,  10.2, 
10.3,  10.7;  anchura  a  traves  de  los  caninos,  8.9,  8.9,  8.5;  anchura  a  traves  de 
M3-M3,  12.5,  12.4,  12.5. 

Ohservaciones. — El  nombre  de  la  subespecie  se  usa  aqui  siguiendo  las 
concliiciones  de  Cabrera  (1935:13).  Este  autor  deja  explicadas  claramente  las 
razones  por  las  que  adopto  el  nombre  Noctilio  leporinus  rufipes.  Evidente- 
mente,  son  animales  de  tamano  mayor  que  N.  I.  mexicanus.  Se  han  hecho 
comparaciones  directas  con  el  material  de  N.  I.  mexicanus  e.xistente  en  las 
colecciones  del  Instituto  de  Biologia  ( UNAM )  y  no  hay  duda  en  cuanto  a  la 
diferencia  de  tamano.  En  los  tres  ejemplares  examinados,  el  color  dorsal  varia 
de  Ochraceous-Tawny  a  Buckthorn  Brown,  con  el  cuello  ligeramente  mas  claro, 
asi  como  la  Hnea  media  dorsal;  el  vientre  es  Orange-Buff;  por  tanto  no  es  tan 
rojizo  como  se  infiere  de  la  descripcion  de  Cabrera.  En  los  ejemplares  que 
tenemos  a  nuestra  disposicion,  todas  hembras,  no  hay  diferencias  marcadas  a 
este  respecto,  Villa-R.  (1966:164-165)  refiriendose  a  N.  I.  mexicanus  e.xplica 
que  en  esta  especie  hay  una  gran  variacion  en  el  color  que  va  desde  el  Brussels 
Brown  hasta  el  Zinc  Orange,  pasando  por  el  Buckthorn  Brown;  este  color,  como 
se  ve,  lo  encontramos  en  los  ejemplares  argentinos,  pero  no  estamos  en  condi- 
ciones  de  aclarar  si  es  un  fenomeno  generalizado,  a  causa  del  exiguo  material 
exaniinado. 

Aimque  no  es  la  primera  vez  que  se  cita  a  N.  I.  rufipes  del  extremo  norte 
de  Argentina,  porque  ya  fue  mencionado  por  Burmeister  que,  segun  Cabrera 
(op.  cit.:  6),  mas  bien  se  referia  a  "Dirias  albivetiter,"  en  esta  ocasion  esta 
representado  de  modo  inequivoco,  de  la  Provdncia  Subtropical  de  Salta  y  de 
Jujuy  en  los  ejemplares  estudiados,  muy  dentro  del  tenitorio  continental,  cerca 
de  rios  y  depositos  dulceacuicolas. 

Familia  Phyllostomatidae 

Tonatia  silvicola  silvicola  (D'Orbigny) 

Murcielago  de  orejas  redondas 

Ejemplares  examinados  (1). — Provincia  de  Misiones:  Cataratas  del 
Igiiazu,  125  m. 

Medidas. — Longitud  total  del  cuerpo,  29.0;  cola  vertebral,  20.0;  pata 
tra.sera,  16.0;  oreja  desde  la  escotadura,  29.0;  antebrazo,  53.7;  tibia,  22.4. 

Ohservaciones. — El  ejemplar  fue  colectado  junto  con  otros  del  genero 
Sturnira,  en  las  cercanias  de  la  caida  de  agua  "Dos  Hermanas"  donde  colocamos 
varias  redes,  en  las  inmediaciones  de  un  restaurante  en  cuyo  interior  se 
encontraban  varios  racimos  de  platanos  suspendidos  de  las  vigas  de  madera 
del  techo.  Esta  especies  tiene  una  distribucion  extensa  desde  Brazil  Central  y 
el  norte  de  Bolivia  hasta  Mexico.  Aparentemente,  con  este  ejemplar  la 
distribucion  debe  agregarse  el  noreste  de  Argentina. 

Fue  el  primer  ejemplar  que  atrapamos  a  las  siete  horas  de  la  tarde.  En  este 
lugar  la  temperatura,  durante  el  dia,  era  de  27°  C.  a  las  tres  horas  de  la  tarde 
con  gran  humedad  deliido  a  la  cercania  de  las  imponentes  Cataratas  del  Iguazu 
que  marcan  la  frontera  entre  Brasil  y  la  Republica  Argentina.  La  vegetacion 
es  caracteristicamente  tropical.  Debemos  aclarar  que  el  craneo  de  este 
murcielago  fue  exaniinado  con  cuidado,  inmediatamente  despues  que  fue 
preparada  la  piel,  pero  por  desgracia,  lo  perdimos  en  el  trayecto.  El  tratamiento 
sistematico  que  se  da  aqui  esta  basado  en  los  caracteres  externos,  principal- 


ViLLA-R.  AND  Villa  Cornejo — Argentine  Bats  415 

mente.  Segun  parece,  es  este  el  primer  registro  de  la  especie  en  territorio 
Argentino.  En  septiemlire  de  1967.  Fornes  cf  al.  (1967:149-152)  describen  otro 
ejemplai"  de  Pabna  Sola,  Provincia  de  Jnjuy. 

Chrotopterus  auritus  australis  Thomas 
Gran  murcielago  carnivore 

Ejemplares  exatninados  (3).- — Provincia  de  Salta:  Mina  Pablo,  Unchime, 
25  km.  E  General  Guemes,  1100  m.,  2;  Rio  Mojo  Toro,  5  km.  N  Salta,  1. 
Vease  tambien  explicacion  bajo  observaciones. 

Medidas. — Ejemplares  nos.  9496  (hembra),  9678  (hembra),  y  9579 
(macho)  sigiiiendo  este  mismo  orden  son  como  sigue:  longitiid  total  del 
cuerpo,  110.0,  110.0,  110.0;  oreja  desde  la  escotadura,  26.0,  29.0,  23.0;  pata 
trasera,  43.0,  50.5,  52.1;  antebrazo,  80.0,  85.0,  78.9;  tibia,  36.8,  39.1,  36.7; 
longitud  mayor  del  craneo,  36.5,  37.0,  36.8;  longitud  condilobasal,  32.1,  32.8, 
32.7;  longitnd  palatal,  17.3,  17.7,  16.9;  anchura  bicigomatica,  19.5,  19.2,  19.7; 
anchura  interorbitaria,  6.3,  6.3,  6.2;  anchura  del  rostro,  8.7,  8.4,  9.3;  anchura 
mastoidea,  15.2,  15.3,  15.4;  anchura  de  la  caja  craneal,  14.0,  14.0,  14.1;  hilera 
superior  de  dientes,  13.5,  13.4,  13.9;  anchura  a  traves  de  los  caninos,  7.4,  7.3, 
7.0;  anchtna  a  traves  de  M3-M3,  12.0,  11.0,  11.0. 

Observaciones. — Los  ejemplares  registrados  en  el  piesente  trabajo  amplian 
su  distribucion  hasta  la  Provincia  de  Salta.  Cinco  ejemplares  vivos  fueron 
obtenidos  el  31  de  julio  de  1965,  en  el  interior  de  la  Mina  Pablo,  a  40  metros 
de  profundidad  medidos  desde  la  entrada.  La  captura  se  hizo  con  una  red  de 
mano,  a  las  cinco  boras  de  la  tarde,  la  temperatura  era  de  20°  C.  con  una 
humedad  relativa  de  80  por  ciento.  Los  murcielagos  formaban  im  grupo 
compacto,  suspendidos  del  techo  con  las  patas  y  con  la  cabeza  hacia  abajo. 
Puede  decirse  de  ellos,  segun  la  clasificacion  de  Villa-R.  (1966:224-226)  que 
son  niTircielagos  litofagos  internos  libres. 

Directamente  abajo  del  grupo,  sobre  el  piso  de  la  mina  fueron  encontrados 
fragmentos  de  esqueletos,  piel  y  pelos  de  pequeiios  mamiferos,  probablemente 
del  genero  Ctenomijs,  indudables  residuos  de  la  dieta  de  los  murcielagos  en 
cuestion.  Los  cinco  animales  fueron  transportados  \ivos  para  e.xhibirse  en 
una  exposicion  ganadera  local  en  Perico  del  Carmen,  Jujuy;  se  les  alimento 
en  la  misma  noche  del  dia  de  su  captura,  con  carne  de  una  ave  conocida 
localmente  con  el  nombre  de  "Chuiia  pata  colorada"  (Carioma  cristata).  En  los 
dias  signientes  se  les  alimento  con  carne  de  \aca.  Posteriormente,  en  marzo 
y  abril  de  1967,  en  una  pequena  oquedad  conocida  con  el  nombre  de  Furna 
de  Zafreire,  en  las  cercanias  de  la  Universidad  Rural  do  Brasil,  Rio  de  Janeiro, 
Brasil,  uno  de  nosotros  ( Villa-R. )  colecto  otro  ejemplar,  ocupando  el  mismo 
refugio  con  Desiuodus  rotundus  lotiindus.  En  cautixerio  se  observe  otro 
ejemplar  en  el  Institute  Biologico  de  Sao  Paulo,  devorando  ejemplares  de 
Desmodus  y  ratones  blancos  de  laboratorio.  De  los  cinco  murcielagos  obtenidos, 
preparamos  como  ejemplares  para  estudio  cientifico,  im  macho  y  una  hembra; 
el  macho  tenia  los  testiculos  escrotados  y  la  hembra  un  embrion  en  los  primeros 
estados  del  desarrollo.  El  otro  ejemplar  fue  capturado  en  las  Hsuras  de  una 
roca  de  los  bancos  del  Rio  Mojo  Toro,  en  las  cercanias  de  la  Ciudad  de  Salta. 

Esta  especies  es  notable  por  su  cuerpo  robusto  y  las  grandes  orejas  redon- 
deadas,  con  la  base  y  el  borde  inferior  interno  cubierto  de  pelos  sedosos  de 
color  Light  Ochraceous-BufF.    Casi  las  tres  cuartas  partes  del  antebrazo  estan 


416  Misc.  Publ.  51,  Uxw.  Kansas  Mus.  Nat.  Hist. 

cubiertas  de  pelo,  dorsal  y  ventralmente.  La  membrana  interfemoral  es  ancha 
y  la  membrana  alar  se  desprende  de  la  base  de  las  falanges,  comprendiendo 
todo  el  borde  externo  del  tarso.  La  coloracion  general  del  dorso  es  Mummy 
Brown,  con  el  cuello  ligeramente  mas  claro,  los  pelos  de  esta  parte  del  cuerpo 
son  francamente  blancos.  El  pelaje  es  largo,  denso  y  sedoso.  La  region 
cercana  a  los  organos  genitales  es  ligeramente  Russet. 

Stumira  lilium  lilium  (E.  Geoffroy  St.-Hilaire) 
Murcielago  de  Charreteras 

Ejemplares  examinados  (15). — Provincia  de  Jujuy:  Finca  El  Remate,  24 
km.  SE  San  Salvador  de  Jujuy,  740  m.,  3;  Arroyo  de  la  Urbana,  45  km.  E  y 
5.4  km.  N  San  Salvador  de  Jujuy,  620  m.,  1;  Palma  Sola,  550  m.,  2.  Provincia 
DE  Mi.sioxES,  Cataratas  del  Iguazu,  125  m.,  9. 

Mcdidas. — El  promedio,  minima  y  maxima  entre  parentesis,  de  las  medidas 
somaticas  son  como  sigue:  longitud  total  del  cuerpo,  64.0  (55.7-72.8);  pata 
trasera,  13.0  (10.0-19.9);  oreja  desde  la  escotadura,  15.0  (16.3).  Las  medidas 
craneales  fueron  tratadas  de  ocho  ejemplares:  longitud  mayor  del  craneo, 
26.4  (22.7-22.9);  longitud  condilobasal,  20.9  (20.5-21.5);  longitud  palatal, 
9.9  (10.3);  anchura  bicigomadca,  13.0  (13.0-13.6);  anchura  interorbitaria, 
5.8  (5.4-6.2);  anchura  del  rostro,  6.3  (6.1-7.1);  anchura  mastoidea,  12.2 
(10.7-13.5);  hilera  superior  de  dientes,  6.5  (6.1-6.9);  anchura  a  traves  de  los 
caninos,  6.5  (6.0-6.6),  7  ejemplares;  anchura  a  traves  de  M3-M3,  8.1  (7.9-8.4). 

Ohservaciones. — Los  murcielagos  de  esta  especies  fueron  los  que  con  mas 
frecuencia  atrapamos  en  las  redes.  En  las  cercanias  de  las  Cataratas  del 
Iguazu,  constituyeron  el  mayor  niimero  del  total  de  animales  atrapados  a  las 
primeras  horas  del  crepusculo  vespertino.  En  esta  localidad,  la  temperatura 
alcanzaba,  en  el  tiempo  de  nuestra  visita,  hasta  27°  C,  con  una  gran  humedad 
rclativa. 

Como  se  ha  dicho  al  tratar  de  Tonatia,  la  vegetacion  es  exuberante,  como 
corresponde  a  la  Provincia  Subtropical  Oriental. 

En  general,  las  medidas  de  S.  /.  lilium  son  difinitivamente  mayores  que  las 
de  la  especies  de  Mexico,  pero  por  sus  caracteres  externos  no  se  pueden 
diferenciar  claramente.  Entre  el  material  que  existe  en  la  coleccion,  se  obser\an 
dos  fases  de  coloracion,  una  marcadamente  rojiza  y  la  otra  cafe,  con  tonalidades 
intermedias  que  van  desde  el  Ochraceous-Tawny,  pasando  por  el  Tawny  hasta 
el  Mars  Brown.  La  mancha  amarilla  de  los  hombros  solo  aparece  bien  notable 
en  dos  ejemplares  machos  y  es  Raw  Sienna.  La  membrana  interfemoral  esta 
profusamente  cubierta  de  pelos  que  se  proyectan  mas  alia  de  su  borde.  Segun 
las  ohservaciones  en  el  lugar  de  su  colecta,  toman  los  frutos  de  la  palma 
datilifera  y  en  las  cercanias  de  las  Cataratas  del  Iguazu  eran  atraidos  por  los 
racimos  de  platanos  ( Musa  sp. )  suspendidos  de  las  vigas  de  madera  del  techo 
del  restaurante  en  cuyas  cercanias  colocamos  las  redes. 

La  hembra  (no.  9533)  se  encontraba  lactando;  los  machos  tenian  los 
testiculos  no  escrotados. 

Artibeus  lituratus  lituratus  (Olfers) 
Gran  murcielago  frutero 
Ejemplares  examinados  (7). — Provixci.\  de  Jujuy:    Finca  El  Remate,  24 
km.  SE  San  Salvador  de  Jujuy,  740  m.,  2;   Finca   La  Carolina,   Los   Perales, 
San  Salvador  de  Jujuy,  1310  m.,  5. 


ViLLA-R.  AND  Villa  Cokxejo — Argentine  Bats  417 

Medidas. — Los  promedios  de  los  siete  ejemplares  con  las  minimas  y 
maximas  correspondientes  entre  parentesis  son  como  sigue:  longitud  total  del 
cuerpo,  88.0  (75.0-98.0);  pata  trasera,  17.0  (12.4-23.0);  oreja  desde  la  esco- 
tadura,  17.0  (18.0-23.0);  antebrazo,  65.1  (64.4-65.0);  tibia,  21.9  (21.3-22.5); 
longitud  mayor  del  craneo,  30.6  (30.2-31.4);  longitud  condilobasal,  27.6  (27.1- 
28.2);  longitud  palatal,  15.1  (14.3-15.6);  anchura  bicigomatica,  19.2  (18.8- 
20.0);  anchura  interorbitaria,  74.0  (73.0-77.0);  anchura  del  rostro,  11.4  (11.0- 
12.0);  anchura  mastoidea,  16.2  (15.8-16.8);  anchura  de  la  caja  craneal,  13.6 
(13.0-13.8);  hilera  superior  de  dientes,  11.3  (11.0-11.9);  anchura  a  traves  de 
los  caninos,  7.8  (7.5-8.1);  anchura  a  traves  de  M3-M3,  13.7   (13.1-14.9). 

Observaciones. — Este  gran  niurcielago  se  distingue  facilmente  de  A. 
jamaicensis  jamaicensis  por  su  gran  tamafio;  el  antebrazo  alcanza  y  sobrepasa 
los  65  mm.;  en  la  Finca  El  Remate,  donde  se  culti\'a  principalmente  algodon, 
encontramos  ima  pequena  fraccion  de  terreno  cercana  a  la  casa  principal,  con 
un  buen  ninnero  de  palmas  en  plena  fructificacion  y  algunos  naranjales,  tam- 
bien  en  floracion;  las  redes  fueron  colocadas  entre  las  palmeras.  Durante  las 
prinieras  horas  del  crepusculo  capturamos  algunos  ejemplares  de  Sttirnira 
lilium  lilium  y  entrada  la  noche  observamos  gran  numero  de  Artiheus  lituratus 
lituratus  en  torno  de  las  frondas  de  las  palmeras  apoderandose  de  los  frutos 
maduros  o  tomando  el  polen  de  las  flores.  Como  acontece  con  los  higos  sil- 
vestres  en  Mexico  (vease  Villa-R.,  1966:298),  algunos  frutos  de  las  palmeras 
se  desprendian  cayendo  al  suelo,  produciendo  un  ruido  que  era  facilmente 
perceptible  por  quienes  atendiamos  las  redes.  A  las  siete  horas  de  la  tarde 
atrapamos  los  ejemplares.  Posteriormente  las  capturas  disminnyeron.  Lo  mismo 
acontecio  en  la  Finca  La  Carolina,  completamente  abandonada. 

En  todos  los  craneos  examinados  se  encuentra  el  tercer  molar  diminuto, 
excepto  en  uno.  Los  ejemplares  registrados  en  este  trabajo  demuestran  la 
presencia  de  este  murcielago  en  la  Provincia  de  Jujuy. 

Familia  Desmodontidae 
Desmodus  rotundus  rotundus  (E.  Geoffroy  St.-Hilaire) 

Vampiro  de  patas  pelonas,  chupador  o  mordedor  de  Azara 
Ejemplares  examinados  (37). — Provincia  de  Salta:  Cueva  de  Murcie- 
lagallo,  15  km.  SSO  Santa  Victoria,  2000  m.,  5;  30  km.  SSO  General  M.  M. 
de  Giiemes,  6;  Toma  de  los  Laureles,  6  km.  SSO  Chicoana,  1400  m.,  4;  Cueva 
del  Indio,  4  km.  O  Cafayate,  1800  m.,  1.  Provincia  de  Jujuy:  Angosto  El 
Duraznal,  15  km.  S  Palma  Sola,  1225  m.,  4;  Cueva  del  Tigre,  74  km.  N  Pampa 
Blanca  (por  carretera),  700  m.,  3;  Finca  Catamontana,  33  km.  SSE  San 
Salvador  de  Jujuy,  925  m.,  2;  Arroyo  la  Urbana,  45  km.  E  y  5.4  km.  N  San 
Salvador  de  Jujuy,  620  m.,  1.  Provincia  del  Chaco:  Colonia  Benitez,  20  km. 
N  Resistencia,  30  m.,  3.  Provincia  de  Misiones:  Colonia  Martires,  Chacra 
Ferreira,  125  m.,  6.  Provincia  de  Cordoba:  Tabaquillo,  15  km.  E  (por  car- 
retera) Cruz  del  Eje,  Departamento  Cruz  del  Eje,  1;  Cordoba,  1. 

Adenitis,  en  el  Museo  de  Historia  Natural  "Bernardino  Rivadavia"  cxisten 
ejemplares  de  la  Provincia  de  Cordoba  y  de  la  Provincia  de  San  Luis.  House 
(1953:13)  hace  saber  que  es  comun  en  la  Provincia  de  Coquimbo  y  que  se 
ha  esparcido  hasta  el  paralelo  34,  al  norte  de  Rancagua,  y  Mann  (1951:3) 
colecto  ejemplares  en  la  Caleta  de  Cuya,  Provincia  de  Tarapaca. 

Observaciones. — Osgood  (1912:63)  encontro  marcadas  diferencias  en  el 
tamano,  comparando  ejemplares  tipicos  de  Desmodus  rotundus  de  Paraguay  y 


418  Misc.  Publ.  51,  Uxiv.  Kansas  Mus.  Nat.  Hist. 

ejemplares  de  Mexico  y  Centro  America.  Esto  ha  servido  de  base  para  usar 
el  nombre  Desmodus  rotundus  miirinus,  para  los  murcielagos  vampiros  de  la 
America  Central  y  de  Mexico,  y  el  de  Desmodus  rotundus  rotundus  para  los 
de  toda  la  America  del  Siir,  segun  las  conclusiones  de  Cabrera  (1958:93). 

Los  limites  de  separacion  entre  estas  dos  razas  no  estan,  sin  embargo, 
delimitados.  Es  casi  seguro  que  la  zona  de  integracion  se  halle  en  el  norte  de 
Sudamerica,  tal  \ez  a  lo  largo  del  flanco  noroeste  de  la  Cordillera  Occidental  de 
Colombia  y  Ecuador.  Es  interesante  notar,  como  expresa  Villa-R.  (1966:327), 
que  estas  subespecies  no  difieren  en  sus  caracteres  externos  marcadamente. 

Haciendo  uso  de  un  exposimetro  Weston  Photronic  Exposure  Meter 
(Modelo  650)  y  tomando  como  fuente  de  luz  la  del  sol  en  el  cenit  de  la 
Ciudad  de  Mexico,  a  las  12  boras  del  dia,  los  resultados  fueron  iguales  asl 
en  los  ejemplares  mexicanos,  como  en  los  argentinos.  La  cifra  obtenida  fue 
in\'ariablemente  de  25,  de  manera  que  no  encontramos  diferencias  en  la  luz 
reflejada  por  el  pelaje  del  dorso,  ni  en  los  machos  ni  en  las  hembras  de  los 
dos  grupos  comparados.  Con  excepcion  de  la  fuente  de  luz,  el  procedimiento 
para  esta  operacion  fue  el  mismo  que  explica  Desha  (1965:233-236).  La 
usamos  con  el  proposito  de  encontrar  diferencias  en  las  tonalidades  de  la 
coloracion.  Por  otra  parte,  las  medidas  de  pata,  antebrazo  y  tibia,  en  los 
ejemplares  mexicanos,  son  francamente  menores  que  las  de  los  ejemplares  de 
Argentina.  Esto  ya  habia  sido  observado  por  Goldman  (1920:209),  quien 
explica  que  en  tanto  que  los  individuos  son  practicamente  indistinguil:)les,  la 
raza  del  sur,  en  promedio,  es  considerablemente  mayor,  siendo  mas  notable 
la  diferencia  en  las  medidas  del  craneo. 

Se  pudo  obser\'ar  en  las  medidas  craneales,  que  excepto  la  longitud  basal, 
la  anchura  interorbitaria,  y  la  anchura  del  rostro,  todas  son  mas  grandes  en 
los  ejemplares  sudamericanos,  tomando  en  cuenta  la  media  aritmctica,  la 
maxima  y  la  minima.  En  los  ejemplares  de  otros  pai'ses  de  Sudamerica, 
aparte  de  Argentina,  que  existen  en  las  colecciones  del  Instituto  de  Biologia, 
Unix'ersidad  Nacional  Autonoma  de  Mexico,  se  observa  la  misma  tendencia. 

En  el  norte  de  Argentina,  el  refugio  diiuno  de  los  \'ampiros  es  el  hueco 
de  los  arboles,  con  frecuencia  en  el  tronco  de  los  "yuchanes,"  CJwrisia  insignis 
y  en  el  de  los  "guayacanes"  Caesalpinia  paraguaijensis  donde  encuentran 
temperatura  y  humedad  relatixa  constantes.  Los  brocales  de  los  pozos,  donde 
la  vegetacion  ha  sido  severamente  pertiubada,  son  tambien  lugares  de  abrigo, 
contaminando  el  agua  con  sus  deyecciones  e  inutilizandola  para  uso  humano. 

Familia  Vespertilionidae 
Myotis  chiloensis  atacamensis  (Lataste) 
Murcielago  insectixoro 
Ejemplares  examinados  (2). — Provixcia  de  Salta:    Finca  La  Rosa,  Cafa- 
yate,  1;  Rio  Blanco,  35  km.  SO  Salta,  1600  m.,  1. 

Medidas. — Las  medidas  somaticas  de  los  ejemplares  nos.  9594  y  9592 
(hembra  y  macho)  son  las  siguientes:  longitud  total  del  cuerpo,  85.3,  83.4; 
cola  vertebral,  40.2,  41.6;  oreja  desde  la  escotadura,  6.8,  7.7;  pata  trasera, 
12.2,  11.0;  antebrazo,  37.4,  37.2;  tibia,  14.3,  14.3. 

Las  medidas  craneales  del  craneo  no.  9592  son  las  siguientes:  longitud 
mayor   del    craneo,    14.9;    longitud    condilobasal,    14.2;    longitud    palatal,    7.3; 


ViLLA-R.  AND  Villa  Cornejo — Argentine  Bats  419 

anchura  bicigomatica,  8.2;  anchura  interorbitaria,  3.1;  anchura  del  rostro,  4.5; 
anchura  mastoidea,  7.2;  anchura  caja  cianeal,  6.4;  anchura  a  traves  de  los 
caninos,  3.4;  anchura  a  traves  de  M3-M3,  5.5. 

Observaciones. — Se  nos  informo  que  los  murcielagos  de  esta  especie  .son 
aliundantes  en  las  casas  abandonadas.  El  que  lleva  el  numero  9594  fue  cap- 
turado  en  el  techo  de  la  Finca  La  Rosa,  alguna  vez  senorial,  pero  en  el 
momento  de  la  captura,  una  ruina,  donde  si  es  cierto  que  observamos  gran 
numero  de  e.xcrementos,  los  murcielagos  habian  desaparecido.  Solo  estaba  el 
que  se  menciona  en  este  trabajo  y  no  tenia  emljrion.  El  numero  9592  ( macho ) 
tenia  los  testiculos  no  escrotados  y  tambien  se  le  observo  aislado. 

Myotis  albescens    (E.  GeofFroy  St.-Hilaire) 
Murcielago  insectivoro 

Ejcmplares  examinados  (1). — Provixcia  de  Salta:    Casa  habitada,  Salta. 

Mcdidas. — Las  medidas  del  ejemplar  (no.  9523,  macho)  son  las  siguientes: 
longitud  total  del  cuerpo,  82.0;  cola  \ertebral,  35.0;  oreja  desde  la  escoladura, 
8.0;  pata  trasera,  12.0;  antebrazo,  35.3;  tibia,  14.1;  longitud  mayor  del  craneo, 
14.1;  longitud  condilobasal,  13.2;  longitud  palatal,  6.4;  anchura  del  rostro,  4.3; 
anchura  mastoidea,  7.3;  anchura  caja  craneal,  7.8;  hilera  superior  de  dientes, 
4.8;  anchura  a  traves  de  los  caninos,  3.6;  anchura  a  tra\es  de  M3-M3,  5.4. 

Observaciones. — Nuestro  linico  ejemplar  es  de  coloracion  dorsal  pardo 
obscuro,  con  el  extremo  del  pelo  amarillento  palido,  de  modo  que  el  dorso, 
en  general,  es  Rood's  Brown,  lustroso.  No  encontramos  la  zona  desprovista  de 
pelo  en  la  nuca  a  que  se  refieren  Miller  y  Allen  (1928:203)  ni  recordamos 
haberle  visto  en  el  animal  recientemente  capturado. 

El  pelaje  de  la  region  ventral,  concuerda  con  la  descripcion  de  Miller  y 
Allen  {op.  cit.:  202),  esto  es,  la  base  de  los  pelos  es  de  color  chocolate  claro, 
con  la  porcion  terminal  blanquecina,  produciendo  una  superficie  mas  palida 
que  la  del  dorso.  Posteriormente  la  extension  blanquecina  de  la  punta  de  los 
pelos  aumenta  en  longitud  y  el  l^orde  del  abdomen  resulta  blanquecino.  La 
region  perianal  es  definiti\  amente  blanca,  presentando  el  mismo  efecto  descrito 
por  Acosta  y  Lara  (1950:7).  De  acuerdo  con  este  autor  {op.  cit.:  6),  la 
localidad  tipica  corresponderia  a  la  estancia  de  San  Solano,  junto  al  estero 
de  Ibera  (Sur  del  Rio  Parana),  en  la  Republica  Argentina  y  no  al  Paraguay. 

Myotis  nigricans  nigricans  (Schinz) 
Murcielago  insecti\oro 

Ejcmplares  examinados  (1). — Provixcia  de  Jujuy:  Arroyo  la  Urbana, 
45  km.  E  y  5.4  km.  N  San  Salvador  de  Jujuy,  620  m. 

Medidas. — Las  medidas  del  ejemplar  (no.  9520,  macho)  son  las  siguientes: 
longitud  total  del  cuerpo,  72.7;  cola  \ertebral,  35.0;  oreja  desde  la  escotadura, 
12.5;  pata  trasera,  6.8;  antebrazo,  .34.2;  tibia,  13.3;  longitud  mayor  del  craneo, 
13.5;  longitud  condilobasal,  12.3;  longitud  palatal,  7.9;  anchura  bicigomatica, 
8.2;  anchura  interorbitaria,  31.1;  anchura  del  rostro,  4.0;  anchura  mastoidea, 
6.6;  anchura  caja  craneal,  6.1;  hilera  superior  de  dientes,  4.9;  anchura  a  traves 
de  los  caninos,  3.2;  anchura  a  traves  de  M3-M3,  5.1. 

Observaciones. — Esta  pequena  especie  de  murcielagos  insectivoros  fue 
atrapada  en  una  red  de  seda,  junto  con  Lasiurus  y  Desmodus.    Durante  el 


420  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

tiempo  en  que  vigilamos  la  red,  extendida  a  traves  de  iin  peqiieno  remanso  del 
arroyo,  obserxamos  que  un  l)uen  numero  esquivaba  con  maestria  la  trampa. 
El  ejemplar  macho  que  se  registra  aqui,  se  atrapo  al  tiempo  en  que  im  ejemplar 
de  Lasiurus  horealis  pugnaba  por  escaparse  de  las  mallas  inferiores  de  la  red 
sumergida  en  el  agua,  dando  muestra  de  una  gran  habilidad  para  nadar.  Los 
frecuentes  chillidos  de  Lasiurus  parecen  haber  atraido  a  M.  n.  nigricans. 

No  lo  encontramos  formando  colonias,  en  ningtin  refugio  de  los  que  pudimos 
explorar;  es  posible  que  sea  abundante,  pero  durante  nuestra  visita  al  norte  de 
Argentina  nos  dejo  la  impresion  de  que  no  lo  era  mucho.  Por  el  contrario, 
en  las  cercanias  de  Rio  de  Janeiro,  Brasil,  en  el  mes  de  abril,  es  de  tal  manera 
abundante  que  penetra  a  las  haliitaciones  y  facilmente  se  les  captiira,  segun 
explica  Villa-R.  quien  los  derribaba  en  el  interior  de  la  habitacion  que  ocupaba 
en  la  Universidad  Rural  de  Brasil  solaniente  con  la  mano,  o  con  una  toalla. 

Eptesicus  innoxius  (Gervais) 
Murcielago  insectivoro 

Ejemplares  examinados  (2). — Provincia  de  Jujuy:    Palma  Sola,  550  m. 

Medidas. — Las  medidas  somaticas  de  los  ejemplares  ( 9532,  macho,  y  9534, 
macho)  son  las  siguientes:  longitud  total  del  cuerpo,  90.0,  9L0;  cola  vertebral, 
36.0,  38.0;  oreja  desde  la  escotadura,  13.0,  10.0;  pata  trasera,  9.0,  8.0;  ante- 
brazo,  38.9,  39.1;  tibia,  14.7,  14.8;  longitiid  mayor  del  craneo,  15.6,  15.5; 
longitiid  condilobasal,  14.9,  15.1;  longitud  palatal,  7.2,  7.1;  anchura  bicigo- 
matica,  9.5;  anchura  interorbitaria,  3.7,  3.6;  anchura  caja  craneal,  7.2,  6.9; 
hilera  superior  de  dientes,  5.6,  5.5;  anchura  a  traves  de  los  caninos,  4.5,  4.3; 
anchura  a  traves  de  M.3-M3,  6.2,  6.3. 

Ohservaciones. — El  tratamiento  sistematico  que  se  da  aqui  a  este  murcielago 
insectivoro,  esta  de  acuerdo  con  las  conclusiones  de  Davis  (1965:229-240)  en 
su  revision  del  complejo  Eptesicus  hrasilicnsis,  eiuien  explica  que  los  miembros 
del  complejo  examinado  por  el,  se  apartan  en  dos  grupos  basandose  en  el 
tamano  de  los  dientes  molariformes. 

E.  innoxius,  forma  parte  del  grupo  de  los  dientes  molariformes  pequeiios, 
con  E.  punicus  Thomas,  E.  furinalis  (D'Orbigny),  E.  chiralensis  Anthony, 
E.  melanopterus  (Jentink),  E.  fidelis  Thomas,  y  segiin  la  propia  e.xpresion 
de  Davis,  "probably  E.  diminutus  Osgood,  and  the  two  forms  in  Middle 
America."  Entre  el  material  que  se  obtuxo  en  el  norte  de  Argentina,  solo 
capturamos  los  dos  ejemplares  a  que  nos  venimos  refiriendo  y  a  no  ser  porque 
el  antebrazo  de  uno  de  ellos  es  de  39.1,  la  hilera  superior  de  dientes  queda 
dentro  de  las  medidas  que  registra  Davis  en  la  clave  sinoptica  correspondiente 
(op.  cit.:  239).  Es,  pues,  exidente,  que  el  area  de  distribucion  de  la  especie 
alcanza  esta  parte  de  la  Republica  Argentina. 

Nuestros  ejemplares  los  atrapamos  en  las  redes  japonesas  de  seda,  cuando 
habia  desaparecido  el  crepusculo  vespertino,  a  las  8  de  la  noche.  Las  redes  se 
colocaron  a  traves  de  im  estancamiento  somero  de  una  pequena  corriente  de 
agua.  El  bosque  mixto  dejaba  espacios  descubiertos  y  nuestro  sitio  de  trabajo 
no  estaba  distante  de  un  camino  que  eventualmente  servia  para  el  paso  de 
vehiculos  de  motor  cargados  con  trozos  de  madera. 

La  distribucion  de  esta  especies,  fue  limitado  al  oeste  de  Ecuador  y  Peru, 
incluyendo  la  isla  de  Puna,  por  Cabrera  (1958:107).  Segun  nuestro  registro, 
a  la  distribucion  conocida  debe  agregarse  el  norte  de  Argentina. 


ViLLA-R.  AND  Villa  Cornejo — Argentlne  Bats  421 

Histiotus  montanus  montanus  (Phillipi  y  Landbeck) 
Mmcielago  orejon 

Ejemplares  cxaminados  (6). — Provixcia  de  Jujuy:  Casa  Club  Nautico, 
Diqiie  la  Cienega,  30  km.  SSO  San  Salvador  de  Jujuy,  1000  m.,  2;  Finca  La 
Toma,  25  km.  SO  San  Salvador  de  Jujuy,  1.  Provincia  de  Salta:  Toma  de 
Los  Laureles,  6  km.  SSO  Chicoana,  1400  m.,  Depto.  Cliicoana,  Salta,  3. 

Medidas. — Las  medidas  de  los  ejemplares  9503  ( hembra ) ,  9504  ( hembra ) , 
9536    (macho),    9584    (macho),    9586    (macho),    y    9585    (macho)    son    las 

siguientes:    longitud  total  del  cuerpo,  150.0,  100.2,  109.4,  104.0,  112.3,  ; 

cola  vertebral,  40.0,  40.2,  50.5,  47.0,  53.7,  56.0;  pata  trasera,  31.7,  30.3,  34.6, 
36.5,  32.7,  37.5;  antebrazo,  46.4,  46.9,  46.4,  45.8,  48.1,  45.5;  tibia,  17.5,  17.5, 

17.0,  17.3,  17.1,  ;  longitud  mayor  del  craneo,  18.6,  17.9,  18.4,  18.5,  18.5, 

;  longitud  condilobasal,  17.4,  17.0,  17.1,  17.5,  17.5,  17.4;  longitud  palatal, 

8.4,  8.0,  7.2,  7.5,  9.1,  8.9;  anchura  bicigomatica,  10.7,  10.3,  10.0,  10.8,  10.5, 
10.6;  anchura  interorbitaria,  3.8,  3.5,  3.8,  3.9,  3.7,  3.7;  anchura  rostro,  6.1,  5.9, 
5.6,  5.9,  6.4,  6.0;  anchura  mastoidea,  9.6,  9.2,  9.3,  9.8,  9.3,  9.3;  anchura  caja 
craneal,  8.5,  7.8,  7.9,  8.3,  8.1,  8.2;  hilera  superior  de  dientes,  6.1,  5.7,  5.7,  5.9, 
5.8,  5.8;  anchura  a  traves  de  los  caninos,  4.8,  4.6,  4.5,  4.9,  4.7,  4.7. 

Observaciones. — Tres  de  los  ejemplares  examinados  aqui  se  les  capture  en 
el  techo  de  la  casa  del  Club  Nautico  de  las  cercanias  de  Perico  del  Carmen, 
en  redes  de  seda,  colocadas  sobre  el  techo  de  lamina,  paralelas  a  un  alero  de 
otro  techo  bajo  cuyas  laminas  se  refugiaban  estos  animales  en  gran  numero, 
en  con\i\encia  con  Tadarida  y  otros  molosidos.  Desde  las  siete  horas  de  la 
tarde  en  que  emergieron  los  primeros  indi\iduos,  la  acti\'idad  de  todos  los  que 
participamos,  fue  constante,  desprendiendo  de  las  redes  a  los  ejemplares  cap- 
turados.  La  temperatura  era  de  5°  C.  con  una  humedad  relativa  de  80  por 
ciento,  lo  que  causaba  una  incomoda  situacion  de  frio.  No  obstante,  los 
murcielagos  se  lanzaban  a  la  persecucion  de  sus  \'ictinias. 

Los  ejemplares  restantes,  tambien  se  obtuvieron  del  techo  de  casas  habi- 
tadas,  formando  grupos  numerosos. 

Lasiurus  borealis  varius  (Poeppig) 
Murcielago  rojizo 

Ejemplares  examinados  (1). — Provixcia  de  Salta:  Finca  La  Florida, 
Rosario  de  Lerma,  25  km.  SSO  Salta. 

Medidas. — Las  medidas  del  ejemplar  (9519,  macho)  son  las  siguientes: 
longitud  total  del  cuerpo,  101.5;  cola  vertebral,  45.8;  oreja  desde  la  escotua- 
dura,  10.0;  pata  trasera,  9.5;  antebrazo,  39.7;  tibia,  15.9;  longitud  mayor  del 
craneo,  116.0;  longitud  condilobasal,  113.0;  longitud  palatal,  5.5;  anchura 
bicigomatica,  8.9;  anchura  rostro,  5.3;  anchma  mastoidea,  7.6;  anchura  caja 
craneal,  7.3;  hilera  superior  de  dientes,  3.9;  anchura  a  traves  de  los  caninos, 
4.6;  anchura  a  traves  de  M3-M3,  5.6. 

Observaciones. — El  pelaje  de  la  membrana  interfemoral  es  denso  en  las 
cercanias  del  cuerpo,  largo  y  sedoso.  En  el  borde  es  ligeramente  mas  corto 
y  ralo;  el  color,  en  general,  es  Hazel  y  aparece  conro  fondo,  con  rayas  grisaseas 
o  Deep  Mouse  Gray;  en  los  hombros  este  color  es  predominante  en  la  super- 
ficie  y,  en  la  parte  superior  de  la  region  cer\'ical,  aparece  salpicando  el  fondo 
de  la  coloracion  superficial;  es  el  resultado  del  color  indi\idual  del  pelo  que 
en  su  base   es   obscuro,   seguido   de   una   amplia   banda   de   Ochraceous-Buff, 


422  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

terminando  en  Hazel  y  la  punta  de  coloracion  Deep  Mouse  Gray.  Ventialmente 
el  antebrazo  esta  recubierto  de  pelos  Ochraceous-Buff,  en  toda  su  longitud; 
en  la  porcion  distal  de  cada  femur,  el  pelo  es  muy  abundante.  La  parte  de  la 
niembrana  alar,  entre  los  costados  y  el  brazo,  tanibien  presentan  abundante 
pelo  como  una  prolongacion  del  pelaje  \'entral;  este  es  de  color  Othraceous- 
Buff.    El  ejemplar  fue  capturado  en  una  red  de  seda. 

Lasiurus  cinereus  villosissimus  (£.  GeofFroy  St.-Hilaire) 
Murcielago  canoso 

Ejemplares  examinados  (1). — Provixcia  de  Salta:  Finca  Belgrano,  Cer- 
rillos,  30  km.  SO  Salta. 

Medidas. — Las  medidas  somaticas  del  ejemplar  (9528,  macho)  son  las 
siguientes:  longitud  total  del  cuerpo,  150.0;  cola  vertebral,  54.5;  oreja  desde 
la  escotadura,  n.4;  pata  trasera,  12.6;  antebrazo,  52.1;  tibia,  20.0. 

Ohservaciones. — El  color  de  este  ejemplar  es  moreno  amarillento  variando 
de  moreno  caoba  salpicado  con  color  plateado,  dando  la  apariencia  de  canoso 
o  escarchado.  Como  expresa  Villa-R.  (1966:409-411),  en  la  base  del  pelo  pre- 
senta  un  color  Mummy  Brown,  en  la  porcion  media,  un  color  moreno  amari- 
llento y  en  la  punta  un  color  bianco  plateado.  El  plagiopatagio  esta  cubierto 
por  un  pelaje  denso,  desde  el  nivel  del  codo,  hasta  cerca  de  la  articulacion 
hrimero-carpal,  cubriendo  la  base  del  quinto  y  cuarto  dedos.  El  color  es  cerca- 
namente  Antimony  Yellow.  La  parte  ventral  es  semejante  a  la  dorsal.  La 
membrana  interfemoral  es  de  color  Cinnamon  Brown;  el  pelo  es  denso,  largo 
y  sedoso. 

La  adscripcion  de  este  material,  a  la  especie  L.  c.  villosissimus  se  establece 
aqui  sobre  la  base  de  precedentes  o  de  costmnbre,  por  las  siguientes  razones: 
En  la  misma  red  que  se  capturo  este  ejemplar,  fue  atrapado  otro  de  la  misma 
apariencia  externa  que,  al  tratar  de  desprenderlo,  se  escape  de  las  manos  de 
uno  de  nosotros  (Villa-R.);  llevaba  una  banda  de  aluminio,  muy  desgastada, 
en  el  antebrazo.  Por  tratar  de  examinar  con  menticulosidad  la  borrosa  ins- 
cripcion  de  esta  banda,  el  animal  se  escaliullo  y  no  hubo  ocasion  de  obtener 
la  informacion  pertinente.  Lo  anterior  nos  planteo  la  interesante  cuestion  de 
saber  si  se  trataba  de  la  misma  especie  neartica  L.  c.  cinereus  emigrado  hasta 
aquella  parte  de  Sudamerica,  tomando  en  cuenta  que  carecemos  de  informacion 
acerca  de  que  algima  persona  recientemente  haya  estado  anillando  con  propo- 
sitos  de  investigacion,  en  aquella  parte  del  Continente.  Por  otro  lado,  no  tene- 
mos  suficiente  material  para  hacer  coniparaciones  detenidas.  Con  los  ejemplares 
obtenidos  en  Mexico  y  almacenados  en  las  colecciones  del  laboratorio  de 
Mastozoologia  del  Institute  de  Biologia,  no  vemos  diferencias  fundamentales. 

Lasiurus  ega  argentinus  (Thomas) 
Murcielago  amarillo 
Ejemplares  examinados   (1). — Provixcia  de  Jujuy:     Arroyo   La   Urbana, 
45  km.  E  y  5.4  km.  N  (por  carretera),  620  m.,  San  Salvador  de  Jujuy. 

Medidas. — Las  medidas  del  ejemplar  (9521,  bembra)  son  las  siguientes: 
longitud  total  del  cuerpo,  121.3;  cola  vertebral,  42.0;  oreja  desde  la  escotadura, 
10.3;  pata  trasera,  17.7;  antebrazo,  48.2;  tibia,  IS.l;  longitud  mayor  del  craneo, 
16.1;  k)ngitud  condilobasal,  6.5;  longitud  palatal,  7.7;  anchura  bicigomatica, 
11.7;  anchura  interorbitaria,  4.2;  anchura  mastoidea,  9.4;   anchura  de  la  caja 


ViLLA-R.  AND  Villa  Cornejo — Argentine  Bats  423 

craneal,  8.4;  hilera  superior  de  dientes,  5.7;  anchura  a  tra\'cs  de  lo.s  caninos, 
6.6;  anchura  a  traves  de  M3-M3,  7.6. 

Obsc'ivaciones. — El  color  dorsal  es  amarillento,  llegando  al  Warm  Buff, 
tambien  se  e.xtiende  en  la  parte  ventral,  pero  en  la  piuita  de  la  membrana 
interfenioral,  el  pelo  se  vuelve  mas  amarillento,  llegando  al  Yellow  Ocher. 
Este  ejemplar  lo  atrapamos  en  una  red;  fue  de  los  primeros  en  caer. 

El  nombre  generico  adaptado  aqui  es  el  que  sigue  Dalquest  (1953:61)  y 
Handley  (1960:473)  por  las  misnias  razones  explicadas  por  Villa-R.  (1966: 
405).  Al  parecer,  estos  murcielagos  se  acomodan  durante  el  dia  entre  las  hojas 
secas  de  las  frondas  de  las  palmeras.  Se  les  ha  encontrado  en  el  techo  de  paja 
o  de  hojas  de  palma  de  las  casas  de  los  campesinos.  La  hembra  que  forma 
parte  de  nuestra  coleccion,  carecia  de  signos  de  reproduccion. 

Familia  Molossidae 

Molossops  temminckii  temminckii  (Burmeister) 

Murcielago  moloso 

Ejei7iplares  examinados   (2). — Provincia  de  Jujtjy:     Palma  Sola,  550  m. 

Medidas. — Las  medidas  de  los  ejemplares  9592  ( hembra )  y  9530  (  hembra ) 
son  las  siguientes:  longitud  total  del  cuerpo,  70.0,  69.0;  cola  vertebral,  22.0, 
27.0;  pata  trasera,  5.7,  6.7,  oreja  desde  la  escotadura,  11.7,  12.0;  antebrazo, 
30.3,  30.5;  tibia,  9.4,  10.0;  longitud  mayor  del  craneo,  12.8,  12.9;  longitud 
condilobasal,  11.0,  10.9;  longitud  palatal,  5.8,  5.6;  anchura  interorbitaria,  5.4, 
5.5;  anchura  del  rostro,  3.3,  3.2;  anchura  niastoidea,  8.0,  8.3;  anchura  caja 
craneal,  6.6,  6.7;  hilera  superior  de  dientes,  4.9,  5.9;  anchura  a  traves  de  los 
caninos,  4.9,  5.9;  anchura  a  traves  de  M3-M3,  5.8,  6.9. 

Observaciones. — Aunque  esta  especie  es  sumamente  de  pequefia  talla,  en 
todos  los  otros  respectos  superficiales  es  un  murcielago  tipico  de  la  familia 
Molossidae.  El  examen  del  craneo  re\el6,  sin  embargo,  su  verdadera  posicion 
sistematica.  Los  incisivos  superiores  aparecen  en  cercano  contacto  uno  con 
otro,  pero  separados  de  los  caninos  por  un  estrecho  espacio;  las  coronas  son 
delgadas,  fuertemente  encorvadas  hacia  adelante,  con  la  cara  anterior  suave- 
mente  convexa;  la  posterior,  ligeramente  concava.  Los  ejemplares  examinados, 
solo  presentan  incisivos  inferiores  mas  bajo  que  el  cingulo  del  canino  y  el  borde 
cortante  profundamente  bifido;  los  dos  dientes  emergen  de  alveolos  colocados 
enfrente  de  la  base  de  los  dos  caninos.  El  premaxilar  esta  complete  y  no 
presenta  espacio  libre  como  en  los  otros  generos  de  la  familia. 

El  25  de  junio  de  1965,  estos  dos  ejemplares  fueron  los  primeros  que 
cayeron  en  la  parte  superior  de  nuestra  red.  Otros  animales  de  la  misma  especie 
aparecian  con  frecuencia  \olando  en  torno,  pero  esquivaban  con  maestria  las 
mallas.  Ninguno  mostraba  signos  de  reproduccion  al  momento  de  su  captura. 
La  temperatura  era  fria  en  el  sitio  en  que  capturamos  a  estos  pequeiios 
murcielagos. 

Tadarida  brasiliensis  brasiliensis  (I.  Geoftroy  St.-Hilaire) 
Murcielago  guanero  o  de  cola  libre 
Ejemplares  examinados   (6). — Provincia  de  Salta:     Finca   La  Cruz,   28 
km.  SSE   (por  carretera)   Salta,  4;  Casa  Club  Nautico,  Dique  La  Cienega,  30 
km.  SSO  San  Sahador  de  Jujuy,  1000  m.,  2. 

Medidas. — Las  medidas  de  los  ejemplares  (machos),  9577,  9583,  9582,  y 
9507,  son  las  siguientes:    longitud  total  del  cueipo,  95.0,  92.0,  96.0,  90.0;  cola 


424  Misc.  PuBL.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

verteliral,  31.0,  32.0,  32.0,  .32.0;  oreja  desde  la  escotadura,  19.0,  19..5,  18.0, 
19.0;  antebra?o,  42.4,  44.7,  48.3,  42.,5;  tibia,  13.3,  12.4,  11.7,  10.9;  longitud 
mayor  del  craneo,  16.7,  17.1,  16.6,  16.6;  longitud  condilobasal,  15.8,  16.5,  15.8, 
16.6;  longitud  palatal,  7.0,  6.9,  6.9,  7.0;  anchura  interorbitaria,  3.8,  3.9,  4.0, 
4.1;  anchura  rostro,  5.8,  5.0,  5.6,  5.7;  anchura  mastoidea,  9.4,  9.0,  8.3,  9.3; 
anchura  caja  craneal,  8.1,  8.6,  8.1,  8.5;  hilera  superior  de  dientes,  5.9,  6.2, 
5.8,  5.9;  anchura  a  trave.s  de  los  caninos,  6.8,  7.1,  6.7,  6.9. 

Ohservaciones. — Tadarida  hrasiliensis  hrasdiensis  es,  sin  duda,  ima  especie 
muy  abundante  en  Sudamerica,  tanto  como  en  el  Sur  de  los  Estados  Unidos  y 
el  Norte  de  Me.xico.  En  este  trabajo,  con  finalidades  puramente  taxonomicas, 
solo  hemos  considerado  la  obser\aci6n  de  los  ejemplares  que  preparamos  en 
piel,  pero  una  gran  cantidad  se  entrego  a  los  laboratories  de  SELSA,  INTA,  y 
CEPAiXZO,  para  estndios  virologicos,  tomando  en  cuenta  que  en  los  Estados 
Unidos  de  Norteanierica  y  en  Mexico  son  los  que  con  mayor  frecuencia  y  en 
alto  porcentaje  se  han  halJado  infectados  con  \irus  rabico,  en  condiciones 
naturales.    Nuestro  proposito  era  que  se  estudiaran  en  estas  tres  instituciones. 

Es  muy  posible  que  en  la  Ciudad  de  Buenos  Aires,  su  poblacion  sea  extra- 
ordinarianiente  abundante  durante  el  verano.  Se  nos  informo  que  en  esta 
estacion  del  ano  constituyen  un  gran  problema  en  las  casas  con  techo  de  teja 
y  en  multitud  de  edificios  de  la  Capital  Federal. 

En  Iguazii,  cerca  de  las  Cataratas  de  este  nombre,  tambien  nos  dieron 
informes  interesantes  al  respecto.  Se  nos  explico  que,  durante  el  verano, 
aparece  una  extraordinaria  cantidad  de  estos  murcielagos,  acomodandose  entre 
los  tejados  de  las  casas.  Debido  a  su  abundancia,  se  les  ve  hasta  en  las  paredes 
del  interior  de  las  habitaciones.  Como  consecuencia,  los  gatos  domesticos  los 
dev'oran  y,  despues,  se  obser\a  que  enferman  mostrando  dificultad  en  mo\er 
el  tren  posterior;  algunas  veces  atacan  a  las  personas.  No  se  nos  explico  que 
se  hayan  presentando  casos  de  rabia  en  humanos,  pero  si  se  nos  informo  que 
posteriormente,  cuando  han  desaparecido  los  murcielagos,  aparecen  perros 
rabiosos.  Las  personas  que  nos  proporcionaron  esta  informacion,  nos  hicieron 
patente  su  extraiieza  per  la  aparicion  en  masa  de  los  murcielagos  en  epoca  de 
calor  y  luego  su  desaparicion  en  epoca  de  otoiio;  lo  curioso,  nos  dijeron,  es  que 
no  se  mueren  niuchos.  Solo  desaparecen.  Esto  puede  explicarse  por  el  hecho 
bien  conocido  en  Norteanierica  de  las  eniigraciones  de  la  subespecie  T.  h. 
mexicana.  Aunque  no  tenemos  informacion  de  que  se  hayan  efectuado  in- 
vestigaciones  acerca  de  los  movimientos  migratorios  de  T.  h.  hrasdiensis  en 
la  America  del  Sur,  es  posible  que  presenten  igual  comportamiento  biologico. 
Para  corroborar  esta  suposicion,  nos  parece  pertinente  citar  aqui,  que  en  el 
diario  "La  Nacion"  del  dia  3  de  abril  de  1968,  se  dio  la  informacion  que  en 
la  Ciudad  de  Neuquen,  un  establecimiento  escolar  fue  clausurado  debido  al 
extraordinario  numero  de  murcielagos  que  se  alojaban  en  "los  mil  y  un 
recovecos  del  \etusto  edificio  inagurado  en  el  ano  de  1910"  segiin  la  informa- 
cion. "Se  trata  de  la  Escuela  Nacional  No.  2 — sigiie  diciendo  el  diario-ubicado 
en  el  corazon  de  la  Ciudad."  Para  lograr  la  eliminacion  de  los  murcielagos, 
se  trabajo  durante  un  dia;  despues  de  tres  de  asueto  por  ese  motivo,  los  esco- 
lares  retornaron  a  la  aulas,  estimandose  que  la  escuela  habia  quedado  libre  de 
estos  aniniales.  El  diario  no  idendfica  a  los  murcielagos,  pero  Villa-R.  obserx-o 
a  niuchos  de  estos,  xolando  en  las  cercanlas  del  Hotel  Llao-Llao,  en  los  ultiiiios 
dias  de  marzo.    Este  Hotel  se  encuentra  a  25  kilometros  al  sur  de  San  Carlos 


ViLLA-R.  AND  Villa  Cornejo — Argentine  Bats  425 

de  Bariloche,   en   el   Parque   Nacional   Nahiial   Hiiapai;   es   posible,   r)or  tanto, 
que  se  trate  de  la  misma  especie. 

Acosta  y  Lara  (1950:42),  por  su  parte,  refiere  que  "Tadaiida  biasiliensis 
habita  tanto  en  las  viejas  y  aisladas  construcciones  de  nuestro  campo — 
Uruguay — como  en  lo  mas  centrico  de  las  ciudades  donde  se  puede  localizar 
sus  guaridas  en  las  torres  y  campanarios  de  varias  iglesias  bien  conocidas." 

Es  significativo  que  este  misnio  autor  refiere  que  la  "despoblacion  de  los 
nidos  puede  ser  motivada  por  el  desarrollo  de  epizootias,  como  sucede  con  otros 
aniniales  cuya  procreacion  es  grande  y  no  esta  contraloreada  por  la  existencia 
de  enemigos  naturales  proporcionalmente  fuertes  y  numerosos."  Es  bien 
conocido  el  hecho  de  que  en  Tadaiida  hrasiliensis  mexicana  se  ha  comprobado 
la  presencia  de  rabia. 

Eumops  perotis  perotis  (Schinz) 
Gran  murcielago  mastin 

Ejemplares  examinados  (3). — Provixcia  de  Salta:  Escuela  149,  30  km. 
NE  Salta,  1;  Dragones,  Chaco  Salteno,  1.  Provixcia  de  Jujuy:  Casa  Club  de 
Pesca,  Dique  La  Cienega,  30  km.  SSO  San  Salvador  de  Jujuy,  1000  m.,  1. 

Medidas. — Las  medidas  de  los  ejemplares  no.  9499  (macho),  no.  9580 
(hembra),  y  no.  9505  (macho)  son  las  siguientes:  longitud  total  del  cuerpo, 
170.7,  180.0,  185.0;  cola  vertebral,  51.8,  50.8,  62.3;  pata  trasera,  16.3,  15.0, 
19.5;  oreja  desde  la  escotadura,  36.8,  42.3,  40.0;  antebrazo,  78.6,  77.5,  79.2; 
tibia,  20.4,  20.6,  16.4;  longitud  mayor  del  craneo,  33.5,  33.4,  32.3;  longitud 
condilobasal,  32.7,  31.6,  31.1;  longitud  palatal,  14.9,  14.2,  13.9;  anchura 
bicigomatica,  19.1,  19.3,  18.9;  anchura  interorbitaria,  5.7,  6.1,  5.6;  anchura 
rostro,  8.9,  8.4,  7.3;  anchura  mastoidea,  15.8,  16.4,  15.5;  anchura  caja  craneal, 
13.5,  13.2,  13.1;  hilera  superior  de  dientes,  12.7,  12.2,  13.0;  anchura  a  traves 
de  los  caninos,  8.4,  8.1,  8.1;  anchura  a  traves  de  M3-M3,  12.9,  13.4,  12.6; 
anchura  del  rostro  entre  los  procesos  lacrimales,  10.2,  11.1,   10.4. 

Ohservacioncs. — El  e.xamen  de  los  ejemplares  mencionados  nos  hace  llegar 
a  la  conclusion  de  que  la  subespecie  Eumops  perotis  perotis  corresponde 
propiamente  a  Sudamerica.  Por  consiguiente,  y  de  acuerdo  con  Cockrum 
(1960:79),  los  murcielagos  mastines  de  Norteamerica,  representan  solo  una 
sola  subespecie,  Eumops  perotis  californicus.  Para  comparacion  solo  tenemos 
un  ejemplar,  colectado  en  el  techo  de  una  casa  habitada  de  la  Universidad 
Rural  de  Brasil  (km.  47),  Rio  de  Janeiro,  y  a  no  ser  por  la  coloracion  del 
pelaje,  que  en  este  es  hgeramente  mas  obscuro,  la  anchura  del  rostro  entre  los 
procesos  lacrimales,  varia  de  10.0  a  11.4  mm.,  como  lo  senala  Sanborn  (1932: 
351). 

Molossus  ater  nigricans  Miller 

Murcielago  moloso 

Ejemplares  examinados  (4). — Provixcia  de  Jujuy:  Casa  Club  Nautico, 
Dique  la  Cienega,  30  km.  SSO  San  Salvador  de  Jujuy,  1000  m.,  2.  Provixcia 
DE  Salta:    Ciudad  de  Salta,  2. 

Medidas. — Las  medidas  de  los  ejemplares  no.  9509  (hembra),  no.  95081 
(macho),  no.  95381  (hembra),  y  no.  9537  (hembra)  son  las  siguientes: 
longitud  total  del  cuerpo,  121.0,  125.0,  135.0,  121.0;  cola  vertebral,  46.0,  40.5, 
42.0,  43.4;  pata  trasera,  15.0,  11.8,  10.0,  13.5;  oreja  desde  la  escotadura,  15.0, 
11.4,  14.5,  14.1;  antebrazo,  48.7,  46.6,  48.4,  48.2;  tibia,  15.5,  15.8,  14.9,  15.1; 
longitud  mayor  del  craneo,  20.8,  21.6,  21.6,  20.4;  longitud  condilobasal,  19.3, 


426  Misc.  Publ.  51,  Uxiv.  Kansas  Mus.  Nat.  Hist. 

19.9,    19.6,    19.1;   longitud   palatal,   7.9,    8.1,   7.8,    8.0;    anchura   bicigomatica, 

12.7, , , ;  anchura  interorbitaria,  4.4,  4.7,  4. .5,  4.2;  anchura  rostro, 

5.0,  6.4,  6.4,  5.8;  anchura  mastoidea,  12.4,  1.3.3,  13.4,  11.8;  anchura  caja 
craneal,  10.4,  10.9,  10.6,  10.7;  hilera  superior  de  dientes,  7.8,  7.9,  8.2,  7.7; 
anchura  a  traves  de  M3-M3,  9.6,  10.1,  10.1,  9.9. 

Observaciones. — Los  dos  ejemplares  capturados  en  el  techo  de  la  casa  del 
Club  Nautico,  Jujuy,  adultos,  son  parte  de  un  numero  mayor  que  fue  atrapado 
en  una  red  de  seda,  junto  con  murcielagos  de  las  especies  Histiotus  moiitamis, 
Eumops  perotis  y  Tadarida  hrasiliensis,  acomodados  debajo  de  las  laminas  de 
zinc  del  techo.  La  captura  se  llevo  al  cabo  a  las  6:30  horas  de  la  tarde,  a  una 
temperatura  de  14.4°  C.  y  con  100  por  ciento  de  humedad  relativa,  produci- 
endo  una  sensacion  de  gran  incomodidad  por  el  frio  invernal  en  aquellas 
latitudes.  Los  otros  dos,  tambien  adultos,  fueron  Uevados  al  laboratorio  de 
\irus  de  SELSA,  por  personas  que  los  consideraban  vanipiros,  a  causa  de  su 
gran  talla. 

En  el  tratamiento  especifico  sigo  las  concluciones  de  Goodwin  (1960:6). 
Al  llevar  al  cabo  las  comparaciones  con  el  material  de  Me.xico  registrado  por 
Villa-R.  (1966:454-455)  y  que  se  conserva  en  las  Colecciones  del  Instituto  de 
Biologia  de  la  Universidad  Nacional  Autonoma  de  Me.xico,  no  he  hallado 
diferencias,  ni  en  los  caracteres  taxonomicos  externos,  ni  en  los  craneales.  Fue 
examinado,  ademas,  un  ejemplar,  el  no.  9817  (macho),  colectado  en  la  Gruta 
de  Limoiero,  200  m.  O  Limoiero,  Estado  de  Espiritu  Santo,  Brasil.  Este 
ejemplar  formaba  parte  de  mas  de  20  animales  que  fueron  obligados  a  dejar 
su  refugio  en  un  agujero  de  la  roca  calcarea  de  la  Gruta,  usando  una  antorcha. 

Comparaciones. — En  el  material  de  Me.xico,  el  dicromatismo  es  bien 
definido;  por  lo  tanto,  encontre  ejemplares  de  coloracion  dorsal  Bister  y  Slate- 
Black  predominantemente,  con  algunos  de  pelambre  color  Hazel  y  Auburn. 
El  material  de  Jujuy  y  Salta  es,  sin  excepcion,  de  coloracion  Bister;  el  ejemplar 
de  la  Gruta  de  Limoiero,  Brasil,  es  Slate-Black;  otros  murcielagos  de  esta 
especie  que  formalian  el  grupo  eran  de  color  Hazel. 

Los  promedios  de  las  medidas  somaticas  y  craneales,  la  minima  y  la 
maxima  entre  parentesis,  con  el  numero  de  ejemplares  tratados  a  continuacion 
de  los  datos  anteriores,  del  material  mexicano,  formado  por  ejemplares  adultos 
son  como  sigue:  Cola  vertebral,  48.0  (43.0-56),  17;  antebrazo,  50.6  (48.5- 
53.5),  20;  tibia,  15.0  (14.2-17.7),  20;  longitud  mayor  del  craneo,  21.8  (21.0- 
23.5),  20;  longitud  condiloba.sal,  20.3  (19.6-21.5),  10;  longitud  palatal,  8.0 
(7.7-8.8),  20;  anchura  bicigomatica,  14.0  (13.4-15.2),  18;  anchura  inter- 
orbitaria, 4.2  (4.1-4.8),  20;  anchura  del  rostro,  7.3  (6.5-8.3),  20;  anchura 
mastoidea,  13.2  (12.2-12.4),  20;  anchura  de  caja  craneal,  106.6  (10.0-11.8), 
20;  hilera  superior  de  dientes,  8.2  (7.8-8.6),  20;  anchura  a  traves  de  los 
caninos,  5.9  (5.7-6.4),  20;  anchura  a  traves  de  M3-M3,  10.0  (9.7-10.1),  21. 
La  longitud  de  la  cabeza  y  la  del  cuerpo  nos  parecen  inadecuadas  para 
propositos  de  tratamiento  estadistico,  porque  el  error  personal  y  la  naturaleza 
misma  de  las  partes  medidas,  les  hace  muy  variables. 

El  ejemplar  no.  9847  obtenido  en  Brasil  mide:  cola  vertebral,  55.5; 
antebrazo,  50.1;  tibia,  16.9;  longitud  mayor  del  craneo,  22.8;  longitud  condilo- 
basal,  21.1;  longitud  palatal,  8.1;  anchura  bicigomatica,  13.9;  anchura  inter- 
orbitaria, 4.8;  anchura  del  rostro,  7.9;  anchura  mastoidea,  14.1;  anchura  de  la 
caja  craneal,  11.2;  hilera  superior  de  dientes,  8.3;  anchura  a  traves  de  los 
caninos,  6.2;  anchura  a  traves  de  M3-M3,  10.0. 


ViLLA-R.  AND  Villa  Cornejo — Argentine  Bats  427 

Las  comparaciones  de  cstas  medidas  y  los  caracteres  externos,  hacen  evi- 
dente  que  el  material  del  norte  de  Argentina  que  colectamos  y  que  hemns 
estudiado,  coiresponde,  sin  duda,  a  la  subespecie  Molossus  ater  nigricans 
Miller,  1902. 

Summary 

In  June,  July,  and  August  of  1965,  the  senior  author  collected 
bats  in  northern  Argentina  in  conjunction  with  a  survey  of  rabies 
in  that  region.  Of  the  IS  species  obtained,  two  (Tonatia  silvicola 
and  Eptesicus  innoxius)  are  here  recorded  for  the  first  time  from 
Argentina;  the  distributions  of  several  other  species  are  extended 
geographically  within  the  country.  The  generalized  vegetational 
zones  of  northern  Argentina  arc  outlined  and  briefly  discussed  in 
an  introductory  section. 

Measurements  and  notes  on  natural  history  are  recorded  for 
each  of  the  18  kinds  of  bats,  and  comments  on  systematics  and 
incidence  of  rabies  are  incorporated  into  several  accounts.  Speci- 
mens of  Molossus  ater  are  referred  to  the  Mexican  subspecies 
(M.  a.  nigricans),  because  we  found  no  fundamental  differences 
between  Mexican  and  Argentinean  specimens. 

LiTERATURA  CiTADA 

AcosTA  Y  Lara,  E.  F. 

1950.  Quiropteros  del  Uruguay,  Com.  Zool.  Mus.  Hist.  Nat.  Montevideo, 
3(58):  1-71,  1  pi. 

1959.  Observaciones  sobre  una  colonia  de  Desmodus  rotundus  (E.  Geoff- 
roy)  en  el  Cerro  Salamanca,  Depto.  de  Maldonado.  Com.  Zool., 
Museo  Hist.  Nat.  Montevideo,  83(4):  1-2,  1  pi. 

Bernstein,  J. 

1952.  Portrait  of  a  vampire.    Nat.  Hist.,  61:82-87,  92-94,  9  figs. 

Cabrera,  A. 

1938.     Sobre  dos  murcielagos  nuevos  para  la  Argentina.    Inst.  Mus.  Univ. 

Nac.  Plata,  3:5-14,  2  figs. 
1958.     Catalogo  de  los  mamiferos  de  la  America  del  Sur.  Mus.  Argentino 

de  Ciencias  Naturales  "Bernardino  Rivadavia,"  4:iv  +   1-307. 

Cabrera,  A.  L. 

1953.  Esquema  fitogeografico  de  la  Republica  Argentina.  Revista  Mus. 
Cd.  Eva  Peron,  nueva  .serie,  8:87-168. 

Castellanos,  a.,  y  R.  A.  Perez-Moreau 

1944.  Los  tipos  de  vegetacion  de  la  Republica  Argentina.  Monog.  Inst. 
Estudios  Geograficos,  Univ.  Nac.  Tucuman,  364:1-154,  30  pis., 
1  map. 

COCKRUM,  E.  L. 

1960.  Distribution,  habitat  and  habits  of  the  mastiff'  bat,  Eumops  perotis, 
in  North  America.    Jour.  Arizona  Acad.  Sci.,  1:79-84,  1  fig. 

Dalquest,  W.  W. 

1953.     Mexican  bats  of  the  genus  Artibeus.    Proc.  Biol.  Soc.  Washington, 
66:61-66. 
Davis,  W.  B. 

1965.  Review  of  the  Eptesicus  hrasiliensis  complex  in  Middle  America 
with  the  description  of  a  new  subspecies  from  Costa  Rica.  Jour. 
Mamm.,  46:229-240,  5  figs. 


428  Misc.  Publ.  51,  Univ.  Kansas  Mus.  Nat.  Hist. 

Desha,  P.  G. 

1965.  A  new  colorimetric  technique  for  small  mammals.  Texas  Jour.  Sci., 
17:233-236,  1  fig. 

FoRXEs,  A.,  Elio  Massoia,  y  Guillermo  Forrest 

1967.  Toiiatia  sylvicola  (D'Orbigny)  nuevo  genero  y  especie  para  la 
Republica  Argentina  (Ghiroptera  Phyllostomidae ) .  Physis,  37:149- 
152. 

GOLUMAX,   E.    A. 

1920.  Mammals  of  Panama.  Smithsonian  Misc.  Coll.,  69:1-309,  24  figs., 
39  pis. 

Goodwin,  G.  G. 

1960.  The  status  of  Vespertilio  auripendulus  Shaw,  1800,  and  Molossus 
ater  GeofFroy,  1805.    Amer.  Mus.  Novit.,  1994:1-6,  1  fig. 

Handley,  C.  O.,  Jr. 

1960.  Descriptions  of  new  bats  from  Panama.  Proc.  U.S.  Nat.  Mus.,  112: 
459-479. 

Hauman,  L. 

1917.  Notes  floristiques.  Ann.  Mus.  Nac.  Hist.  Nat.,  Buenos  Aires,  29: 
391-444. 

House,  F. 

1953.     Animales  salvajes  de  Chile  en  su  clasificacion  moderna.    Edic.  Univ. 
Chile,  vi  +  7-189  pp. 
Mann,  F.  G. 

1951.     Biologia  del  vampiro.    Inst.  Biol.,  Fac.  Biol,  y  Ciencias  Med.,  Chile, 
nos.  12-13,  pp.  3-24,  12  figs.,  4  pis. 
Miller,  G.  S.,  Jr.,  and  G.  M.  Allen 

1928.     The  American  bats  of  the  genera  Myotis  and  Pizonvx.    Bull.  U.S. 
Nat.  Mus.,  44:viii  +  1-218,  1  fig.,  13  maps. 
Osgood,  W.  H. 

1912.  Mammals  from  western  Venezuela  and  eastern  Colombia.  Field 
Mus.  Nat.  Hist.,  Zool.  Ser.,  10:33-66. 

Ridgway,  R. 

1912.  Color  standards  and  color  nomenclature.  Privately  published, 
Wa.shington,  D.C.,  iv  -|-  44  pp.,  53  pis. 

Sanborn,  C.  C. 

1932.     The  bats  of  the  genus  Eumops.   Jour.  Mamm.,  13:347-357. 

ViLLA-R.,   B. 

1966.  Los  murcielagos  de  Me.xico.  .  .  .  Univ.  Nac.  Aut.  Mexico,  Inst. 
Biol.,  xvi  +  3-491,  171  figs.,  98  maps. 


3  2044  093  361    608 


Date  Due 


-^('fHo